diff options
| author | Linus Torvalds <torvalds@linux-foundation.org> | 2014-06-09 18:10:34 -0700 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2014-06-09 18:10:34 -0700 |
| commit | 82abb273d838318424644d8f02825db0fbbd400a (patch) | |
| tree | e1ea8a92db4ba68f347249986ffe3a25ffbf8219 | |
| parent | 9b651cc2277b5e4883012ebab0fea2bcda4cbafa (diff) | |
| parent | f8647b506d7116a1a3accd8d618184096e85f50b (diff) | |
Merge branch 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus
Pull MIPS updates from Ralf Baechle:
- three fixes for 3.15 that didn't make it in time
- limited Octeon 3 support.
- paravirtualization support
- improvment to platform support for Netlogix SOCs.
- add support for powering down the Malta eval board in software
- add many instructions to the in-kernel microassembler.
- add support for the BPF JIT.
- minor cleanups of the BCM47xx code.
- large cleanup of math emu code resulting in significant code size
reduction, better readability of the code and more accurate
emulation.
- improvments to the MIPS CPS code.
- support C3 power status for the R4k count/compare clock device.
- improvments to the GIO support for older SGI workstations.
- increase number of supported CPUs to 256; this can be reached on
certain embedded multithreaded ccNUMA configurations.
- various small cleanups, updates and fixes
* 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus: (173 commits)
MIPS: IP22/IP28: Improve GIO support
MIPS: Octeon: Add twsi interrupt initialization for OCTEON 3XXX, 5XXX, 63XX
DEC: Document the R4k MB ASIC mini interrupt controller
DEC: Add self as the maintainer
MIPS: Add microMIPS MSA support.
MIPS: Replace calls to obsolete strict_strto call with kstrto* equivalents.
MIPS: Replace obsolete strict_strto call with kstrto
MIPS: BFP: Simplify code slightly.
MIPS: Call find_vma with the mmap_sem held
MIPS: Fix 'write_msa_##' inline macro.
MIPS: Fix MSA toolchain support detection.
mips: Update the email address of Geert Uytterhoeven
MIPS: Add minimal defconfig for mips_paravirt
MIPS: Enable build for new system 'paravirt'
MIPS: paravirt: Add pci controller for virtio
MIPS: Add code for new system 'paravirt'
MIPS: Add functions for hypervisor call
MIPS: OCTEON: Add OCTEON3 to __get_cpu_type
MIPS: Add function get_ebase_cpunum
MIPS: Add minimal support for OCTEON3 to c-r4k.c
...
256 files changed, 8562 insertions, 7905 deletions
diff --git a/MAINTAINERS b/MAINTAINERS index 454c054b179..77556fc91cd 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2700,6 +2700,15 @@ S: Orphan F: Documentation/networking/decnet.txt F: net/decnet/ +DECSTATION PLATFORM SUPPORT +M: "Maciej W. Rozycki" <macro@linux-mips.org> +L: linux-mips@linux-mips.org +W: http://www.linux-mips.org/wiki/DECstation +S: Maintained +F: arch/mips/dec/ +F: arch/mips/include/asm/dec/ +F: arch/mips/include/asm/mach-dec/ + DEFXX FDDI NETWORK DRIVER M: "Maciej W. Rozycki" <macro@linux-mips.org> S: Maintained diff --git a/arch/mips/Kbuild b/arch/mips/Kbuild index d2cfe45f332..dd295335891 100644 --- a/arch/mips/Kbuild +++ b/arch/mips/Kbuild @@ -16,7 +16,7 @@ obj- := $(platform-) obj-y += kernel/ obj-y += mm/ -obj-y += math-emu/ +obj-y += net/ ifdef CONFIG_KVM obj-y += kvm/ diff --git a/arch/mips/Kbuild.platforms b/arch/mips/Kbuild.platforms index 6e239123d6f..f5e18bf3275 100644 --- a/arch/mips/Kbuild.platforms +++ b/arch/mips/Kbuild.platforms @@ -18,6 +18,7 @@ platforms += loongson1 platforms += mti-malta platforms += mti-sead3 platforms += netlogic +platforms += paravirt platforms += pmcs-msp71xx platforms += pnx833x platforms += ralink diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index 5e0014e864f..7a469acee33 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -12,6 +12,7 @@ config MIPS select HAVE_ARCH_KGDB select HAVE_ARCH_SECCOMP_FILTER select HAVE_ARCH_TRACEHOOK + select HAVE_BPF_JIT if !CPU_MICROMIPS select ARCH_HAVE_CUSTOM_GPIO_H select HAVE_FUNCTION_TRACER select HAVE_FUNCTION_TRACE_MCOUNT_TEST @@ -50,6 +51,8 @@ config MIPS select CLONE_BACKWARDS select HAVE_DEBUG_STACKOVERFLOW select HAVE_CC_STACKPROTECTOR + select CPU_PM if CPU_IDLE + select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST menu "Machine selection" @@ -83,6 +86,7 @@ config AR7 select SYS_HAS_EARLY_PRINTK select SYS_SUPPORTS_32BIT_KERNEL select SYS_SUPPORTS_LITTLE_ENDIAN + select SYS_SUPPORTS_MIPS16 select SYS_SUPPORTS_ZBOOT_UART16550 select ARCH_REQUIRE_GPIOLIB select VLYNQ @@ -106,6 +110,7 @@ config ATH79 select SYS_HAS_EARLY_PRINTK select SYS_SUPPORTS_32BIT_KERNEL select SYS_SUPPORTS_BIG_ENDIAN + select SYS_SUPPORTS_MIPS16 help Support for the Atheros AR71XX/AR724X/AR913X SoCs. @@ -122,6 +127,7 @@ config BCM47XX select NO_EXCEPT_FILL select SYS_SUPPORTS_32BIT_KERNEL select SYS_SUPPORTS_LITTLE_ENDIAN + select SYS_SUPPORTS_MIPS16 select SYS_HAS_EARLY_PRINTK select USE_GENERIC_EARLY_PRINTK_8250 help @@ -168,9 +174,9 @@ config MACH_DECSTATION bool "DECstations" select BOOT_ELF32 select CEVT_DS1287 - select CEVT_R4K + select CEVT_R4K if CPU_R4X00 select CSRC_IOASIC - select CSRC_R4K + select CSRC_R4K if CPU_R4X00 select CPU_DADDI_WORKAROUNDS if 64BIT select CPU_R4000_WORKAROUNDS if 64BIT select CPU_R4400_WORKAROUNDS if 64BIT @@ -248,6 +254,7 @@ config LANTIQ select SYS_HAS_CPU_MIPS32_R2 select SYS_SUPPORTS_BIG_ENDIAN select SYS_SUPPORTS_32BIT_KERNEL + select SYS_SUPPORTS_MIPS16 select SYS_SUPPORTS_MULTITHREADING select SYS_HAS_EARLY_PRINTK select ARCH_REQUIRE_GPIOLIB @@ -330,6 +337,7 @@ config MIPS_MALTA select SYS_SUPPORTS_LITTLE_ENDIAN select SYS_SUPPORTS_MIPS_CMP select SYS_SUPPORTS_MIPS_CPS + select SYS_SUPPORTS_MIPS16 select SYS_SUPPORTS_MULTITHREADING select SYS_SUPPORTS_SMARTMIPS select SYS_SUPPORTS_ZBOOT @@ -361,6 +369,7 @@ config MIPS_SEAD3 select SYS_SUPPORTS_LITTLE_ENDIAN select SYS_SUPPORTS_SMARTMIPS select SYS_SUPPORTS_MICROMIPS + select SYS_SUPPORTS_MIPS16 select USB_EHCI_BIG_ENDIAN_DESC select USB_EHCI_BIG_ENDIAN_MMIO select USE_OF @@ -380,6 +389,7 @@ config MACH_VR41XX select CEVT_R4K select CSRC_R4K select SYS_HAS_CPU_VR41XX + select SYS_SUPPORTS_MIPS16 select ARCH_REQUIRE_GPIOLIB config NXP_STB220 @@ -407,6 +417,7 @@ config PMC_MSP select SYS_HAS_CPU_MIPS32_R2 select SYS_SUPPORTS_32BIT_KERNEL select SYS_SUPPORTS_BIG_ENDIAN + select SYS_SUPPORTS_MIPS16 select IRQ_CPU select SERIAL_8250 select SERIAL_8250_CONSOLE @@ -430,6 +441,7 @@ config RALINK select SYS_HAS_CPU_MIPS32_R2 select SYS_SUPPORTS_32BIT_KERNEL select SYS_SUPPORTS_LITTLE_ENDIAN + select SYS_SUPPORTS_MIPS16 select SYS_HAS_EARLY_PRINTK select HAVE_MACH_CLKDEV select CLKDEV_LOOKUP @@ -674,7 +686,6 @@ config SNI_RM select SYS_SUPPORTS_BIG_ENDIAN select SYS_SUPPORTS_HIGHMEM select SYS_SUPPORTS_LITTLE_ENDIAN - select USE_GENERIC_EARLY_PRINTK_8250 help The SNI RM200/300/400 are MIPS-based machines manufactured by Siemens Nixdorf Informationssysteme (SNI), parent company of Pyramid @@ -721,6 +732,11 @@ config CAVIUM_OCTEON_SOC select ZONE_DMA32 select HOLES_IN_ZONE select ARCH_REQUIRE_GPIOLIB + select LIBFDT + select USE_OF + select ARCH_SPARSEMEM_ENABLE + select SYS_SUPPORTS_SMP + select NR_CPUS_DEFAULT_16 help This option supports all of the Octeon reference boards from Cavium Networks. It builds a kernel that dynamically determines the Octeon @@ -789,6 +805,25 @@ config NLM_XLP_BOARD This board is based on Netlogic XLP Processor. Say Y here if you have a XLP based board. +config MIPS_PARAVIRT + bool "Para-Virtualized guest system" + select CEVT_R4K + select CSRC_R4K + select DMA_COHERENT + select SYS_SUPPORTS_64BIT_KERNEL + select SYS_SUPPORTS_32BIT_KERNEL + select SYS_SUPPORTS_BIG_ENDIAN + select SYS_SUPPORTS_SMP + select NR_CPUS_DEFAULT_4 + select SYS_HAS_EARLY_PRINTK + select SYS_HAS_CPU_MIPS32_R2 + select SYS_HAS_CPU_MIPS64_R2 + select SYS_HAS_CPU_CAVIUM_OCTEON + select HW_HAS_PCI + select SWAP_IO_SPACE + help + This option supports guest running under ???? + endchoice source "arch/mips/alchemy/Kconfig" @@ -809,6 +844,7 @@ source "arch/mips/cavium-octeon/Kconfig" source "arch/mips/loongson/Kconfig" source "arch/mips/loongson1/Kconfig" source "arch/mips/netlogic/Kconfig" +source "arch/mips/paravirt/Kconfig" endmenu @@ -1059,6 +1095,7 @@ config SOC_PNX833X select SYS_SUPPORTS_32BIT_KERNEL select SYS_SUPPORTS_LITTLE_ENDIAN select SYS_SUPPORTS_BIG_ENDIAN + select SYS_SUPPORTS_MIPS16 select CPU_MIPSR2_IRQ_VI config SOC_PNX8335 @@ -1398,16 +1435,11 @@ config CPU_SB1 config CPU_CAVIUM_OCTEON bool "Cavium Octeon processor" depends on SYS_HAS_CPU_CAVIUM_OCTEON - select ARCH_SPARSEMEM_ENABLE select CPU_HAS_PREFETCH select CPU_SUPPORTS_64BIT_KERNEL - select SYS_SUPPORTS_SMP - select NR_CPUS_DEFAULT_16 select WEAK_ORDERING select CPU_SUPPORTS_HIGHMEM select CPU_SUPPORTS_HUGEPAGES - select LIBFDT - select USE_OF select USB_EHCI_BIG_ENDIAN_MMIO select MIPS_L1_CACHE_SHIFT_7 help @@ -1659,6 +1691,12 @@ config SYS_HAS_CPU_XLR config SYS_HAS_CPU_XLP bool +config MIPS_MALTA_PM + depends on MIPS_MALTA + depends on PCI + bool + default y + # # CPU may reorder R->R, R->W, W->R, W->W # Reordering beyond LL and SC is handled in WEAK_REORDERING_BEYOND_LLSC @@ -1842,7 +1880,7 @@ config FORCE_MAX_ZONEORDER config CEVT_GIC bool "Use GIC global counter for clock events" - depends on IRQ_GIC && !(MIPS_SEAD3 || MIPS_MT_SMTC) + depends on IRQ_GIC && !MIPS_SEAD3 help Use the GIC global counter for the clock events. The R4K clock event driver is always present, so if the platform ends up not @@ -1895,19 +1933,8 @@ config CPU_R4K_CACHE_TLB bool default y if !(CPU_R3000 || CPU_R8000 || CPU_SB1 || CPU_TX39XX || CPU_CAVIUM_OCTEON) -choice - prompt "MIPS MT options" - -config MIPS_MT_DISABLED - bool "Disable multithreading support" - help - Use this option if your platform does not support the MT ASE - which is hardware multithreading support. On systems without - an MT-enabled processor, this will be the only option that is - available in this menu. - config MIPS_MT_SMP - bool "Use 1 TC on each available VPE for SMP" + bool "MIPS MT SMP support (1 TC on each available VPE)" depends on SYS_SUPPORTS_MULTITHREADING select CPU_MIPSR2_IRQ_VI select CPU_MIPSR2_IRQ_EI @@ -1926,26 +1953,6 @@ config MIPS_MT_SMP Intel Hyperthreading feature. For further information go to <http://www.imgtec.com/mips/mips-multithreading.asp>. -config MIPS_MT_SMTC - bool "Use all TCs on all VPEs for SMP (DEPRECATED)" - depends on CPU_MIPS32_R2 - depends on SYS_SUPPORTS_MULTITHREADING - depends on !MIPS_CPS - select CPU_MIPSR2_IRQ_VI - select CPU_MIPSR2_IRQ_EI - select MIPS_MT - select SMP - select SMP_UP - select SYS_SUPPORTS_SMP - select NR_CPUS_DEFAULT_8 - help - This is a kernel model which is known as SMTC. This is - supported on cores with the MT ASE and presents all TCs - available on all VPEs to support SMP. For further - information see <http://www.linux-mips.org/wiki/34K#SMTC>. - -endchoice - config MIPS_MT bool @@ -1967,7 +1974,7 @@ config SYS_SUPPORTS_MULTITHREADING config MIPS_MT_FPAFF bool "Dynamic FPU affinity for FP-intensive threads" default y - depends on MIPS_MT_SMP || MIPS_MT_SMTC + depends on MIPS_MT_SMP config MIPS_VPE_LOADER bool "VPE loader support." @@ -1989,29 +1996,6 @@ config MIPS_VPE_LOADER_MT default "y" depends on MIPS_VPE_LOADER && !MIPS_CMP -config MIPS_MT_SMTC_IM_BACKSTOP - bool "Use per-TC register bits as backstop for inhibited IM bits" - depends on MIPS_MT_SMTC - default n - help - To support multiple TC microthreads acting as "CPUs" within - a VPE, VPE-wide interrupt mask bits must be specially manipulated - during interrupt handling. To support legacy drivers and interrupt - controller management code, SMTC has a "backstop" to track and - if necessary restore the interrupt mask. This has some performance - impact on interrupt service overhead. - -config MIPS_MT_SMTC_IRQAFF - bool "Support IRQ affinity API" - depends on MIPS_MT_SMTC - default n - help - Enables SMP IRQ affinity API (/proc/irq/*/smp_affinity, etc.) - for SMTC Linux kernel. Requires platform support, of which - an example can be found in the MIPS kernel i8259 and Malta - platform code. Adds some overhead to interrupt dispatch, and - should be used only if you know what you are doing. - config MIPS_VPE_LOADER_TOM bool "Load VPE program into memory hidden from linux" depends on MIPS_VPE_LOADER @@ -2039,7 +2023,7 @@ config MIPS_VPE_APSP_API_MT config MIPS_CMP bool "MIPS CMP framework support (DEPRECATED)" - depends on SYS_SUPPORTS_MIPS_CMP && !MIPS_MT_SMTC + depends on SYS_SUPPORTS_MIPS_CMP select MIPS_GIC_IPI select SYNC_R4K select WEAK_ORDERING @@ -2057,9 +2041,11 @@ config MIPS_CPS depends on SYS_SUPPORTS_MIPS_CPS select MIPS_CM select MIPS_CPC + select MIPS_CPS_PM if HOTPLUG_CPU select MIPS_GIC_IPI select SMP select SYNC_R4K if (CEVT_R4K || CSRC_R4K) + select SYS_SUPPORTS_HOTPLUG_CPU select SYS_SUPPORTS_SMP select WEAK_ORDERING help @@ -2069,6 +2055,9 @@ config MIPS_CPS no external assistance. It is safe to enable this when hardware support is unavailable. +config MIPS_CPS_PM + bool + config MIPS_GIC_IPI bool @@ -2199,6 +2188,13 @@ config SYS_SUPPORTS_SMARTMIPS config SYS_SUPPORTS_MICROMIPS bool +config SYS_SUPPORTS_MIPS16 + bool + help + This option must be set if a kernel might be executed on a MIPS16- + enabled CPU even if MIPS16 is not actually being used. In other + words, it makes the kernel MIPS16-tolerant. + config CPU_SUPPORTS_MSA bool @@ -2239,7 +2235,7 @@ config NODES_SHIFT config HW_PERF_EVENTS bool "Enable hardware performance counter support for perf events" - depends on PERF_EVENTS && !MIPS_MT_SMTC && OPROFILE=n && (CPU_MIPS32 || CPU_MIPS64 || CPU_R10000 || CPU_SB1 || CPU_CAVIUM_OCTEON || CPU_XLP) + depends on PERF_EVENTS && OPROFILE=n && (CPU_MIPS32 || CPU_MIPS64 || CPU_R10000 || CPU_SB1 || CPU_CAVIUM_OCTEON || CPU_XLP) default y help Enable hardware performance counter support for perf events. If @@ -2297,8 +2293,8 @@ config NR_CPUS_DEFAULT_64 bool config NR_CPUS - int "Maximum number of CPUs (2-64)" - range 2 64 + int "Maximum number of CPUs (2-256)" + range 2 256 depends on SMP default "4" if NR_CPUS_DEFAULT_4 default "8" if NR_CPUS_DEFAULT_8 @@ -2671,12 +2667,16 @@ endmenu config MIPS_EXTERNAL_TIMER bool -if CPU_SUPPORTS_CPUFREQ && MIPS_EXTERNAL_TIMER menu "CPU Power Management" + +if CPU_SUPPORTS_CPUFREQ && MIPS_EXTERNAL_TIMER source "drivers/cpufreq/Kconfig" -endmenu endif +source "drivers/cpuidle/Kconfig" + +endmenu + source "net/Kconfig" source "drivers/Kconfig" diff --git a/arch/mips/Kconfig.debug b/arch/mips/Kconfig.debug index 25de29211d7..3a2b775e845 100644 --- a/arch/mips/Kconfig.debug +++ b/arch/mips/Kconfig.debug @@ -79,15 +79,6 @@ config CMDLINE_OVERRIDE Normally, you will choose 'N' here. -config SMTC_IDLE_HOOK_DEBUG - bool "Enable additional debug checks before going into CPU idle loop" - depends on DEBUG_KERNEL && MIPS_MT_SMTC - help - This option enables Enable additional debug checks before going into - CPU idle loop. For details on these checks, see - arch/mips/kernel/smtc.c. This debugging option result in significant - overhead so should be disabled in production kernels. - config SB1XXX_CORELIS bool "Corelis Debugger" depends on SIBYTE_SB1xxx_SOC diff --git a/arch/mips/Makefile b/arch/mips/Makefile index 60a359cfa32..a8521de1479 100644 --- a/arch/mips/Makefile +++ b/arch/mips/Makefile @@ -120,7 +120,7 @@ cflags-$(CONFIG_SB1XXX_CORELIS) += $(call cc-option,-mno-sched-prolog) \ -fno-omit-frame-pointer ifeq ($(CONFIG_CPU_HAS_MSA),y) -toolchain-msa := $(call cc-option-yn,-mhard-float -mfp64 -mmsa) +toolchain-msa := $(call cc-option-yn,-mhard-float -mfp64 -Wa$(comma)-mmsa) cflags-$(toolchain-msa) += -DTOOLCHAIN_SUPPORTS_MSA endif @@ -251,6 +251,7 @@ OBJCOPYFLAGS += --remove-section=.reginfo head-y := arch/mips/kernel/head.o libs-y += arch/mips/lib/ +libs-y += arch/mips/math-emu/ # See arch/mips/Kbuild for content of core part of the kernel core-y += arch/mips/ diff --git a/arch/mips/alchemy/board-xxs1500.c b/arch/mips/alchemy/board-xxs1500.c index bd551365029..3fb814be0e9 100644 --- a/arch/mips/alchemy/board-xxs1500.c +++ b/arch/mips/alchemy/board-xxs1500.c @@ -49,7 +49,7 @@ void __init prom_init(void) prom_init_cmdline(); memsize_str = prom_getenv("memsize"); - if (!memsize_str || strict_strtoul(memsize_str, 0, &memsize)) + if (!memsize_str || kstrtoul(memsize_str, 0, &memsize)) memsize = 0x04000000; add_memory_region(0, memsize, BOOT_MEM_RAM); diff --git a/arch/mips/alchemy/common/setup.c b/arch/mips/alchemy/common/setup.c index 566a1743f68..8267e3c9772 100644 --- a/arch/mips/alchemy/common/setup.c +++ b/arch/mips/alchemy/common/setup.c @@ -67,6 +67,12 @@ void __init plat_mem_setup(void) case ALCHEMY_CPU_AU1500: case ALCHEMY_CPU_AU1100: coherentio = 0; + break; + case ALCHEMY_CPU_AU1200: + /* Au1200 AB USB does not support coherent memory */ + if (0 == (read_c0_prid() & PRID_REV_MASK)) + coherentio = 0; + break; } board_setup(); /* board specific setup */ diff --git a/arch/mips/alchemy/common/usb.c b/arch/mips/alchemy/common/usb.c index 2adc7edda49..d193dbea84a 100644 --- a/arch/mips/alchemy/common/usb.c +++ b/arch/mips/alchemy/common/usb.c @@ -355,47 +355,25 @@ static inline void __au1200_udc_control(void __iomem *base, int enable) } } -static inline int au1200_coherency_bug(void) -{ -#if defined(CONFIG_DMA_COHERENT) - /* Au1200 AB USB does not support coherent memory */ - if (!(read_c0_prid() & PRID_REV_MASK)) { - printk(KERN_INFO "Au1200 USB: this is chip revision AB !!\n"); - printk(KERN_INFO "Au1200 USB: update your board or re-configure" - " the kernel\n"); - return -ENODEV; - } -#endif - return 0; -} - static inline int au1200_usb_control(int block, int enable) { void __iomem *base = (void __iomem *)KSEG1ADDR(AU1200_USB_CTL_PHYS_ADDR); - int ret = 0; switch (block) { case ALCHEMY_USB_OHCI0: - ret = au1200_coherency_bug(); - if (ret && enable) - goto out; __au1200_ohci_control(base, enable); break; case ALCHEMY_USB_UDC0: __au1200_udc_control(base, enable); break; case ALCHEMY_USB_EHCI0: - ret = au1200_coherency_bug(); - if (ret && enable) - goto out; __au1200_ehci_control(base, enable); break; default: - ret = -ENODEV; + return -ENODEV; } -out: - return ret; + return 0; } diff --git a/arch/mips/alchemy/devboards/pm.c b/arch/mips/alchemy/devboards/pm.c index b86bff31d1d..61e90fe9eab 100644 --- a/arch/mips/alchemy/devboards/pm.c +++ b/arch/mips/alchemy/devboards/pm.c @@ -158,7 +158,7 @@ static ssize_t db1x_pmattr_store(struct kobject *kobj, int tmp; if (ATTRCMP(timer_timeout)) { - tmp = strict_strtoul(instr, 0, &l); + tmp = kstrtoul(instr, 0, &l); if (tmp) return tmp; @@ -181,7 +181,7 @@ static ssize_t db1x_pmattr_store(struct kobject *kobj, } } else if (ATTRCMP(wakemsk)) { - tmp = strict_strtoul(instr, 0, &l); + tmp = kstrtoul(instr, 0, &l); if (tmp) return tmp; diff --git a/arch/mips/bcm47xx/prom.c b/arch/mips/bcm47xx/prom.c index 0af808dfd1c..1a03a2f4349 100644 --- a/arch/mips/bcm47xx/prom.c +++ b/arch/mips/bcm47xx/prom.c @@ -69,15 +69,18 @@ static __init void prom_init_mem(void) * BCM47XX uses 128MB for addressing the ram, if the system contains * less that that amount of ram it remaps the ram more often into the * available space. - * Accessing memory after 128MB will cause an exception. - * max contains the biggest possible address supported by the platform. - * If the method wants to try something above we assume 128MB ram. */ - off = (unsigned long)prom_init; - max = off | ((128 << 20) - 1); - for (mem = (1 << 20); mem < (128 << 20); mem += (1 << 20)) { - if ((off + mem) > max) { - mem = (128 << 20); + + /* Physical address, without mapping to any kernel segment */ + off = CPHYSADDR((unsigned long)prom_init); + + /* Accessing memory after 128 MiB will cause an exception */ + max = 128 << 20; + + for (mem = 1 << 20; mem < max; mem += 1 << 20) { + /* Loop condition may be not enough, off may be over 1 MiB */ + if (off + mem >= max) { + mem = max; printk(KERN_DEBUG "assume 128MB RAM\n"); break; } diff --git a/arch/mips/cavium-octeon/Kconfig b/arch/mips/cavium-octeon/Kconfig index 227705d9d5a..60286665793 100644 --- a/arch/mips/cavium-octeon/Kconfig +++ b/arch/mips/cavium-octeon/Kconfig @@ -10,6 +10,17 @@ config CAVIUM_CN63XXP1 non-CN63XXP1 hardware, so it is recommended to select "n" unless it is known the workarounds are needed. +config CAVIUM_OCTEON_CVMSEG_SIZE + int "Number of L1 cache lines reserved for CVMSEG memory" + range 0 54 + default 1 + help + CVMSEG LM is a segment that accesses portions of the dcache as a + local memory; the larger CVMSEG is, the smaller the cache is. + This selects the size of CVMSEG LM, which is in cache blocks. The + legally range is from zero to 54 cache blocks (i.e. CVMSEG LM is + between zero and 6192 bytes). + endif # CPU_CAVIUM_OCTEON if CAVIUM_OCTEON_SOC @@ -23,17 +34,6 @@ config CAVIUM_OCTEON_2ND_KERNEL with this option to be run at the same time as one built without this option. -config CAVIUM_OCTEON_CVMSEG_SIZE - int "Number of L1 cache lines reserved for CVMSEG memory" - range 0 54 - default 1 - help - CVMSEG LM is a segment that accesses portions of the dcache as a - local memory; the larger CVMSEG is, the smaller the cache is. - This selects the size of CVMSEG LM, which is in cache blocks. The - legally range is from zero to 54 cache blocks (i.e. CVMSEG LM is - between zero and 6192 bytes). - config CAVIUM_OCTEON_LOCK_L2 bool "Lock often used kernel code in the L2" default "y" @@ -86,7 +86,6 @@ config SWIOTLB select IOMMU_HELPER select NEED_SG_DMA_LENGTH - config OCTEON_ILM tristate "Module to measure interrupt latency using Octeon CIU Timer" help diff --git a/arch/mips/cavium-octeon/executive/cvmx-helper.c b/arch/mips/cavium-octeon/executive/cvmx-helper.c index 8553ad5c72b..7e5cf7a5e2f 100644 --- a/arch/mips/cavium-octeon/executive/cvmx-helper.c +++ b/arch/mips/cavium-octeon/executive/cvmx-helper.c @@ -106,6 +106,158 @@ int cvmx_helper_ports_on_interface(int interface) EXPORT_SYMBOL_GPL(cvmx_helper_ports_on_interface); /** + * @INTERNAL + * Return interface mode for CN68xx. + */ +static cvmx_helper_interface_mode_t __cvmx_get_mode_cn68xx(int interface) +{ + union cvmx_mio_qlmx_cfg qlm_cfg; + switch (interface) { + case 0: + qlm_cfg.u64 = cvmx_read_csr(CVMX_MIO_QLMX_CFG(0)); + /* QLM is disabled when QLM SPD is 15. */ + if (qlm_cfg.s.qlm_spd == 15) + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + + if (qlm_cfg.s.qlm_cfg == 2) + return CVMX_HELPER_INTERFACE_MODE_SGMII; + else if (qlm_cfg.s.qlm_cfg == 3) + return CVMX_HELPER_INTERFACE_MODE_XAUI; + else + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + case 2: + case 3: + case 4: + qlm_cfg.u64 = cvmx_read_csr(CVMX_MIO_QLMX_CFG(interface)); + /* QLM is disabled when QLM SPD is 15. */ + if (qlm_cfg.s.qlm_spd == 15) + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + + if (qlm_cfg.s.qlm_cfg == 2) + return CVMX_HELPER_INTERFACE_MODE_SGMII; + else if (qlm_cfg.s.qlm_cfg == 3) + return CVMX_HELPER_INTERFACE_MODE_XAUI; + else + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + case 7: + qlm_cfg.u64 = cvmx_read_csr(CVMX_MIO_QLMX_CFG(3)); + /* QLM is disabled when QLM SPD is 15. */ + if (qlm_cfg.s.qlm_spd == 15) { + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + } else if (qlm_cfg.s.qlm_cfg != 0) { + qlm_cfg.u64 = cvmx_read_csr(CVMX_MIO_QLMX_CFG(1)); + if (qlm_cfg.s.qlm_cfg != 0) + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + } + return CVMX_HELPER_INTERFACE_MODE_NPI; + case 8: + return CVMX_HELPER_INTERFACE_MODE_LOOP; + default: + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + } +} + +/** + * @INTERNAL + * Return interface mode for an Octeon II + */ +static cvmx_helper_interface_mode_t __cvmx_get_mode_octeon2(int interface) +{ + union cvmx_gmxx_inf_mode mode; + + if (OCTEON_IS_MODEL(OCTEON_CN68XX)) + return __cvmx_get_mode_cn68xx(interface); + + if (interface == 2) + return CVMX_HELPER_INTERFACE_MODE_NPI; + + if (interface == 3) + return CVMX_HELPER_INTERFACE_MODE_LOOP; + + /* Only present in CN63XX & CN66XX Octeon model */ + if ((OCTEON_IS_MODEL(OCTEON_CN63XX) && + (interface == 4 || interface == 5)) || + (OCTEON_IS_MODEL(OCTEON_CN66XX) && + interface >= 4 && interface <= 7)) { + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + } + + if (OCTEON_IS_MODEL(OCTEON_CN66XX)) { + union cvmx_mio_qlmx_cfg mio_qlm_cfg; + + /* QLM2 is SGMII0 and QLM1 is SGMII1 */ + if (interface == 0) + mio_qlm_cfg.u64 = cvmx_read_csr(CVMX_MIO_QLMX_CFG(2)); + else if (interface == 1) + mio_qlm_cfg.u64 = cvmx_read_csr(CVMX_MIO_QLMX_CFG(1)); + else + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + + if (mio_qlm_cfg.s.qlm_spd == 15) + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + + if (mio_qlm_cfg.s.qlm_cfg == 9) + return CVMX_HELPER_INTERFACE_MODE_SGMII; + else if (mio_qlm_cfg.s.qlm_cfg == 11) + return CVMX_HELPER_INTERFACE_MODE_XAUI; + else + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + } else if (OCTEON_IS_MODEL(OCTEON_CN61XX)) { + union cvmx_mio_qlmx_cfg qlm_cfg; + + if (interface == 0) { + qlm_cfg.u64 = cvmx_read_csr(CVMX_MIO_QLMX_CFG(2)); + if (qlm_cfg.s.qlm_cfg == 2) + return CVMX_HELPER_INTERFACE_MODE_SGMII; + else if (qlm_cfg.s.qlm_cfg == 3) + return CVMX_HELPER_INTERFACE_MODE_XAUI; + else + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + } else if (interface == 1) { + qlm_cfg.u64 = cvmx_read_csr(CVMX_MIO_QLMX_CFG(0)); + if (qlm_cfg.s.qlm_cfg == 2) + return CVMX_HELPER_INTERFACE_MODE_SGMII; + else if (qlm_cfg.s.qlm_cfg == 3) + return CVMX_HELPER_INTERFACE_MODE_XAUI; + else + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + } + } else if (OCTEON_IS_MODEL(OCTEON_CNF71XX)) { + if (interface == 0) { + union cvmx_mio_qlmx_cfg qlm_cfg; + qlm_cfg.u64 = cvmx_read_csr(CVMX_MIO_QLMX_CFG(0)); + if (qlm_cfg.s.qlm_cfg == 2) + return CVMX_HELPER_INTERFACE_MODE_SGMII; + } + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + } + + if (interface == 1 && OCTEON_IS_MODEL(OCTEON_CN63XX)) + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + + mode.u64 = cvmx_read_csr(CVMX_GMXX_INF_MODE(interface)); + + if (OCTEON_IS_MODEL(OCTEON_CN63XX)) { + switch (mode.cn63xx.mode) { + case 0: + return CVMX_HELPER_INTERFACE_MODE_SGMII; + case 1: + return CVMX_HELPER_INTERFACE_MODE_XAUI; + default: + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + } + } else { + if (!mode.s.en) + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + + if (mode.s.type) + return CVMX_HELPER_INTERFACE_MODE_GMII; + else + return CVMX_HELPER_INTERFACE_MODE_RGMII; + } +} + +/** * Get the operating mode of an interface. Depending on the Octeon * chip and configuration, this function returns an enumeration * of the type of packet I/O supported by an interface. @@ -118,6 +270,20 @@ EXPORT_SYMBOL_GPL(cvmx_helper_ports_on_interface); cvmx_helper_interface_mode_t cvmx_helper_interface_get_mode(int interface) { union cvmx_gmxx_inf_mode mode; + + if (interface < 0 || + interface >= cvmx_helper_get_number_of_interfaces()) + return CVMX_HELPER_INTERFACE_MODE_DISABLED; + + /* + * Octeon II models + */ + if (OCTEON_IS_MODEL(OCTEON_CN6XXX) || OCTEON_IS_MODEL(OCTEON_CNF71XX)) + return __cvmx_get_mode_octeon2(interface); + + /* + * Octeon and Octeon Plus models + */ if (interface == 2) return CVMX_HELPER_INTERFACE_MODE_NPI; diff --git a/arch/mips/cavium-octeon/octeon-irq.c b/arch/mips/cavium-octeon/octeon-irq.c index 3aa5b46b2d4..1b82ac6921e 100644 --- a/arch/mips/cavium-octeon/octeon-irq.c +++ b/arch/mips/cavium-octeon/octeon-irq.c @@ -1260,11 +1260,13 @@ static void __init octeon_irq_init_ciu(void) for (i = 0; i < 4; i++) octeon_irq_force_ciu_mapping(ciu_domain, i + OCTEON_IRQ_PCI_MSI0, 0, i + 40); + octeon_irq_force_ciu_mapping(ciu_domain, OCTEON_IRQ_TWSI, 0, 45); octeon_irq_force_ciu_mapping(ciu_domain, OCTEON_IRQ_RML, 0, 46); for (i = 0; i < 4; i++) octeon_irq_force_ciu_mapping(ciu_domain, i + OCTEON_IRQ_TIMER0, 0, i + 52); octeon_irq_force_ciu_mapping(ciu_domain, OCTEON_IRQ_USB0, 0, 56); + octeon_irq_force_ciu_mapping(ciu_domain, OCTEON_IRQ_TWSI2, 0, 59); /* CIU_1 */ for (i = 0; i < 16; i++) diff --git a/arch/mips/cavium-octeon/setup.c b/arch/mips/cavium-octeon/setup.c index f1bec00d5a8..008e9c8b8ea 100644 --- a/arch/mips/cavium-octeon/setup.c +++ b/arch/mips/cavium-octeon/setup.c @@ -729,17 +729,6 @@ void __init prom_init(void) octeon_write_lcd("Linux"); #endif -#ifdef CONFIG_CAVIUM_GDB - /* - * When debugging the linux kernel, force the cores to enter - * the debug exception handler to break in. - */ - if (octeon_get_boot_debug_flag()) { - cvmx_write_csr(CVMX_CIU_DINT, 1 << cvmx_get_core_num()); - cvmx_read_csr(CVMX_CIU_DINT); - } -#endif - octeon_setup_delays(); /* @@ -779,12 +768,6 @@ void __init prom_init(void) MAX_MEMORY = 32ull << 30; if (*p == '@') RESERVE_LOW_MEM = memparse(p + 1, &p); - } else if (strcmp(arg, "ecc_verbose") == 0) { -#ifdef CONFIG_CAVIUM_REPORT_SINGLE_BIT_ECC - __cvmx_interrupt_ecc_report_single_bit_errors = 1; - pr_notice("Reporting of single bit ECC errors is " - "turned on\n"); -#endif #ifdef CONFIG_KEXEC } else if (strncmp(arg, "crashkernel=", 12) == 0) { crashk_size = memparse(arg+12, &p); diff --git a/arch/mips/cavium-octeon/smp.c b/arch/mips/cavium-octeon/smp.c index 67a078ffc46..a7b3ae104d8 100644 --- a/arch/mips/cavium-octeon/smp.c +++ b/arch/mips/cavium-octeon/smp.c @@ -218,15 +218,6 @@ void octeon_prepare_cpus(unsigned int max_cpus) */ static void octeon_smp_finish(void) { -#ifdef CONFIG_CAVIUM_GDB - unsigned long tmp; - /* Pulse MCD0 signal on Ctrl-C to stop all the cores. Also set the MCD0 - to be not masked by this core so we know the signal is received by - someone */ - asm volatile ("dmfc0 %0, $22\n" - "ori %0, %0, 0x9100\n" "dmtc0 %0, $22\n" : "=r" (tmp)); -#endif - octeon_user_io_init(); /* to generate the first CPU timer interrupt */ @@ -234,21 +225,6 @@ static void octeon_smp_finish(void) local_irq_enable(); } -/** - * Hook for after all CPUs are online - */ -static void octeon_cpus_done(void) -{ -#ifdef CONFIG_CAVIUM_GDB - unsigned long tmp; - /* Pulse MCD0 signal on Ctrl-C to stop all the cores. Also set the MCD0 - to be not masked by this core so we know the signal is received by - someone */ - asm volatile ("dmfc0 %0, $22\n" - "ori %0, %0, 0x9100\n" "dmtc0 %0, $22\n" : "=r" (tmp)); -#endif -} - #ifdef CONFIG_HOTPLUG_CPU /* State of each CPU. */ @@ -405,7 +381,6 @@ struct plat_smp_ops octeon_smp_ops = { .send_ipi_mask = octeon_send_ipi_mask, .init_secondary = octeon_init_secondary, .smp_finish = octeon_smp_finish, - .cpus_done = octeon_cpus_done, .boot_secondary = octeon_boot_secondary, .smp_setup = octeon_smp_setup, .prepare_cpus = octeon_prepare_cpus, diff --git a/arch/mips/configs/ath79_defconfig b/arch/mips/configs/ath79_defconfig index e3a3836508e..134879c1310 100644 --- a/arch/mips/configs/ath79_defconfig +++ b/arch/mips/configs/ath79_defconfig @@ -46,7 +46,6 @@ CONFIG_MTD=y CONFIG_MTD_REDBOOT_PARTS=y CONFIG_MTD_REDBOOT_DIRECTORY_BLOCK=-2 CONFIG_MTD_CMDLINE_PARTS=y -CONFIG_MTD_CHAR=y CONFIG_MTD_BLOCK=y CONFIG_MTD_CFI=y CONFIG_MTD_JEDECPROBE=y @@ -54,7 +53,7 @@ CONFIG_MTD_CFI_AMDSTD=y CONFIG_MTD_COMPLEX_MAPPINGS=y CONFIG_MTD_PHYSMAP=y CONFIG_MTD_M25P80=y -# CONFIG_M25PXX_USE_FAST_READ is not set +CONFIG_MTD_SPI_NOR=y CONFIG_NETDEVICES=y # CONFIG_NET_PACKET_ENGINE is not set CONFIG_ATH_COMMON=m diff --git a/arch/mips/configs/db1xxx_defconfig b/arch/mips/configs/db1xxx_defconfig index c99b6eeda90..a64b30b96a0 100644 --- a/arch/mips/configs/db1xxx_defconfig +++ b/arch/mips/configs/db1xxx_defconfig @@ -113,6 +113,7 @@ CONFIG_MTD_NAND=y CONFIG_MTD_NAND_ECC_BCH=y CONFIG_MTD_NAND_AU1550=y CONFIG_MTD_NAND_PLATFORM=y +CONFIG_MTD_SPI_NOR=y CONFIG_EEPROM_AT24=y CONFIG_EEPROM_AT25=y CONFIG_SCSI_TGT=y diff --git a/arch/mips/configs/maltasmtc_defconfig b/arch/mips/configs/maltasmtc_defconfig deleted file mode 100644 index eb316447588..00000000000 --- a/arch/mips/configs/maltasmtc_defconfig +++ /dev/null @@ -1,196 +0,0 @@ -CONFIG_MIPS_MALTA=y -CONFIG_CPU_LITTLE_ENDIAN=y -CONFIG_CPU_MIPS32_R2=y -CONFIG_PAGE_SIZE_16KB=y -CONFIG_MIPS_MT_SMTC=y -# CONFIG_MIPS_MT_FPAFF is not set -CONFIG_NR_CPUS=9 -CONFIG_HZ_48=y -CONFIG_LOCALVERSION="smtc" -CONFIG_SYSVIPC=y -CONFIG_POSIX_MQUEUE=y -CONFIG_AUDIT=y -CONFIG_IKCONFIG=y -CONFIG_IKCONFIG_PROC=y -CONFIG_LOG_BUF_SHIFT=15 -CONFIG_SYSCTL_SYSCALL=y -CONFIG_EMBEDDED=y -CONFIG_SLAB=y -CONFIG_MODULES=y -CONFIG_MODULE_UNLOAD=y -CONFIG_MODVERSIONS=y -CONFIG_MODULE_SRCVERSION_ALL=y -# CONFIG_BLK_DEV_BSG is not set -CONFIG_PCI=y -# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set -CONFIG_NET=y -CONFIG_PACKET=y -CONFIG_UNIX=y -CONFIG_XFRM_USER=m -CONFIG_NET_KEY=y -CONFIG_INET=y -CONFIG_IP_MULTICAST=y -CONFIG_IP_ADVANCED_ROUTER=y -CONFIG_IP_MULTIPLE_TABLES=y -CONFIG_IP_ROUTE_MULTIPATH=y -CONFIG_IP_ROUTE_VERBOSE=y -CONFIG_IP_PNP=y -CONFIG_IP_PNP_DHCP=y -CONFIG_IP_PNP_BOOTP=y -CONFIG_NET_IPIP=m -CONFIG_IP_MROUTE=y -CONFIG_IP_PIMSM_V1=y -CONFIG_IP_PIMSM_V2=y -CONFIG_SYN_COOKIES=y -CONFIG_INET_AH=m -CONFIG_INET_ESP=m -CONFIG_INET_IPCOMP=m -# CONFIG_INET_LRO is not set -CONFIG_INET6_AH=m -CONFIG_INET6_ESP=m -CONFIG_INET6_IPCOMP=m -CONFIG_IPV6_TUNNEL=m -CONFIG_BRIDGE=m -CONFIG_VLAN_8021Q=m -CONFIG_ATALK=m -CONFIG_DEV_APPLETALK=m -CONFIG_IPDDP=m -CONFIG_IPDDP_ENCAP=y -CONFIG_NET_SCHED=y -CONFIG_NET_SCH_CBQ=m -CONFIG_NET_SCH_HTB=m -CONFIG_NET_SCH_HFSC=m -CONFIG_NET_SCH_PRIO=m -CONFIG_NET_SCH_RED=m -CONFIG_NET_SCH_SFQ=m -CONFIG_NET_SCH_TEQL=m -CONFIG_NET_SCH_TBF=m -CONFIG_NET_SCH_GRED=m -CONFIG_NET_SCH_DSMARK=m -CONFIG_NET_SCH_NETEM=m -CONFIG_NET_SCH_INGRESS=m -CONFIG_NET_CLS_BASIC=m -CONFIG_NET_CLS_TCINDEX=m -CONFIG_NET_CLS_ROUTE4=m -CONFIG_NET_CLS_FW=m -CONFIG_NET_CLS_U32=m -CONFIG_NET_CLS_RSVP=m -CONFIG_NET_CLS_RSVP6=m -CONFIG_NET_CLS_ACT=y -CONFIG_NET_ACT_POLICE=y -CONFIG_NET_CLS_IND=y -# CONFIG_WIRELESS is not set -CONFIG_DEVTMPFS=y -CONFIG_BLK_DEV_LOOP=y -CONFIG_BLK_DEV_CRYPTOLOOP=m -CONFIG_IDE=y -# CONFIG_IDE_PROC_FS is not set -# CONFIG_IDEPCI_PCIBUS_ORDER is not set -CONFIG_BLK_DEV_GENERIC=y -CONFIG_BLK_DEV_PIIX=y -CONFIG_SCSI=y -CONFIG_BLK_DEV_SD=y -CONFIG_CHR_DEV_SG=y -# CONFIG_SCSI_LOWLEVEL is not set -CONFIG_NETDEVICES=y -# CONFIG_NET_VENDOR_3COM is not set -# CONFIG_NET_VENDOR_ADAPTEC is not set -# CONFIG_NET_VENDOR_ALTEON is not set -CONFIG_PCNET32=y -# CONFIG_NET_VENDOR_ATHEROS is not set -# CONFIG_NET_VENDOR_BROADCOM is not set -# CONFIG_NET_VENDOR_BROCADE is not set -# CONFIG_NET_VENDOR_CHELSIO is not set -# CONFIG_NET_VENDOR_CISCO is not set -# CONFIG_NET_VENDOR_DEC is not set -# CONFIG_NET_VENDOR_DLINK is not set -# CONFIG_NET_VENDOR_EMULEX is not set -# CONFIG_NET_VENDOR_EXAR is not set -# CONFIG_NET_VENDOR_HP is not set -# CONFIG_NET_VENDOR_INTEL is not set -# CONFIG_NET_VENDOR_MARVELL is not set -# CONFIG_NET_VENDOR_MELLANOX is not set -# CONFIG_NET_VENDOR_MICREL is not set -# CONFIG_NET_VENDOR_MYRI is not set -# CONFIG_NET_VENDOR_NATSEMI is not set -# CONFIG_NET_VENDOR_NVIDIA is not set -# CONFIG_NET_VENDOR_OKI is not set -# CONFIG_NET_PACKET_ENGINE is not set -# CONFIG_NET_VENDOR_QLOGIC is not set -# CONFIG_NET_VENDOR_REALTEK is not set -# CONFIG_NET_VENDOR_RDC is not set -# CONFIG_NET_VENDOR_SEEQ is not set -# CONFIG_NET_VENDOR_SILAN is not set -# CONFIG_NET_VENDOR_SIS is not set -# CONFIG_NET_VENDOR_SMSC is not set -# CONFIG_NET_VENDOR_STMICRO is not set -# CONFIG_NET_VENDOR_SUN is not set -# CONFIG_NET_VENDOR_TEHUTI is not set -# CONFIG_NET_VENDOR_TI is not set -# CONFIG_NET_VENDOR_TOSHIBA is not set -# CONFIG_NET_VENDOR_VIA is not set -# CONFIG_WLAN is not set -# CONFIG_VT is not set -CONFIG_LEGACY_PTY_COUNT=16 -CONFIG_SERIAL_8250=y -CONFIG_SERIAL_8250_CONSOLE=y -CONFIG_HW_RANDOM=y -# CONFIG_HWMON is not set -CONFIG_VIDEO_OUTPUT_CONTROL=m -CONFIG_FB=y -CONFIG_FIRMWARE_EDID=y -CONFIG_FB_MATROX=y -CONFIG_FB_MATROX_G=y -CONFIG_USB=y -CONFIG_USB_EHCI_HCD=y -# CONFIG_USB_EHCI_TT_NEWSCHED is not set -CONFIG_USB_UHCI_HCD=y -CONFIG_USB_STORAGE=y -CONFIG_NEW_LEDS=y -CONFIG_LEDS_CLASS=y -CONFIG_LEDS_TRIGGERS=y -CONFIG_LEDS_TRIGGER_TIMER=y -CONFIG_LEDS_TRIGGER_IDE_DISK=y -CONFIG_LEDS_TRIGGER_HEARTBEAT=y -CONFIG_LEDS_TRIGGER_BACKLIGHT=y -CONFIG_LEDS_TRIGGER_DEFAULT_ON=y -CONFIG_RTC_CLASS=y -CONFIG_RTC_DRV_CMOS=y -CONFIG_EXT2_FS=y -CONFIG_EXT3_FS=y -# CONFIG_EXT3_DEFAULTS_TO_ORDERED is not set -CONFIG_XFS_FS=y -CONFIG_XFS_QUOTA=y -CONFIG_XFS_POSIX_ACL=y -CONFIG_QUOTA=y -CONFIG_QFMT_V2=y -CONFIG_MSDOS_FS=m -CONFIG_VFAT_FS=m -CONFIG_PROC_KCORE=y -CONFIG_TMPFS=y -CONFIG_NFS_FS=y -CONFIG_ROOT_NFS=y -CONFIG_CIFS=m -CONFIG_CIFS_WEAK_PW_HASH=y -CONFIG_CIFS_XATTR=y -CONFIG_CIFS_POSIX=y -CONFIG_NLS_CODEPAGE_437=m -CONFIG_NLS_ISO8859_1=m -# CONFIG_FTRACE is not set -CONFIG_CRYPTO_NULL=m -CONFIG_CRYPTO_PCBC=m -CONFIG_CRYPTO_HMAC=y -CONFIG_CRYPTO_MICHAEL_MIC=m -CONFIG_CRYPTO_SHA512=m -CONFIG_CRYPTO_TGR192=m -CONFIG_CRYPTO_WP512=m -CONFIG_CRYPTO_ANUBIS=m -CONFIG_CRYPTO_BLOWFISH=m -CONFIG_CRYPTO_CAST5=m -CONFIG_CRYPTO_CAST6=m -CONFIG_CRYPTO_KHAZAD=m -CONFIG_CRYPTO_SERPENT=m -CONFIG_CRYPTO_TEA=m -CONFIG_CRYPTO_TWOFISH=m -# CONFIG_CRYPTO_ANSI_CPRNG is not set -# CONFIG_CRYPTO_HW is not set diff --git a/arch/mips/configs/maltasmvp_defconfig b/arch/mips/configs/maltasmvp_defconfig index 10ef3bed5f4..f8a32315bb3 100644 --- a/arch/mips/configs/maltasmvp_defconfig +++ b/arch/mips/configs/maltasmvp_defconfig @@ -4,10 +4,9 @@ CONFIG_CPU_MIPS32_R2=y CONFIG_PAGE_SIZE_16KB=y CONFIG_MIPS_MT_SMP=y CONFIG_SCHED_SMT=y -CONFIG_MIPS_CMP=y +CONFIG_MIPS_CPS=y CONFIG_NR_CPUS=8 CONFIG_HZ_100=y -CONFIG_LOCALVERSION="cmp" CONFIG_SYSVIPC=y CONFIG_POSIX_MQUEUE=y CONFIG_AUDIT=y diff --git a/arch/mips/configs/maltasmvp_eva_defconfig b/arch/mips/configs/maltasmvp_eva_defconfig index 2d3002cba10..c83338a3991 100644 --- a/arch/mips/configs/maltasmvp_eva_defconfig +++ b/arch/mips/configs/maltasmvp_eva_defconfig @@ -5,10 +5,9 @@ CONFIG_CPU_MIPS32_3_5_FEATURES=y CONFIG_PAGE_SIZE_16KB=y CONFIG_MIPS_MT_SMP=y CONFIG_SCHED_SMT=y -CONFIG_MIPS_CMP=y +CONFIG_MIPS_CPS=y CONFIG_NR_CPUS=8 CONFIG_HZ_100=y -CONFIG_LOCALVERSION="cmp" CONFIG_SYSVIPC=y CONFIG_POSIX_MQUEUE=y CONFIG_AUDIT=y diff --git a/arch/mips/configs/mips_paravirt_defconfig b/arch/mips/configs/mips_paravirt_defconfig new file mode 100644 index 00000000000..84cfcb4bf2e --- /dev/null +++ b/arch/mips/configs/mips_paravirt_defconfig @@ -0,0 +1,103 @@ +CONFIG_MIPS_PARAVIRT=y +CONFIG_CPU_MIPS64_R2=y +CONFIG_64BIT=y +CONFIG_TRANSPARENT_HUGEPAGE=y +CONFIG_SMP=y +CONFIG_HZ_1000=y +CONFIG_PREEMPT=y +CONFIG_SYSVIPC=y +CONFIG_BSD_PROCESS_ACCT=y +CONFIG_BSD_PROCESS_ACCT_V3=y +CONFIG_IKCONFIG=y +CONFIG_IKCONFIG_PROC=y +CONFIG_LOG_BUF_SHIFT=14 +CONFIG_RELAY=y +CONFIG_BLK_DEV_INITRD=y +CONFIG_EXPERT=y +CONFIG_SLAB=y +CONFIG_MODULES=y +CONFIG_MODULE_UNLOAD=y +# CONFIG_BLK_DEV_BSG is not set +CONFIG_PCI=y +CONFIG_MIPS32_COMPAT=y +CONFIG_MIPS32_O32=y +CONFIG_MIPS32_N32=y +CONFIG_NET=y +CONFIG_PACKET=y +CONFIG_UNIX=y +CONFIG_INET=y +CONFIG_IP_MULTICAST=y +CONFIG_IP_ADVANCED_ROUTER=y +CONFIG_IP_MULTIPLE_TABLES=y +CONFIG_IP_ROUTE_MULTIPATH=y +CONFIG_IP_ROUTE_VERBOSE=y +CONFIG_IP_PNP=y +CONFIG_IP_PNP_DHCP=y +CONFIG_IP_PNP_BOOTP=y +CONFIG_IP_PNP_RARP=y +CONFIG_IP_MROUTE=y +CONFIG_IP_PIMSM_V1=y +CONFIG_IP_PIMSM_V2=y +CONFIG_SYN_COOKIES=y +# CONFIG_INET_LRO is not set +CONFIG_IPV6=y +# CONFIG_WIRELESS is not set +CONFIG_UEVENT_HELPER_PATH="/sbin/hotplug" +# CONFIG_FW_LOADER is not set +CONFIG_BLK_DEV_LOOP=y +CONFIG_VIRTIO_BLK=y +CONFIG_SCSI=y +CONFIG_BLK_DEV_SD=y +CONFIG_NETDEVICES=y +CONFIG_VIRTIO_NET=y +# CONFIG_NET_VENDOR_BROADCOM is not set +# CONFIG_NET_VENDOR_INTEL is not set +# CONFIG_NET_VENDOR_MARVELL is not set +# CONFIG_NET_VENDOR_MICREL is not set +# CONFIG_NET_VENDOR_NATSEMI is not set +# CONFIG_NET_VENDOR_SMSC is not set +# CONFIG_NET_VENDOR_STMICRO is not set +# CONFIG_NET_VENDOR_WIZNET is not set +CONFIG_PHYLIB=y +CONFIG_MARVELL_PHY=y +CONFIG_BROADCOM_PHY=y +CONFIG_BCM87XX_PHY=y +# CONFIG_WLAN is not set +# CONFIG_INPUT is not set +# CONFIG_SERIO is not set +# CONFIG_VT is not set +CONFIG_VIRTIO_CONSOLE=y +# CONFIG_HW_RANDOM is not set +# CONFIG_HWMON is not set +# CONFIG_USB_SUPPORT is not set +CONFIG_VIRTIO_PCI=y +CONFIG_VIRTIO_BALLOON=y +CONFIG_VIRTIO_MMIO=y +# CONFIG_IOMMU_SUPPORT is not set +CONFIG_EXT4_FS=y +CONFIG_EXT4_FS_POSIX_ACL=y +CONFIG_EXT4_FS_SECURITY=y +CONFIG_MSDOS_FS=y +CONFIG_VFAT_FS=y +CONFIG_PROC_KCORE=y +CONFIG_TMPFS=y +CONFIG_HUGETLBFS=y +# CONFIG_MISC_FILESYSTEMS is not set +CONFIG_NFS_FS=y +CONFIG_NFS_V4=y +CONFIG_NFS_V4_1=y +CONFIG_ROOT_NFS=y +CONFIG_NLS_CODEPAGE_437=y +CONFIG_NLS_ASCII=y +CONFIG_NLS_ISO8859_1=y +CONFIG_NLS_UTF8=y +CONFIG_DEBUG_INFO=y +CONFIG_DEBUG_FS=y +CONFIG_MAGIC_SYSRQ=y +# CONFIG_SCHED_DEBUG is not set +# CONFIG_FTRACE is not set +CONFIG_CRYPTO_CBC=y +CONFIG_CRYPTO_HMAC=y +CONFIG_CRYPTO_MD5=y +CONFIG_CRYPTO_DES=y +# CONFIG_CRYPTO_ANSI_CPRNG is not set diff --git a/arch/mips/configs/rt305x_defconfig b/arch/mips/configs/rt305x_defconfig index d1741bcf894..d14ae2fa7d1 100644 --- a/arch/mips/configs/rt305x_defconfig +++ b/arch/mips/configs/rt305x_defconfig @@ -81,7 +81,6 @@ CONFIG_UEVENT_HELPER_PATH="/sbin/hotplug" # CONFIG_FIRMWARE_IN_KERNEL is not set CONFIG_MTD=y CONFIG_MTD_CMDLINE_PARTS=y -CONFIG_MTD_CHAR=y CONFIG_MTD_BLOCK=y CONFIG_MTD_CFI=y CONFIG_MTD_CFI_AMDSTD=y @@ -89,6 +88,7 @@ CONFIG_MTD_COMPLEX_MAPPINGS=y CONFIG_MTD_PHYSMAP=y CONFIG_MTD_PHYSMAP_OF=y CONFIG_MTD_M25P80=y +CONFIG_MTD_SPI_NOR=y CONFIG_EEPROM_93CX6=m CONFIG_SCSI=y CONFIG_BLK_DEV_SD=y diff --git a/arch/mips/dec/setup.c b/arch/mips/dec/setup.c index 56e6e2c2368..41bbffd9cc0 100644 --- a/arch/mips/dec/setup.c +++ b/arch/mips/dec/setup.c @@ -23,6 +23,7 @@ #include <asm/bootinfo.h> #include <asm/cpu.h> #include <asm/cpu-features.h> +#include <asm/cpu-type.h> #include <asm/irq.h> #include <asm/irq_cpu.h> #include <asm/mipsregs.h> @@ -748,6 +749,10 @@ void __init arch_init_irq(void) cpu_fpu_mask = 0; dec_interrupt[DEC_IRQ_FPU] = -1; } + /* Free the halt interrupt unused on R4k systems. */ + if (current_cpu_type() == CPU_R4000SC || + current_cpu_type() == CPU_R4400SC) + dec_interrupt[DEC_IRQ_HALT] = -1; /* Register board interrupts: FPU and cascade. */ if (dec_interrupt[DEC_IRQ_FPU] >= 0) diff --git a/arch/mips/include/asm/asmmacro.h b/arch/mips/include/asm/asmmacro.h index b464b8b1147..935543f1453 100644 --- a/arch/mips/include/asm/asmmacro.h +++ b/arch/mips/include/asm/asmmacro.h @@ -17,26 +17,8 @@ #ifdef CONFIG_64BIT #include <asm/asmmacro-64.h> #endif -#ifdef CONFIG_MIPS_MT_SMTC -#include <asm/mipsmtregs.h> -#endif - -#ifdef CONFIG_MIPS_MT_SMTC - .macro local_irq_enable reg=t0 - mfc0 \reg, CP0_TCSTATUS - ori \reg, \reg, TCSTATUS_IXMT - xori \reg, \reg, TCSTATUS_IXMT - mtc0 \reg, CP0_TCSTATUS - _ehb - .endm - .macro local_irq_disable reg=t0 - mfc0 \reg, CP0_TCSTATUS - ori \reg, \reg, TCSTATUS_IXMT - mtc0 \reg, CP0_TCSTATUS - _ehb - .endm -#elif defined(CONFIG_CPU_MIPSR2) +#ifdef CONFIG_CPU_MIPSR2 .macro local_irq_enable reg=t0 ei irq_enable_hazard @@ -71,7 +53,7 @@ sw \reg, TI_PRE_COUNT($28) #endif .endm -#endif /* CONFIG_MIPS_MT_SMTC */ +#endif /* CONFIG_CPU_MIPSR2 */ .macro fpu_save_16even thread tmp=t0 cfc1 \tmp, fcr31 @@ -267,13 +249,35 @@ .set pop .endm #else + +#ifdef CONFIG_CPU_MICROMIPS +#define CFC_MSA_INSN 0x587e0056 +#define CTC_MSA_INSN 0x583e0816 +#define LDD_MSA_INSN 0x58000837 +#define STD_MSA_INSN 0x5800083f +#define COPY_UW_MSA_INSN 0x58f00056 +#define COPY_UD_MSA_INSN 0x58f80056 +#define INSERT_W_MSA_INSN 0x59300816 +#define INSERT_D_MSA_INSN 0x59380816 +#else +#define CFC_MSA_INSN 0x787e0059 +#define CTC_MSA_INSN 0x783e0819 +#define LDD_MSA_INSN 0x78000823 +#define STD_MSA_INSN 0x78000827 +#define COPY_UW_MSA_INSN 0x78f00059 +#define COPY_UD_MSA_INSN 0x78f80059 +#define INSERT_W_MSA_INSN 0x79300819 +#define INSERT_D_MSA_INSN 0x79380819 +#endif + /* * Temporary until all toolchains in use include MSA support. */ .macro cfcmsa rd, cs .set push .set noat - .word 0x787e0059 | (\cs << 11) + .insn + .word CFC_MSA_INSN | (\cs << 11) move \rd, $1 .set pop .endm @@ -282,7 +286,7 @@ .set push .set noat move $1, \rs - .word 0x783e0819 | (\cd << 6) + .word CTC_MSA_INSN | (\cd << 6) .set pop .endm @@ -290,7 +294,7 @@ .set push .set noat add $1, \base, \off - .word 0x78000823 | (\wd << 6) + .word LDD_MSA_INSN | (\wd << 6) .set pop .endm @@ -298,14 +302,15 @@ .set push .set noat add $1, \base, \off - .word 0x78000827 | (\wd << 6) + .word STD_MSA_INSN | (\wd << 6) .set pop .endm .macro copy_u_w rd, ws, n .set push .set noat - .word 0x78f00059 | (\n << 16) | (\ws << 11) + .insn + .word COPY_UW_MSA_INSN | (\n << 16) | (\ws << 11) /* move triggers an assembler bug... */ or \rd, $1, zero .set pop @@ -314,7 +319,8 @@ .macro copy_u_d rd, ws, n .set push .set noat - .word 0x78f80059 | (\n << 16) | (\ws << 11) + .insn + .word COPY_UD_MSA_INSN | (\n << 16) | (\ws << 11) /* move triggers an assembler bug... */ or \rd, $1, zero .set pop @@ -325,7 +331,7 @@ .set noat /* move triggers an assembler bug... */ or $1, \rs, zero - .word 0x79300819 | (\n << 16) | (\wd << 6) + .word INSERT_W_MSA_INSN | (\n << 16) | (\wd << 6) .set pop .endm @@ -334,7 +340,7 @@ .set noat /* move triggers an assembler bug... */ or $1, \rs, zero - .word 0x79380819 | (\n << 16) | (\wd << 6) + .word INSERT_D_MSA_INSN | (\n << 16) | (\wd << 6) .set pop .endm #endif diff --git a/arch/mips/include/asm/branch.h b/arch/mips/include/asm/branch.h index e28a3e0eb3c..de781cf54bc 100644 --- a/arch/mips/include/asm/branch.h +++ b/arch/mips/include/asm/branch.h @@ -8,6 +8,8 @@ #ifndef _ASM_BRANCH_H #define _ASM_BRANCH_H +#include <asm/cpu-features.h> +#include <asm/mipsregs.h> #include <asm/ptrace.h> #include <asm/inst.h> @@ -18,12 +20,40 @@ extern int __compute_return_epc_for_insn(struct pt_regs *regs, extern int __microMIPS_compute_return_epc(struct pt_regs *regs); extern int __MIPS16e_compute_return_epc(struct pt_regs *regs); +/* + * microMIPS bitfields + */ +#define MM_POOL32A_MINOR_MASK 0x3f +#define MM_POOL32A_MINOR_SHIFT 0x6 +#define MM_MIPS32_COND_FC 0x30 + +extern int __mm_isBranchInstr(struct pt_regs *regs, + struct mm_decoded_insn dec_insn, unsigned long *contpc); + +static inline int mm_isBranchInstr(struct pt_regs *regs, + struct mm_decoded_insn dec_insn, unsigned long *contpc) +{ + if (!cpu_has_mmips) + return 0; + + return __mm_isBranchInstr(regs, dec_insn, contpc); +} static inline int delay_slot(struct pt_regs *regs) { return regs->cp0_cause & CAUSEF_BD; } +static inline void clear_delay_slot(struct pt_regs *regs) +{ + regs->cp0_cause &= ~CAUSEF_BD; +} + +static inline void set_delay_slot(struct pt_regs *regs) +{ + regs->cp0_cause |= CAUSEF_BD; +} + static inline unsigned long exception_epc(struct pt_regs *regs) { if (likely(!delay_slot(regs))) diff --git a/arch/mips/include/asm/cacheflush.h b/arch/mips/include/asm/cacheflush.h index 69468ded282..e08381a37f8 100644 --- a/arch/mips/include/asm/cacheflush.h +++ b/arch/mips/include/asm/cacheflush.h @@ -113,6 +113,12 @@ unsigned long run_uncached(void *func); extern void *kmap_coherent(struct page *page, unsigned long addr); extern void kunmap_coherent(void); +extern void *kmap_noncoherent(struct page *page, unsigned long addr); + +static inline void kunmap_noncoherent(void) +{ + kunmap_coherent(); +} #define ARCH_HAS_FLUSH_KERNEL_DCACHE_PAGE static inline void flush_kernel_dcache_page(struct page *page) diff --git a/arch/mips/include/asm/cmp.h b/arch/mips/include/asm/cmp.h index 89a73fb93ae..033d97303c8 100644 --- a/arch/mips/include/asm/cmp.h +++ b/arch/mips/include/asm/cmp.h @@ -10,7 +10,6 @@ extern void cmp_smp_setup(void); extern void cmp_smp_finish(void); extern void cmp_boot_secondary(int cpu, struct task_struct *t); extern void cmp_init_secondary(void); -extern void cmp_cpus_done(void); extern void cmp_prepare_cpus(unsigned int max_cpus); /* This is platform specific */ diff --git a/arch/mips/include/asm/cpu-features.h b/arch/mips/include/asm/cpu-features.h index f56cc975b92..c7d8c997d93 100644 --- a/arch/mips/include/asm/cpu-features.h +++ b/arch/mips/include/asm/cpu-features.h @@ -110,9 +110,15 @@ #ifndef cpu_has_smartmips #define cpu_has_smartmips (cpu_data[0].ases & MIPS_ASE_SMARTMIPS) #endif + #ifndef cpu_has_rixi -#define cpu_has_rixi (cpu_data[0].options & MIPS_CPU_RIXI) +# ifdef CONFIG_64BIT +# define cpu_has_rixi (cpu_data[0].options & MIPS_CPU_RIXI) +# else /* CONFIG_32BIT */ +# define cpu_has_rixi ((cpu_data[0].options & MIPS_CPU_RIXI) && !cpu_has_64bits) +# endif #endif + #ifndef cpu_has_mmips # ifdef CONFIG_SYS_SUPPORTS_MICROMIPS # define cpu_has_mmips (cpu_data[0].options & MIPS_CPU_MICROMIPS) @@ -120,6 +126,7 @@ # define cpu_has_mmips 0 # endif #endif + #ifndef cpu_has_vtag_icache #define cpu_has_vtag_icache (cpu_data[0].icache.flags & MIPS_CACHE_VTAG) #endif @@ -183,6 +190,17 @@ /* * Shortcuts ... */ +#define cpu_has_mips_2_3_4_5 (cpu_has_mips_2 | cpu_has_mips_3_4_5) +#define cpu_has_mips_3_4_5 (cpu_has_mips_3 | cpu_has_mips_4_5) +#define cpu_has_mips_4_5 (cpu_has_mips_4 | cpu_has_mips_5) + +#define cpu_has_mips_2_3_4_5_r (cpu_has_mips_2 | cpu_has_mips_3_4_5_r) +#define cpu_has_mips_3_4_5_r (cpu_has_mips_3 | cpu_has_mips_4_5_r) +#define cpu_has_mips_4_5_r (cpu_has_mips_4 | cpu_has_mips_5_r) +#define cpu_has_mips_5_r (cpu_has_mips_5 | cpu_has_mips_r) + +#define cpu_has_mips_4_5_r2 (cpu_has_mips_4_5 | cpu_has_mips_r2) + #define cpu_has_mips32 (cpu_has_mips32r1 | cpu_has_mips32r2) #define cpu_has_mips64 (cpu_has_mips64r1 | cpu_has_mips64r2) #define cpu_has_mips_r1 (cpu_has_mips32r1 | cpu_has_mips64r1) diff --git a/arch/mips/include/asm/cpu-info.h b/arch/mips/include/asm/cpu-info.h index ff2707ab329..47d5967ce7e 100644 --- a/arch/mips/include/asm/cpu-info.h +++ b/arch/mips/include/asm/cpu-info.h @@ -65,18 +65,13 @@ struct cpuinfo_mips { #ifdef CONFIG_64BIT int vmbits; /* Virtual memory size in bits */ #endif -#if defined(CONFIG_MIPS_MT_SMP) || defined(CONFIG_MIPS_MT_SMTC) +#ifdef CONFIG_MIPS_MT_SMP /* - * In the MIPS MT "SMTC" model, each TC is considered - * to be a "CPU" for the purposes of scheduling, but - * exception resources, ASID spaces, etc, are common - * to all TCs within the same VPE. + * There is not necessarily a 1:1 mapping of VPE num to CPU number + * in particular on multi-core systems. */ int vpe_id; /* Virtual Processor number */ #endif -#ifdef CONFIG_MIPS_MT_SMTC - int tc_id; /* Thread Context number */ -#endif void *data; /* Additional data */ unsigned int watch_reg_count; /* Number that exist */ unsigned int watch_reg_use_cnt; /* Usable by ptrace */ @@ -117,7 +112,7 @@ struct proc_cpuinfo_notifier_args { unsigned long n; }; -#if defined(CONFIG_MIPS_MT_SMP) || defined(CONFIG_MIPS_MT_SMTC) +#ifdef CONFIG_MIPS_MT_SMP # define cpu_vpe_id(cpuinfo) ((cpuinfo)->vpe_id) #else # define cpu_vpe_id(cpuinfo) 0 diff --git a/arch/mips/include/asm/cpu-type.h b/arch/mips/include/asm/cpu-type.h index 721906130a5..b4e2bd87df5 100644 --- a/arch/mips/include/asm/cpu-type.h +++ b/arch/mips/include/asm/cpu-type.h @@ -155,9 +155,6 @@ static inline int __pure __get_cpu_type(const int cpu_type) case CPU_RM7000: case CPU_SR71000: #endif -#ifdef CONFIG_SYS_HAS_CPU_RM9000 - case CPU_RM9000: -#endif #ifdef CONFIG_SYS_HAS_CPU_SB1 case CPU_SB1: case CPU_SB1A: @@ -166,6 +163,7 @@ static inline int __pure __get_cpu_type(const int cpu_type) case CPU_CAVIUM_OCTEON: case CPU_CAVIUM_OCTEON_PLUS: case CPU_CAVIUM_OCTEON2: + case CPU_CAVIUM_OCTEON3: #endif #if defined(CONFIG_SYS_HAS_CPU_BMIPS32_3300) || \ diff --git a/arch/mips/include/asm/cpu.h b/arch/mips/include/asm/cpu.h index 530eb8b3a68..129d08701e9 100644 --- a/arch/mips/include/asm/cpu.h +++ b/arch/mips/include/asm/cpu.h @@ -201,6 +201,7 @@ #define PRID_IMP_NETLOGIC_XLP3XX 0x1100 #define PRID_IMP_NETLOGIC_XLP2XX 0x1200 #define PRID_IMP_NETLOGIC_XLP9XX 0x1500 +#define PRID_IMP_NETLOGIC_XLP5XX 0x1300 /* * Particular Revision values for bits 7:0 of the PRId register. @@ -281,7 +282,7 @@ enum cpu_type_enum { CPU_R4700, CPU_R5000, CPU_R5500, CPU_NEVADA, CPU_R5432, CPU_R10000, CPU_R12000, CPU_R14000, CPU_VR41XX, CPU_VR4111, CPU_VR4121, CPU_VR4122, CPU_VR4131, CPU_VR4133, CPU_VR4181, CPU_VR4181A, CPU_RM7000, - CPU_SR71000, CPU_RM9000, CPU_TX49XX, + CPU_SR71000, CPU_TX49XX, /* * R8000 class processors diff --git a/arch/mips/include/asm/dec/kn05.h b/arch/mips/include/asm/dec/kn05.h index 56d22dc8803..8e14f677e5e 100644 --- a/arch/mips/include/asm/dec/kn05.h +++ b/arch/mips/include/asm/dec/kn05.h @@ -49,12 +49,20 @@ #define KN4K_RES_15 (15*IOASIC_SLOT_SIZE) /* unused? */ /* + * MB ASIC interrupt bits. + */ +#define KN4K_MB_INR_MB 4 /* ??? */ +#define KN4K_MB_INR_MT 3 /* memory, I/O bus read/write errors */ +#define KN4K_MB_INR_RES_2 2 /* unused */ +#define KN4K_MB_INR_RTC 1 /* RTC */ +#define KN4K_MB_INR_TC 0 /* I/O ASIC cascade */ + +/* * Bits for the MB interrupt register. * The register appears read-only. */ -#define KN4K_MB_INT_TC (1<<0) /* TURBOchannel? */ -#define KN4K_MB_INT_RTC (1<<1) /* RTC? */ -#define KN4K_MB_INT_MT (1<<3) /* I/O ASIC cascade */ +#define KN4K_MB_INT_IRQ (0x1f<<0) /* CPU Int[4:0] status. */ +#define KN4K_MB_INT_IRQ_N(n) (1<<(n)) /* Individual status bits. */ /* * Bits for the MB control & status register. @@ -70,6 +78,7 @@ #define KN4K_MB_CSR_NC (1<<14) /* ??? */ #define KN4K_MB_CSR_EE (1<<15) /* (bus) Exception Enable? */ #define KN4K_MB_CSR_MSK (0x1f<<16) /* CPU Int[4:0] mask */ +#define KN4K_MB_CSR_MSK_N(n) (1<<((n)+16)) /* Individual mask bits. */ #define KN4K_MB_CSR_FW (1<<21) /* ??? */ #define KN4K_MB_CSR_W (1<<31) /* ??? */ diff --git a/arch/mips/include/asm/fixmap.h b/arch/mips/include/asm/fixmap.h index 8c012af2f45..6842ffafd1e 100644 --- a/arch/mips/include/asm/fixmap.h +++ b/arch/mips/include/asm/fixmap.h @@ -48,11 +48,7 @@ enum fixed_addresses { #define FIX_N_COLOURS 8 FIX_CMAP_BEGIN, -#ifdef CONFIG_MIPS_MT_SMTC - FIX_CMAP_END = FIX_CMAP_BEGIN + (FIX_N_COLOURS * NR_CPUS * 2), -#else FIX_CMAP_END = FIX_CMAP_BEGIN + (FIX_N_COLOURS * 2), -#endif #ifdef CONFIG_HIGHMEM /* reserved pte's for temporary kernel mappings */ FIX_KMAP_BEGIN = FIX_CMAP_END + 1, diff --git a/arch/mips/include/asm/fpu.h b/arch/mips/include/asm/fpu.h index 4d86b72750c..a939574f829 100644 --- a/arch/mips/include/asm/fpu.h +++ b/arch/mips/include/asm/fpu.h @@ -17,6 +17,7 @@ #include <asm/mipsregs.h> #include <asm/cpu.h> #include <asm/cpu-features.h> +#include <asm/fpu_emulator.h> #include <asm/hazards.h> #include <asm/processor.h> #include <asm/current.h> @@ -28,7 +29,6 @@ struct sigcontext; struct sigcontext32; -extern void fpu_emulator_init_fpu(void); extern void _init_fpu(void); extern void _save_fp(struct task_struct *); extern void _restore_fp(struct task_struct *); @@ -156,15 +156,16 @@ static inline int init_fpu(void) int ret = 0; preempt_disable(); + if (cpu_has_fpu) { ret = __own_fpu(); if (!ret) _init_fpu(); - } else { + } else fpu_emulator_init_fpu(); - } preempt_enable(); + return ret; } diff --git a/arch/mips/include/asm/fpu_emulator.h b/arch/mips/include/asm/fpu_emulator.h index 2abb587d5ab..0195745b4b1 100644 --- a/arch/mips/include/asm/fpu_emulator.h +++ b/arch/mips/include/asm/fpu_emulator.h @@ -23,9 +23,12 @@ #ifndef _ASM_FPU_EMULATOR_H #define _ASM_FPU_EMULATOR_H +#include <linux/sched.h> #include <asm/break.h> +#include <asm/thread_info.h> #include <asm/inst.h> #include <asm/local.h> +#include <asm/processor.h> #ifdef CONFIG_DEBUG_FS @@ -36,6 +39,11 @@ struct mips_fpu_emulator_stats { local_t cp1ops; local_t cp1xops; local_t errors; + local_t ieee754_inexact; + local_t ieee754_underflow; + local_t ieee754_overflow; + local_t ieee754_zerodiv; + local_t ieee754_invalidop; }; DECLARE_PER_CPU(struct mips_fpu_emulator_stats, fpuemustats); @@ -71,4 +79,17 @@ int mm_isBranchInstr(struct pt_regs *regs, struct mm_decoded_insn dec_insn, */ #define BREAK_MATH (0x0000000d | (BRK_MEMU << 16)) +#define SIGNALLING_NAN 0x7ff800007ff80000LL + +static inline void fpu_emulator_init_fpu(void) +{ + struct task_struct *t = current; + int i; + + t->thread.fpu.fcr31 = 0; + + for (i = 0; i < 32; i++) + set_fpr64(&t->thread.fpu.fpr[i], 0, SIGNALLING_NAN); +} + #endif /* _ASM_FPU_EMULATOR_H */ diff --git a/arch/mips/include/asm/gic.h b/arch/mips/include/asm/gic.h index 08271669058..10f6a99f92c 100644 --- a/arch/mips/include/asm/gic.h +++ b/arch/mips/include/asm/gic.h @@ -380,6 +380,7 @@ extern unsigned int gic_compare_int (void); extern cycle_t gic_read_count(void); extern cycle_t gic_read_compare(void); extern void gic_write_compare(cycle_t cnt); +extern void gic_write_cpu_compare(cycle_t cnt, int cpu); extern void gic_send_ipi(unsigned int intr); extern unsigned int plat_ipi_call_int_xlate(unsigned int); extern unsigned int plat_ipi_resched_int_xlate(unsigned int); diff --git a/arch/mips/include/asm/gio_device.h b/arch/mips/include/asm/gio_device.h index 0878701712f..4be1a57cdbb 100644 --- a/arch/mips/include/asm/gio_device.h +++ b/arch/mips/include/asm/gio_device.h @@ -50,7 +50,7 @@ static inline void gio_device_free(struct gio_device *dev) extern int gio_register_driver(struct gio_driver *); extern void gio_unregister_driver(struct gio_driver *); -#define gio_get_drvdata(_dev) drv_get_drvdata(&(_dev)->dev) -#define gio_set_drvdata(_dev, data) drv_set_drvdata(&(_dev)->dev, (data)) +#define gio_get_drvdata(_dev) dev_get_drvdata(&(_dev)->dev) +#define gio_set_drvdata(_dev, data) dev_set_drvdata(&(_dev)->dev, (data)) extern void gio_set_master(struct gio_device *); diff --git a/arch/mips/include/asm/idle.h b/arch/mips/include/asm/idle.h index d192158886b..d9f932de80e 100644 --- a/arch/mips/include/asm/idle.h +++ b/arch/mips/include/asm/idle.h @@ -1,6 +1,7 @@ #ifndef __ASM_IDLE_H #define __ASM_IDLE_H +#include <linux/cpuidle.h> #include <linux/linkage.h> extern void (*cpu_wait)(void); @@ -20,4 +21,17 @@ static inline int address_is_in_r4k_wait_irqoff(unsigned long addr) addr < (unsigned long)__pastwait; } +extern int mips_cpuidle_wait_enter(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index); + +#define MIPS_CPUIDLE_WAIT_STATE {\ + .enter = mips_cpuidle_wait_enter,\ + .exit_latency = 1,\ + .target_residency = 1,\ + .power_usage = UINT_MAX,\ + .flags = CPUIDLE_FLAG_TIME_VALID,\ + .name = "wait",\ + .desc = "MIPS wait",\ +} + #endif /* __ASM_IDLE_H */ diff --git a/arch/mips/include/asm/irq.h b/arch/mips/include/asm/irq.h index 7bc2cdb3505..ae1f7b24dd1 100644 --- a/arch/mips/include/asm/irq.h +++ b/arch/mips/include/asm/irq.h @@ -26,104 +26,8 @@ static inline int irq_canonicalize(int irq) #define irq_canonicalize(irq) (irq) /* Sane hardware, sane code ... */ #endif -#ifdef CONFIG_MIPS_MT_SMTC - -struct irqaction; - -extern unsigned long irq_hwmask[]; -extern int setup_irq_smtc(unsigned int irq, struct irqaction * new, - unsigned long hwmask); - -static inline void smtc_im_ack_irq(unsigned int irq) -{ - if (irq_hwmask[irq] & ST0_IM) - set_c0_status(irq_hwmask[irq] & ST0_IM); -} - -#else - -static inline void smtc_im_ack_irq(unsigned int irq) -{ -} - -#endif /* CONFIG_MIPS_MT_SMTC */ - -#ifdef CONFIG_MIPS_MT_SMTC_IRQAFF -#include <linux/cpumask.h> - -extern int plat_set_irq_affinity(struct irq_data *d, - const struct cpumask *affinity, bool force); -extern void smtc_forward_irq(struct irq_data *d); - -/* - * IRQ affinity hook invoked at the beginning of interrupt dispatch - * if option is enabled. - * - * Up through Linux 2.6.22 (at least) cpumask operations are very - * inefficient on MIPS. Initial prototypes of SMTC IRQ affinity - * used a "fast path" per-IRQ-descriptor cache of affinity information - * to reduce latency. As there is a project afoot to optimize the - * cpumask implementations, this version is optimistically assuming - * that cpumask.h macro overhead is reasonable during interrupt dispatch. - */ -static inline int handle_on_other_cpu(unsigned int irq) -{ - struct irq_data *d = irq_get_irq_data(irq); - - if (cpumask_test_cpu(smp_processor_id(), d->affinity)) - return 0; - smtc_forward_irq(d); - return 1; -} - -#else /* Not doing SMTC affinity */ - -static inline int handle_on_other_cpu(unsigned int irq) { return 0; } - -#endif /* CONFIG_MIPS_MT_SMTC_IRQAFF */ - -#ifdef CONFIG_MIPS_MT_SMTC_IM_BACKSTOP - -static inline void smtc_im_backstop(unsigned int irq) -{ - if (irq_hwmask[irq] & 0x0000ff00) - write_c0_tccontext(read_c0_tccontext() & - ~(irq_hwmask[irq] & 0x0000ff00)); -} - -/* - * Clear interrupt mask handling "backstop" if irq_hwmask - * entry so indicates. This implies that the ack() or end() - * functions will take over re-enabling the low-level mask. - * Otherwise it will be done on return from exception. - */ -static inline int smtc_handle_on_other_cpu(unsigned int irq) -{ - int ret = handle_on_other_cpu(irq); - - if (!ret) - smtc_im_backstop(irq); - return ret; -} - -#else - -static inline void smtc_im_backstop(unsigned int irq) { } -static inline int smtc_handle_on_other_cpu(unsigned int irq) -{ - return handle_on_other_cpu(irq); -} - -#endif - extern void do_IRQ(unsigned int irq); -#ifdef CONFIG_MIPS_MT_SMTC_IRQAFF - -extern void do_IRQ_no_affinity(unsigned int irq); - -#endif /* CONFIG_MIPS_MT_SMTC_IRQAFF */ - extern void arch_init_irq(void); extern void spurious_interrupt(void); diff --git a/arch/mips/include/asm/irqflags.h b/arch/mips/include/asm/irqflags.h index 45c00951888..0fa5fdcd1f0 100644 --- a/arch/mips/include/asm/irqflags.h +++ b/arch/mips/include/asm/irqflags.h @@ -17,7 +17,7 @@ #include <linux/stringify.h> #include <asm/hazards.h> -#if defined(CONFIG_CPU_MIPSR2) && !defined(CONFIG_MIPS_MT_SMTC) +#ifdef CONFIG_CPU_MIPSR2 static inline void arch_local_irq_disable(void) { @@ -118,30 +118,15 @@ void arch_local_irq_disable(void); unsigned long arch_local_irq_save(void); void arch_local_irq_restore(unsigned long flags); void __arch_local_irq_restore(unsigned long flags); -#endif /* if defined(CONFIG_CPU_MIPSR2) && !defined(CONFIG_MIPS_MT_SMTC) */ - - -extern void smtc_ipi_replay(void); +#endif /* CONFIG_CPU_MIPSR2 */ static inline void arch_local_irq_enable(void) { -#ifdef CONFIG_MIPS_MT_SMTC - /* - * SMTC kernel needs to do a software replay of queued - * IPIs, at the cost of call overhead on each local_irq_enable() - */ - smtc_ipi_replay(); -#endif __asm__ __volatile__( " .set push \n" " .set reorder \n" " .set noat \n" -#ifdef CONFIG_MIPS_MT_SMTC - " mfc0 $1, $2, 1 # SMTC - clear TCStatus.IXMT \n" - " ori $1, 0x400 \n" - " xori $1, 0x400 \n" - " mtc0 $1, $2, 1 \n" -#elif defined(CONFIG_CPU_MIPSR2) +#if defined(CONFIG_CPU_MIPSR2) " ei \n" #else " mfc0 $1,$12 \n" @@ -163,11 +148,7 @@ static inline unsigned long arch_local_save_flags(void) asm __volatile__( " .set push \n" " .set reorder \n" -#ifdef CONFIG_MIPS_MT_SMTC - " mfc0 %[flags], $2, 1 \n" -#else " mfc0 %[flags], $12 \n" -#endif " .set pop \n" : [flags] "=r" (flags)); @@ -177,14 +158,7 @@ static inline unsigned long arch_local_save_flags(void) static inline int arch_irqs_disabled_flags(unsigned long flags) { -#ifdef CONFIG_MIPS_MT_SMTC - /* - * SMTC model uses TCStatus.IXMT to disable interrupts for a thread/CPU - */ - return flags & 0x400; -#else return !(flags & 1); -#endif } #endif /* #ifndef __ASSEMBLY__ */ diff --git a/arch/mips/include/asm/kvm_para.h b/arch/mips/include/asm/kvm_para.h new file mode 100644 index 00000000000..5a9aa918abe --- /dev/null +++ b/arch/mips/include/asm/kvm_para.h @@ -0,0 +1,109 @@ +#ifndef _ASM_MIPS_KVM_PARA_H +#define _ASM_MIPS_KVM_PARA_H + +#include <uapi/asm/kvm_para.h> + +#define KVM_HYPERCALL ".word 0x42000028" + +/* + * Hypercalls for KVM. + * + * Hypercall number is passed in v0. + * Return value will be placed in v0. + * Up to 3 arguments are passed in a0, a1, and a2. + */ +static inline unsigned long kvm_hypercall0(unsigned long num) +{ + register unsigned long n asm("v0"); + register unsigned long r asm("v0"); + + n = num; + __asm__ __volatile__( + KVM_HYPERCALL + : "=r" (r) : "r" (n) : "memory" + ); + + return r; +} + +static inline unsigned long kvm_hypercall1(unsigned long num, + unsigned long arg0) +{ + register unsigned long n asm("v0"); + register unsigned long r asm("v0"); + register unsigned long a0 asm("a0"); + + n = num; + a0 = arg0; + __asm__ __volatile__( + KVM_HYPERCALL + : "=r" (r) : "r" (n), "r" (a0) : "memory" + ); + + return r; +} + +static inline unsigned long kvm_hypercall2(unsigned long num, + unsigned long arg0, unsigned long arg1) +{ + register unsigned long n asm("v0"); + register unsigned long r asm("v0"); + register unsigned long a0 asm("a0"); + register unsigned long a1 asm("a1"); + + n = num; + a0 = arg0; + a1 = arg1; + __asm__ __volatile__( + KVM_HYPERCALL + : "=r" (r) : "r" (n), "r" (a0), "r" (a1) : "memory" + ); + + return r; +} + +static inline unsigned long kvm_hypercall3(unsigned long num, + unsigned long arg0, unsigned long arg1, unsigned long arg2) +{ + register unsigned long n asm("v0"); + register unsigned long r asm("v0"); + register unsigned long a0 asm("a0"); + register unsigned long a1 asm("a1"); + register unsigned long a2 asm("a2"); + + n = num; + a0 = arg0; + a1 = arg1; + a2 = arg2; + __asm__ __volatile__( + KVM_HYPERCALL + : "=r" (r) : "r" (n), "r" (a0), "r" (a1), "r" (a2) : "memory" + ); + + return r; +} + +static inline bool kvm_check_and_clear_guest_paused(void) +{ + return false; +} + +static inline unsigned int kvm_arch_para_features(void) +{ + return 0; +} + +#ifdef CONFIG_MIPS_PARAVIRT +static inline bool kvm_para_available(void) +{ + return true; +} +#else +static inline bool kvm_para_available(void) +{ + return false; +} +#endif + + +#endif /* _ASM_MIPS_KVM_PARA_H */ diff --git a/arch/mips/include/asm/mach-cavium-octeon/cpu-feature-overrides.h b/arch/mips/include/asm/mach-cavium-octeon/cpu-feature-overrides.h index 94ed063eec9..cf802287289 100644 --- a/arch/mips/include/asm/mach-cavium-octeon/cpu-feature-overrides.h +++ b/arch/mips/include/asm/mach-cavium-octeon/cpu-feature-overrides.h @@ -22,7 +22,6 @@ #define cpu_has_3k_cache 0 #define cpu_has_4k_cache 0 #define cpu_has_tx39_cache 0 -#define cpu_has_fpu 0 #define cpu_has_counter 1 #define cpu_has_watch 1 #define cpu_has_divec 1 diff --git a/arch/mips/include/asm/mach-cavium-octeon/irq.h b/arch/mips/include/asm/mach-cavium-octeon/irq.h index 60fc4c347c4..cceae32a073 100644 --- a/arch/mips/include/asm/mach-cavium-octeon/irq.h +++ b/arch/mips/include/asm/mach-cavium-octeon/irq.h @@ -35,6 +35,8 @@ enum octeon_irq { OCTEON_IRQ_PCI_MSI2, OCTEON_IRQ_PCI_MSI3, + OCTEON_IRQ_TWSI, + OCTEON_IRQ_TWSI2, OCTEON_IRQ_RML, OCTEON_IRQ_TIMER0, OCTEON_IRQ_TIMER1, diff --git a/arch/mips/include/asm/mach-ip22/cpu-feature-overrides.h b/arch/mips/include/asm/mach-ip22/cpu-feature-overrides.h index 1bcb6421205..1dfe47453ea 100644 --- a/arch/mips/include/asm/mach-ip22/cpu-feature-overrides.h +++ b/arch/mips/include/asm/mach-ip22/cpu-feature-overrides.h @@ -39,6 +39,10 @@ #define cpu_has_nofpuex 0 #define cpu_has_64bits 1 +#define cpu_has_mips_2 1 +#define cpu_has_mips_3 1 +#define cpu_has_mips_5 0 + #define cpu_has_mips32r1 0 #define cpu_has_mips32r2 0 #define cpu_has_mips64r1 0 diff --git a/arch/mips/include/asm/mach-malta/kernel-entry-init.h b/arch/mips/include/asm/mach-malta/kernel-entry-init.h index 7c5e17a1784..77eeda77e73 100644 --- a/arch/mips/include/asm/mach-malta/kernel-entry-init.h +++ b/arch/mips/include/asm/mach-malta/kernel-entry-init.h @@ -80,36 +80,6 @@ .endm .macro kernel_entry_setup -#ifdef CONFIG_MIPS_MT_SMTC - mfc0 t0, CP0_CONFIG - bgez t0, 9f - mfc0 t0, CP0_CONFIG, 1 - bgez t0, 9f - mfc0 t0, CP0_CONFIG, 2 - bgez t0, 9f - mfc0 t0, CP0_CONFIG, 3 - and t0, 1<<2 - bnez t0, 0f -9: - /* Assume we came from YAMON... */ - PTR_LA v0, 0x9fc00534 /* YAMON print */ - lw v0, (v0) - move a0, zero - PTR_LA a1, nonmt_processor - jal v0 - - PTR_LA v0, 0x9fc00520 /* YAMON exit */ - lw v0, (v0) - li a0, 1 - jal v0 - -1: b 1b - - __INITDATA -nonmt_processor: - .asciz "SMTC kernel requires the MT ASE to run\n" - __FINIT -#endif #ifdef CONFIG_EVA sync diff --git a/arch/mips/include/asm/mach-malta/malta-pm.h b/arch/mips/include/asm/mach-malta/malta-pm.h new file mode 100644 index 00000000000..c2c2e201013 --- /dev/null +++ b/arch/mips/include/asm/mach-malta/malta-pm.h @@ -0,0 +1,37 @@ +/* + * Copyright (C) 2014 Imagination Technologies + * Author: Paul Burton <paul.burton@imgtec.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + */ + +#ifndef __ASM_MIPS_MACH_MALTA_PM_H__ +#define __ASM_MIPS_MACH_MALTA_PM_H__ + +#include <asm/mips-boards/piix4.h> + +#ifdef CONFIG_MIPS_MALTA_PM + +/** + * mips_pm_suspend - enter a suspend state + * @state: the state to enter, one of PIIX4_FUNC3IO_PMCNTRL_SUS_TYP_* + * + * Enters a suspend state via the Malta's PIIX4. If the state to be entered + * is one which loses context (eg. SOFF) then this function will never + * return. + */ +extern int mips_pm_suspend(unsigned state); + +#else /* !CONFIG_MIPS_MALTA_PM */ + +static inline int mips_pm_suspend(unsigned state) +{ + return -EINVAL; +} + +#endif /* !CONFIG_MIPS_MALTA_PM */ + +#endif /* __ASM_MIPS_MACH_MALTA_PM_H__ */ diff --git a/arch/mips/include/asm/mach-netlogic/topology.h b/arch/mips/include/asm/mach-netlogic/topology.h index 0da99fa11c3..ceeb1f5e712 100644 --- a/arch/mips/include/asm/mach-netlogic/topology.h +++ b/arch/mips/include/asm/mach-netlogic/topology.h @@ -10,10 +10,12 @@ #include <asm/mach-netlogic/multi-node.h> +#ifdef CONFIG_SMP #define topology_physical_package_id(cpu) cpu_to_node(cpu) #define topology_core_id(cpu) (cpu_logical_map(cpu) / NLM_THREADS_PER_CORE) #define topology_thread_cpumask(cpu) (&cpu_sibling_map[cpu]) #define topology_core_cpumask(cpu) cpumask_of_node(cpu_to_node(cpu)) +#endif #include <asm-generic/topology.h> diff --git a/arch/mips/include/asm/mach-paravirt/cpu-feature-overrides.h b/arch/mips/include/asm/mach-paravirt/cpu-feature-overrides.h new file mode 100644 index 00000000000..725e1ed83f6 --- /dev/null +++ b/arch/mips/include/asm/mach-paravirt/cpu-feature-overrides.h @@ -0,0 +1,36 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2013 Cavium, Inc. + */ +#ifndef __ASM_MACH_PARAVIRT_CPU_FEATURE_OVERRIDES_H +#define __ASM_MACH_PARAVIRT_CPU_FEATURE_OVERRIDES_H + +#define cpu_has_4kex 1 +#define cpu_has_3k_cache 0 +#define cpu_has_tx39_cache 0 +#define cpu_has_counter 1 +#define cpu_has_llsc 1 +/* + * We Disable LL/SC on non SMP systems as it is faster to disable + * interrupts for atomic access than a LL/SC. + */ +#ifdef CONFIG_SMP +# define kernel_uses_llsc 1 +#else +# define kernel_uses_llsc 0 +#endif + +#ifdef CONFIG_CPU_CAVIUM_OCTEON +#define cpu_dcache_line_size() 128 +#define cpu_icache_line_size() 128 +#define cpu_has_octeon_cache 1 +#define cpu_has_4k_cache 0 +#else +#define cpu_has_octeon_cache 0 +#define cpu_has_4k_cache 1 +#endif + +#endif /* __ASM_MACH_PARAVIRT_CPU_FEATURE_OVERRIDES_H */ diff --git a/arch/mips/include/asm/mach-paravirt/irq.h b/arch/mips/include/asm/mach-paravirt/irq.h new file mode 100644 index 00000000000..9b4d35eca97 --- /dev/null +++ b/arch/mips/include/asm/mach-paravirt/irq.h @@ -0,0 +1,19 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2013 Cavium, Inc. + */ +#ifndef __ASM_MACH_PARAVIRT_IRQ_H__ +#define __ASM_MACH_PARAVIRT_IRQ_H__ + +#define NR_IRQS 64 +#define MIPS_CPU_IRQ_BASE 1 + +#define MIPS_IRQ_PCIA (MIPS_CPU_IRQ_BASE + 8) + +#define MIPS_IRQ_MBOX0 (MIPS_CPU_IRQ_BASE + 32) +#define MIPS_IRQ_MBOX1 (MIPS_CPU_IRQ_BASE + 33) + +#endif /* __ASM_MACH_PARAVIRT_IRQ_H__ */ diff --git a/arch/mips/include/asm/mach-paravirt/kernel-entry-init.h b/arch/mips/include/asm/mach-paravirt/kernel-entry-init.h new file mode 100644 index 00000000000..2f82bfa3a77 --- /dev/null +++ b/arch/mips/include/asm/mach-paravirt/kernel-entry-init.h @@ -0,0 +1,50 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2013 Cavium, Inc + */ +#ifndef __ASM_MACH_PARAVIRT_KERNEL_ENTRY_H +#define __ASM_MACH_PARAVIRT_KERNEL_ENTRY_H + +#define CP0_EBASE $15, 1 + + .macro kernel_entry_setup + mfc0 t0, CP0_EBASE + andi t0, t0, 0x3ff # CPUNum + beqz t0, 1f + # CPUs other than zero goto smp_bootstrap + j smp_bootstrap + +1: + .endm + +/* + * Do SMP slave processor setup necessary before we can safely execute + * C code. + */ + .macro smp_slave_setup + mfc0 t0, CP0_EBASE + andi t0, t0, 0x3ff # CPUNum + slti t1, t0, NR_CPUS + bnez t1, 1f +2: + di + wait + b 2b # Unknown CPU, loop forever. +1: + PTR_LA t1, paravirt_smp_sp + PTR_SLL t0, PTR_SCALESHIFT + PTR_ADDU t1, t1, t0 +3: + PTR_L sp, 0(t1) + beqz sp, 3b # Spin until told to proceed. + + PTR_LA t1, paravirt_smp_gp + PTR_ADDU t1, t1, t0 + sync + PTR_L gp, 0(t1) + .endm + +#endif /* __ASM_MACH_PARAVIRT_KERNEL_ENTRY_H */ diff --git a/arch/mips/include/asm/mach-paravirt/war.h b/arch/mips/include/asm/mach-paravirt/war.h new file mode 100644 index 00000000000..36d3afb9845 --- /dev/null +++ b/arch/mips/include/asm/mach-paravirt/war.h @@ -0,0 +1,25 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2002, 2004, 2007 by Ralf Baechle <ralf@linux-mips.org> + * Copyright (C) 2013 Cavium Networks <support@caviumnetworks.com> + */ +#ifndef __ASM_MIPS_MACH_PARAVIRT_WAR_H +#define __ASM_MIPS_MACH_PARAVIRT_WAR_H + +#define R4600_V1_INDEX_ICACHEOP_WAR 0 +#define R4600_V1_HIT_CACHEOP_WAR 0 +#define R4600_V2_HIT_CACHEOP_WAR 0 +#define R5432_CP0_INTERRUPT_WAR 0 +#define BCM1250_M3_WAR 0 +#define SIBYTE_1956_WAR 0 +#define MIPS4K_ICACHE_REFILL_WAR 0 +#define MIPS_CACHE_SYNC_WAR 0 +#define TX49XX_ICACHE_INDEX_INV_WAR 0 +#define ICACHE_REFILLS_WORKAROUND_WAR 0 +#define R10000_LLSC_WAR 0 +#define MIPS34K_MISSED_ITLB_WAR 0 + +#endif /* __ASM_MIPS_MACH_PARAVIRT_WAR_H */ diff --git a/arch/mips/include/asm/mach-pmcs-msp71xx/msp_usb.h b/arch/mips/include/asm/mach-pmcs-msp71xx/msp_usb.h index aa45e6a0712..fe1566f2913 100644 --- a/arch/mips/include/asm/mach-pmcs-msp71xx/msp_usb.h +++ b/arch/mips/include/asm/mach-pmcs-msp71xx/msp_usb.h @@ -25,11 +25,7 @@ #ifndef MSP_USB_H_ #define MSP_USB_H_ -#ifdef CONFIG_MSP_HAS_DUAL_USB -#define NUM_USB_DEVS 2 -#else #define NUM_USB_DEVS 1 -#endif /* Register spaces for USB host 0 */ #define MSP_USB0_MAB_START (MSP_USB0_BASE + 0x0) diff --git a/arch/mips/include/asm/mach-ralink/war.h b/arch/mips/include/asm/mach-ralink/war.h index a7b712cf2d2..c074b5dc1f8 100644 --- a/arch/mips/include/asm/mach-ralink/war.h +++ b/arch/mips/include/asm/mach-ralink/war.h @@ -17,7 +17,6 @@ #define MIPS4K_ICACHE_REFILL_WAR 0 #define MIPS_CACHE_SYNC_WAR 0 #define TX49XX_ICACHE_INDEX_INV_WAR 0 -#define RM9000_CDEX_SMP_WAR 0 #define ICACHE_REFILLS_WORKAROUND_WAR 0 #define R10000_LLSC_WAR 0 #define MIPS34K_MISSED_ITLB_WAR 0 diff --git a/arch/mips/include/asm/mach-sead3/kernel-entry-init.h b/arch/mips/include/asm/mach-sead3/kernel-entry-init.h index 3dfbd8e7947..6cccd4d558d 100644 --- a/arch/mips/include/asm/mach-sead3/kernel-entry-init.h +++ b/arch/mips/include/asm/mach-sead3/kernel-entry-init.h @@ -10,37 +10,6 @@ #define __ASM_MACH_MIPS_KERNEL_ENTRY_INIT_H .macro kernel_entry_setup -#ifdef CONFIG_MIPS_MT_SMTC - mfc0 t0, CP0_CONFIG - bgez t0, 9f - mfc0 t0, CP0_CONFIG, 1 - bgez t0, 9f - mfc0 t0, CP0_CONFIG, 2 - bgez t0, 9f - mfc0 t0, CP0_CONFIG, 3 - and t0, 1<<2 - bnez t0, 0f -9 : - /* Assume we came from YAMON... */ - PTR_LA v0, 0x9fc00534 /* YAMON print */ - lw v0, (v0) - move a0, zero - PTR_LA a1, nonmt_processor - jal v0 - - PTR_LA v0, 0x9fc00520 /* YAMON exit */ - lw v0, (v0) - li a0, 1 - jal v0 - -1 : b 1b - - __INITDATA -nonmt_processor : - .asciz "SMTC kernel requires the MT ASE to run\n" - __FINIT -0 : -#endif .endm /* diff --git a/arch/mips/include/asm/mips-boards/piix4.h b/arch/mips/include/asm/mips-boards/piix4.h index 9cf54041d41..9e340be52a5 100644 --- a/arch/mips/include/asm/mips-boards/piix4.h +++ b/arch/mips/include/asm/mips-boards/piix4.h @@ -55,4 +55,16 @@ #define PIIX4_FUNC3_PMREGMISC 0x80 #define PIIX4_FUNC3_PMREGMISC_EN (1 << 0) +/* Power Management IO Space */ +#define PIIX4_FUNC3IO_PMSTS 0x00 +#define PIIX4_FUNC3IO_PMSTS_PWRBTN_STS (1 << 8) +#define PIIX4_FUNC3IO_PMCNTRL 0x04 +#define PIIX4_FUNC3IO_PMCNTRL_SUS_EN (1 << 13) +#define PIIX4_FUNC3IO_PMCNTRL_SUS_TYP (0x7 << 10) +#define PIIX4_FUNC3IO_PMCNTRL_SUS_TYP_SOFF (0x0 << 10) +#define PIIX4_FUNC3IO_PMCNTRL_SUS_TYP_STR (0x1 << 10) + +/* Data for magic special PCI cycle */ +#define PIIX4_SUSPEND_MAGIC 0x00120002 + #endif /* __ASM_MIPS_BOARDS_PIIX4_H */ diff --git a/arch/mips/include/asm/mips-cpc.h b/arch/mips/include/asm/mips-cpc.h index 988507e46d4..e139a534e0f 100644 --- a/arch/mips/include/asm/mips-cpc.h +++ b/arch/mips/include/asm/mips-cpc.h @@ -72,7 +72,12 @@ static inline bool mips_cpc_present(void) #define MIPS_CPC_COCB_OFS 0x4000 /* Macros to ease the creation of register access functions */ -#define BUILD_CPC_R_(name, off) \ +#define BUILD_CPC_R_(name, off) \ +static inline u32 *addr_cpc_##name(void) \ +{ \ + return (u32 *)(mips_cpc_base + (off)); \ +} \ + \ static inline u32 read_cpc_##name(void) \ { \ return __raw_readl(mips_cpc_base + (off)); \ @@ -147,4 +152,31 @@ BUILD_CPC_Cx_RW(other, 0x10) #define CPC_Cx_OTHER_CORENUM_SHF 16 #define CPC_Cx_OTHER_CORENUM_MSK (_ULCAST_(0xff) << 16) +#ifdef CONFIG_MIPS_CPC + +/** + * mips_cpc_lock_other - lock access to another core + * core: the other core to be accessed + * + * Call before operating upon a core via the 'other' register region in + * order to prevent the region being moved during access. Must be followed + * by a call to mips_cpc_unlock_other. + */ +extern void mips_cpc_lock_other(unsigned int core); + +/** + * mips_cpc_unlock_other - unlock access to another core + * + * Call after operating upon another core via the 'other' register region. + * Must be called after mips_cpc_lock_other. + */ +extern void mips_cpc_unlock_other(void); + +#else /* !CONFIG_MIPS_CPC */ + +static inline void mips_cpc_lock_other(unsigned int core) { } +static inline void mips_cpc_unlock_other(void) { } + +#endif /* !CONFIG_MIPS_CPC */ + #endif /* __MIPS_ASM_MIPS_CPC_H__ */ diff --git a/arch/mips/include/asm/mips_mt.h b/arch/mips/include/asm/mips_mt.h index a3df0c3faa0..f6ba004a771 100644 --- a/arch/mips/include/asm/mips_mt.h +++ b/arch/mips/include/asm/mips_mt.h @@ -1,7 +1,6 @@ /* - * Definitions and decalrations for MIPS MT support - * that are common between SMTC, VSMP, and/or AP/SP - * kernel models. + * Definitions and decalrations for MIPS MT support that are common between + * the VSMP, and AP/SP kernel models. */ #ifndef __ASM_MIPS_MT_H #define __ASM_MIPS_MT_H diff --git a/arch/mips/include/asm/mipsmtregs.h b/arch/mips/include/asm/mipsmtregs.h index 6efa79a27b6..5f8052ce43b 100644 --- a/arch/mips/include/asm/mipsmtregs.h +++ b/arch/mips/include/asm/mipsmtregs.h @@ -36,6 +36,8 @@ #define read_c0_tcbind() __read_32bit_c0_register($2, 2) +#define write_c0_tchalt(val) __write_32bit_c0_register($2, 4, val) + #define read_c0_tccontext() __read_32bit_c0_register($2, 5) #define write_c0_tccontext(val) __write_32bit_c0_register($2, 5, val) diff --git a/arch/mips/include/asm/mipsregs.h b/arch/mips/include/asm/mipsregs.h index 3e025b5311d..98e9754a4b6 100644 --- a/arch/mips/include/asm/mipsregs.h +++ b/arch/mips/include/asm/mipsregs.h @@ -709,11 +709,18 @@ #ifndef __ASSEMBLY__ /* - * Macros for handling the ISA mode bit for microMIPS. + * Macros for handling the ISA mode bit for MIPS16 and microMIPS. */ +#if defined(CONFIG_SYS_SUPPORTS_MIPS16) || \ + defined(CONFIG_SYS_SUPPORTS_MICROMIPS) #define get_isa16_mode(x) ((x) & 0x1) #define msk_isa16_mode(x) ((x) & ~0x1) #define set_isa16_mode(x) do { (x) |= 0x1; } while(0) +#else +#define get_isa16_mode(x) 0 +#define msk_isa16_mode(x) (x) +#define set_isa16_mode(x) do { } while(0) +#endif /* * microMIPS instructions can be 16-bit or 32-bit in length. This @@ -1007,19 +1014,8 @@ do { \ #define write_c0_compare3(val) __write_32bit_c0_register($11, 7, val) #define read_c0_status() __read_32bit_c0_register($12, 0) -#ifdef CONFIG_MIPS_MT_SMTC -#define write_c0_status(val) \ -do { \ - __write_32bit_c0_register($12, 0, val); \ - __ehb(); \ -} while (0) -#else -/* - * Legacy non-SMTC code, which may be hazardous - * but which might not support EHB - */ + #define write_c0_status(val) __write_32bit_c0_register($12, 0, val) -#endif /* CONFIG_MIPS_MT_SMTC */ #define read_c0_cause() __read_32bit_c0_register($13, 0) #define write_c0_cause(val) __write_32bit_c0_register($13, 0, val) @@ -1743,11 +1739,6 @@ static inline void tlb_write_random(void) /* * Manipulate bits in a c0 register. */ -#ifndef CONFIG_MIPS_MT_SMTC -/* - * SMTC Linux requires shutting-down microthread scheduling - * during CP0 register read-modify-write sequences. - */ #define __BUILD_SET_C0(name) \ static inline unsigned int \ set_c0_##name(unsigned int set) \ @@ -1786,121 +1777,6 @@ change_c0_##name(unsigned int change, unsigned int val) \ return res; \ } -#else /* SMTC versions that manage MT scheduling */ - -#include <linux/irqflags.h> - -/* - * This is a duplicate of dmt() in mipsmtregs.h to avoid problems with - * header file recursion. - */ -static inline unsigned int __dmt(void) -{ - int res; - - __asm__ __volatile__( - " .set push \n" - " .set mips32r2 \n" - " .set noat \n" - " .word 0x41610BC1 # dmt $1 \n" - " ehb \n" - " move %0, $1 \n" - " .set pop \n" - : "=r" (res)); - - instruction_hazard(); - - return res; -} - -#define __VPECONTROL_TE_SHIFT 15 -#define __VPECONTROL_TE (1UL << __VPECONTROL_TE_SHIFT) - -#define __EMT_ENABLE __VPECONTROL_TE - -static inline void __emt(unsigned int previous) -{ - if ((previous & __EMT_ENABLE)) - __asm__ __volatile__( - " .set mips32r2 \n" - " .word 0x41600be1 # emt \n" - " ehb \n" - " .set mips0 \n"); -} - -static inline void __ehb(void) -{ - __asm__ __volatile__( - " .set mips32r2 \n" - " ehb \n" " .set mips0 \n"); -} - -/* - * Note that local_irq_save/restore affect TC-specific IXMT state, - * not Status.IE as in non-SMTC kernel. - */ - -#define __BUILD_SET_C0(name) \ -static inline unsigned int \ -set_c0_##name(unsigned int set) \ -{ \ - unsigned int res; \ - unsigned int new; \ - unsigned int omt; \ - unsigned long flags; \ - \ - local_irq_save(flags); \ - omt = __dmt(); \ - res = read_c0_##name(); \ - new = res | set; \ - write_c0_##name(new); \ - __emt(omt); \ - local_irq_restore(flags); \ - \ - return res; \ -} \ - \ -static inline unsigned int \ -clear_c0_##name(unsigned int clear) \ -{ \ - unsigned int res; \ - unsigned int new; \ - unsigned int omt; \ - unsigned long flags; \ - \ - local_irq_save(flags); \ - omt = __dmt(); \ - res = read_c0_##name(); \ - new = res & ~clear; \ - write_c0_##name(new); \ - __emt(omt); \ - local_irq_restore(flags); \ - \ - return res; \ -} \ - \ -static inline unsigned int \ -change_c0_##name(unsigned int change, unsigned int newbits) \ -{ \ - unsigned int res; \ - unsigned int new; \ - unsigned int omt; \ - unsigned long flags; \ - \ - local_irq_save(flags); \ - \ - omt = __dmt(); \ - res = read_c0_##name(); \ - new = res & ~change; \ - new |= (newbits & change); \ - write_c0_##name(new); \ - __emt(omt); \ - local_irq_restore(flags); \ - \ - return res; \ -} -#endif - __BUILD_SET_C0(status) __BUILD_SET_C0(cause) __BUILD_SET_C0(config) @@ -1916,6 +1792,15 @@ __BUILD_SET_C0(brcm_cmt_ctrl) __BUILD_SET_C0(brcm_config) __BUILD_SET_C0(brcm_mode) +/* + * Return low 10 bits of ebase. + * Note that under KVM (MIPSVZ) this returns vcpu id. + */ +static inline unsigned int get_ebase_cpunum(void) +{ + return read_c0_ebase() & 0x3ff; +} + #endif /* !__ASSEMBLY__ */ #endif /* _ASM_MIPSREGS_H */ diff --git a/arch/mips/include/asm/mmu_context.h b/arch/mips/include/asm/mmu_context.h index e277bbad287..2e373da5f8e 100644 --- a/arch/mips/include/asm/mmu_context.h +++ b/arch/mips/include/asm/mmu_context.h @@ -18,10 +18,6 @@ #include <asm/cacheflush.h> #include <asm/hazards.h> #include <asm/tlbflush.h> -#ifdef CONFIG_MIPS_MT_SMTC -#include <asm/mipsmtregs.h> -#include <asm/smtc.h> -#endif /* SMTC */ #include <asm-generic/mm_hooks.h> #define TLBMISS_HANDLER_SETUP_PGD(pgd) \ @@ -31,11 +27,15 @@ do { \ } while (0) #ifdef CONFIG_MIPS_PGD_C0_CONTEXT + +#define TLBMISS_HANDLER_RESTORE() \ + write_c0_xcontext((unsigned long) smp_processor_id() << \ + SMP_CPUID_REGSHIFT) + #define TLBMISS_HANDLER_SETUP() \ do { \ TLBMISS_HANDLER_SETUP_PGD(swapper_pg_dir); \ - write_c0_xcontext((unsigned long) smp_processor_id() << \ - SMP_CPUID_REGSHIFT); \ + TLBMISS_HANDLER_RESTORE(); \ } while (0) #else /* !CONFIG_MIPS_PGD_C0_CONTEXT: using pgd_current*/ @@ -47,9 +47,12 @@ do { \ */ extern unsigned long pgd_current[]; -#define TLBMISS_HANDLER_SETUP() \ +#define TLBMISS_HANDLER_RESTORE() \ write_c0_context((unsigned long) smp_processor_id() << \ - SMP_CPUID_REGSHIFT); \ + SMP_CPUID_REGSHIFT) + +#define TLBMISS_HANDLER_SETUP() \ + TLBMISS_HANDLER_RESTORE(); \ back_to_back_c0_hazard(); \ TLBMISS_HANDLER_SETUP_PGD(swapper_pg_dir) #endif /* CONFIG_MIPS_PGD_C0_CONTEXT*/ @@ -63,13 +66,6 @@ extern unsigned long pgd_current[]; #define ASID_INC 0x10 #define ASID_MASK 0xff0 -#elif defined(CONFIG_MIPS_MT_SMTC) - -#define ASID_INC 0x1 -extern unsigned long smtc_asid_mask; -#define ASID_MASK (smtc_asid_mask) -#define HW_ASID_MASK 0xff -/* End SMTC/34K debug hack */ #else /* FIXME: not correct for R6000 */ #define ASID_INC 0x1 @@ -92,7 +88,6 @@ static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk) #define ASID_VERSION_MASK ((unsigned long)~(ASID_MASK|(ASID_MASK-1))) #define ASID_FIRST_VERSION ((unsigned long)(~ASID_VERSION_MASK) + 1) -#ifndef CONFIG_MIPS_MT_SMTC /* Normal, classic MIPS get_new_mmu_context */ static inline void get_new_mmu_context(struct mm_struct *mm, unsigned long cpu) @@ -115,12 +110,6 @@ get_new_mmu_context(struct mm_struct *mm, unsigned long cpu) cpu_context(cpu, mm) = asid_cache(cpu) = asid; } -#else /* CONFIG_MIPS_MT_SMTC */ - -#define get_new_mmu_context(mm, cpu) smtc_get_new_mmu_context((mm), (cpu)) - -#endif /* CONFIG_MIPS_MT_SMTC */ - /* * Initialize the context related info for a new mm_struct * instance. @@ -141,46 +130,12 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, { unsigned int cpu = smp_processor_id(); unsigned long flags; -#ifdef CONFIG_MIPS_MT_SMTC - unsigned long oldasid; - unsigned long mtflags; - int mytlb = (smtc_status & SMTC_TLB_SHARED) ? 0 : cpu_data[cpu].vpe_id; local_irq_save(flags); - mtflags = dvpe(); -#else /* Not SMTC */ - local_irq_save(flags); -#endif /* CONFIG_MIPS_MT_SMTC */ /* Check if our ASID is of an older version and thus invalid */ if ((cpu_context(cpu, next) ^ asid_cache(cpu)) & ASID_VERSION_MASK) get_new_mmu_context(next, cpu); -#ifdef CONFIG_MIPS_MT_SMTC - /* - * If the EntryHi ASID being replaced happens to be - * the value flagged at ASID recycling time as having - * an extended life, clear the bit showing it being - * in use by this "CPU", and if that's the last bit, - * free up the ASID value for use and flush any old - * instances of it from the TLB. - */ - oldasid = (read_c0_entryhi() & ASID_MASK); - if(smtc_live_asid[mytlb][oldasid]) { - smtc_live_asid[mytlb][oldasid] &= ~(0x1 << cpu); - if(smtc_live_asid[mytlb][oldasid] == 0) - smtc_flush_tlb_asid(oldasid); - } - /* - * Tread softly on EntryHi, and so long as we support - * having ASID_MASK smaller than the hardware maximum, - * make sure no "soft" bits become "hard"... - */ - write_c0_entryhi((read_c0_entryhi() & ~HW_ASID_MASK) | - cpu_asid(cpu, next)); - ehb(); /* Make sure it propagates to TCStatus */ - evpe(mtflags); -#else write_c0_entryhi(cpu_asid(cpu, next)); -#endif /* CONFIG_MIPS_MT_SMTC */ TLBMISS_HANDLER_SETUP_PGD(next->pgd); /* @@ -213,34 +168,12 @@ activate_mm(struct mm_struct *prev, struct mm_struct *next) unsigned long flags; unsigned int cpu = smp_processor_id(); -#ifdef CONFIG_MIPS_MT_SMTC - unsigned long oldasid; - unsigned long mtflags; - int mytlb = (smtc_status & SMTC_TLB_SHARED) ? 0 : cpu_data[cpu].vpe_id; -#endif /* CONFIG_MIPS_MT_SMTC */ - local_irq_save(flags); /* Unconditionally get a new ASID. */ get_new_mmu_context(next, cpu); -#ifdef CONFIG_MIPS_MT_SMTC - /* See comments for similar code above */ - mtflags = dvpe(); - oldasid = read_c0_entryhi() & ASID_MASK; - if(smtc_live_asid[mytlb][oldasid]) { - smtc_live_asid[mytlb][oldasid] &= ~(0x1 << cpu); - if(smtc_live_asid[mytlb][oldasid] == 0) - smtc_flush_tlb_asid(oldasid); - } - /* See comments for similar code above */ - write_c0_entryhi((read_c0_entryhi() & ~HW_ASID_MASK) | - cpu_asid(cpu, next)); - ehb(); /* Make sure it propagates to TCStatus */ - evpe(mtflags); -#else write_c0_entryhi(cpu_asid(cpu, next)); -#endif /* CONFIG_MIPS_MT_SMTC */ TLBMISS_HANDLER_SETUP_PGD(next->pgd); /* mark mmu ownership change */ @@ -258,48 +191,15 @@ static inline void drop_mmu_context(struct mm_struct *mm, unsigned cpu) { unsigned long flags; -#ifdef CONFIG_MIPS_MT_SMTC - unsigned long oldasid; - /* Can't use spinlock because called from TLB flush within DVPE */ - unsigned int prevvpe; - int mytlb = (smtc_status & SMTC_TLB_SHARED) ? 0 : cpu_data[cpu].vpe_id; -#endif /* CONFIG_MIPS_MT_SMTC */ local_irq_save(flags); if (cpumask_test_cpu(cpu, mm_cpumask(mm))) { get_new_mmu_context(mm, cpu); -#ifdef CONFIG_MIPS_MT_SMTC - /* See comments for similar code above */ - prevvpe = dvpe(); - oldasid = (read_c0_entryhi() & ASID_MASK); - if (smtc_live_asid[mytlb][oldasid]) { - smtc_live_asid[mytlb][oldasid] &= ~(0x1 << cpu); - if(smtc_live_asid[mytlb][oldasid] == 0) - smtc_flush_tlb_asid(oldasid); - } - /* See comments for similar code above */ - write_c0_entryhi((read_c0_entryhi() & ~HW_ASID_MASK) - | cpu_asid(cpu, mm)); - ehb(); /* Make sure it propagates to TCStatus */ - evpe(prevvpe); -#else /* not CONFIG_MIPS_MT_SMTC */ write_c0_entryhi(cpu_asid(cpu, mm)); -#endif /* CONFIG_MIPS_MT_SMTC */ } else { /* will get a new context next time */ -#ifndef CONFIG_MIPS_MT_SMTC cpu_context(cpu, mm) = 0; -#else /* SMTC */ - int i; - - /* SMTC shares the TLB (and ASIDs) across VPEs */ - for_each_online_cpu(i) { - if((smtc_status & SMTC_TLB_SHARED) - || (cpu_data[i].vpe_id == cpu_data[cpu].vpe_id)) - cpu_context(i, mm) = 0; - } -#endif /* CONFIG_MIPS_MT_SMTC */ } local_irq_restore(flags); } diff --git a/arch/mips/include/asm/module.h b/arch/mips/include/asm/module.h index c2edae382d5..800fe578dc9 100644 --- a/arch/mips/include/asm/module.h +++ b/arch/mips/include/asm/module.h @@ -144,13 +144,7 @@ search_module_dbetables(unsigned long addr) #define MODULE_KERNEL_TYPE "64BIT " #endif -#ifdef CONFIG_MIPS_MT_SMTC -#define MODULE_KERNEL_SMTC "MT_SMTC " -#else -#define MODULE_KERNEL_SMTC "" -#endif - #define MODULE_ARCH_VERMAGIC \ - MODULE_PROC_FAMILY MODULE_KERNEL_TYPE MODULE_KERNEL_SMTC + MODULE_PROC_FAMILY MODULE_KERNEL_TYPE #endif /* _ASM_MODULE_H */ diff --git a/arch/mips/include/asm/msa.h b/arch/mips/include/asm/msa.h index a2aba6c3ec0..538f6d482db 100644 --- a/arch/mips/include/asm/msa.h +++ b/arch/mips/include/asm/msa.h @@ -84,7 +84,7 @@ static inline void write_msa_##name(unsigned int val) \ __asm__ __volatile__( \ " .set push\n" \ " .set msa\n" \ - " cfcmsa $" #cs ", %0\n" \ + " ctcmsa $" #cs ", %0\n" \ " .set pop\n" \ : : "r"(val)); \ } @@ -96,6 +96,13 @@ static inline void write_msa_##name(unsigned int val) \ * allow compilation with toolchains that do not support MSA. Once all * toolchains in use support MSA these can be removed. */ +#ifdef CONFIG_CPU_MICROMIPS +#define CFC_MSA_INSN 0x587e0056 +#define CTC_MSA_INSN 0x583e0816 +#else +#define CFC_MSA_INSN 0x787e0059 +#define CTC_MSA_INSN 0x783e0819 +#endif #define __BUILD_MSA_CTL_REG(name, cs) \ static inline unsigned int read_msa_##name(void) \ @@ -104,7 +111,8 @@ static inline unsigned int read_msa_##name(void) \ __asm__ __volatile__( \ " .set push\n" \ " .set noat\n" \ - " .word 0x787e0059 | (" #cs " << 11)\n" \ + " .insn\n" \ + " .word #CFC_MSA_INSN | (" #cs " << 11)\n" \ " move %0, $1\n" \ " .set pop\n" \ : "=r"(reg)); \ @@ -117,7 +125,8 @@ static inline void write_msa_##name(unsigned int val) \ " .set push\n" \ " .set noat\n" \ " move $1, %0\n" \ - " .word 0x783e0819 | (" #cs " << 6)\n" \ + " .insn\n" \ + " .word #CTC_MSA_INSN | (" #cs " << 6)\n" \ " .set pop\n" \ : : "r"(val)); \ } diff --git a/arch/mips/include/asm/netlogic/mips-extns.h b/arch/mips/include/asm/netlogic/mips-extns.h index de9aada6f4c..06f1f75bfa9 100644 --- a/arch/mips/include/asm/netlogic/mips-extns.h +++ b/arch/mips/include/asm/netlogic/mips-extns.h @@ -146,9 +146,10 @@ static inline int hard_smp_processor_id(void) static inline int nlm_nodeid(void) { - uint32_t prid = read_c0_prid(); + uint32_t prid = read_c0_prid() & PRID_IMP_MASK; - if ((prid & 0xff00) == PRID_IMP_NETLOGIC_XLP9XX) + if ((prid == PRID_IMP_NETLOGIC_XLP9XX) || + (prid == PRID_IMP_NETLOGIC_XLP5XX)) return (__read_32bit_c0_register($15, 1) >> 7) & 0x7; else return (__read_32bit_c0_register($15, 1) >> 5) & 0x3; diff --git a/arch/mips/include/asm/netlogic/xlp-hal/iomap.h b/arch/mips/include/asm/netlogic/xlp-hal/iomap.h index 1f23dfaa716..805bfd21f33 100644 --- a/arch/mips/include/asm/netlogic/xlp-hal/iomap.h +++ b/arch/mips/include/asm/netlogic/xlp-hal/iomap.h @@ -74,6 +74,8 @@ #define XLP_IO_USB_OHCI2_OFFSET(node) XLP_HDR_OFFSET(node, 0, 2, 4) #define XLP_IO_USB_OHCI3_OFFSET(node) XLP_HDR_OFFSET(node, 0, 2, 5) +#define XLP_IO_SATA_OFFSET(node) XLP_HDR_OFFSET(node, 0, 3, 2) + /* XLP2xx has an updated USB block */ #define XLP2XX_IO_USB_OFFSET(node, i) XLP_HDR_OFFSET(node, 0, 4, i) #define XLP2XX_IO_USB_XHCI0_OFFSET(node) XLP_HDR_OFFSET(node, 0, 4, 1) @@ -103,13 +105,11 @@ #define XLP_IO_SYS_OFFSET(node) XLP_HDR_OFFSET(node, 0, 6, 5) #define XLP_IO_JTAG_OFFSET(node) XLP_HDR_OFFSET(node, 0, 6, 6) +/* Flash */ #define XLP_IO_NOR_OFFSET(node) XLP_HDR_OFFSET(node, 0, 7, 0) #define XLP_IO_NAND_OFFSET(node) XLP_HDR_OFFSET(node, 0, 7, 1) #define XLP_IO_SPI_OFFSET(node) XLP_HDR_OFFSET(node, 0, 7, 2) -/* SD flash */ -#define XLP_IO_SD_OFFSET(node) XLP_HDR_OFFSET(node, 0, 7, 3) -#define XLP_IO_MMC_OFFSET(node, slot) \ - ((XLP_IO_SD_OFFSET(node))+(slot*0x100)+XLP_IO_PCI_HDRSZ) +#define XLP_IO_MMC_OFFSET(node) XLP_HDR_OFFSET(node, 0, 7, 3) /* Things have changed drastically in XLP 9XX */ #define XLP9XX_HDR_OFFSET(n, d, f) \ @@ -120,6 +120,8 @@ #define XLP9XX_IO_UART_OFFSET(node) XLP9XX_HDR_OFFSET(node, 2, 2) #define XLP9XX_IO_SYS_OFFSET(node) XLP9XX_HDR_OFFSET(node, 6, 0) #define XLP9XX_IO_FUSE_OFFSET(node) XLP9XX_HDR_OFFSET(node, 6, 1) +#define XLP9XX_IO_CLOCK_OFFSET(node) XLP9XX_HDR_OFFSET(node, 6, 2) +#define XLP9XX_IO_POWER_OFFSET(node) XLP9XX_HDR_OFFSET(node, 6, 3) #define XLP9XX_IO_JTAG_OFFSET(node) XLP9XX_HDR_OFFSET(node, 6, 4) #define XLP9XX_IO_PCIE_OFFSET(node, i) XLP9XX_HDR_OFFSET(node, 1, i) @@ -135,11 +137,11 @@ /* XLP9XX on-chip SATA controller */ #define XLP9XX_IO_SATA_OFFSET(node) XLP9XX_HDR_OFFSET(node, 3, 2) +/* Flash */ #define XLP9XX_IO_NOR_OFFSET(node) XLP9XX_HDR_OFFSET(node, 7, 0) #define XLP9XX_IO_NAND_OFFSET(node) XLP9XX_HDR_OFFSET(node, 7, 1) #define XLP9XX_IO_SPI_OFFSET(node) XLP9XX_HDR_OFFSET(node, 7, 2) -/* SD flash */ -#define XLP9XX_IO_MMCSD_OFFSET(node) XLP9XX_HDR_OFFSET(node, 7, 3) +#define XLP9XX_IO_MMC_OFFSET(node) XLP9XX_HDR_OFFSET(node, 7, 3) /* PCI config header register id's */ #define XLP_PCI_CFGREG0 0x00 @@ -186,8 +188,10 @@ #define PCI_DEVICE_ID_NLM_NOR 0x1015 #define PCI_DEVICE_ID_NLM_NAND 0x1016 #define PCI_DEVICE_ID_NLM_MMC 0x1018 -#define PCI_DEVICE_ID_NLM_XHCI 0x101d +#define PCI_DEVICE_ID_NLM_SATA 0x101A +#define PCI_DEVICE_ID_NLM_XHCI 0x101D +#define PCI_DEVICE_ID_XLP9XX_MMC 0x9018 #define PCI_DEVICE_ID_XLP9XX_SATA 0x901A #define PCI_DEVICE_ID_XLP9XX_XHCI 0x901D diff --git a/arch/mips/include/asm/netlogic/xlp-hal/pcibus.h b/arch/mips/include/asm/netlogic/xlp-hal/pcibus.h index d4deb87ad06..91540f41e1e 100644 --- a/arch/mips/include/asm/netlogic/xlp-hal/pcibus.h +++ b/arch/mips/include/asm/netlogic/xlp-hal/pcibus.h @@ -69,6 +69,20 @@ #define PCIE_9XX_BYTE_SWAP_IO_BASE 0x25e #define PCIE_9XX_BYTE_SWAP_IO_LIM 0x25f +#define PCIE_9XX_BRIDGE_MSIX_ADDR_BASE 0x264 +#define PCIE_9XX_BRIDGE_MSIX_ADDR_LIMIT 0x265 +#define PCIE_9XX_MSI_STATUS 0x283 +#define PCIE_9XX_MSI_EN 0x284 +/* 128 MSIX vectors available in 9xx */ +#define PCIE_9XX_MSIX_STATUS0 0x286 +#define PCIE_9XX_MSIX_STATUSX(n) (n + 0x286) +#define PCIE_9XX_MSIX_VEC 0x296 +#define PCIE_9XX_MSIX_VECX(n) (n + 0x296) +#define PCIE_9XX_INT_STATUS0 0x397 +#define PCIE_9XX_INT_STATUS1 0x398 +#define PCIE_9XX_INT_EN0 0x399 +#define PCIE_9XX_INT_EN1 0x39a + /* other */ #define PCIE_NLINKS 4 diff --git a/arch/mips/include/asm/netlogic/xlp-hal/pic.h b/arch/mips/include/asm/netlogic/xlp-hal/pic.h index f10bf3bba58..41cefe94f0c 100644 --- a/arch/mips/include/asm/netlogic/xlp-hal/pic.h +++ b/arch/mips/include/asm/netlogic/xlp-hal/pic.h @@ -199,6 +199,10 @@ #define PIC_IRT_PCIE_LINK_3_INDEX 81 #define PIC_IRT_PCIE_LINK_INDEX(num) ((num) + PIC_IRT_PCIE_LINK_0_INDEX) +#define PIC_9XX_IRT_PCIE_LINK_0_INDEX 191 +#define PIC_9XX_IRT_PCIE_LINK_INDEX(num) \ + ((num) + PIC_9XX_IRT_PCIE_LINK_0_INDEX) + #define PIC_CLOCK_TIMER 7 #if !defined(LOCORE) && !defined(__ASSEMBLY__) diff --git a/arch/mips/include/asm/netlogic/xlp-hal/sys.h b/arch/mips/include/asm/netlogic/xlp-hal/sys.h index d9b107ffca9..bc7bddf25be 100644 --- a/arch/mips/include/asm/netlogic/xlp-hal/sys.h +++ b/arch/mips/include/asm/netlogic/xlp-hal/sys.h @@ -118,6 +118,10 @@ #define SYS_SCRTCH3 0x4c /* PLL registers XLP2XX */ +#define SYS_CPU_PLL_CTRL0(core) (0x1c0 + (core * 4)) +#define SYS_CPU_PLL_CTRL1(core) (0x1c1 + (core * 4)) +#define SYS_CPU_PLL_CTRL2(core) (0x1c2 + (core * 4)) +#define SYS_CPU_PLL_CTRL3(core) (0x1c3 + (core * 4)) #define SYS_PLL_CTRL0 0x240 #define SYS_PLL_CTRL1 0x241 #define SYS_PLL_CTRL2 0x242 @@ -147,6 +151,32 @@ #define SYS_SYS_PLL_MEM_REQ 0x2a3 #define SYS_PLL_MEM_STAT 0x2a4 +/* PLL registers XLP9XX */ +#define SYS_9XX_CPU_PLL_CTRL0(core) (0xc0 + (core * 4)) +#define SYS_9XX_CPU_PLL_CTRL1(core) (0xc1 + (core * 4)) +#define SYS_9XX_CPU_PLL_CTRL2(core) (0xc2 + (core * 4)) +#define SYS_9XX_CPU_PLL_CTRL3(core) (0xc3 + (core * 4)) +#define SYS_9XX_DMC_PLL_CTRL0 0x140 +#define SYS_9XX_DMC_PLL_CTRL1 0x141 +#define SYS_9XX_DMC_PLL_CTRL2 0x142 +#define SYS_9XX_DMC_PLL_CTRL3 0x143 +#define SYS_9XX_PLL_CTRL0 0x144 +#define SYS_9XX_PLL_CTRL1 0x145 +#define SYS_9XX_PLL_CTRL2 0x146 +#define SYS_9XX_PLL_CTRL3 0x147 + +#define SYS_9XX_PLL_CTRL0_DEVX(x) (0x148 + (x) * 4) +#define SYS_9XX_PLL_CTRL1_DEVX(x) (0x149 + (x) * 4) +#define SYS_9XX_PLL_CTRL2_DEVX(x) (0x14a + (x) * 4) +#define SYS_9XX_PLL_CTRL3_DEVX(x) (0x14b + (x) * 4) + +#define SYS_9XX_CPU_PLL_CHG_CTRL 0x188 +#define SYS_9XX_PLL_CHG_CTRL 0x189 +#define SYS_9XX_CLK_DEV_DIS 0x18a +#define SYS_9XX_CLK_DEV_SEL 0x18b +#define SYS_9XX_CLK_DEV_DIV 0x18d +#define SYS_9XX_CLK_DEV_CHG 0x18f + /* Registers changed on 9XX */ #define SYS_9XX_POWER_ON_RESET_CFG 0x00 #define SYS_9XX_CHIP_RESET 0x01 @@ -170,6 +200,11 @@ #define nlm_get_fuse_regbase(node) \ (nlm_get_fuse_pcibase(node) + XLP_IO_PCI_HDRSZ) +#define nlm_get_clock_pcibase(node) \ + nlm_pcicfg_base(XLP9XX_IO_CLOCK_OFFSET(node)) +#define nlm_get_clock_regbase(node) \ + (nlm_get_clock_pcibase(node) + XLP_IO_PCI_HDRSZ) + unsigned int nlm_get_pic_frequency(int node); #endif #endif diff --git a/arch/mips/include/asm/netlogic/xlp-hal/xlp.h b/arch/mips/include/asm/netlogic/xlp-hal/xlp.h index 2b0c9599ebe..a862b93223c 100644 --- a/arch/mips/include/asm/netlogic/xlp-hal/xlp.h +++ b/arch/mips/include/asm/netlogic/xlp-hal/xlp.h @@ -58,6 +58,10 @@ #define PIC_I2C_1_IRQ 31 #define PIC_I2C_2_IRQ 32 #define PIC_I2C_3_IRQ 33 +#define PIC_SPI_IRQ 34 +#define PIC_NAND_IRQ 37 +#define PIC_SATA_IRQ 38 +#define PIC_GPIO_IRQ 39 #define PIC_PCIE_LINK_MSI_IRQ_BASE 44 /* 44 - 47 MSI IRQ */ #define PIC_PCIE_LINK_MSI_IRQ(i) (44 + (i)) @@ -66,8 +70,9 @@ #define PIC_PCIE_MSIX_IRQ_BASE 48 /* 48 - 51 MSI-X IRQ */ #define PIC_PCIE_MSIX_IRQ(i) (48 + (i)) -#define NLM_MSIX_VEC_BASE 96 /* 96 - 127 - MSIX mapped */ -#define NLM_MSI_VEC_BASE 128 /* 128 -255 - MSI mapped */ +/* XLP9xx and XLP8xx has 128 and 32 MSIX vectors respectively */ +#define NLM_MSIX_VEC_BASE 96 /* 96 - 223 - MSIX mapped */ +#define NLM_MSI_VEC_BASE 224 /* 224 -351 - MSI mapped */ #define NLM_PIC_INDIRECT_VEC_BASE 512 #define NLM_GPIO_VEC_BASE 768 @@ -95,17 +100,19 @@ void *xlp_dt_init(void *fdtp); static inline int cpu_is_xlpii(void) { - int chip = read_c0_prid() & 0xff00; + int chip = read_c0_prid() & PRID_IMP_MASK; return chip == PRID_IMP_NETLOGIC_XLP2XX || - chip == PRID_IMP_NETLOGIC_XLP9XX; + chip == PRID_IMP_NETLOGIC_XLP9XX || + chip == PRID_IMP_NETLOGIC_XLP5XX; } static inline int cpu_is_xlp9xx(void) { - int chip = read_c0_prid() & 0xff00; + int chip = read_c0_prid() & PRID_IMP_MASK; - return chip == PRID_IMP_NETLOGIC_XLP9XX; + return chip == PRID_IMP_NETLOGIC_XLP9XX || + chip == PRID_IMP_NETLOGIC_XLP5XX; } #endif /* !__ASSEMBLY__ */ #endif /* _ASM_NLM_XLP_H */ diff --git a/arch/mips/include/asm/nile4.h b/arch/mips/include/asm/nile4.h index 2e2436d0e94..99e97f8bfbc 100644 --- a/arch/mips/include/asm/nile4.h +++ b/arch/mips/include/asm/nile4.h @@ -1,7 +1,7 @@ /* * asm-mips/nile4.h -- NEC Vrc-5074 Nile 4 definitions * - * Copyright (C) 2000 Geert Uytterhoeven <geert@sonycom.com> + * Copyright (C) 2000 Geert Uytterhoeven <geert@linux-m68k.org> * Sony Software Development Center Europe (SDCE), Brussels * * This file is based on the following documentation: diff --git a/arch/mips/include/asm/octeon/octeon.h b/arch/mips/include/asm/octeon/octeon.h index f5d77b91537..d781f9e6688 100644 --- a/arch/mips/include/asm/octeon/octeon.h +++ b/arch/mips/include/asm/octeon/octeon.h @@ -211,7 +211,6 @@ union octeon_cvmemctl { extern void octeon_write_lcd(const char *s); extern void octeon_check_cpu_bist(void); -extern int octeon_get_boot_debug_flag(void); extern int octeon_get_boot_uart(void); struct uart_port; diff --git a/arch/mips/include/asm/pgtable.h b/arch/mips/include/asm/pgtable.h index 008324d1c26..539ddd148bb 100644 --- a/arch/mips/include/asm/pgtable.h +++ b/arch/mips/include/asm/pgtable.h @@ -32,6 +32,8 @@ struct vm_area_struct; _page_cachable_default) #define PAGE_KERNEL __pgprot(_PAGE_PRESENT | __READABLE | __WRITEABLE | \ _PAGE_GLOBAL | _page_cachable_default) +#define PAGE_KERNEL_NC __pgprot(_PAGE_PRESENT | __READABLE | __WRITEABLE | \ + _PAGE_GLOBAL | _CACHE_CACHABLE_NONCOHERENT) #define PAGE_USERIO __pgprot(_PAGE_PRESENT | (cpu_has_rixi ? 0 : _PAGE_READ) | _PAGE_WRITE | \ _page_cachable_default) #define PAGE_KERNEL_UNCACHED __pgprot(_PAGE_PRESENT | __READABLE | \ diff --git a/arch/mips/include/asm/pm-cps.h b/arch/mips/include/asm/pm-cps.h new file mode 100644 index 00000000000..625eda53d57 --- /dev/null +++ b/arch/mips/include/asm/pm-cps.h @@ -0,0 +1,51 @@ +/* + * Copyright (C) 2014 Imagination Technologies + * Author: Paul Burton <paul.burton@imgtec.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + */ + +#ifndef __MIPS_ASM_PM_CPS_H__ +#define __MIPS_ASM_PM_CPS_H__ + +/* + * The CM & CPC can only handle coherence & power control on a per-core basis, + * thus in an MT system the VPEs within each core are coupled and can only + * enter or exit states requiring CM or CPC assistance in unison. + */ +#ifdef CONFIG_MIPS_MT +# define coupled_coherence cpu_has_mipsmt +#else +# define coupled_coherence 0 +#endif + +/* Enumeration of possible PM states */ +enum cps_pm_state { + CPS_PM_NC_WAIT, /* MIPS wait instruction, non-coherent */ + CPS_PM_CLOCK_GATED, /* Core clock gated */ + CPS_PM_POWER_GATED, /* Core power gated */ + CPS_PM_STATE_COUNT, +}; + +/** + * cps_pm_support_state - determine whether the system supports a PM state + * @state: the state to test for support + * + * Returns true if the system supports the given state, otherwise false. + */ +extern bool cps_pm_support_state(enum cps_pm_state state); + +/** + * cps_pm_enter_state - enter a PM state + * @state: the state to enter + * + * Enter the given PM state. If coupled_coherence is non-zero then it is + * expected that this function be called at approximately the same time on + * each coupled CPU. Returns 0 on successful entry & exit, otherwise -errno. + */ +extern int cps_pm_enter_state(enum cps_pm_state state); + +#endif /* __MIPS_ASM_PM_CPS_H__ */ diff --git a/arch/mips/include/asm/pm.h b/arch/mips/include/asm/pm.h new file mode 100644 index 00000000000..7c03469e043 --- /dev/null +++ b/arch/mips/include/asm/pm.h @@ -0,0 +1,159 @@ +/* + * Copyright (C) 2014 Imagination Technologies Ltd + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * PM helper macros for CPU power off (e.g. Suspend-to-RAM). + */ + +#ifndef __ASM_PM_H +#define __ASM_PM_H + +#ifdef __ASSEMBLY__ + +#include <asm/asm-offsets.h> +#include <asm/asm.h> +#include <asm/mipsregs.h> +#include <asm/regdef.h> + +/* Save CPU state to stack for suspend to RAM */ +.macro SUSPEND_SAVE_REGS + subu sp, PT_SIZE + /* Call preserved GPRs */ + LONG_S $16, PT_R16(sp) + LONG_S $17, PT_R17(sp) + LONG_S $18, PT_R18(sp) + LONG_S $19, PT_R19(sp) + LONG_S $20, PT_R20(sp) + LONG_S $21, PT_R21(sp) + LONG_S $22, PT_R22(sp) + LONG_S $23, PT_R23(sp) + LONG_S $28, PT_R28(sp) + LONG_S $30, PT_R30(sp) + LONG_S $31, PT_R31(sp) + /* A couple of CP0 registers with space in pt_regs */ + mfc0 k0, CP0_STATUS + LONG_S k0, PT_STATUS(sp) +.endm + +/* Restore CPU state from stack after resume from RAM */ +.macro RESUME_RESTORE_REGS_RETURN + .set push + .set noreorder + /* A couple of CP0 registers with space in pt_regs */ + LONG_L k0, PT_STATUS(sp) + mtc0 k0, CP0_STATUS + /* Call preserved GPRs */ + LONG_L $16, PT_R16(sp) + LONG_L $17, PT_R17(sp) + LONG_L $18, PT_R18(sp) + LONG_L $19, PT_R19(sp) + LONG_L $20, PT_R20(sp) + LONG_L $21, PT_R21(sp) + LONG_L $22, PT_R22(sp) + LONG_L $23, PT_R23(sp) + LONG_L $28, PT_R28(sp) + LONG_L $30, PT_R30(sp) + LONG_L $31, PT_R31(sp) + /* Pop and return */ + jr ra + addiu sp, PT_SIZE + .set pop +.endm + +/* Get address of static suspend state into t1 */ +.macro LA_STATIC_SUSPEND + la t1, mips_static_suspend_state +.endm + +/* Save important CPU state for early restoration to global data */ +.macro SUSPEND_SAVE_STATIC +#ifdef CONFIG_EVA + /* + * Segment configuration is saved in global data where it can be easily + * reloaded without depending on the segment configuration. + */ + mfc0 k0, CP0_PAGEMASK, 2 /* SegCtl0 */ + LONG_S k0, SSS_SEGCTL0(t1) + mfc0 k0, CP0_PAGEMASK, 3 /* SegCtl1 */ + LONG_S k0, SSS_SEGCTL1(t1) + mfc0 k0, CP0_PAGEMASK, 4 /* SegCtl2 */ + LONG_S k0, SSS_SEGCTL2(t1) +#endif + /* save stack pointer (pointing to GPRs) */ + LONG_S sp, SSS_SP(t1) +.endm + +/* Restore important CPU state early from global data */ +.macro RESUME_RESTORE_STATIC +#ifdef CONFIG_EVA + /* + * Segment configuration must be restored prior to any access to + * allocated memory, as it may reside outside of the legacy kernel + * segments. + */ + LONG_L k0, SSS_SEGCTL0(t1) + mtc0 k0, CP0_PAGEMASK, 2 /* SegCtl0 */ + LONG_L k0, SSS_SEGCTL1(t1) + mtc0 k0, CP0_PAGEMASK, 3 /* SegCtl1 */ + LONG_L k0, SSS_SEGCTL2(t1) + mtc0 k0, CP0_PAGEMASK, 4 /* SegCtl2 */ + tlbw_use_hazard +#endif + /* restore stack pointer (pointing to GPRs) */ + LONG_L sp, SSS_SP(t1) +.endm + +/* flush caches to make sure context has reached memory */ +.macro SUSPEND_CACHE_FLUSH + .extern __wback_cache_all + .set push + .set noreorder + la t1, __wback_cache_all + LONG_L t0, 0(t1) + jalr t0 + nop + .set pop + .endm + +/* Save suspend state and flush data caches to RAM */ +.macro SUSPEND_SAVE + SUSPEND_SAVE_REGS + LA_STATIC_SUSPEND + SUSPEND_SAVE_STATIC + SUSPEND_CACHE_FLUSH +.endm + +/* Restore saved state after resume from RAM and return */ +.macro RESUME_RESTORE_RETURN + LA_STATIC_SUSPEND + RESUME_RESTORE_STATIC + RESUME_RESTORE_REGS_RETURN +.endm + +#else /* __ASSEMBLY__ */ + +/** + * struct mips_static_suspend_state - Core saved CPU state across S2R. + * @segctl: CP0 Segment control registers. + * @sp: Stack frame where GP register context is saved. + * + * This structure contains minimal CPU state that must be saved in static kernel + * data in order to be able to restore the rest of the state. This includes + * segmentation configuration in the case of EVA being enabled, as they must be + * restored prior to any kmalloc'd memory being referenced (even the stack + * pointer). + */ +struct mips_static_suspend_state { +#ifdef CONFIG_EVA + unsigned long segctl[3]; +#endif + unsigned long sp; +}; + +#endif /* !__ASSEMBLY__ */ + +#endif /* __ASM_PM_HELPERS_H */ diff --git a/arch/mips/include/asm/ptrace.h b/arch/mips/include/asm/ptrace.h index bf1ac8d3578..7e6e682aece 100644 --- a/arch/mips/include/asm/ptrace.h +++ b/arch/mips/include/asm/ptrace.h @@ -39,9 +39,6 @@ struct pt_regs { unsigned long cp0_badvaddr; unsigned long cp0_cause; unsigned long cp0_epc; -#ifdef CONFIG_MIPS_MT_SMTC - unsigned long cp0_tcstatus; -#endif /* CONFIG_MIPS_MT_SMTC */ #ifdef CONFIG_CPU_CAVIUM_OCTEON unsigned long long mpl[3]; /* MTM{0,1,2} */ unsigned long long mtp[3]; /* MTP{0,1,2} */ diff --git a/arch/mips/include/asm/r4kcache.h b/arch/mips/include/asm/r4kcache.h index ca64cbe4449..0b8bd28a0df 100644 --- a/arch/mips/include/asm/r4kcache.h +++ b/arch/mips/include/asm/r4kcache.h @@ -43,11 +43,10 @@ : "i" (op), "R" (*(unsigned char *)(addr))) #ifdef CONFIG_MIPS_MT + /* - * Temporary hacks for SMTC debug. Optionally force single-threaded - * execution during I-cache flushes. + * Optionally force single-threaded execution during I-cache flushes. */ - #define PROTECT_CACHE_FLUSHES 1 #ifdef PROTECT_CACHE_FLUSHES @@ -524,6 +523,8 @@ __BUILD_BLAST_CACHE(s, scache, Index_Writeback_Inv_SD, Hit_Writeback_Inv_SD, 32, __BUILD_BLAST_CACHE(d, dcache, Index_Writeback_Inv_D, Hit_Writeback_Inv_D, 64, ) __BUILD_BLAST_CACHE(i, icache, Index_Invalidate_I, Hit_Invalidate_I, 64, ) __BUILD_BLAST_CACHE(s, scache, Index_Writeback_Inv_SD, Hit_Writeback_Inv_SD, 64, ) +__BUILD_BLAST_CACHE(d, dcache, Index_Writeback_Inv_D, Hit_Writeback_Inv_D, 128, ) +__BUILD_BLAST_CACHE(i, icache, Index_Invalidate_I, Hit_Invalidate_I, 128, ) __BUILD_BLAST_CACHE(s, scache, Index_Writeback_Inv_SD, Hit_Writeback_Inv_SD, 128, ) __BUILD_BLAST_CACHE(inv_d, dcache, Index_Writeback_Inv_D, Hit_Invalidate_D, 16, ) diff --git a/arch/mips/include/asm/sgi/ip22.h b/arch/mips/include/asm/sgi/ip22.h index 8db1a3588cf..87ec9eaa04e 100644 --- a/arch/mips/include/asm/sgi/ip22.h +++ b/arch/mips/include/asm/sgi/ip22.h @@ -69,6 +69,8 @@ #define SGI_EISA_IRQ SGINT_LOCAL2 + 3 /* EISA interrupts */ #define SGI_KEYBD_IRQ SGINT_LOCAL2 + 4 /* keyboard */ #define SGI_SERIAL_IRQ SGINT_LOCAL2 + 5 /* onboard serial */ +#define SGI_GIOEXP0_IRQ (SGINT_LOCAL2 + 6) /* Indy GIO EXP0 */ +#define SGI_GIOEXP1_IRQ (SGINT_LOCAL2 + 7) /* Indy GIO EXP1 */ #define ip22_is_fullhouse() (sgioc->sysid & SGIOC_SYSID_FULLHOUSE) diff --git a/arch/mips/include/asm/smp-cps.h b/arch/mips/include/asm/smp-cps.h index d60d1a2180d..a06a08a9afc 100644 --- a/arch/mips/include/asm/smp-cps.h +++ b/arch/mips/include/asm/smp-cps.h @@ -13,17 +13,28 @@ #ifndef __ASSEMBLY__ -struct boot_config { - unsigned int core; - unsigned int vpe; +struct vpe_boot_config { unsigned long pc; unsigned long sp; unsigned long gp; }; -extern struct boot_config mips_cps_bootcfg; +struct core_boot_config { + atomic_t vpe_mask; + struct vpe_boot_config *vpe_config; +}; + +extern struct core_boot_config *mips_cps_core_bootcfg; extern void mips_cps_core_entry(void); +extern void mips_cps_core_init(void); + +extern struct vpe_boot_config *mips_cps_boot_vpes(void); + +extern bool mips_cps_smp_in_use(void); + +extern void mips_cps_pm_save(void); +extern void mips_cps_pm_restore(void); #else /* __ASSEMBLY__ */ diff --git a/arch/mips/include/asm/smp-ops.h b/arch/mips/include/asm/smp-ops.h index 73d35b18fb6..6ba1fb8b11e 100644 --- a/arch/mips/include/asm/smp-ops.h +++ b/arch/mips/include/asm/smp-ops.h @@ -26,7 +26,6 @@ struct plat_smp_ops { void (*send_ipi_mask)(const struct cpumask *mask, unsigned int action); void (*init_secondary)(void); void (*smp_finish)(void); - void (*cpus_done)(void); void (*boot_secondary)(int cpu, struct task_struct *idle); void (*smp_setup)(void); void (*prepare_cpus)(unsigned int max_cpus); diff --git a/arch/mips/include/asm/smp.h b/arch/mips/include/asm/smp.h index efa02acd3dd..b037334fca2 100644 --- a/arch/mips/include/asm/smp.h +++ b/arch/mips/include/asm/smp.h @@ -46,6 +46,9 @@ extern int __cpu_logical_map[NR_CPUS]; extern volatile cpumask_t cpu_callin_map; +/* Mask of CPUs which are currently definitely operating coherently */ +extern cpumask_t cpu_coherent_mask; + extern void asmlinkage smp_bootstrap(void); /* diff --git a/arch/mips/include/asm/smtc.h b/arch/mips/include/asm/smtc.h deleted file mode 100644 index e56b439b787..00000000000 --- a/arch/mips/include/asm/smtc.h +++ /dev/null @@ -1,78 +0,0 @@ -#ifndef _ASM_SMTC_MT_H -#define _ASM_SMTC_MT_H - -/* - * Definitions for SMTC multitasking on MIPS MT cores - */ - -#include <asm/mips_mt.h> -#include <asm/smtc_ipi.h> - -/* - * System-wide SMTC status information - */ - -extern unsigned int smtc_status; - -#define SMTC_TLB_SHARED 0x00000001 -#define SMTC_MTC_ACTIVE 0x00000002 - -/* - * TLB/ASID Management information - */ - -#define MAX_SMTC_TLBS 2 -#define MAX_SMTC_ASIDS 256 -#if NR_CPUS <= 8 -typedef char asiduse; -#else -#if NR_CPUS <= 16 -typedef short asiduse; -#else -typedef long asiduse; -#endif -#endif - -/* - * VPE Management information - */ - -#define MAX_SMTC_VPES MAX_SMTC_TLBS /* FIXME: May not always be true. */ - -extern asiduse smtc_live_asid[MAX_SMTC_TLBS][MAX_SMTC_ASIDS]; - -struct mm_struct; -struct task_struct; - -void smtc_get_new_mmu_context(struct mm_struct *mm, unsigned long cpu); -void self_ipi(struct smtc_ipi *); -void smtc_flush_tlb_asid(unsigned long asid); -extern int smtc_build_cpu_map(int startslot); -extern void smtc_prepare_cpus(int cpus); -extern void smtc_smp_finish(void); -extern void smtc_boot_secondary(int cpu, struct task_struct *t); -extern void smtc_cpus_done(void); -extern void smtc_init_secondary(void); - - -/* - * Sharing the TLB between multiple VPEs means that the - * "random" index selection function is not allowed to - * select the current value of the Index register. To - * avoid additional TLB pressure, the Index registers - * are "parked" with an non-Valid value. - */ - -#define PARKED_INDEX ((unsigned int)0x80000000) - -/* - * Define low-level interrupt mask for IPIs, if necessary. - * By default, use SW interrupt 1, which requires no external - * hardware support, but which works only for single-core - * MIPS MT systems. - */ -#ifndef MIPS_CPU_IPI_IRQ -#define MIPS_CPU_IPI_IRQ 1 -#endif - -#endif /* _ASM_SMTC_MT_H */ diff --git a/arch/mips/include/asm/smtc_ipi.h b/arch/mips/include/asm/smtc_ipi.h deleted file mode 100644 index 15278dbd7e7..00000000000 --- a/arch/mips/include/asm/smtc_ipi.h +++ /dev/null @@ -1,129 +0,0 @@ -/* - * Definitions used in MIPS MT SMTC "Interprocessor Interrupt" code. - */ -#ifndef __ASM_SMTC_IPI_H -#define __ASM_SMTC_IPI_H - -#include <linux/spinlock.h> - -//#define SMTC_IPI_DEBUG - -#ifdef SMTC_IPI_DEBUG -#include <asm/mipsregs.h> -#include <asm/mipsmtregs.h> -#endif /* SMTC_IPI_DEBUG */ - -/* - * An IPI "message" - */ - -struct smtc_ipi { - struct smtc_ipi *flink; - int type; - void *arg; - int dest; -#ifdef SMTC_IPI_DEBUG - int sender; - long stamp; -#endif /* SMTC_IPI_DEBUG */ -}; - -/* - * Defined IPI Types - */ - -#define LINUX_SMP_IPI 1 -#define SMTC_CLOCK_TICK 2 -#define IRQ_AFFINITY_IPI 3 - -/* - * A queue of IPI messages - */ - -struct smtc_ipi_q { - struct smtc_ipi *head; - spinlock_t lock; - struct smtc_ipi *tail; - int depth; - int resched_flag; /* reschedule already queued */ -}; - -static inline void smtc_ipi_nq(struct smtc_ipi_q *q, struct smtc_ipi *p) -{ - unsigned long flags; - - spin_lock_irqsave(&q->lock, flags); - if (q->head == NULL) - q->head = q->tail = p; - else - q->tail->flink = p; - p->flink = NULL; - q->tail = p; - q->depth++; -#ifdef SMTC_IPI_DEBUG - p->sender = read_c0_tcbind(); - p->stamp = read_c0_count(); -#endif /* SMTC_IPI_DEBUG */ - spin_unlock_irqrestore(&q->lock, flags); -} - -static inline struct smtc_ipi *__smtc_ipi_dq(struct smtc_ipi_q *q) -{ - struct smtc_ipi *p; - - if (q->head == NULL) - p = NULL; - else { - p = q->head; - q->head = q->head->flink; - q->depth--; - /* Arguably unnecessary, but leaves queue cleaner */ - if (q->head == NULL) - q->tail = NULL; - } - - return p; -} - -static inline struct smtc_ipi *smtc_ipi_dq(struct smtc_ipi_q *q) -{ - unsigned long flags; - struct smtc_ipi *p; - - spin_lock_irqsave(&q->lock, flags); - p = __smtc_ipi_dq(q); - spin_unlock_irqrestore(&q->lock, flags); - - return p; -} - -static inline void smtc_ipi_req(struct smtc_ipi_q *q, struct smtc_ipi *p) -{ - unsigned long flags; - - spin_lock_irqsave(&q->lock, flags); - if (q->head == NULL) { - q->head = q->tail = p; - p->flink = NULL; - } else { - p->flink = q->head; - q->head = p; - } - q->depth++; - spin_unlock_irqrestore(&q->lock, flags); -} - -static inline int smtc_ipi_qdepth(struct smtc_ipi_q *q) -{ - unsigned long flags; - int retval; - - spin_lock_irqsave(&q->lock, flags); - retval = q->depth; - spin_unlock_irqrestore(&q->lock, flags); - return retval; -} - -extern void smtc_send_ipi(int cpu, int type, unsigned int action); - -#endif /* __ASM_SMTC_IPI_H */ diff --git a/arch/mips/include/asm/smtc_proc.h b/arch/mips/include/asm/smtc_proc.h deleted file mode 100644 index 25da651f1f5..00000000000 --- a/arch/mips/include/asm/smtc_proc.h +++ /dev/null @@ -1,23 +0,0 @@ -/* - * Definitions for SMTC /proc entries - * Copyright(C) 2005 MIPS Technologies Inc. - */ -#ifndef __ASM_SMTC_PROC_H -#define __ASM_SMTC_PROC_H - -/* - * per-"CPU" statistics - */ - -struct smtc_cpu_proc { - unsigned long timerints; - unsigned long selfipis; -}; - -extern struct smtc_cpu_proc smtc_cpu_stats[NR_CPUS]; - -/* Count of number of recoveries of "stolen" FPU access rights on 34K */ - -extern atomic_t smtc_fpu_recoveries; - -#endif /* __ASM_SMTC_PROC_H */ diff --git a/arch/mips/include/asm/stackframe.h b/arch/mips/include/asm/stackframe.h index d301e108d5b..b188c797565 100644 --- a/arch/mips/include/asm/stackframe.h +++ b/arch/mips/include/asm/stackframe.h @@ -19,22 +19,12 @@ #include <asm/asm-offsets.h> #include <asm/thread_info.h> -/* - * For SMTC kernel, global IE should be left set, and interrupts - * controlled exclusively via IXMT. - */ -#ifdef CONFIG_MIPS_MT_SMTC -#define STATMASK 0x1e -#elif defined(CONFIG_CPU_R3000) || defined(CONFIG_CPU_TX39XX) +#if defined(CONFIG_CPU_R3000) || defined(CONFIG_CPU_TX39XX) #define STATMASK 0x3f #else #define STATMASK 0x1f #endif -#ifdef CONFIG_MIPS_MT_SMTC -#include <asm/mipsmtregs.h> -#endif /* CONFIG_MIPS_MT_SMTC */ - .macro SAVE_AT .set push .set noat @@ -186,16 +176,6 @@ mfc0 v1, CP0_STATUS LONG_S $2, PT_R2(sp) LONG_S v1, PT_STATUS(sp) -#ifdef CONFIG_MIPS_MT_SMTC - /* - * Ideally, these instructions would be shuffled in - * to cover the pipeline delay. - */ - .set mips32 - mfc0 k0, CP0_TCSTATUS - .set mips0 - LONG_S k0, PT_TCSTATUS(sp) -#endif /* CONFIG_MIPS_MT_SMTC */ LONG_S $4, PT_R4(sp) mfc0 v1, CP0_CAUSE LONG_S $5, PT_R5(sp) @@ -321,36 +301,6 @@ .set push .set reorder .set noat -#ifdef CONFIG_MIPS_MT_SMTC - .set mips32r2 - /* - * We need to make sure the read-modify-write - * of Status below isn't perturbed by an interrupt - * or cross-TC access, so we need to do at least a DMT, - * protected by an interrupt-inhibit. But setting IXMT - * also creates a few-cycle window where an IPI could - * be queued and not be detected before potentially - * returning to a WAIT or user-mode loop. It must be - * replayed. - * - * We're in the middle of a context switch, and - * we can't dispatch it directly without trashing - * some registers, so we'll try to detect this unlikely - * case and program a software interrupt in the VPE, - * as would be done for a cross-VPE IPI. To accommodate - * the handling of that case, we're doing a DVPE instead - * of just a DMT here to protect against other threads. - * This is a lot of cruft to cover a tiny window. - * If you can find a better design, implement it! - * - */ - mfc0 v0, CP0_TCSTATUS - ori v0, TCSTATUS_IXMT - mtc0 v0, CP0_TCSTATUS - _ehb - DVPE 5 # dvpe a1 - jal mips_ihb -#endif /* CONFIG_MIPS_MT_SMTC */ mfc0 a0, CP0_STATUS ori a0, STATMASK xori a0, STATMASK @@ -362,59 +312,6 @@ and v0, v1 or v0, a0 mtc0 v0, CP0_STATUS -#ifdef CONFIG_MIPS_MT_SMTC -/* - * Only after EXL/ERL have been restored to status can we - * restore TCStatus.IXMT. - */ - LONG_L v1, PT_TCSTATUS(sp) - _ehb - mfc0 a0, CP0_TCSTATUS - andi v1, TCSTATUS_IXMT - bnez v1, 0f - -/* - * We'd like to detect any IPIs queued in the tiny window - * above and request an software interrupt to service them - * when we ERET. - * - * Computing the offset into the IPIQ array of the executing - * TC's IPI queue in-line would be tedious. We use part of - * the TCContext register to hold 16 bits of offset that we - * can add in-line to find the queue head. - */ - mfc0 v0, CP0_TCCONTEXT - la a2, IPIQ - srl v0, v0, 16 - addu a2, a2, v0 - LONG_L v0, 0(a2) - beqz v0, 0f -/* - * If we have a queue, provoke dispatch within the VPE by setting C_SW1 - */ - mfc0 v0, CP0_CAUSE - ori v0, v0, C_SW1 - mtc0 v0, CP0_CAUSE -0: - /* - * This test should really never branch but - * let's be prudent here. Having atomized - * the shared register modifications, we can - * now EVPE, and must do so before interrupts - * are potentially re-enabled. - */ - andi a1, a1, MVPCONTROL_EVP - beqz a1, 1f - evpe -1: - /* We know that TCStatua.IXMT should be set from above */ - xori a0, a0, TCSTATUS_IXMT - or a0, a0, v1 - mtc0 a0, CP0_TCSTATUS - _ehb - - .set mips0 -#endif /* CONFIG_MIPS_MT_SMTC */ LONG_L v1, PT_EPC(sp) MTC0 v1, CP0_EPC LONG_L $31, PT_R31(sp) @@ -467,33 +364,11 @@ * Set cp0 enable bit as sign that we're running on the kernel stack */ .macro CLI -#if !defined(CONFIG_MIPS_MT_SMTC) mfc0 t0, CP0_STATUS li t1, ST0_CU0 | STATMASK or t0, t1 xori t0, STATMASK mtc0 t0, CP0_STATUS -#else /* CONFIG_MIPS_MT_SMTC */ - /* - * For SMTC, we need to set privilege - * and disable interrupts only for the - * current TC, using the TCStatus register. - */ - mfc0 t0, CP0_TCSTATUS - /* Fortunately CU 0 is in the same place in both registers */ - /* Set TCU0, TMX, TKSU (for later inversion) and IXMT */ - li t1, ST0_CU0 | 0x08001c00 - or t0, t1 - /* Clear TKSU, leave IXMT */ - xori t0, 0x00001800 - mtc0 t0, CP0_TCSTATUS - _ehb - /* We need to leave the global IE bit set, but clear EXL...*/ - mfc0 t0, CP0_STATUS - ori t0, ST0_EXL | ST0_ERL - xori t0, ST0_EXL | ST0_ERL - mtc0 t0, CP0_STATUS -#endif /* CONFIG_MIPS_MT_SMTC */ irq_disable_hazard .endm @@ -502,35 +377,11 @@ * Set cp0 enable bit as sign that we're running on the kernel stack */ .macro STI -#if !defined(CONFIG_MIPS_MT_SMTC) mfc0 t0, CP0_STATUS li t1, ST0_CU0 | STATMASK or t0, t1 xori t0, STATMASK & ~1 mtc0 t0, CP0_STATUS -#else /* CONFIG_MIPS_MT_SMTC */ - /* - * For SMTC, we need to set privilege - * and enable interrupts only for the - * current TC, using the TCStatus register. - */ - _ehb - mfc0 t0, CP0_TCSTATUS - /* Fortunately CU 0 is in the same place in both registers */ - /* Set TCU0, TKSU (for later inversion) and IXMT */ - li t1, ST0_CU0 | 0x08001c00 - or t0, t1 - /* Clear TKSU *and* IXMT */ - xori t0, 0x00001c00 - mtc0 t0, CP0_TCSTATUS - _ehb - /* We need to leave the global IE bit set, but clear EXL...*/ - mfc0 t0, CP0_STATUS - ori t0, ST0_EXL - xori t0, ST0_EXL - mtc0 t0, CP0_STATUS - /* irq_enable_hazard below should expand to EHB for 24K/34K cpus */ -#endif /* CONFIG_MIPS_MT_SMTC */ irq_enable_hazard .endm @@ -540,32 +391,6 @@ * Set cp0 enable bit as sign that we're running on the kernel stack */ .macro KMODE -#ifdef CONFIG_MIPS_MT_SMTC - /* - * This gets baroque in SMTC. We want to - * protect the non-atomic clearing of EXL - * with DMT/EMT, but we don't want to take - * an interrupt while DMT is still in effect. - */ - - /* KMODE gets invoked from both reorder and noreorder code */ - .set push - .set mips32r2 - .set noreorder - mfc0 v0, CP0_TCSTATUS - andi v1, v0, TCSTATUS_IXMT - ori v0, TCSTATUS_IXMT - mtc0 v0, CP0_TCSTATUS - _ehb - DMT 2 # dmt v0 - /* - * We don't know a priori if ra is "live" - */ - move t0, ra - jal mips_ihb - nop /* delay slot */ - move ra, t0 -#endif /* CONFIG_MIPS_MT_SMTC */ mfc0 t0, CP0_STATUS li t1, ST0_CU0 | (STATMASK & ~1) #if defined(CONFIG_CPU_R3000) || defined(CONFIG_CPU_TX39XX) @@ -576,25 +401,6 @@ or t0, t1 xori t0, STATMASK & ~1 mtc0 t0, CP0_STATUS -#ifdef CONFIG_MIPS_MT_SMTC - _ehb - andi v0, v0, VPECONTROL_TE - beqz v0, 2f - nop /* delay slot */ - emt -2: - mfc0 v0, CP0_TCSTATUS - /* Clear IXMT, then OR in previous value */ - ori v0, TCSTATUS_IXMT - xori v0, TCSTATUS_IXMT - or v0, v1, v0 - mtc0 v0, CP0_TCSTATUS - /* - * irq_disable_hazard below should expand to EHB - * on 24K/34K CPUS - */ - .set pop -#endif /* CONFIG_MIPS_MT_SMTC */ irq_disable_hazard .endm diff --git a/arch/mips/include/asm/thread_info.h b/arch/mips/include/asm/thread_info.h index d2d961d6cb8..7de865805de 100644 --- a/arch/mips/include/asm/thread_info.h +++ b/arch/mips/include/asm/thread_info.h @@ -159,11 +159,7 @@ static inline struct thread_info *current_thread_info(void) * We stash processor id into a COP0 register to retrieve it fast * at kernel exception entry. */ -#if defined(CONFIG_MIPS_MT_SMTC) -#define SMP_CPUID_REG 2, 2 /* TCBIND */ -#define ASM_SMP_CPUID_REG $2, 2 -#define SMP_CPUID_PTRSHIFT 19 -#elif defined(CONFIG_MIPS_PGD_C0_CONTEXT) +#if defined(CONFIG_MIPS_PGD_C0_CONTEXT) #define SMP_CPUID_REG 20, 0 /* XCONTEXT */ #define ASM_SMP_CPUID_REG $20 #define SMP_CPUID_PTRSHIFT 48 @@ -179,13 +175,8 @@ static inline struct thread_info *current_thread_info(void) #define SMP_CPUID_REGSHIFT (SMP_CPUID_PTRSHIFT + 2) #endif -#ifdef CONFIG_MIPS_MT_SMTC -#define ASM_CPUID_MFC0 mfc0 -#define UASM_i_CPUID_MFC0 uasm_i_mfc0 -#else #define ASM_CPUID_MFC0 MFC0 #define UASM_i_CPUID_MFC0 UASM_i_MFC0 -#endif #endif /* __KERNEL__ */ #endif /* _ASM_THREAD_INFO_H */ diff --git a/arch/mips/include/asm/time.h b/arch/mips/include/asm/time.h index 24f534a7fbc..8f3047d611e 100644 --- a/arch/mips/include/asm/time.h +++ b/arch/mips/include/asm/time.h @@ -52,14 +52,11 @@ extern int (*perf_irq)(void); */ extern unsigned int __weak get_c0_compare_int(void); extern int r4k_clockevent_init(void); -extern int smtc_clockevent_init(void); extern int gic_clockevent_init(void); static inline int mips_clockevent_init(void) { -#ifdef CONFIG_MIPS_MT_SMTC - return smtc_clockevent_init(); -#elif defined(CONFIG_CEVT_GIC) +#if defined(CONFIG_CEVT_GIC) return (gic_clockevent_init() | r4k_clockevent_init()); #elif defined(CONFIG_CEVT_R4K) return r4k_clockevent_init(); diff --git a/arch/mips/include/asm/timex.h b/arch/mips/include/asm/timex.h index c5424757da6..b05bb70a2e4 100644 --- a/arch/mips/include/asm/timex.h +++ b/arch/mips/include/asm/timex.h @@ -4,12 +4,16 @@ * for more details. * * Copyright (C) 1998, 1999, 2003 by Ralf Baechle + * Copyright (C) 2014 by Maciej W. Rozycki */ #ifndef _ASM_TIMEX_H #define _ASM_TIMEX_H #ifdef __KERNEL__ +#include <linux/compiler.h> + +#include <asm/cpu.h> #include <asm/cpu-features.h> #include <asm/mipsregs.h> #include <asm/cpu-type.h> @@ -45,29 +49,54 @@ typedef unsigned int cycles_t; * However for now the implementaton of this function doesn't get these * fine details right. */ -static inline cycles_t get_cycles(void) +static inline int can_use_mips_counter(unsigned int prid) { - switch (boot_cpu_type()) { - case CPU_R4400PC: - case CPU_R4400SC: - case CPU_R4400MC: - if ((read_c0_prid() & 0xff) >= 0x0050) - return read_c0_count(); - break; + int comp = (prid & PRID_COMP_MASK) != PRID_COMP_LEGACY; - case CPU_R4000PC: - case CPU_R4000SC: - case CPU_R4000MC: - break; + if (__builtin_constant_p(cpu_has_counter) && !cpu_has_counter) + return 0; + else if (__builtin_constant_p(cpu_has_mips_r) && cpu_has_mips_r) + return 1; + else if (likely(!__builtin_constant_p(cpu_has_mips_r) && comp)) + return 1; + /* Make sure we don't peek at cpu_data[0].options in the fast path! */ + if (!__builtin_constant_p(cpu_has_counter)) + asm volatile("" : "=m" (cpu_data[0].options)); + if (likely(cpu_has_counter && + prid >= (PRID_IMP_R4000 | PRID_REV_ENCODE_44(5, 0)))) + return 1; + else + return 0; +} - default: - if (cpu_has_counter) - return read_c0_count(); - break; - } +static inline cycles_t get_cycles(void) +{ + if (can_use_mips_counter(read_c0_prid())) + return read_c0_count(); + else + return 0; /* no usable counter */ +} + +/* + * Like get_cycles - but where c0_count is not available we desperately + * use c0_random in an attempt to get at least a little bit of entropy. + * + * R6000 and R6000A neither have a count register nor a random register. + * That leaves no entropy source in the CPU itself. + */ +static inline unsigned long random_get_entropy(void) +{ + unsigned int prid = read_c0_prid(); + unsigned int imp = prid & PRID_IMP_MASK; - return 0; /* no usable counter */ + if (can_use_mips_counter(prid)) + return read_c0_count(); + else if (likely(imp != PRID_IMP_R6000 && imp != PRID_IMP_R6000A)) + return read_c0_random(); + else + return 0; /* no usable register */ } +#define random_get_entropy random_get_entropy #endif /* __KERNEL__ */ diff --git a/arch/mips/include/asm/uasm.h b/arch/mips/include/asm/uasm.h index c33a9564fb4..f8d63b3b40b 100644 --- a/arch/mips/include/asm/uasm.h +++ b/arch/mips/include/asm/uasm.h @@ -55,6 +55,9 @@ void ISAOPC(op)(u32 **buf, unsigned int a, unsigned int b, unsigned int c) #define Ip_u2u1u3(op) \ void ISAOPC(op)(u32 **buf, unsigned int a, unsigned int b, unsigned int c) +#define Ip_u3u2u1(op) \ +void ISAOPC(op)(u32 **buf, unsigned int a, unsigned int b, unsigned int c) + #define Ip_u3u1u2(op) \ void ISAOPC(op)(u32 **buf, unsigned int a, unsigned int b, unsigned int c) @@ -74,6 +77,9 @@ void ISAOPC(op)(u32 **buf, unsigned int a, unsigned int b, unsigned int c, \ #define Ip_u1u2(op) \ void ISAOPC(op)(u32 **buf, unsigned int a, unsigned int b) +#define Ip_u2u1(op) \ +void ISAOPC(op)(u32 **buf, unsigned int a, unsigned int b) + #define Ip_u1s2(op) \ void ISAOPC(op)(u32 **buf, unsigned int a, signed int b) @@ -99,6 +105,7 @@ Ip_u2u1s3(_daddiu); Ip_u3u1u2(_daddu); Ip_u2u1msbu3(_dins); Ip_u2u1msbu3(_dinsm); +Ip_u1u2(_divu); Ip_u1u2u3(_dmfc0); Ip_u1u2u3(_dmtc0); Ip_u2u1u3(_drotr); @@ -114,16 +121,22 @@ Ip_u2u1msbu3(_ext); Ip_u2u1msbu3(_ins); Ip_u1(_j); Ip_u1(_jal); +Ip_u2u1(_jalr); Ip_u1(_jr); +Ip_u2s3u1(_lb); Ip_u2s3u1(_ld); Ip_u3u1u2(_ldx); +Ip_u2s3u1(_lh); Ip_u2s3u1(_ll); Ip_u2s3u1(_lld); Ip_u1s2(_lui); Ip_u2s3u1(_lw); Ip_u3u1u2(_lwx); Ip_u1u2u3(_mfc0); +Ip_u1(_mfhi); +Ip_u1(_mflo); Ip_u1u2u3(_mtc0); +Ip_u3u1u2(_mul); Ip_u3u1u2(_or); Ip_u2u1u3(_ori); Ip_u2s3u1(_pref); @@ -133,17 +146,25 @@ Ip_u2s3u1(_sc); Ip_u2s3u1(_scd); Ip_u2s3u1(_sd); Ip_u2u1u3(_sll); +Ip_u3u2u1(_sllv); +Ip_u2u1s3(_sltiu); +Ip_u3u1u2(_sltu); Ip_u2u1u3(_sra); Ip_u2u1u3(_srl); +Ip_u3u2u1(_srlv); Ip_u3u1u2(_subu); Ip_u2s3u1(_sw); +Ip_u1(_sync); Ip_u1(_syscall); Ip_0(_tlbp); Ip_0(_tlbr); Ip_0(_tlbwi); Ip_0(_tlbwr); +Ip_u1(_wait); +Ip_u2u1(_wsbh); Ip_u3u1u2(_xor); Ip_u2u1u3(_xori); +Ip_u2u1(_yield); /* Handle labels. */ @@ -264,6 +285,8 @@ void uasm_il_bbit0(u32 **p, struct uasm_reloc **r, unsigned int reg, unsigned int bit, int lid); void uasm_il_bbit1(u32 **p, struct uasm_reloc **r, unsigned int reg, unsigned int bit, int lid); +void uasm_il_beq(u32 **p, struct uasm_reloc **r, unsigned int r1, + unsigned int r2, int lid); void uasm_il_beqz(u32 **p, struct uasm_reloc **r, unsigned int reg, int lid); void uasm_il_beqzl(u32 **p, struct uasm_reloc **r, unsigned int reg, int lid); void uasm_il_bgezl(u32 **p, struct uasm_reloc **r, unsigned int reg, int lid); diff --git a/arch/mips/include/uapi/asm/Kbuild b/arch/mips/include/uapi/asm/Kbuild index be7196eacb8..96fe7395ed8 100644 --- a/arch/mips/include/uapi/asm/Kbuild +++ b/arch/mips/include/uapi/asm/Kbuild @@ -4,6 +4,7 @@ include include/uapi/asm-generic/Kbuild.asm generic-y += auxvec.h generic-y += ipcbuf.h +header-y += bitfield.h header-y += bitsperlong.h header-y += break.h header-y += byteorder.h diff --git a/arch/mips/include/uapi/asm/bitfield.h b/arch/mips/include/uapi/asm/bitfield.h new file mode 100644 index 00000000000..ad9861359ce --- /dev/null +++ b/arch/mips/include/uapi/asm/bitfield.h @@ -0,0 +1,29 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2014 by Ralf Baechle <ralf@linux-mips.org> + */ +#ifndef __UAPI_ASM_BITFIELD_H +#define __UAPI_ASM_BITFIELD_H + +/* + * * Damn ... bitfields depend from byteorder :-( + * */ +#ifdef __MIPSEB__ +#define __BITFIELD_FIELD(field, more) \ + field; \ + more + +#elif defined(__MIPSEL__) + +#define __BITFIELD_FIELD(field, more) \ + more \ + field; + +#else /* !defined (__MIPSEB__) && !defined (__MIPSEL__) */ +#error "MIPS but neither __MIPSEL__ nor __MIPSEB__?" +#endif + +#endif /* __UAPI_ASM_BITFIELD_H */ diff --git a/arch/mips/include/uapi/asm/inst.h b/arch/mips/include/uapi/asm/inst.h index 3125797f2a8..4b716025929 100644 --- a/arch/mips/include/uapi/asm/inst.h +++ b/arch/mips/include/uapi/asm/inst.h @@ -13,6 +13,8 @@ #ifndef _UAPI_ASM_INST_H #define _UAPI_ASM_INST_H +#include <asm/bitfield.h> + /* * Major opcodes; before MIPS IV cop1x was called cop3. */ @@ -74,16 +76,17 @@ enum spec2_op { enum spec3_op { ext_op, dextm_op, dextu_op, dext_op, ins_op, dinsm_op, dinsu_op, dins_op, - lx_op = 0x0a, lwle_op = 0x19, - lwre_op = 0x1a, cachee_op = 0x1b, - sbe_op = 0x1c, she_op = 0x1d, - sce_op = 0x1e, swe_op = 0x1f, - bshfl_op = 0x20, swle_op = 0x21, - swre_op = 0x22, prefe_op = 0x23, - dbshfl_op = 0x24, lbue_op = 0x28, - lhue_op = 0x29, lbe_op = 0x2c, - lhe_op = 0x2d, lle_op = 0x2e, - lwe_op = 0x2f, rdhwr_op = 0x3b + yield_op = 0x09, lx_op = 0x0a, + lwle_op = 0x19, lwre_op = 0x1a, + cachee_op = 0x1b, sbe_op = 0x1c, + she_op = 0x1d, sce_op = 0x1e, + swe_op = 0x1f, bshfl_op = 0x20, + swle_op = 0x21, swre_op = 0x22, + prefe_op = 0x23, dbshfl_op = 0x24, + lbue_op = 0x28, lhue_op = 0x29, + lbe_op = 0x2c, lhe_op = 0x2d, + lle_op = 0x2e, lwe_op = 0x2f, + rdhwr_op = 0x3b }; /* @@ -125,7 +128,8 @@ enum bcop_op { enum cop0_coi_func { tlbr_op = 0x01, tlbwi_op = 0x02, tlbwr_op = 0x06, tlbp_op = 0x08, - rfe_op = 0x10, eret_op = 0x18 + rfe_op = 0x10, eret_op = 0x18, + wait_op = 0x20, }; /* @@ -202,6 +206,16 @@ enum lx_func { }; /* + * BSHFL opcodes + */ +enum bshfl_func { + wsbh_op = 0x2, + dshd_op = 0x5, + seb_op = 0x10, + seh_op = 0x18, +}; + +/* * (microMIPS) Major opcodes. */ enum mm_major_op { @@ -244,17 +258,22 @@ enum mm_32i_minor_op { enum mm_32a_minor_op { mm_sll32_op = 0x000, mm_ins_op = 0x00c, + mm_sllv32_op = 0x010, mm_ext_op = 0x02c, mm_pool32axf_op = 0x03c, mm_srl32_op = 0x040, mm_sra_op = 0x080, + mm_srlv32_op = 0x090, mm_rotr_op = 0x0c0, mm_lwxs_op = 0x118, mm_addu32_op = 0x150, mm_subu32_op = 0x1d0, + mm_wsbh_op = 0x1ec, + mm_mul_op = 0x210, mm_and_op = 0x250, mm_or32_op = 0x290, mm_xor32_op = 0x310, + mm_sltu_op = 0x390, }; /* @@ -294,15 +313,20 @@ enum mm_32axf_minor_op { mm_mfc0_op = 0x003, mm_mtc0_op = 0x00b, mm_tlbp_op = 0x00d, + mm_mfhi32_op = 0x035, mm_jalr_op = 0x03c, mm_tlbr_op = 0x04d, + mm_mflo32_op = 0x075, mm_jalrhb_op = 0x07c, mm_tlbwi_op = 0x08d, mm_tlbwr_op = 0x0cd, mm_jalrs_op = 0x13c, mm_jalrshb_op = 0x17c, + mm_sync_op = 0x1ad, mm_syscall_op = 0x22d, + mm_wait_op = 0x24d, mm_eret_op = 0x3cd, + mm_divu_op = 0x5dc, }; /* @@ -480,24 +504,6 @@ enum MIPS6e_i8_func { */ #define MM_NOP16 0x0c00 -/* - * Damn ... bitfields depend from byteorder :-( - */ -#ifdef __MIPSEB__ -#define __BITFIELD_FIELD(field, more) \ - field; \ - more - -#elif defined(__MIPSEL__) - -#define __BITFIELD_FIELD(field, more) \ - more \ - field; - -#else /* !defined (__MIPSEB__) && !defined (__MIPSEL__) */ -#error "MIPS but neither __MIPSEL__ nor __MIPSEB__?" -#endif - struct j_format { __BITFIELD_FIELD(unsigned int opcode : 6, /* Jump format */ __BITFIELD_FIELD(unsigned int target : 26, diff --git a/arch/mips/include/uapi/asm/kvm_para.h b/arch/mips/include/uapi/asm/kvm_para.h index 14fab8f0b95..7e16d7c42e6 100644 --- a/arch/mips/include/uapi/asm/kvm_para.h +++ b/arch/mips/include/uapi/asm/kvm_para.h @@ -1 +1,5 @@ -#include <asm-generic/kvm_para.h> +#ifndef _UAPI_ASM_MIPS_KVM_PARA_H +#define _UAPI_ASM_MIPS_KVM_PARA_H + + +#endif /* _UAPI_ASM_MIPS_KVM_PARA_H */ diff --git a/arch/mips/include/uapi/asm/types.h b/arch/mips/include/uapi/asm/types.h index 7ac9d0baad8..f3dd9ff0cc0 100644 --- a/arch/mips/include/uapi/asm/types.h +++ b/arch/mips/include/uapi/asm/types.h @@ -14,9 +14,12 @@ /* * We don't use int-l64.h for the kernel anymore but still use it for * userspace to avoid code changes. + * + * However, some user programs (e.g. perf) may not want this. They can + * flag __SANE_USERSPACE_TYPES__ to get int-ll64.h here. */ #ifndef __KERNEL__ -# if _MIPS_SZLONG == 64 +# if _MIPS_SZLONG == 64 && !defined(__SANE_USERSPACE_TYPES__) # include <asm-generic/int-l64.h> # else # include <asm-generic/int-ll64.h> diff --git a/arch/mips/kernel/Makefile b/arch/mips/kernel/Makefile index 277dab301ce..008a2fed058 100644 --- a/arch/mips/kernel/Makefile +++ b/arch/mips/kernel/Makefile @@ -17,7 +17,6 @@ endif obj-$(CONFIG_CEVT_BCM1480) += cevt-bcm1480.o obj-$(CONFIG_CEVT_R4K) += cevt-r4k.o -obj-$(CONFIG_MIPS_MT_SMTC) += cevt-smtc.o obj-$(CONFIG_CEVT_DS1287) += cevt-ds1287.o obj-$(CONFIG_CEVT_GIC) += cevt-gic.o obj-$(CONFIG_CEVT_GT641XX) += cevt-gt641xx.o @@ -42,7 +41,7 @@ obj-$(CONFIG_CPU_R4K_FPU) += r4k_fpu.o r4k_switch.o obj-$(CONFIG_CPU_R3000) += r2300_fpu.o r2300_switch.o obj-$(CONFIG_CPU_R6000) += r6000_fpu.o r4k_switch.o obj-$(CONFIG_CPU_TX39XX) += r2300_fpu.o r2300_switch.o -obj-$(CONFIG_CPU_CAVIUM_OCTEON) += octeon_switch.o +obj-$(CONFIG_CPU_CAVIUM_OCTEON) += r4k_fpu.o octeon_switch.o obj-$(CONFIG_SMP) += smp.o obj-$(CONFIG_SMP_UP) += smp-up.o @@ -50,7 +49,6 @@ obj-$(CONFIG_CPU_BMIPS) += smp-bmips.o bmips_vec.o obj-$(CONFIG_MIPS_MT) += mips-mt.o obj-$(CONFIG_MIPS_MT_FPAFF) += mips-mt-fpaff.o -obj-$(CONFIG_MIPS_MT_SMTC) += smtc.o smtc-asm.o smtc-proc.o obj-$(CONFIG_MIPS_MT_SMP) += smp-mt.o obj-$(CONFIG_MIPS_CMP) += smp-cmp.o obj-$(CONFIG_MIPS_CPS) += smp-cps.o cps-vec.o @@ -107,6 +105,9 @@ obj-$(CONFIG_JUMP_LABEL) += jump_label.o obj-$(CONFIG_MIPS_CM) += mips-cm.o obj-$(CONFIG_MIPS_CPC) += mips-cpc.o +obj-$(CONFIG_CPU_PM) += pm.o +obj-$(CONFIG_MIPS_CPS_PM) += pm-cps.o + # # DSP ASE supported for MIPS32 or MIPS64 Release 2 cores only. It is not # safe to unconditionnaly use the assembler -mdsp / -mdspr2 switches diff --git a/arch/mips/kernel/asm-offsets.c b/arch/mips/kernel/asm-offsets.c index 0ea75c244b4..02f075df8f2 100644 --- a/arch/mips/kernel/asm-offsets.c +++ b/arch/mips/kernel/asm-offsets.c @@ -14,6 +14,7 @@ #include <linux/mm.h> #include <linux/kbuild.h> #include <linux/suspend.h> +#include <asm/pm.h> #include <asm/ptrace.h> #include <asm/processor.h> #include <asm/smp-cps.h> @@ -64,9 +65,6 @@ void output_ptreg_defines(void) OFFSET(PT_BVADDR, pt_regs, cp0_badvaddr); OFFSET(PT_STATUS, pt_regs, cp0_status); OFFSET(PT_CAUSE, pt_regs, cp0_cause); -#ifdef CONFIG_MIPS_MT_SMTC - OFFSET(PT_TCSTATUS, pt_regs, cp0_tcstatus); -#endif /* CONFIG_MIPS_MT_SMTC */ #ifdef CONFIG_CPU_CAVIUM_OCTEON OFFSET(PT_MPL, pt_regs, mpl); OFFSET(PT_MTP, pt_regs, mtp); @@ -404,6 +402,20 @@ void output_pbe_defines(void) } #endif +#ifdef CONFIG_CPU_PM +void output_pm_defines(void) +{ + COMMENT(" PM offsets. "); +#ifdef CONFIG_EVA + OFFSET(SSS_SEGCTL0, mips_static_suspend_state, segctl[0]); + OFFSET(SSS_SEGCTL1, mips_static_suspend_state, segctl[1]); + OFFSET(SSS_SEGCTL2, mips_static_suspend_state, segctl[2]); +#endif + OFFSET(SSS_SP, mips_static_suspend_state, sp); + BLANK(); +} +#endif + void output_kvm_defines(void) { COMMENT(" KVM/MIPS Specfic offsets. "); @@ -472,10 +484,14 @@ void output_kvm_defines(void) void output_cps_defines(void) { COMMENT(" MIPS CPS offsets. "); - OFFSET(BOOTCFG_CORE, boot_config, core); - OFFSET(BOOTCFG_VPE, boot_config, vpe); - OFFSET(BOOTCFG_PC, boot_config, pc); - OFFSET(BOOTCFG_SP, boot_config, sp); - OFFSET(BOOTCFG_GP, boot_config, gp); + + OFFSET(COREBOOTCFG_VPEMASK, core_boot_config, vpe_mask); + OFFSET(COREBOOTCFG_VPECONFIG, core_boot_config, vpe_config); + DEFINE(COREBOOTCFG_SIZE, sizeof(struct core_boot_config)); + + OFFSET(VPEBOOTCFG_PC, vpe_boot_config, pc); + OFFSET(VPEBOOTCFG_SP, vpe_boot_config, sp); + OFFSET(VPEBOOTCFG_GP, vpe_boot_config, gp); + DEFINE(VPEBOOTCFG_SIZE, sizeof(struct vpe_boot_config)); } #endif diff --git a/arch/mips/kernel/branch.c b/arch/mips/kernel/branch.c index 76122ff5cb5..7b2df224f04 100644 --- a/arch/mips/kernel/branch.c +++ b/arch/mips/kernel/branch.c @@ -48,6 +48,202 @@ int __isa_exception_epc(struct pt_regs *regs) return epc; } +/* (microMIPS) Convert 16-bit register encoding to 32-bit register encoding. */ +static const unsigned int reg16to32map[8] = {16, 17, 2, 3, 4, 5, 6, 7}; + +int __mm_isBranchInstr(struct pt_regs *regs, struct mm_decoded_insn dec_insn, + unsigned long *contpc) +{ + union mips_instruction insn = (union mips_instruction)dec_insn.insn; + int bc_false = 0; + unsigned int fcr31; + unsigned int bit; + + if (!cpu_has_mmips) + return 0; + + switch (insn.mm_i_format.opcode) { + case mm_pool32a_op: + if ((insn.mm_i_format.simmediate & MM_POOL32A_MINOR_MASK) == + mm_pool32axf_op) { + switch (insn.mm_i_format.simmediate >> + MM_POOL32A_MINOR_SHIFT) { + case mm_jalr_op: + case mm_jalrhb_op: + case mm_jalrs_op: + case mm_jalrshb_op: + if (insn.mm_i_format.rt != 0) /* Not mm_jr */ + regs->regs[insn.mm_i_format.rt] = + regs->cp0_epc + + dec_insn.pc_inc + + dec_insn.next_pc_inc; + *contpc = regs->regs[insn.mm_i_format.rs]; + return 1; + } + } + break; + case mm_pool32i_op: + switch (insn.mm_i_format.rt) { + case mm_bltzals_op: + case mm_bltzal_op: + regs->regs[31] = regs->cp0_epc + + dec_insn.pc_inc + + dec_insn.next_pc_inc; + /* Fall through */ + case mm_bltz_op: + if ((long)regs->regs[insn.mm_i_format.rs] < 0) + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + (insn.mm_i_format.simmediate << 1); + else + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + dec_insn.next_pc_inc; + return 1; + case mm_bgezals_op: + case mm_bgezal_op: + regs->regs[31] = regs->cp0_epc + + dec_insn.pc_inc + + dec_insn.next_pc_inc; + /* Fall through */ + case mm_bgez_op: + if ((long)regs->regs[insn.mm_i_format.rs] >= 0) + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + (insn.mm_i_format.simmediate << 1); + else + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + dec_insn.next_pc_inc; + return 1; + case mm_blez_op: + if ((long)regs->regs[insn.mm_i_format.rs] <= 0) + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + (insn.mm_i_format.simmediate << 1); + else + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + dec_insn.next_pc_inc; + return 1; + case mm_bgtz_op: + if ((long)regs->regs[insn.mm_i_format.rs] <= 0) + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + (insn.mm_i_format.simmediate << 1); + else + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + dec_insn.next_pc_inc; + return 1; + case mm_bc2f_op: + case mm_bc1f_op: + bc_false = 1; + /* Fall through */ + case mm_bc2t_op: + case mm_bc1t_op: + preempt_disable(); + if (is_fpu_owner()) + asm volatile("cfc1\t%0,$31" : "=r" (fcr31)); + else + fcr31 = current->thread.fpu.fcr31; + preempt_enable(); + + if (bc_false) + fcr31 = ~fcr31; + + bit = (insn.mm_i_format.rs >> 2); + bit += (bit != 0); + bit += 23; + if (fcr31 & (1 << bit)) + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + (insn.mm_i_format.simmediate << 1); + else + *contpc = regs->cp0_epc + + dec_insn.pc_inc + dec_insn.next_pc_inc; + return 1; + } + break; + case mm_pool16c_op: + switch (insn.mm_i_format.rt) { + case mm_jalr16_op: + case mm_jalrs16_op: + regs->regs[31] = regs->cp0_epc + + dec_insn.pc_inc + dec_insn.next_pc_inc; + /* Fall through */ + case mm_jr16_op: + *contpc = regs->regs[insn.mm_i_format.rs]; + return 1; + } + break; + case mm_beqz16_op: + if ((long)regs->regs[reg16to32map[insn.mm_b1_format.rs]] == 0) + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + (insn.mm_b1_format.simmediate << 1); + else + *contpc = regs->cp0_epc + + dec_insn.pc_inc + dec_insn.next_pc_inc; + return 1; + case mm_bnez16_op: + if ((long)regs->regs[reg16to32map[insn.mm_b1_format.rs]] != 0) + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + (insn.mm_b1_format.simmediate << 1); + else + *contpc = regs->cp0_epc + + dec_insn.pc_inc + dec_insn.next_pc_inc; + return 1; + case mm_b16_op: + *contpc = regs->cp0_epc + dec_insn.pc_inc + + (insn.mm_b0_format.simmediate << 1); + return 1; + case mm_beq32_op: + if (regs->regs[insn.mm_i_format.rs] == + regs->regs[insn.mm_i_format.rt]) + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + (insn.mm_i_format.simmediate << 1); + else + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + dec_insn.next_pc_inc; + return 1; + case mm_bne32_op: + if (regs->regs[insn.mm_i_format.rs] != + regs->regs[insn.mm_i_format.rt]) + *contpc = regs->cp0_epc + + dec_insn.pc_inc + + (insn.mm_i_format.simmediate << 1); + else + *contpc = regs->cp0_epc + + dec_insn.pc_inc + dec_insn.next_pc_inc; + return 1; + case mm_jalx32_op: + regs->regs[31] = regs->cp0_epc + + dec_insn.pc_inc + dec_insn.next_pc_inc; + *contpc = regs->cp0_epc + dec_insn.pc_inc; + *contpc >>= 28; + *contpc <<= 28; + *contpc |= (insn.j_format.target << 2); + return 1; + case mm_jals32_op: + case mm_jal32_op: + regs->regs[31] = regs->cp0_epc + + dec_insn.pc_inc + dec_insn.next_pc_inc; + /* Fall through */ + case mm_j32_op: + *contpc = regs->cp0_epc + dec_insn.pc_inc; + *contpc >>= 27; + *contpc <<= 27; + *contpc |= (insn.j_format.target << 1); + set_isa16_mode(*contpc); + return 1; + } + return 0; +} + /* * Compute return address and emulate branch in microMIPS mode after an * exception only. It does not handle compact branches/jumps and cannot @@ -366,7 +562,11 @@ int __compute_return_epc_for_insn(struct pt_regs *regs, case cop1_op: preempt_disable(); if (is_fpu_owner()) - asm volatile("cfc1\t%0,$31" : "=r" (fcr31)); + asm volatile( + ".set push\n" + "\t.set mips1\n" + "\tcfc1\t%0,$31\n" + "\t.set pop" : "=r" (fcr31)); else fcr31 = current->thread.fpu.fcr31; preempt_enable(); diff --git a/arch/mips/kernel/cevt-gic.c b/arch/mips/kernel/cevt-gic.c index 594cbbf16d6..6093716980b 100644 --- a/arch/mips/kernel/cevt-gic.c +++ b/arch/mips/kernel/cevt-gic.c @@ -26,7 +26,7 @@ static int gic_next_event(unsigned long delta, struct clock_event_device *evt) cnt = gic_read_count(); cnt += (u64)delta; - gic_write_compare(cnt); + gic_write_cpu_compare(cnt, cpumask_first(evt->cpumask)); res = ((int)(gic_read_count() - cnt) >= 0) ? -ETIME : 0; return res; } @@ -73,7 +73,8 @@ int gic_clockevent_init(void) cd = &per_cpu(gic_clockevent_device, cpu); cd->name = "MIPS GIC"; - cd->features = CLOCK_EVT_FEAT_ONESHOT; + cd->features = CLOCK_EVT_FEAT_ONESHOT | + CLOCK_EVT_FEAT_C3STOP; clockevent_set_clock(cd, gic_frequency); diff --git a/arch/mips/kernel/cevt-r4k.c b/arch/mips/kernel/cevt-r4k.c index 50d3f5a8d6b..bc127e22fda 100644 --- a/arch/mips/kernel/cevt-r4k.c +++ b/arch/mips/kernel/cevt-r4k.c @@ -12,17 +12,10 @@ #include <linux/smp.h> #include <linux/irq.h> -#include <asm/smtc_ipi.h> #include <asm/time.h> #include <asm/cevt-r4k.h> #include <asm/gic.h> -/* - * The SMTC Kernel for the 34K, 1004K, et. al. replaces several - * of these routines with SMTC-specific variants. - */ - -#ifndef CONFIG_MIPS_MT_SMTC static int mips_next_event(unsigned long delta, struct clock_event_device *evt) { @@ -36,8 +29,6 @@ static int mips_next_event(unsigned long delta, return res; } -#endif /* CONFIG_MIPS_MT_SMTC */ - void mips_set_clock_mode(enum clock_event_mode mode, struct clock_event_device *evt) { @@ -47,7 +38,6 @@ void mips_set_clock_mode(enum clock_event_mode mode, DEFINE_PER_CPU(struct clock_event_device, mips_clockevent_device); int cp0_timer_irq_installed; -#ifndef CONFIG_MIPS_MT_SMTC irqreturn_t c0_compare_interrupt(int irq, void *dev_id) { const int r2 = cpu_has_mips_r2; @@ -72,9 +62,6 @@ irqreturn_t c0_compare_interrupt(int irq, void *dev_id) /* Clear Count/Compare Interrupt */ write_c0_compare(read_c0_compare()); cd = &per_cpu(mips_clockevent_device, cpu); -#ifdef CONFIG_CEVT_GIC - if (!gic_present) -#endif cd->event_handler(cd); } @@ -82,8 +69,6 @@ out: return IRQ_HANDLED; } -#endif /* Not CONFIG_MIPS_MT_SMTC */ - struct irqaction c0_compare_irqaction = { .handler = c0_compare_interrupt, .flags = IRQF_PERCPU | IRQF_TIMER, @@ -170,7 +155,6 @@ int c0_compare_int_usable(void) return 1; } -#ifndef CONFIG_MIPS_MT_SMTC int r4k_clockevent_init(void) { unsigned int cpu = smp_processor_id(); @@ -195,7 +179,9 @@ int r4k_clockevent_init(void) cd = &per_cpu(mips_clockevent_device, cpu); cd->name = "MIPS"; - cd->features = CLOCK_EVT_FEAT_ONESHOT; + cd->features = CLOCK_EVT_FEAT_ONESHOT | + CLOCK_EVT_FEAT_C3STOP | + CLOCK_EVT_FEAT_PERCPU; clockevent_set_clock(cd, mips_hpt_frequency); @@ -210,9 +196,6 @@ int r4k_clockevent_init(void) cd->set_mode = mips_set_clock_mode; cd->event_handler = mips_event_handler; -#ifdef CONFIG_CEVT_GIC - if (!gic_present) -#endif clockevents_register_device(cd); if (cp0_timer_irq_installed) @@ -225,4 +208,3 @@ int r4k_clockevent_init(void) return 0; } -#endif /* Not CONFIG_MIPS_MT_SMTC */ diff --git a/arch/mips/kernel/cevt-smtc.c b/arch/mips/kernel/cevt-smtc.c deleted file mode 100644 index b6cf0a60d89..00000000000 --- a/arch/mips/kernel/cevt-smtc.c +++ /dev/null @@ -1,324 +0,0 @@ -/* - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - * - * Copyright (C) 2007 MIPS Technologies, Inc. - * Copyright (C) 2007 Ralf Baechle <ralf@linux-mips.org> - * Copyright (C) 2008 Kevin D. Kissell, Paralogos sarl - */ -#include <linux/clockchips.h> -#include <linux/interrupt.h> -#include <linux/percpu.h> -#include <linux/smp.h> -#include <linux/irq.h> - -#include <asm/smtc_ipi.h> -#include <asm/time.h> -#include <asm/cevt-r4k.h> - -/* - * Variant clock event timer support for SMTC on MIPS 34K, 1004K - * or other MIPS MT cores. - * - * Notes on SMTC Support: - * - * SMTC has multiple microthread TCs pretending to be Linux CPUs. - * But there's only one Count/Compare pair per VPE, and Compare - * interrupts are taken opportunisitically by available TCs - * bound to the VPE with the Count register. The new timer - * framework provides for global broadcasts, but we really - * want VPE-level multicasts for best behavior. So instead - * of invoking the high-level clock-event broadcast code, - * this version of SMTC support uses the historical SMTC - * multicast mechanisms "under the hood", appearing to the - * generic clock layer as if the interrupts are per-CPU. - * - * The approach taken here is to maintain a set of NR_CPUS - * virtual timers, and track which "CPU" needs to be alerted - * at each event. - * - * It's unlikely that we'll see a MIPS MT core with more than - * 2 VPEs, but we *know* that we won't need to handle more - * VPEs than we have "CPUs". So NCPUs arrays of NCPUs elements - * is always going to be overkill, but always going to be enough. - */ - -unsigned long smtc_nexttime[NR_CPUS][NR_CPUS]; -static int smtc_nextinvpe[NR_CPUS]; - -/* - * Timestamps stored are absolute values to be programmed - * into Count register. Valid timestamps will never be zero. - * If a Zero Count value is actually calculated, it is converted - * to be a 1, which will introduce 1 or two CPU cycles of error - * roughly once every four billion events, which at 1000 HZ means - * about once every 50 days. If that's actually a problem, one - * could alternate squashing 0 to 1 and to -1. - */ - -#define MAKEVALID(x) (((x) == 0L) ? 1L : (x)) -#define ISVALID(x) ((x) != 0L) - -/* - * Time comparison is subtle, as it's really truncated - * modular arithmetic. - */ - -#define IS_SOONER(a, b, reference) \ - (((a) - (unsigned long)(reference)) < ((b) - (unsigned long)(reference))) - -/* - * CATCHUP_INCREMENT, used when the function falls behind the counter. - * Could be an increasing function instead of a constant; - */ - -#define CATCHUP_INCREMENT 64 - -static int mips_next_event(unsigned long delta, - struct clock_event_device *evt) -{ - unsigned long flags; - unsigned int mtflags; - unsigned long timestamp, reference, previous; - unsigned long nextcomp = 0L; - int vpe = current_cpu_data.vpe_id; - int cpu = smp_processor_id(); - local_irq_save(flags); - mtflags = dmt(); - - /* - * Maintain the per-TC virtual timer - * and program the per-VPE shared Count register - * as appropriate here... - */ - reference = (unsigned long)read_c0_count(); - timestamp = MAKEVALID(reference + delta); - /* - * To really model the clock, we have to catch the case - * where the current next-in-VPE timestamp is the old - * timestamp for the calling CPE, but the new value is - * in fact later. In that case, we have to do a full - * scan and discover the new next-in-VPE CPU id and - * timestamp. - */ - previous = smtc_nexttime[vpe][cpu]; - if (cpu == smtc_nextinvpe[vpe] && ISVALID(previous) - && IS_SOONER(previous, timestamp, reference)) { - int i; - int soonest = cpu; - - /* - * Update timestamp array here, so that new - * value gets considered along with those of - * other virtual CPUs on the VPE. - */ - smtc_nexttime[vpe][cpu] = timestamp; - for_each_online_cpu(i) { - if (ISVALID(smtc_nexttime[vpe][i]) - && IS_SOONER(smtc_nexttime[vpe][i], - smtc_nexttime[vpe][soonest], reference)) { - soonest = i; - } - } - smtc_nextinvpe[vpe] = soonest; - nextcomp = smtc_nexttime[vpe][soonest]; - /* - * Otherwise, we don't have to process the whole array rank, - * we just have to see if the event horizon has gotten closer. - */ - } else { - if (!ISVALID(smtc_nexttime[vpe][smtc_nextinvpe[vpe]]) || - IS_SOONER(timestamp, - smtc_nexttime[vpe][smtc_nextinvpe[vpe]], reference)) { - smtc_nextinvpe[vpe] = cpu; - nextcomp = timestamp; - } - /* - * Since next-in-VPE may me the same as the executing - * virtual CPU, we update the array *after* checking - * its value. - */ - smtc_nexttime[vpe][cpu] = timestamp; - } - - /* - * It may be that, in fact, we don't need to update Compare, - * but if we do, we want to make sure we didn't fall into - * a crack just behind Count. - */ - if (ISVALID(nextcomp)) { - write_c0_compare(nextcomp); - ehb(); - /* - * We never return an error, we just make sure - * that we trigger the handlers as quickly as - * we can if we fell behind. - */ - while ((nextcomp - (unsigned long)read_c0_count()) - > (unsigned long)LONG_MAX) { - nextcomp += CATCHUP_INCREMENT; - write_c0_compare(nextcomp); - ehb(); - } - } - emt(mtflags); - local_irq_restore(flags); - return 0; -} - - -void smtc_distribute_timer(int vpe) -{ - unsigned long flags; - unsigned int mtflags; - int cpu; - struct clock_event_device *cd; - unsigned long nextstamp; - unsigned long reference; - - -repeat: - nextstamp = 0L; - for_each_online_cpu(cpu) { - /* - * Find virtual CPUs within the current VPE who have - * unserviced timer requests whose time is now past. - */ - local_irq_save(flags); - mtflags = dmt(); - if (cpu_data[cpu].vpe_id == vpe && - ISVALID(smtc_nexttime[vpe][cpu])) { - reference = (unsigned long)read_c0_count(); - if ((smtc_nexttime[vpe][cpu] - reference) - > (unsigned long)LONG_MAX) { - smtc_nexttime[vpe][cpu] = 0L; - emt(mtflags); - local_irq_restore(flags); - /* - * We don't send IPIs to ourself. - */ - if (cpu != smp_processor_id()) { - smtc_send_ipi(cpu, SMTC_CLOCK_TICK, 0); - } else { - cd = &per_cpu(mips_clockevent_device, cpu); - cd->event_handler(cd); - } - } else { - /* Local to VPE but Valid Time not yet reached. */ - if (!ISVALID(nextstamp) || - IS_SOONER(smtc_nexttime[vpe][cpu], nextstamp, - reference)) { - smtc_nextinvpe[vpe] = cpu; - nextstamp = smtc_nexttime[vpe][cpu]; - } - emt(mtflags); - local_irq_restore(flags); - } - } else { - emt(mtflags); - local_irq_restore(flags); - - } - } - /* Reprogram for interrupt at next soonest timestamp for VPE */ - if (ISVALID(nextstamp)) { - write_c0_compare(nextstamp); - ehb(); - if ((nextstamp - (unsigned long)read_c0_count()) - > (unsigned long)LONG_MAX) - goto repeat; - } -} - - -irqreturn_t c0_compare_interrupt(int irq, void *dev_id) -{ - int cpu = smp_processor_id(); - - /* If we're running SMTC, we've got MIPS MT and therefore MIPS32R2 */ - handle_perf_irq(1); - - if (read_c0_cause() & (1 << 30)) { - /* Clear Count/Compare Interrupt */ - write_c0_compare(read_c0_compare()); - smtc_distribute_timer(cpu_data[cpu].vpe_id); - } - return IRQ_HANDLED; -} - - -int smtc_clockevent_init(void) -{ - uint64_t mips_freq = mips_hpt_frequency; - unsigned int cpu = smp_processor_id(); - struct clock_event_device *cd; - unsigned int irq; - int i; - int j; - - if (!cpu_has_counter || !mips_hpt_frequency) - return -ENXIO; - if (cpu == 0) { - for (i = 0; i < num_possible_cpus(); i++) { - smtc_nextinvpe[i] = 0; - for (j = 0; j < num_possible_cpus(); j++) - smtc_nexttime[i][j] = 0L; - } - /* - * SMTC also can't have the usablility test - * run by secondary TCs once Compare is in use. - */ - if (!c0_compare_int_usable()) - return -ENXIO; - } - - /* - * With vectored interrupts things are getting platform specific. - * get_c0_compare_int is a hook to allow a platform to return the - * interrupt number of it's liking. - */ - irq = MIPS_CPU_IRQ_BASE + cp0_compare_irq; - if (get_c0_compare_int) - irq = get_c0_compare_int(); - - cd = &per_cpu(mips_clockevent_device, cpu); - - cd->name = "MIPS"; - cd->features = CLOCK_EVT_FEAT_ONESHOT; - - /* Calculate the min / max delta */ - cd->mult = div_sc((unsigned long) mips_freq, NSEC_PER_SEC, 32); - cd->shift = 32; - cd->max_delta_ns = clockevent_delta2ns(0x7fffffff, cd); - cd->min_delta_ns = clockevent_delta2ns(0x300, cd); - - cd->rating = 300; - cd->irq = irq; - cd->cpumask = cpumask_of(cpu); - cd->set_next_event = mips_next_event; - cd->set_mode = mips_set_clock_mode; - cd->event_handler = mips_event_handler; - - clockevents_register_device(cd); - - /* - * On SMTC we only want to do the data structure - * initialization and IRQ setup once. - */ - if (cpu) - return 0; - /* - * And we need the hwmask associated with the c0_compare - * vector to be initialized. - */ - irq_hwmask[irq] = (0x100 << cp0_compare_irq); - if (cp0_timer_irq_installed) - return 0; - - cp0_timer_irq_installed = 1; - - setup_irq(irq, &c0_compare_irqaction); - - return 0; -} diff --git a/arch/mips/kernel/cps-vec.S b/arch/mips/kernel/cps-vec.S index f7a46db4b16..6f4f739dad9 100644 --- a/arch/mips/kernel/cps-vec.S +++ b/arch/mips/kernel/cps-vec.S @@ -14,19 +14,43 @@ #include <asm/asmmacro.h> #include <asm/cacheops.h> #include <asm/mipsregs.h> +#include <asm/mipsmtregs.h> +#include <asm/pm.h> -#define GCR_CL_COHERENCE_OFS 0x2008 +#define GCR_CL_COHERENCE_OFS 0x2008 +#define GCR_CL_ID_OFS 0x2028 + +.extern mips_cm_base + +.set noreorder + + /* + * Set dest to non-zero if the core supports the MT ASE, else zero. If + * MT is not supported then branch to nomt. + */ + .macro has_mt dest, nomt + mfc0 \dest, CP0_CONFIG + bgez \dest, \nomt + mfc0 \dest, CP0_CONFIG, 1 + bgez \dest, \nomt + mfc0 \dest, CP0_CONFIG, 2 + bgez \dest, \nomt + mfc0 \dest, CP0_CONFIG, 3 + andi \dest, \dest, MIPS_CONF3_MT + beqz \dest, \nomt + .endm .section .text.cps-vec .balign 0x1000 -.set noreorder LEAF(mips_cps_core_entry) /* - * These first 8 bytes will be patched by cps_smp_setup to load the - * base address of the CM GCRs into register v1. + * These first 12 bytes will be patched by cps_smp_setup to load the + * base address of the CM GCRs into register v1 and the CCA to use into + * register s0. */ .quad 0 + .word 0 /* Check whether we're here due to an NMI */ mfc0 k0, CP0_STATUS @@ -117,10 +141,11 @@ icache_done: add a0, a0, t0 dcache_done: - /* Set Kseg0 cacheable, coherent, write-back, write-allocate */ + /* Set Kseg0 CCA to that in s0 */ mfc0 t0, CP0_CONFIG ori t0, 0x7 - xori t0, 0x2 + xori t0, 0x7 + or t0, t0, s0 mtc0 t0, CP0_CONFIG ehb @@ -134,21 +159,24 @@ dcache_done: jr t0 nop -1: /* We're up, cached & coherent */ + /* + * We're up, cached & coherent. Perform any further required core-level + * initialisation. + */ +1: jal mips_cps_core_init + nop /* - * TODO: We should check the VPE number we intended to boot here, and - * if non-zero we should start that VPE and stop this one. For - * the moment this doesn't matter since CPUs are brought up - * sequentially and in order, but once hotplug is implemented - * this will need revisiting. + * Boot any other VPEs within this core that should be online, and + * deactivate this VPE if it should be offline. */ + jal mips_cps_boot_vpes + nop /* Off we go! */ - la t0, mips_cps_bootcfg - lw t1, BOOTCFG_PC(t0) - lw gp, BOOTCFG_GP(t0) - lw sp, BOOTCFG_SP(t0) + lw t1, VPEBOOTCFG_PC(v0) + lw gp, VPEBOOTCFG_GP(v0) + lw sp, VPEBOOTCFG_SP(v0) jr t1 nop END(mips_cps_core_entry) @@ -189,3 +217,271 @@ LEAF(excep_ejtag) jr k0 nop END(excep_ejtag) + +LEAF(mips_cps_core_init) +#ifdef CONFIG_MIPS_MT + /* Check that the core implements the MT ASE */ + has_mt t0, 3f + nop + + .set push + .set mt + + /* Only allow 1 TC per VPE to execute... */ + dmt + + /* ...and for the moment only 1 VPE */ + dvpe + la t1, 1f + jr.hb t1 + nop + + /* Enter VPE configuration state */ +1: mfc0 t0, CP0_MVPCONTROL + ori t0, t0, MVPCONTROL_VPC + mtc0 t0, CP0_MVPCONTROL + + /* Retrieve the number of VPEs within the core */ + mfc0 t0, CP0_MVPCONF0 + srl t0, t0, MVPCONF0_PVPE_SHIFT + andi t0, t0, (MVPCONF0_PVPE >> MVPCONF0_PVPE_SHIFT) + addi t7, t0, 1 + + /* If there's only 1, we're done */ + beqz t0, 2f + nop + + /* Loop through each VPE within this core */ + li t5, 1 + +1: /* Operate on the appropriate TC */ + mtc0 t5, CP0_VPECONTROL + ehb + + /* Bind TC to VPE (1:1 TC:VPE mapping) */ + mttc0 t5, CP0_TCBIND + + /* Set exclusive TC, non-active, master */ + li t0, VPECONF0_MVP + sll t1, t5, VPECONF0_XTC_SHIFT + or t0, t0, t1 + mttc0 t0, CP0_VPECONF0 + + /* Set TC non-active, non-allocatable */ + mttc0 zero, CP0_TCSTATUS + + /* Set TC halted */ + li t0, TCHALT_H + mttc0 t0, CP0_TCHALT + + /* Next VPE */ + addi t5, t5, 1 + slt t0, t5, t7 + bnez t0, 1b + nop + + /* Leave VPE configuration state */ +2: mfc0 t0, CP0_MVPCONTROL + xori t0, t0, MVPCONTROL_VPC + mtc0 t0, CP0_MVPCONTROL + +3: .set pop +#endif + jr ra + nop + END(mips_cps_core_init) + +LEAF(mips_cps_boot_vpes) + /* Retrieve CM base address */ + la t0, mips_cm_base + lw t0, 0(t0) + + /* Calculate a pointer to this cores struct core_boot_config */ + lw t0, GCR_CL_ID_OFS(t0) + li t1, COREBOOTCFG_SIZE + mul t0, t0, t1 + la t1, mips_cps_core_bootcfg + lw t1, 0(t1) + addu t0, t0, t1 + + /* Calculate this VPEs ID. If the core doesn't support MT use 0 */ + has_mt t6, 1f + li t9, 0 + + /* Find the number of VPEs present in the core */ + mfc0 t1, CP0_MVPCONF0 + srl t1, t1, MVPCONF0_PVPE_SHIFT + andi t1, t1, MVPCONF0_PVPE >> MVPCONF0_PVPE_SHIFT + addi t1, t1, 1 + + /* Calculate a mask for the VPE ID from EBase.CPUNum */ + clz t1, t1 + li t2, 31 + subu t1, t2, t1 + li t2, 1 + sll t1, t2, t1 + addiu t1, t1, -1 + + /* Retrieve the VPE ID from EBase.CPUNum */ + mfc0 t9, $15, 1 + and t9, t9, t1 + +1: /* Calculate a pointer to this VPEs struct vpe_boot_config */ + li t1, VPEBOOTCFG_SIZE + mul v0, t9, t1 + lw t7, COREBOOTCFG_VPECONFIG(t0) + addu v0, v0, t7 + +#ifdef CONFIG_MIPS_MT + + /* If the core doesn't support MT then return */ + bnez t6, 1f + nop + jr ra + nop + + .set push + .set mt + +1: /* Enter VPE configuration state */ + dvpe + la t1, 1f + jr.hb t1 + nop +1: mfc0 t1, CP0_MVPCONTROL + ori t1, t1, MVPCONTROL_VPC + mtc0 t1, CP0_MVPCONTROL + ehb + + /* Loop through each VPE */ + lw t6, COREBOOTCFG_VPEMASK(t0) + move t8, t6 + li t5, 0 + + /* Check whether the VPE should be running. If not, skip it */ +1: andi t0, t6, 1 + beqz t0, 2f + nop + + /* Operate on the appropriate TC */ + mfc0 t0, CP0_VPECONTROL + ori t0, t0, VPECONTROL_TARGTC + xori t0, t0, VPECONTROL_TARGTC + or t0, t0, t5 + mtc0 t0, CP0_VPECONTROL + ehb + + /* Skip the VPE if its TC is not halted */ + mftc0 t0, CP0_TCHALT + beqz t0, 2f + nop + + /* Calculate a pointer to the VPEs struct vpe_boot_config */ + li t0, VPEBOOTCFG_SIZE + mul t0, t0, t5 + addu t0, t0, t7 + + /* Set the TC restart PC */ + lw t1, VPEBOOTCFG_PC(t0) + mttc0 t1, CP0_TCRESTART + + /* Set the TC stack pointer */ + lw t1, VPEBOOTCFG_SP(t0) + mttgpr t1, sp + + /* Set the TC global pointer */ + lw t1, VPEBOOTCFG_GP(t0) + mttgpr t1, gp + + /* Copy config from this VPE */ + mfc0 t0, CP0_CONFIG + mttc0 t0, CP0_CONFIG + + /* Ensure no software interrupts are pending */ + mttc0 zero, CP0_CAUSE + mttc0 zero, CP0_STATUS + + /* Set TC active, not interrupt exempt */ + mftc0 t0, CP0_TCSTATUS + li t1, ~TCSTATUS_IXMT + and t0, t0, t1 + ori t0, t0, TCSTATUS_A + mttc0 t0, CP0_TCSTATUS + + /* Clear the TC halt bit */ + mttc0 zero, CP0_TCHALT + + /* Set VPE active */ + mftc0 t0, CP0_VPECONF0 + ori t0, t0, VPECONF0_VPA + mttc0 t0, CP0_VPECONF0 + + /* Next VPE */ +2: srl t6, t6, 1 + addi t5, t5, 1 + bnez t6, 1b + nop + + /* Leave VPE configuration state */ + mfc0 t1, CP0_MVPCONTROL + xori t1, t1, MVPCONTROL_VPC + mtc0 t1, CP0_MVPCONTROL + ehb + evpe + + /* Check whether this VPE is meant to be running */ + li t0, 1 + sll t0, t0, t9 + and t0, t0, t8 + bnez t0, 2f + nop + + /* This VPE should be offline, halt the TC */ + li t0, TCHALT_H + mtc0 t0, CP0_TCHALT + la t0, 1f +1: jr.hb t0 + nop + +2: .set pop + +#endif /* CONFIG_MIPS_MT */ + + /* Return */ + jr ra + nop + END(mips_cps_boot_vpes) + +#if defined(CONFIG_MIPS_CPS_PM) && defined(CONFIG_CPU_PM) + + /* Calculate a pointer to this CPUs struct mips_static_suspend_state */ + .macro psstate dest + .set push + .set noat + lw $1, TI_CPU(gp) + sll $1, $1, LONGLOG + la \dest, __per_cpu_offset + addu $1, $1, \dest + lw $1, 0($1) + la \dest, cps_cpu_state + addu \dest, \dest, $1 + .set pop + .endm + +LEAF(mips_cps_pm_save) + /* Save CPU state */ + SUSPEND_SAVE_REGS + psstate t1 + SUSPEND_SAVE_STATIC + jr v0 + nop + END(mips_cps_pm_save) + +LEAF(mips_cps_pm_restore) + /* Restore CPU state */ + psstate t1 + RESUME_RESTORE_STATIC + RESUME_RESTORE_REGS_RETURN + END(mips_cps_pm_restore) + +#endif /* CONFIG_MIPS_CPS_PM && CONFIG_CPU_PM */ diff --git a/arch/mips/kernel/cpu-probe.c b/arch/mips/kernel/cpu-probe.c index 6e8fb85ce7c..d74f957c561 100644 --- a/arch/mips/kernel/cpu-probe.c +++ b/arch/mips/kernel/cpu-probe.c @@ -62,7 +62,7 @@ static inline void check_errata(void) case CPU_34K: /* * Erratum "RPS May Cause Incorrect Instruction Execution" - * This code only handles VPE0, any SMP/SMTC/RTOS code + * This code only handles VPE0, any SMP/RTOS code * making use of VPE1 will be responsable for that VPE. */ if ((c->processor_id & PRID_REV_MASK) <= PRID_REV_34K_V1_0_2) @@ -423,7 +423,7 @@ static void decode_configs(struct cpuinfo_mips *c) #ifndef CONFIG_MIPS_CPS if (cpu_has_mips_r2) { - c->core = read_c0_ebase() & 0x3ff; + c->core = get_ebase_cpunum(); if (cpu_has_mipsmt) c->core >>= fls(core_nvpes()) - 1; } @@ -684,21 +684,6 @@ static inline void cpu_probe_legacy(struct cpuinfo_mips *c, unsigned int cpu) */ c->tlbsize = (read_c0_info() & (1 << 29)) ? 64 : 48; break; - case PRID_IMP_RM9000: - c->cputype = CPU_RM9000; - __cpu_name[cpu] = "RM9000"; - set_isa(c, MIPS_CPU_ISA_IV); - c->options = R4K_OPTS | MIPS_CPU_FPU | MIPS_CPU_32FPR | - MIPS_CPU_LLSC; - /* - * Bit 29 in the info register of the RM9000 - * indicates if the TLB has 48 or 64 entries. - * - * 29 1 => 64 entry JTLB - * 0 => 48 entry JTLB - */ - c->tlbsize = (read_c0_info() & (1 << 29)) ? 64 : 48; - break; case PRID_IMP_R8000: c->cputype = CPU_R8000; __cpu_name[cpu] = "RM8000"; @@ -1041,6 +1026,7 @@ static inline void cpu_probe_ingenic(struct cpuinfo_mips *c, unsigned int cpu) decode_configs(c); /* JZRISC does not implement the CP0 counter. */ c->options &= ~MIPS_CPU_COUNTER; + BUG_ON(!__builtin_constant_p(cpu_has_counter) || cpu_has_counter); switch (c->processor_id & PRID_IMP_MASK) { case PRID_IMP_JZRISC: c->cputype = CPU_JZRISC; @@ -1074,6 +1060,7 @@ static inline void cpu_probe_netlogic(struct cpuinfo_mips *c, int cpu) switch (c->processor_id & PRID_IMP_MASK) { case PRID_IMP_NETLOGIC_XLP2XX: case PRID_IMP_NETLOGIC_XLP9XX: + case PRID_IMP_NETLOGIC_XLP5XX: c->cputype = CPU_XLP; __cpu_name[cpu] = "Broadcom XLPII"; break; diff --git a/arch/mips/kernel/entry.S b/arch/mips/kernel/entry.S index e5786858cdb..4353d323f01 100644 --- a/arch/mips/kernel/entry.S +++ b/arch/mips/kernel/entry.S @@ -16,9 +16,6 @@ #include <asm/isadep.h> #include <asm/thread_info.h> #include <asm/war.h> -#ifdef CONFIG_MIPS_MT_SMTC -#include <asm/mipsmtregs.h> -#endif #ifndef CONFIG_PREEMPT #define resume_kernel restore_all @@ -89,41 +86,6 @@ FEXPORT(syscall_exit) bnez t0, syscall_exit_work restore_all: # restore full frame -#ifdef CONFIG_MIPS_MT_SMTC -#ifdef CONFIG_MIPS_MT_SMTC_IM_BACKSTOP -/* Re-arm any temporarily masked interrupts not explicitly "acked" */ - mfc0 v0, CP0_TCSTATUS - ori v1, v0, TCSTATUS_IXMT - mtc0 v1, CP0_TCSTATUS - andi v0, TCSTATUS_IXMT - _ehb - mfc0 t0, CP0_TCCONTEXT - DMT 9 # dmt t1 - jal mips_ihb - mfc0 t2, CP0_STATUS - andi t3, t0, 0xff00 - or t2, t2, t3 - mtc0 t2, CP0_STATUS - _ehb - andi t1, t1, VPECONTROL_TE - beqz t1, 1f - EMT -1: - mfc0 v1, CP0_TCSTATUS - /* We set IXMT above, XOR should clear it here */ - xori v1, v1, TCSTATUS_IXMT - or v1, v0, v1 - mtc0 v1, CP0_TCSTATUS - _ehb - xor t0, t0, t3 - mtc0 t0, CP0_TCCONTEXT -#endif /* CONFIG_MIPS_MT_SMTC_IM_BACKSTOP */ -/* Detect and execute deferred IPI "interrupts" */ - LONG_L s0, TI_REGS($28) - LONG_S sp, TI_REGS($28) - jal deferred_smtc_ipi - LONG_S s0, TI_REGS($28) -#endif /* CONFIG_MIPS_MT_SMTC */ .set noat RESTORE_TEMP RESTORE_AT diff --git a/arch/mips/kernel/genex.S b/arch/mips/kernel/genex.S index a9ce3408be2..ac35e12cb1f 100644 --- a/arch/mips/kernel/genex.S +++ b/arch/mips/kernel/genex.S @@ -21,20 +21,6 @@ #include <asm/war.h> #include <asm/thread_info.h> -#ifdef CONFIG_MIPS_MT_SMTC -#define PANIC_PIC(msg) \ - .set push; \ - .set nomicromips; \ - .set reorder; \ - PTR_LA a0,8f; \ - .set noat; \ - PTR_LA AT, panic; \ - jr AT; \ -9: b 9b; \ - .set pop; \ - TEXT(msg) -#endif - __INIT /* @@ -251,15 +237,6 @@ NESTED(except_vec_vi, 0, sp) SAVE_AT .set push .set noreorder -#ifdef CONFIG_MIPS_MT_SMTC - /* - * To keep from blindly blocking *all* interrupts - * during service by SMTC kernel, we also want to - * pass the IM value to be cleared. - */ -FEXPORT(except_vec_vi_mori) - ori a0, $0, 0 -#endif /* CONFIG_MIPS_MT_SMTC */ PTR_LA v1, except_vec_vi_handler FEXPORT(except_vec_vi_lui) lui v0, 0 /* Patched */ @@ -277,37 +254,10 @@ EXPORT(except_vec_vi_end) NESTED(except_vec_vi_handler, 0, sp) SAVE_TEMP SAVE_STATIC -#ifdef CONFIG_MIPS_MT_SMTC - /* - * SMTC has an interesting problem that interrupts are level-triggered, - * and the CLI macro will clear EXL, potentially causing a duplicate - * interrupt service invocation. So we need to clear the associated - * IM bit of Status prior to doing CLI, and restore it after the - * service routine has been invoked - we must assume that the - * service routine will have cleared the state, and any active - * level represents a new or otherwised unserviced event... - */ - mfc0 t1, CP0_STATUS - and t0, a0, t1 -#ifdef CONFIG_MIPS_MT_SMTC_IM_BACKSTOP - mfc0 t2, CP0_TCCONTEXT - or t2, t0, t2 - mtc0 t2, CP0_TCCONTEXT -#endif /* CONFIG_MIPS_MT_SMTC_IM_BACKSTOP */ - xor t1, t1, t0 - mtc0 t1, CP0_STATUS - _ehb -#endif /* CONFIG_MIPS_MT_SMTC */ CLI #ifdef CONFIG_TRACE_IRQFLAGS move s0, v0 -#ifdef CONFIG_MIPS_MT_SMTC - move s1, a0 -#endif TRACE_IRQS_OFF -#ifdef CONFIG_MIPS_MT_SMTC - move a0, s1 -#endif move v0, s0 #endif @@ -496,9 +446,6 @@ NESTED(nmi_handler, PT_SIZE, sp) .align 5 LEAF(handle_ri_rdhwr_vivt) -#ifdef CONFIG_MIPS_MT_SMTC - PANIC_PIC("handle_ri_rdhwr_vivt called") -#else .set push .set noat .set noreorder @@ -517,7 +464,6 @@ NESTED(nmi_handler, PT_SIZE, sp) .set pop bltz k1, handle_ri /* slow path */ /* fall thru */ -#endif END(handle_ri_rdhwr_vivt) LEAF(handle_ri_rdhwr) diff --git a/arch/mips/kernel/head.S b/arch/mips/kernel/head.S index e712dcf18b2..95afd663cd4 100644 --- a/arch/mips/kernel/head.S +++ b/arch/mips/kernel/head.S @@ -35,33 +35,12 @@ */ .macro setup_c0_status set clr .set push -#ifdef CONFIG_MIPS_MT_SMTC - /* - * For SMTC, we need to set privilege and disable interrupts only for - * the current TC, using the TCStatus register. - */ - mfc0 t0, CP0_TCSTATUS - /* Fortunately CU 0 is in the same place in both registers */ - /* Set TCU0, TMX, TKSU (for later inversion) and IXMT */ - li t1, ST0_CU0 | 0x08001c00 - or t0, t1 - /* Clear TKSU, leave IXMT */ - xori t0, 0x00001800 - mtc0 t0, CP0_TCSTATUS - _ehb - /* We need to leave the global IE bit set, but clear EXL...*/ - mfc0 t0, CP0_STATUS - or t0, ST0_CU0 | ST0_EXL | ST0_ERL | \set | \clr - xor t0, ST0_EXL | ST0_ERL | \clr - mtc0 t0, CP0_STATUS -#else mfc0 t0, CP0_STATUS or t0, ST0_CU0|\set|0x1f|\clr xor t0, 0x1f|\clr mtc0 t0, CP0_STATUS .set noreorder sll zero,3 # ehb -#endif .set pop .endm @@ -115,24 +94,6 @@ NESTED(kernel_entry, 16, sp) # kernel entry point jr t0 0: -#ifdef CONFIG_MIPS_MT_SMTC - /* - * In SMTC kernel, "CLI" is thread-specific, in TCStatus. - * We still need to enable interrupts globally in Status, - * and clear EXL/ERL. - * - * TCContext is used to track interrupt levels under - * service in SMTC kernel. Clear for boot TC before - * allowing any interrupts. - */ - mtc0 zero, CP0_TCCONTEXT - - mfc0 t0, CP0_STATUS - ori t0, t0, 0xff1f - xori t0, t0, 0x001e - mtc0 t0, CP0_STATUS -#endif /* CONFIG_MIPS_MT_SMTC */ - PTR_LA t0, __bss_start # clear .bss LONG_S zero, (t0) PTR_LA t1, __bss_stop - LONGSIZE @@ -164,25 +125,8 @@ NESTED(kernel_entry, 16, sp) # kernel entry point * function after setting up the stack and gp registers. */ NESTED(smp_bootstrap, 16, sp) -#ifdef CONFIG_MIPS_MT_SMTC - /* - * Read-modify-writes of Status must be atomic, and this - * is one case where CLI is invoked without EXL being - * necessarily set. The CLI and setup_c0_status will - * in fact be redundant for all but the first TC of - * each VPE being booted. - */ - DMT 10 # dmt t2 /* t0, t1 are used by CLI and setup_c0_status() */ - jal mips_ihb -#endif /* CONFIG_MIPS_MT_SMTC */ smp_slave_setup setup_c0_status_sec -#ifdef CONFIG_MIPS_MT_SMTC - andi t2, t2, VPECONTROL_TE - beqz t2, 2f - EMT # emt -2: -#endif /* CONFIG_MIPS_MT_SMTC */ j start_secondary END(smp_bootstrap) #endif /* CONFIG_SMP */ diff --git a/arch/mips/kernel/i8259.c b/arch/mips/kernel/i8259.c index 2b91fe80c43..50b364897dd 100644 --- a/arch/mips/kernel/i8259.c +++ b/arch/mips/kernel/i8259.c @@ -42,9 +42,6 @@ static struct irq_chip i8259A_chip = { .irq_disable = disable_8259A_irq, .irq_unmask = enable_8259A_irq, .irq_mask_ack = mask_and_ack_8259A, -#ifdef CONFIG_MIPS_MT_SMTC_IRQAFF - .irq_set_affinity = plat_set_irq_affinity, -#endif /* CONFIG_MIPS_MT_SMTC_IRQAFF */ }; /* @@ -180,7 +177,6 @@ handle_real_irq: outb(cached_master_mask, PIC_MASTER_IMR); outb(0x60+irq, PIC_MASTER_CMD); /* 'Specific EOI to master */ } - smtc_im_ack_irq(irq); raw_spin_unlock_irqrestore(&i8259A_lock, flags); return; diff --git a/arch/mips/kernel/idle.c b/arch/mips/kernel/idle.c index 837ff27950b..09ce4598075 100644 --- a/arch/mips/kernel/idle.c +++ b/arch/mips/kernel/idle.c @@ -224,29 +224,26 @@ void __init check_wait(void) cpu_wait = r4k_wait; */ break; - case CPU_RM9000: - if ((c->processor_id & 0x00ff) >= 0x40) - cpu_wait = r4k_wait; - break; default: break; } } -static void smtc_idle_hook(void) -{ -#ifdef CONFIG_MIPS_MT_SMTC - void smtc_idle_loop_hook(void); - - smtc_idle_loop_hook(); -#endif -} - void arch_cpu_idle(void) { - smtc_idle_hook(); if (cpu_wait) cpu_wait(); else local_irq_enable(); } + +#ifdef CONFIG_CPU_IDLE + +int mips_cpuidle_wait_enter(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) +{ + arch_cpu_idle(); + return index; +} + +#endif diff --git a/arch/mips/kernel/irq-gic.c b/arch/mips/kernel/irq-gic.c index 8520dad6d4e..88e4c323382 100644 --- a/arch/mips/kernel/irq-gic.c +++ b/arch/mips/kernel/irq-gic.c @@ -54,6 +54,21 @@ void gic_write_compare(cycle_t cnt) (int)(cnt & 0xffffffff)); } +void gic_write_cpu_compare(cycle_t cnt, int cpu) +{ + unsigned long flags; + + local_irq_save(flags); + + GICWRITE(GIC_REG(VPE_LOCAL, GIC_VPE_OTHER_ADDR), cpu); + GICWRITE(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_HI), + (int)(cnt >> 32)); + GICWRITE(GIC_REG(VPE_OTHER, GIC_VPE_COMPARE_LO), + (int)(cnt & 0xffffffff)); + + local_irq_restore(flags); +} + cycle_t gic_read_compare(void) { unsigned int hi, lo; diff --git a/arch/mips/kernel/irq-msc01.c b/arch/mips/kernel/irq-msc01.c index fab40f7d2e0..4858642d543 100644 --- a/arch/mips/kernel/irq-msc01.c +++ b/arch/mips/kernel/irq-msc01.c @@ -53,13 +53,9 @@ static inline void unmask_msc_irq(struct irq_data *d) */ static void level_mask_and_ack_msc_irq(struct irq_data *d) { - unsigned int irq = d->irq; - mask_msc_irq(d); if (!cpu_has_veic) MSCIC_WRITE(MSC01_IC_EOI, 0); - /* This actually needs to be a call into platform code */ - smtc_im_ack_irq(irq); } /* @@ -78,7 +74,6 @@ static void edge_mask_and_ack_msc_irq(struct irq_data *d) MSCIC_WRITE(MSC01_IC_SUP+irq*8, r | ~MSC01_IC_SUP_EDGE_BIT); MSCIC_WRITE(MSC01_IC_SUP+irq*8, r); } - smtc_im_ack_irq(irq); } /* diff --git a/arch/mips/kernel/irq.c b/arch/mips/kernel/irq.c index 1818da4dbb8..d2bfbc2e899 100644 --- a/arch/mips/kernel/irq.c +++ b/arch/mips/kernel/irq.c @@ -73,7 +73,6 @@ void free_irqno(unsigned int irq) */ void ack_bad_irq(unsigned int irq) { - smtc_im_ack_irq(irq); printk("unexpected IRQ # %d\n", irq); } @@ -142,23 +141,7 @@ void __irq_entry do_IRQ(unsigned int irq) { irq_enter(); check_stack_overflow(); - if (!smtc_handle_on_other_cpu(irq)) - generic_handle_irq(irq); - irq_exit(); -} - -#ifdef CONFIG_MIPS_MT_SMTC_IRQAFF -/* - * To avoid inefficient and in some cases pathological re-checking of - * IRQ affinity, we have this variant that skips the affinity check. - */ - -void __irq_entry do_IRQ_no_affinity(unsigned int irq) -{ - irq_enter(); - smtc_im_backstop(irq); generic_handle_irq(irq); irq_exit(); } -#endif /* CONFIG_MIPS_MT_SMTC_IRQAFF */ diff --git a/arch/mips/kernel/mips-cpc.c b/arch/mips/kernel/mips-cpc.c index c9dc6740296..ba473608a34 100644 --- a/arch/mips/kernel/mips-cpc.c +++ b/arch/mips/kernel/mips-cpc.c @@ -9,12 +9,18 @@ */ #include <linux/errno.h> +#include <linux/percpu.h> +#include <linux/spinlock.h> #include <asm/mips-cm.h> #include <asm/mips-cpc.h> void __iomem *mips_cpc_base; +static DEFINE_PER_CPU_ALIGNED(spinlock_t, cpc_core_lock); + +static DEFINE_PER_CPU_ALIGNED(unsigned long, cpc_core_lock_flags); + phys_t __weak mips_cpc_phys_base(void) { u32 cpc_base; @@ -39,6 +45,10 @@ phys_t __weak mips_cpc_phys_base(void) int mips_cpc_probe(void) { phys_t addr; + unsigned cpu; + + for_each_possible_cpu(cpu) + spin_lock_init(&per_cpu(cpc_core_lock, cpu)); addr = mips_cpc_phys_base(); if (!addr) @@ -50,3 +60,21 @@ int mips_cpc_probe(void) return 0; } + +void mips_cpc_lock_other(unsigned int core) +{ + unsigned curr_core; + preempt_disable(); + curr_core = current_cpu_data.core; + spin_lock_irqsave(&per_cpu(cpc_core_lock, curr_core), + per_cpu(cpc_core_lock_flags, curr_core)); + write_cpc_cl_other(core << CPC_Cx_OTHER_CORENUM_SHF); +} + +void mips_cpc_unlock_other(void) +{ + unsigned curr_core = current_cpu_data.core; + spin_unlock_irqrestore(&per_cpu(cpc_core_lock, curr_core), + per_cpu(cpc_core_lock_flags, curr_core)); + preempt_enable(); +} diff --git a/arch/mips/kernel/mips-mt-fpaff.c b/arch/mips/kernel/mips-mt-fpaff.c index cb098628aee..362bb3707e6 100644 --- a/arch/mips/kernel/mips-mt-fpaff.c +++ b/arch/mips/kernel/mips-mt-fpaff.c @@ -1,5 +1,5 @@ /* - * General MIPS MT support routines, usable in AP/SP, SMVP, or SMTC kernels + * General MIPS MT support routines, usable in AP/SP and SMVP. * Copyright (C) 2005 Mips Technologies, Inc */ #include <linux/cpu.h> diff --git a/arch/mips/kernel/mips-mt.c b/arch/mips/kernel/mips-mt.c index 6ded9bd1489..88b1ef5f868 100644 --- a/arch/mips/kernel/mips-mt.c +++ b/arch/mips/kernel/mips-mt.c @@ -1,5 +1,5 @@ /* - * General MIPS MT support routines, usable in AP/SP, SMVP, or SMTC kernels + * General MIPS MT support routines, usable in AP/SP and SMVP. * Copyright (C) 2005 Mips Technologies, Inc */ @@ -57,9 +57,6 @@ void mips_mt_regdump(unsigned long mvpctl) int tc; unsigned long haltval; unsigned long tcstatval; -#ifdef CONFIG_MIPS_MT_SMTC - void smtc_soft_dump(void); -#endif /* CONFIG_MIPT_MT_SMTC */ local_irq_save(flags); vpflags = dvpe(); @@ -116,9 +113,6 @@ void mips_mt_regdump(unsigned long mvpctl) if (!haltval) write_tc_c0_tchalt(0); } -#ifdef CONFIG_MIPS_MT_SMTC - smtc_soft_dump(); -#endif /* CONFIG_MIPT_MT_SMTC */ printk("===========================\n"); evpe(vpflags); local_irq_restore(flags); @@ -295,21 +289,11 @@ void mips_mt_set_cpuoptions(void) void mt_cflush_lockdown(void) { -#ifdef CONFIG_MIPS_MT_SMTC - void smtc_cflush_lockdown(void); - - smtc_cflush_lockdown(); -#endif /* CONFIG_MIPS_MT_SMTC */ /* FILL IN VSMP and AP/SP VERSIONS HERE */ } void mt_cflush_release(void) { -#ifdef CONFIG_MIPS_MT_SMTC - void smtc_cflush_release(void); - - smtc_cflush_release(); -#endif /* CONFIG_MIPS_MT_SMTC */ /* FILL IN VSMP and AP/SP VERSIONS HERE */ } diff --git a/arch/mips/kernel/octeon_switch.S b/arch/mips/kernel/octeon_switch.S index 029e002a4ea..f6547680c81 100644 --- a/arch/mips/kernel/octeon_switch.S +++ b/arch/mips/kernel/octeon_switch.S @@ -10,24 +10,12 @@ * Copyright (C) 2000 MIPS Technologies, Inc. * written by Carsten Langgaard, carstenl@mips.com */ -#include <asm/asm.h> -#include <asm/cachectl.h> -#include <asm/fpregdef.h> -#include <asm/mipsregs.h> -#include <asm/asm-offsets.h> -#include <asm/pgtable-bits.h> -#include <asm/regdef.h> -#include <asm/stackframe.h> -#include <asm/thread_info.h> - -#include <asm/asmmacro.h> - -/* - * Offset to the current process status flags, the first 32 bytes of the - * stack are not used. - */ -#define ST_OFF (_THREAD_SIZE - 32 - PT_SIZE + PT_STATUS) +#define USE_ALTERNATE_RESUME_IMPL 1 + .set push + .set arch=mips64r2 +#include "r4k_switch.S" + .set pop /* * task_struct *resume(task_struct *prev, task_struct *next, * struct thread_info *next_ti, int usedfpu) @@ -40,6 +28,61 @@ cpu_save_nonscratch a0 LONG_S ra, THREAD_REG31(a0) + /* + * check if we need to save FPU registers + */ + PTR_L t3, TASK_THREAD_INFO(a0) + LONG_L t0, TI_FLAGS(t3) + li t1, _TIF_USEDFPU + and t2, t0, t1 + beqz t2, 1f + nor t1, zero, t1 + + and t0, t0, t1 + LONG_S t0, TI_FLAGS(t3) + + /* + * clear saved user stack CU1 bit + */ + LONG_L t0, ST_OFF(t3) + li t1, ~ST0_CU1 + and t0, t0, t1 + LONG_S t0, ST_OFF(t3) + + .set push + .set arch=mips64r2 + fpu_save_double a0 t0 t1 # c0_status passed in t0 + # clobbers t1 + .set pop +1: + + /* check if we need to save COP2 registers */ + PTR_L t2, TASK_THREAD_INFO(a0) + LONG_L t0, ST_OFF(t2) + bbit0 t0, 30, 1f + + /* Disable COP2 in the stored process state */ + li t1, ST0_CU2 + xor t0, t1 + LONG_S t0, ST_OFF(t2) + + /* Enable COP2 so we can save it */ + mfc0 t0, CP0_STATUS + or t0, t1 + mtc0 t0, CP0_STATUS + + /* Save COP2 */ + daddu a0, THREAD_CP2 + jal octeon_cop2_save + dsubu a0, THREAD_CP2 + + /* Disable COP2 now that we are done */ + mfc0 t0, CP0_STATUS + li t1, ST0_CU2 + xor t0, t1 + mtc0 t0, CP0_STATUS + +1: #if CONFIG_CAVIUM_OCTEON_CVMSEG_SIZE > 0 /* Check if we need to store CVMSEG state */ mfc0 t0, $11,7 /* CvmMemCtl */ @@ -85,12 +128,7 @@ move $28, a2 cpu_restore_nonscratch a1 -#if (_THREAD_SIZE - 32) < 0x8000 - PTR_ADDIU t0, $28, _THREAD_SIZE - 32 -#else - PTR_LI t0, _THREAD_SIZE - 32 - PTR_ADDU t0, $28 -#endif + PTR_ADDU t0, $28, _THREAD_SIZE - 32 set_saved_sp t0, t1, t2 mfc0 t1, CP0_STATUS /* Do we really need this? */ diff --git a/arch/mips/kernel/pm-cps.c b/arch/mips/kernel/pm-cps.c new file mode 100644 index 00000000000..5aa4c6f8cf8 --- /dev/null +++ b/arch/mips/kernel/pm-cps.c @@ -0,0 +1,716 @@ +/* + * Copyright (C) 2014 Imagination Technologies + * Author: Paul Burton <paul.burton@imgtec.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + */ + +#include <linux/init.h> +#include <linux/percpu.h> +#include <linux/slab.h> + +#include <asm/asm-offsets.h> +#include <asm/cacheflush.h> +#include <asm/cacheops.h> +#include <asm/idle.h> +#include <asm/mips-cm.h> +#include <asm/mips-cpc.h> +#include <asm/mipsmtregs.h> +#include <asm/pm.h> +#include <asm/pm-cps.h> +#include <asm/smp-cps.h> +#include <asm/uasm.h> + +/* + * cps_nc_entry_fn - type of a generated non-coherent state entry function + * @online: the count of online coupled VPEs + * @nc_ready_count: pointer to a non-coherent mapping of the core ready_count + * + * The code entering & exiting non-coherent states is generated at runtime + * using uasm, in order to ensure that the compiler cannot insert a stray + * memory access at an unfortunate time and to allow the generation of optimal + * core-specific code particularly for cache routines. If coupled_coherence + * is non-zero and this is the entry function for the CPS_PM_NC_WAIT state, + * returns the number of VPEs that were in the wait state at the point this + * VPE left it. Returns garbage if coupled_coherence is zero or this is not + * the entry function for CPS_PM_NC_WAIT. + */ +typedef unsigned (*cps_nc_entry_fn)(unsigned online, u32 *nc_ready_count); + +/* + * The entry point of the generated non-coherent idle state entry/exit + * functions. Actually per-core rather than per-CPU. + */ +static DEFINE_PER_CPU_READ_MOSTLY(cps_nc_entry_fn[CPS_PM_STATE_COUNT], + nc_asm_enter); + +/* Bitmap indicating which states are supported by the system */ +DECLARE_BITMAP(state_support, CPS_PM_STATE_COUNT); + +/* + * Indicates the number of coupled VPEs ready to operate in a non-coherent + * state. Actually per-core rather than per-CPU. + */ +static DEFINE_PER_CPU_ALIGNED(u32*, ready_count); +static DEFINE_PER_CPU_ALIGNED(void*, ready_count_alloc); + +/* Indicates online CPUs coupled with the current CPU */ +static DEFINE_PER_CPU_ALIGNED(cpumask_t, online_coupled); + +/* + * Used to synchronize entry to deep idle states. Actually per-core rather + * than per-CPU. + */ +static DEFINE_PER_CPU_ALIGNED(atomic_t, pm_barrier); + +/* Saved CPU state across the CPS_PM_POWER_GATED state */ +DEFINE_PER_CPU_ALIGNED(struct mips_static_suspend_state, cps_cpu_state); + +/* A somewhat arbitrary number of labels & relocs for uasm */ +static struct uasm_label labels[32] __initdata; +static struct uasm_reloc relocs[32] __initdata; + +/* CPU dependant sync types */ +static unsigned stype_intervention; +static unsigned stype_memory; +static unsigned stype_ordering; + +enum mips_reg { + zero, at, v0, v1, a0, a1, a2, a3, + t0, t1, t2, t3, t4, t5, t6, t7, + s0, s1, s2, s3, s4, s5, s6, s7, + t8, t9, k0, k1, gp, sp, fp, ra, +}; + +bool cps_pm_support_state(enum cps_pm_state state) +{ + return test_bit(state, state_support); +} + +static void coupled_barrier(atomic_t *a, unsigned online) +{ + /* + * This function is effectively the same as + * cpuidle_coupled_parallel_barrier, which can't be used here since + * there's no cpuidle device. + */ + + if (!coupled_coherence) + return; + + smp_mb__before_atomic_inc(); + atomic_inc(a); + + while (atomic_read(a) < online) + cpu_relax(); + + if (atomic_inc_return(a) == online * 2) { + atomic_set(a, 0); + return; + } + + while (atomic_read(a) > online) + cpu_relax(); +} + +int cps_pm_enter_state(enum cps_pm_state state) +{ + unsigned cpu = smp_processor_id(); + unsigned core = current_cpu_data.core; + unsigned online, left; + cpumask_t *coupled_mask = this_cpu_ptr(&online_coupled); + u32 *core_ready_count, *nc_core_ready_count; + void *nc_addr; + cps_nc_entry_fn entry; + struct core_boot_config *core_cfg; + struct vpe_boot_config *vpe_cfg; + + /* Check that there is an entry function for this state */ + entry = per_cpu(nc_asm_enter, core)[state]; + if (!entry) + return -EINVAL; + + /* Calculate which coupled CPUs (VPEs) are online */ +#ifdef CONFIG_MIPS_MT + if (cpu_online(cpu)) { + cpumask_and(coupled_mask, cpu_online_mask, + &cpu_sibling_map[cpu]); + online = cpumask_weight(coupled_mask); + cpumask_clear_cpu(cpu, coupled_mask); + } else +#endif + { + cpumask_clear(coupled_mask); + online = 1; + } + + /* Setup the VPE to run mips_cps_pm_restore when started again */ + if (config_enabled(CONFIG_CPU_PM) && state == CPS_PM_POWER_GATED) { + core_cfg = &mips_cps_core_bootcfg[core]; + vpe_cfg = &core_cfg->vpe_config[current_cpu_data.vpe_id]; + vpe_cfg->pc = (unsigned long)mips_cps_pm_restore; + vpe_cfg->gp = (unsigned long)current_thread_info(); + vpe_cfg->sp = 0; + } + + /* Indicate that this CPU might not be coherent */ + cpumask_clear_cpu(cpu, &cpu_coherent_mask); + smp_mb__after_clear_bit(); + + /* Create a non-coherent mapping of the core ready_count */ + core_ready_count = per_cpu(ready_count, core); + nc_addr = kmap_noncoherent(virt_to_page(core_ready_count), + (unsigned long)core_ready_count); + nc_addr += ((unsigned long)core_ready_count & ~PAGE_MASK); + nc_core_ready_count = nc_addr; + + /* Ensure ready_count is zero-initialised before the assembly runs */ + ACCESS_ONCE(*nc_core_ready_count) = 0; + coupled_barrier(&per_cpu(pm_barrier, core), online); + + /* Run the generated entry code */ + left = entry(online, nc_core_ready_count); + + /* Remove the non-coherent mapping of ready_count */ + kunmap_noncoherent(); + + /* Indicate that this CPU is definitely coherent */ + cpumask_set_cpu(cpu, &cpu_coherent_mask); + + /* + * If this VPE is the first to leave the non-coherent wait state then + * it needs to wake up any coupled VPEs still running their wait + * instruction so that they return to cpuidle, which can then complete + * coordination between the coupled VPEs & provide the governor with + * a chance to reflect on the length of time the VPEs were in the + * idle state. + */ + if (coupled_coherence && (state == CPS_PM_NC_WAIT) && (left == online)) + arch_send_call_function_ipi_mask(coupled_mask); + + return 0; +} + +static void __init cps_gen_cache_routine(u32 **pp, struct uasm_label **pl, + struct uasm_reloc **pr, + const struct cache_desc *cache, + unsigned op, int lbl) +{ + unsigned cache_size = cache->ways << cache->waybit; + unsigned i; + const unsigned unroll_lines = 32; + + /* If the cache isn't present this function has it easy */ + if (cache->flags & MIPS_CACHE_NOT_PRESENT) + return; + + /* Load base address */ + UASM_i_LA(pp, t0, (long)CKSEG0); + + /* Calculate end address */ + if (cache_size < 0x8000) + uasm_i_addiu(pp, t1, t0, cache_size); + else + UASM_i_LA(pp, t1, (long)(CKSEG0 + cache_size)); + + /* Start of cache op loop */ + uasm_build_label(pl, *pp, lbl); + + /* Generate the cache ops */ + for (i = 0; i < unroll_lines; i++) + uasm_i_cache(pp, op, i * cache->linesz, t0); + + /* Update the base address */ + uasm_i_addiu(pp, t0, t0, unroll_lines * cache->linesz); + + /* Loop if we haven't reached the end address yet */ + uasm_il_bne(pp, pr, t0, t1, lbl); + uasm_i_nop(pp); +} + +static int __init cps_gen_flush_fsb(u32 **pp, struct uasm_label **pl, + struct uasm_reloc **pr, + const struct cpuinfo_mips *cpu_info, + int lbl) +{ + unsigned i, fsb_size = 8; + unsigned num_loads = (fsb_size * 3) / 2; + unsigned line_stride = 2; + unsigned line_size = cpu_info->dcache.linesz; + unsigned perf_counter, perf_event; + unsigned revision = cpu_info->processor_id & PRID_REV_MASK; + + /* + * Determine whether this CPU requires an FSB flush, and if so which + * performance counter/event reflect stalls due to a full FSB. + */ + switch (__get_cpu_type(cpu_info->cputype)) { + case CPU_INTERAPTIV: + perf_counter = 1; + perf_event = 51; + break; + + case CPU_PROAPTIV: + /* Newer proAptiv cores don't require this workaround */ + if (revision >= PRID_REV_ENCODE_332(1, 1, 0)) + return 0; + + /* On older ones it's unavailable */ + return -1; + + /* CPUs which do not require the workaround */ + case CPU_P5600: + return 0; + + default: + WARN_ONCE(1, "pm-cps: FSB flush unsupported for this CPU\n"); + return -1; + } + + /* + * Ensure that the fill/store buffer (FSB) is not holding the results + * of a prefetch, since if it is then the CPC sequencer may become + * stuck in the D3 (ClrBus) state whilst entering a low power state. + */ + + /* Preserve perf counter setup */ + uasm_i_mfc0(pp, t2, 25, (perf_counter * 2) + 0); /* PerfCtlN */ + uasm_i_mfc0(pp, t3, 25, (perf_counter * 2) + 1); /* PerfCntN */ + + /* Setup perf counter to count FSB full pipeline stalls */ + uasm_i_addiu(pp, t0, zero, (perf_event << 5) | 0xf); + uasm_i_mtc0(pp, t0, 25, (perf_counter * 2) + 0); /* PerfCtlN */ + uasm_i_ehb(pp); + uasm_i_mtc0(pp, zero, 25, (perf_counter * 2) + 1); /* PerfCntN */ + uasm_i_ehb(pp); + + /* Base address for loads */ + UASM_i_LA(pp, t0, (long)CKSEG0); + + /* Start of clear loop */ + uasm_build_label(pl, *pp, lbl); + + /* Perform some loads to fill the FSB */ + for (i = 0; i < num_loads; i++) + uasm_i_lw(pp, zero, i * line_size * line_stride, t0); + + /* + * Invalidate the new D-cache entries so that the cache will need + * refilling (via the FSB) if the loop is executed again. + */ + for (i = 0; i < num_loads; i++) { + uasm_i_cache(pp, Hit_Invalidate_D, + i * line_size * line_stride, t0); + uasm_i_cache(pp, Hit_Writeback_Inv_SD, + i * line_size * line_stride, t0); + } + + /* Completion barrier */ + uasm_i_sync(pp, stype_memory); + uasm_i_ehb(pp); + + /* Check whether the pipeline stalled due to the FSB being full */ + uasm_i_mfc0(pp, t1, 25, (perf_counter * 2) + 1); /* PerfCntN */ + + /* Loop if it didn't */ + uasm_il_beqz(pp, pr, t1, lbl); + uasm_i_nop(pp); + + /* Restore perf counter 1. The count may well now be wrong... */ + uasm_i_mtc0(pp, t2, 25, (perf_counter * 2) + 0); /* PerfCtlN */ + uasm_i_ehb(pp); + uasm_i_mtc0(pp, t3, 25, (perf_counter * 2) + 1); /* PerfCntN */ + uasm_i_ehb(pp); + + return 0; +} + +static void __init cps_gen_set_top_bit(u32 **pp, struct uasm_label **pl, + struct uasm_reloc **pr, + unsigned r_addr, int lbl) +{ + uasm_i_lui(pp, t0, uasm_rel_hi(0x80000000)); + uasm_build_label(pl, *pp, lbl); + uasm_i_ll(pp, t1, 0, r_addr); + uasm_i_or(pp, t1, t1, t0); + uasm_i_sc(pp, t1, 0, r_addr); + uasm_il_beqz(pp, pr, t1, lbl); + uasm_i_nop(pp); +} + +static void * __init cps_gen_entry_code(unsigned cpu, enum cps_pm_state state) +{ + struct uasm_label *l = labels; + struct uasm_reloc *r = relocs; + u32 *buf, *p; + const unsigned r_online = a0; + const unsigned r_nc_count = a1; + const unsigned r_pcohctl = t7; + const unsigned max_instrs = 256; + unsigned cpc_cmd; + int err; + enum { + lbl_incready = 1, + lbl_poll_cont, + lbl_secondary_hang, + lbl_disable_coherence, + lbl_flush_fsb, + lbl_invicache, + lbl_flushdcache, + lbl_hang, + lbl_set_cont, + lbl_secondary_cont, + lbl_decready, + }; + + /* Allocate a buffer to hold the generated code */ + p = buf = kcalloc(max_instrs, sizeof(u32), GFP_KERNEL); + if (!buf) + return NULL; + + /* Clear labels & relocs ready for (re)use */ + memset(labels, 0, sizeof(labels)); + memset(relocs, 0, sizeof(relocs)); + + if (config_enabled(CONFIG_CPU_PM) && state == CPS_PM_POWER_GATED) { + /* + * Save CPU state. Note the non-standard calling convention + * with the return address placed in v0 to avoid clobbering + * the ra register before it is saved. + */ + UASM_i_LA(&p, t0, (long)mips_cps_pm_save); + uasm_i_jalr(&p, v0, t0); + uasm_i_nop(&p); + } + + /* + * Load addresses of required CM & CPC registers. This is done early + * because they're needed in both the enable & disable coherence steps + * but in the coupled case the enable step will only run on one VPE. + */ + UASM_i_LA(&p, r_pcohctl, (long)addr_gcr_cl_coherence()); + + if (coupled_coherence) { + /* Increment ready_count */ + uasm_i_sync(&p, stype_ordering); + uasm_build_label(&l, p, lbl_incready); + uasm_i_ll(&p, t1, 0, r_nc_count); + uasm_i_addiu(&p, t2, t1, 1); + uasm_i_sc(&p, t2, 0, r_nc_count); + uasm_il_beqz(&p, &r, t2, lbl_incready); + uasm_i_addiu(&p, t1, t1, 1); + + /* Ordering barrier */ + uasm_i_sync(&p, stype_ordering); + + /* + * If this is the last VPE to become ready for non-coherence + * then it should branch below. + */ + uasm_il_beq(&p, &r, t1, r_online, lbl_disable_coherence); + uasm_i_nop(&p); + + if (state < CPS_PM_POWER_GATED) { + /* + * Otherwise this is not the last VPE to become ready + * for non-coherence. It needs to wait until coherence + * has been disabled before proceeding, which it will do + * by polling for the top bit of ready_count being set. + */ + uasm_i_addiu(&p, t1, zero, -1); + uasm_build_label(&l, p, lbl_poll_cont); + uasm_i_lw(&p, t0, 0, r_nc_count); + uasm_il_bltz(&p, &r, t0, lbl_secondary_cont); + uasm_i_ehb(&p); + uasm_i_yield(&p, zero, t1); + uasm_il_b(&p, &r, lbl_poll_cont); + uasm_i_nop(&p); + } else { + /* + * The core will lose power & this VPE will not continue + * so it can simply halt here. + */ + uasm_i_addiu(&p, t0, zero, TCHALT_H); + uasm_i_mtc0(&p, t0, 2, 4); + uasm_build_label(&l, p, lbl_secondary_hang); + uasm_il_b(&p, &r, lbl_secondary_hang); + uasm_i_nop(&p); + } + } + + /* + * This is the point of no return - this VPE will now proceed to + * disable coherence. At this point we *must* be sure that no other + * VPE within the core will interfere with the L1 dcache. + */ + uasm_build_label(&l, p, lbl_disable_coherence); + + /* Invalidate the L1 icache */ + cps_gen_cache_routine(&p, &l, &r, &cpu_data[cpu].icache, + Index_Invalidate_I, lbl_invicache); + + /* Writeback & invalidate the L1 dcache */ + cps_gen_cache_routine(&p, &l, &r, &cpu_data[cpu].dcache, + Index_Writeback_Inv_D, lbl_flushdcache); + + /* Completion barrier */ + uasm_i_sync(&p, stype_memory); + uasm_i_ehb(&p); + + /* + * Disable all but self interventions. The load from COHCTL is defined + * by the interAptiv & proAptiv SUMs as ensuring that the operation + * resulting from the preceeding store is complete. + */ + uasm_i_addiu(&p, t0, zero, 1 << cpu_data[cpu].core); + uasm_i_sw(&p, t0, 0, r_pcohctl); + uasm_i_lw(&p, t0, 0, r_pcohctl); + + /* Sync to ensure previous interventions are complete */ + uasm_i_sync(&p, stype_intervention); + uasm_i_ehb(&p); + + /* Disable coherence */ + uasm_i_sw(&p, zero, 0, r_pcohctl); + uasm_i_lw(&p, t0, 0, r_pcohctl); + + if (state >= CPS_PM_CLOCK_GATED) { + err = cps_gen_flush_fsb(&p, &l, &r, &cpu_data[cpu], + lbl_flush_fsb); + if (err) + goto out_err; + + /* Determine the CPC command to issue */ + switch (state) { + case CPS_PM_CLOCK_GATED: + cpc_cmd = CPC_Cx_CMD_CLOCKOFF; + break; + case CPS_PM_POWER_GATED: + cpc_cmd = CPC_Cx_CMD_PWRDOWN; + break; + default: + BUG(); + goto out_err; + } + + /* Issue the CPC command */ + UASM_i_LA(&p, t0, (long)addr_cpc_cl_cmd()); + uasm_i_addiu(&p, t1, zero, cpc_cmd); + uasm_i_sw(&p, t1, 0, t0); + + if (state == CPS_PM_POWER_GATED) { + /* If anything goes wrong just hang */ + uasm_build_label(&l, p, lbl_hang); + uasm_il_b(&p, &r, lbl_hang); + uasm_i_nop(&p); + + /* + * There's no point generating more code, the core is + * powered down & if powered back up will run from the + * reset vector not from here. + */ + goto gen_done; + } + + /* Completion barrier */ + uasm_i_sync(&p, stype_memory); + uasm_i_ehb(&p); + } + + if (state == CPS_PM_NC_WAIT) { + /* + * At this point it is safe for all VPEs to proceed with + * execution. This VPE will set the top bit of ready_count + * to indicate to the other VPEs that they may continue. + */ + if (coupled_coherence) + cps_gen_set_top_bit(&p, &l, &r, r_nc_count, + lbl_set_cont); + + /* + * VPEs which did not disable coherence will continue + * executing, after coherence has been disabled, from this + * point. + */ + uasm_build_label(&l, p, lbl_secondary_cont); + + /* Now perform our wait */ + uasm_i_wait(&p, 0); + } + + /* + * Re-enable coherence. Note that for CPS_PM_NC_WAIT all coupled VPEs + * will run this. The first will actually re-enable coherence & the + * rest will just be performing a rather unusual nop. + */ + uasm_i_addiu(&p, t0, zero, CM_GCR_Cx_COHERENCE_COHDOMAINEN_MSK); + uasm_i_sw(&p, t0, 0, r_pcohctl); + uasm_i_lw(&p, t0, 0, r_pcohctl); + + /* Completion barrier */ + uasm_i_sync(&p, stype_memory); + uasm_i_ehb(&p); + + if (coupled_coherence && (state == CPS_PM_NC_WAIT)) { + /* Decrement ready_count */ + uasm_build_label(&l, p, lbl_decready); + uasm_i_sync(&p, stype_ordering); + uasm_i_ll(&p, t1, 0, r_nc_count); + uasm_i_addiu(&p, t2, t1, -1); + uasm_i_sc(&p, t2, 0, r_nc_count); + uasm_il_beqz(&p, &r, t2, lbl_decready); + uasm_i_andi(&p, v0, t1, (1 << fls(smp_num_siblings)) - 1); + + /* Ordering barrier */ + uasm_i_sync(&p, stype_ordering); + } + + if (coupled_coherence && (state == CPS_PM_CLOCK_GATED)) { + /* + * At this point it is safe for all VPEs to proceed with + * execution. This VPE will set the top bit of ready_count + * to indicate to the other VPEs that they may continue. + */ + cps_gen_set_top_bit(&p, &l, &r, r_nc_count, lbl_set_cont); + + /* + * This core will be reliant upon another core sending a + * power-up command to the CPC in order to resume operation. + * Thus an arbitrary VPE can't trigger the core leaving the + * idle state and the one that disables coherence might as well + * be the one to re-enable it. The rest will continue from here + * after that has been done. + */ + uasm_build_label(&l, p, lbl_secondary_cont); + + /* Ordering barrier */ + uasm_i_sync(&p, stype_ordering); + } + + /* The core is coherent, time to return to C code */ + uasm_i_jr(&p, ra); + uasm_i_nop(&p); + +gen_done: + /* Ensure the code didn't exceed the resources allocated for it */ + BUG_ON((p - buf) > max_instrs); + BUG_ON((l - labels) > ARRAY_SIZE(labels)); + BUG_ON((r - relocs) > ARRAY_SIZE(relocs)); + + /* Patch branch offsets */ + uasm_resolve_relocs(relocs, labels); + + /* Flush the icache */ + local_flush_icache_range((unsigned long)buf, (unsigned long)p); + + return buf; +out_err: + kfree(buf); + return NULL; +} + +static int __init cps_gen_core_entries(unsigned cpu) +{ + enum cps_pm_state state; + unsigned core = cpu_data[cpu].core; + unsigned dlinesz = cpu_data[cpu].dcache.linesz; + void *entry_fn, *core_rc; + + for (state = CPS_PM_NC_WAIT; state < CPS_PM_STATE_COUNT; state++) { + if (per_cpu(nc_asm_enter, core)[state]) + continue; + if (!test_bit(state, state_support)) + continue; + + entry_fn = cps_gen_entry_code(cpu, state); + if (!entry_fn) { + pr_err("Failed to generate core %u state %u entry\n", + core, state); + clear_bit(state, state_support); + } + + per_cpu(nc_asm_enter, core)[state] = entry_fn; + } + + if (!per_cpu(ready_count, core)) { + core_rc = kmalloc(dlinesz * 2, GFP_KERNEL); + if (!core_rc) { + pr_err("Failed allocate core %u ready_count\n", core); + return -ENOMEM; + } + per_cpu(ready_count_alloc, core) = core_rc; + + /* Ensure ready_count is aligned to a cacheline boundary */ + core_rc += dlinesz - 1; + core_rc = (void *)((unsigned long)core_rc & ~(dlinesz - 1)); + per_cpu(ready_count, core) = core_rc; + } + + return 0; +} + +static int __init cps_pm_init(void) +{ + unsigned cpu; + int err; + + /* Detect appropriate sync types for the system */ + switch (current_cpu_data.cputype) { + case CPU_INTERAPTIV: + case CPU_PROAPTIV: + case CPU_M5150: + case CPU_P5600: + stype_intervention = 0x2; + stype_memory = 0x3; + stype_ordering = 0x10; + break; + + default: + pr_warn("Power management is using heavyweight sync 0\n"); + } + + /* A CM is required for all non-coherent states */ + if (!mips_cm_present()) { + pr_warn("pm-cps: no CM, non-coherent states unavailable\n"); + goto out; + } + + /* + * If interrupts were enabled whilst running a wait instruction on a + * non-coherent core then the VPE may end up processing interrupts + * whilst non-coherent. That would be bad. + */ + if (cpu_wait == r4k_wait_irqoff) + set_bit(CPS_PM_NC_WAIT, state_support); + else + pr_warn("pm-cps: non-coherent wait unavailable\n"); + + /* Detect whether a CPC is present */ + if (mips_cpc_present()) { + /* Detect whether clock gating is implemented */ + if (read_cpc_cl_stat_conf() & CPC_Cx_STAT_CONF_CLKGAT_IMPL_MSK) + set_bit(CPS_PM_CLOCK_GATED, state_support); + else + pr_warn("pm-cps: CPC does not support clock gating\n"); + + /* Power gating is available with CPS SMP & any CPC */ + if (mips_cps_smp_in_use()) + set_bit(CPS_PM_POWER_GATED, state_support); + else + pr_warn("pm-cps: CPS SMP not in use, power gating unavailable\n"); + } else { + pr_warn("pm-cps: no CPC, clock & power gating unavailable\n"); + } + + for_each_present_cpu(cpu) { + err = cps_gen_core_entries(cpu); + if (err) + return err; + } +out: + return 0; +} +arch_initcall(cps_pm_init); diff --git a/arch/mips/kernel/pm.c b/arch/mips/kernel/pm.c new file mode 100644 index 00000000000..fefdf39d3df --- /dev/null +++ b/arch/mips/kernel/pm.c @@ -0,0 +1,99 @@ +/* + * Copyright (C) 2014 Imagination Technologies Ltd. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * CPU PM notifiers for saving/restoring general CPU state. + */ + +#include <linux/cpu_pm.h> +#include <linux/init.h> + +#include <asm/dsp.h> +#include <asm/fpu.h> +#include <asm/mmu_context.h> +#include <asm/pm.h> +#include <asm/watch.h> + +/* Used by PM helper macros in asm/pm.h */ +struct mips_static_suspend_state mips_static_suspend_state; + +/** + * mips_cpu_save() - Save general CPU state. + * Ensures that general CPU context is saved, notably FPU and DSP. + */ +static int mips_cpu_save(void) +{ + /* Save FPU state */ + lose_fpu(1); + + /* Save DSP state */ + save_dsp(current); + + return 0; +} + +/** + * mips_cpu_restore() - Restore general CPU state. + * Restores important CPU context. + */ +static void mips_cpu_restore(void) +{ + unsigned int cpu = smp_processor_id(); + + /* Restore ASID */ + if (current->mm) + write_c0_entryhi(cpu_asid(cpu, current->mm)); + + /* Restore DSP state */ + restore_dsp(current); + + /* Restore UserLocal */ + if (cpu_has_userlocal) + write_c0_userlocal(current_thread_info()->tp_value); + + /* Restore watch registers */ + __restore_watch(); +} + +/** + * mips_pm_notifier() - Notifier for preserving general CPU context. + * @self: Notifier block. + * @cmd: CPU PM event. + * @v: Private data (unused). + * + * This is called when a CPU power management event occurs, and is used to + * ensure that important CPU context is preserved across a CPU power down. + */ +static int mips_pm_notifier(struct notifier_block *self, unsigned long cmd, + void *v) +{ + int ret; + + switch (cmd) { + case CPU_PM_ENTER: + ret = mips_cpu_save(); + if (ret) + return NOTIFY_STOP; + break; + case CPU_PM_ENTER_FAILED: + case CPU_PM_EXIT: + mips_cpu_restore(); + break; + } + + return NOTIFY_OK; +} + +static struct notifier_block mips_pm_notifier_block = { + .notifier_call = mips_pm_notifier, +}; + +static int __init mips_pm_init(void) +{ + return cpu_pm_register_notifier(&mips_pm_notifier_block); +} +arch_initcall(mips_pm_init); diff --git a/arch/mips/kernel/process.c b/arch/mips/kernel/process.c index 60e39dc7f1e..0a1ec0f3bef 100644 --- a/arch/mips/kernel/process.c +++ b/arch/mips/kernel/process.c @@ -140,13 +140,6 @@ int copy_thread(unsigned long clone_flags, unsigned long usp, */ childregs->cp0_status &= ~(ST0_CU2|ST0_CU1); -#ifdef CONFIG_MIPS_MT_SMTC - /* - * SMTC restores TCStatus after Status, and the CU bits - * are aliased there. - */ - childregs->cp0_tcstatus &= ~(ST0_CU2|ST0_CU1); -#endif clear_tsk_thread_flag(p, TIF_USEDFPU); #ifdef CONFIG_MIPS_MT_FPAFF diff --git a/arch/mips/kernel/r4k_switch.S b/arch/mips/kernel/r4k_switch.S index abacac7c33e..81ca3f70fe2 100644 --- a/arch/mips/kernel/r4k_switch.S +++ b/arch/mips/kernel/r4k_switch.S @@ -28,6 +28,7 @@ */ #define ST_OFF (_THREAD_SIZE - 32 - PT_SIZE + PT_STATUS) +#ifndef USE_ALTERNATE_RESUME_IMPL /* * task_struct *resume(task_struct *prev, task_struct *next, * struct thread_info *next_ti, s32 fp_save) @@ -87,18 +88,6 @@ PTR_ADDU t0, $28, _THREAD_SIZE - 32 set_saved_sp t0, t1, t2 -#ifdef CONFIG_MIPS_MT_SMTC - /* Read-modify-writes of Status must be atomic on a VPE */ - mfc0 t2, CP0_TCSTATUS - ori t1, t2, TCSTATUS_IXMT - mtc0 t1, CP0_TCSTATUS - andi t2, t2, TCSTATUS_IXMT - _ehb - DMT 8 # dmt t0 - move t1,ra - jal mips_ihb - move ra,t1 -#endif /* CONFIG_MIPS_MT_SMTC */ mfc0 t1, CP0_STATUS /* Do we really need this? */ li a3, 0xff01 and t1, a3 @@ -107,22 +96,12 @@ and a2, a3 or a2, t1 mtc0 a2, CP0_STATUS -#ifdef CONFIG_MIPS_MT_SMTC - _ehb - andi t0, t0, VPECONTROL_TE - beqz t0, 1f - emt -1: - mfc0 t1, CP0_TCSTATUS - xori t1, t1, TCSTATUS_IXMT - or t1, t1, t2 - mtc0 t1, CP0_TCSTATUS - _ehb -#endif /* CONFIG_MIPS_MT_SMTC */ move v0, a0 jr ra END(resume) +#endif /* USE_ALTERNATE_RESUME_IMPL */ + /* * Save a thread's fp context. */ @@ -176,19 +155,10 @@ LEAF(_restore_msa) #define FPU_DEFAULT 0x00000000 LEAF(_init_fpu) -#ifdef CONFIG_MIPS_MT_SMTC - /* Rather than manipulate per-VPE Status, set per-TC bit in TCStatus */ - mfc0 t0, CP0_TCSTATUS - /* Bit position is the same for Status, TCStatus */ - li t1, ST0_CU1 - or t0, t1 - mtc0 t0, CP0_TCSTATUS -#else /* Normal MIPS CU1 enable */ mfc0 t0, CP0_STATUS li t1, ST0_CU1 or t0, t1 mtc0 t0, CP0_STATUS -#endif /* CONFIG_MIPS_MT_SMTC */ enable_fpu_hazard li t1, FPU_DEFAULT diff --git a/arch/mips/kernel/rtlx-mt.c b/arch/mips/kernel/rtlx-mt.c index 9c1aca00fd5..5a66b975989 100644 --- a/arch/mips/kernel/rtlx-mt.c +++ b/arch/mips/kernel/rtlx-mt.c @@ -36,7 +36,6 @@ static irqreturn_t rtlx_interrupt(int irq, void *dev_id) unsigned long flags; int i; - /* Ought not to be strictly necessary for SMTC builds */ local_irq_save(flags); vpeflags = dvpe(); set_c0_status(0x100 << MIPS_CPU_RTLX_IRQ); diff --git a/arch/mips/kernel/smp-bmips.c b/arch/mips/kernel/smp-bmips.c index ea4c2dc3169..df9e2bd9b2c 100644 --- a/arch/mips/kernel/smp-bmips.c +++ b/arch/mips/kernel/smp-bmips.c @@ -281,13 +281,6 @@ static void bmips_smp_finish(void) } /* - * Runs on CPU0 after all CPUs have been booted - */ -static void bmips_cpus_done(void) -{ -} - -/* * BMIPS5000 raceless IPIs * * Each CPU has two inbound SW IRQs which are independent of all other CPUs. @@ -434,7 +427,6 @@ struct plat_smp_ops bmips43xx_smp_ops = { .boot_secondary = bmips_boot_secondary, .smp_finish = bmips_smp_finish, .init_secondary = bmips_init_secondary, - .cpus_done = bmips_cpus_done, .send_ipi_single = bmips43xx_send_ipi_single, .send_ipi_mask = bmips43xx_send_ipi_mask, #ifdef CONFIG_HOTPLUG_CPU @@ -449,7 +441,6 @@ struct plat_smp_ops bmips5000_smp_ops = { .boot_secondary = bmips_boot_secondary, .smp_finish = bmips_smp_finish, .init_secondary = bmips_init_secondary, - .cpus_done = bmips_cpus_done, .send_ipi_single = bmips5000_send_ipi_single, .send_ipi_mask = bmips5000_send_ipi_mask, #ifdef CONFIG_HOTPLUG_CPU diff --git a/arch/mips/kernel/smp-cmp.c b/arch/mips/kernel/smp-cmp.c index 3ef55fb7ac0..fc8a5155342 100644 --- a/arch/mips/kernel/smp-cmp.c +++ b/arch/mips/kernel/smp-cmp.c @@ -49,14 +49,11 @@ static void cmp_init_secondary(void) /* Enable per-cpu interrupts: platform specific */ -#if defined(CONFIG_MIPS_MT_SMP) || defined(CONFIG_MIPS_MT_SMTC) +#ifdef CONFIG_MIPS_MT_SMP if (cpu_has_mipsmt) c->vpe_id = (read_c0_tcbind() >> TCBIND_CURVPE_SHIFT) & TCBIND_CURVPE; #endif -#ifdef CONFIG_MIPS_MT_SMTC - c->tc_id = (read_c0_tcbind() & TCBIND_CURTC) >> TCBIND_CURTC_SHIFT; -#endif } static void cmp_smp_finish(void) @@ -75,11 +72,6 @@ static void cmp_smp_finish(void) local_irq_enable(); } -static void cmp_cpus_done(void) -{ - pr_debug("SMPCMP: CPU%d: %s\n", smp_processor_id(), __func__); -} - /* * Setup the PC, SP, and GP of a secondary processor and start it running * smp_bootstrap is the place to resume from @@ -135,10 +127,6 @@ void __init cmp_smp_setup(void) unsigned int mvpconf0 = read_c0_mvpconf0(); nvpe = ((mvpconf0 & MVPCONF0_PVPE) >> MVPCONF0_PVPE_SHIFT) + 1; -#elif defined(CONFIG_MIPS_MT_SMTC) - unsigned int mvpconf0 = read_c0_mvpconf0(); - - nvpe = ((mvpconf0 & MVPCONF0_PTC) >> MVPCONF0_PTC_SHIFT) + 1; #endif smp_num_siblings = nvpe; } @@ -165,7 +153,6 @@ struct plat_smp_ops cmp_smp_ops = { .send_ipi_mask = gic_send_ipi_mask, .init_secondary = cmp_init_secondary, .smp_finish = cmp_smp_finish, - .cpus_done = cmp_cpus_done, .boot_secondary = cmp_boot_secondary, .smp_setup = cmp_smp_setup, .prepare_cpus = cmp_prepare_cpus, diff --git a/arch/mips/kernel/smp-cps.c b/arch/mips/kernel/smp-cps.c index 536eec0d21b..df0598d9bfd 100644 --- a/arch/mips/kernel/smp-cps.c +++ b/arch/mips/kernel/smp-cps.c @@ -20,104 +20,43 @@ #include <asm/mips-cpc.h> #include <asm/mips_mt.h> #include <asm/mipsregs.h> +#include <asm/pm-cps.h> #include <asm/smp-cps.h> #include <asm/time.h> #include <asm/uasm.h> static DECLARE_BITMAP(core_power, NR_CPUS); -struct boot_config mips_cps_bootcfg; +struct core_boot_config *mips_cps_core_bootcfg; -static void init_core(void) +static unsigned core_vpe_count(unsigned core) { - unsigned int nvpes, t; - u32 mvpconf0, vpeconf0, vpecontrol, tcstatus, tcbind, status; + unsigned cfg; - if (!cpu_has_mipsmt) - return; - - /* Enter VPE configuration state */ - dvpe(); - set_c0_mvpcontrol(MVPCONTROL_VPC); - - /* Retrieve the count of VPEs in this core */ - mvpconf0 = read_c0_mvpconf0(); - nvpes = ((mvpconf0 & MVPCONF0_PVPE) >> MVPCONF0_PVPE_SHIFT) + 1; - smp_num_siblings = nvpes; - - for (t = 1; t < nvpes; t++) { - /* Use a 1:1 mapping of TC index to VPE index */ - settc(t); - - /* Bind 1 TC to this VPE */ - tcbind = read_tc_c0_tcbind(); - tcbind &= ~TCBIND_CURVPE; - tcbind |= t << TCBIND_CURVPE_SHIFT; - write_tc_c0_tcbind(tcbind); - - /* Set exclusive TC, non-active, master */ - vpeconf0 = read_vpe_c0_vpeconf0(); - vpeconf0 &= ~(VPECONF0_XTC | VPECONF0_VPA); - vpeconf0 |= t << VPECONF0_XTC_SHIFT; - vpeconf0 |= VPECONF0_MVP; - write_vpe_c0_vpeconf0(vpeconf0); - - /* Declare TC non-active, non-allocatable & interrupt exempt */ - tcstatus = read_tc_c0_tcstatus(); - tcstatus &= ~(TCSTATUS_A | TCSTATUS_DA); - tcstatus |= TCSTATUS_IXMT; - write_tc_c0_tcstatus(tcstatus); - - /* Halt the TC */ - write_tc_c0_tchalt(TCHALT_H); - - /* Allow only 1 TC to execute */ - vpecontrol = read_vpe_c0_vpecontrol(); - vpecontrol &= ~VPECONTROL_TE; - write_vpe_c0_vpecontrol(vpecontrol); - - /* Copy (most of) Status from VPE 0 */ - status = read_c0_status(); - status &= ~(ST0_IM | ST0_IE | ST0_KSU); - status |= ST0_CU0; - write_vpe_c0_status(status); - - /* Copy Config from VPE 0 */ - write_vpe_c0_config(read_c0_config()); - write_vpe_c0_config7(read_c0_config7()); - - /* Ensure no software interrupts are pending */ - write_vpe_c0_cause(0); - - /* Sync Count */ - write_vpe_c0_count(read_c0_count()); - } + if (!config_enabled(CONFIG_MIPS_MT_SMP) || !cpu_has_mipsmt) + return 1; - /* Leave VPE configuration state */ - clear_c0_mvpcontrol(MVPCONTROL_VPC); + write_gcr_cl_other(core << CM_GCR_Cx_OTHER_CORENUM_SHF); + cfg = read_gcr_co_config() & CM_GCR_Cx_CONFIG_PVPE_MSK; + return (cfg >> CM_GCR_Cx_CONFIG_PVPE_SHF) + 1; } static void __init cps_smp_setup(void) { unsigned int ncores, nvpes, core_vpes; int c, v; - u32 core_cfg, *entry_code; /* Detect & record VPE topology */ ncores = mips_cm_numcores(); pr_info("VPE topology "); for (c = nvpes = 0; c < ncores; c++) { - if (cpu_has_mipsmt && config_enabled(CONFIG_MIPS_MT_SMP)) { - write_gcr_cl_other(c << CM_GCR_Cx_OTHER_CORENUM_SHF); - core_cfg = read_gcr_co_config(); - core_vpes = ((core_cfg & CM_GCR_Cx_CONFIG_PVPE_MSK) >> - CM_GCR_Cx_CONFIG_PVPE_SHF) + 1; - } else { - core_vpes = 1; - } - + core_vpes = core_vpe_count(c); pr_cont("%c%u", c ? ',' : '{', core_vpes); + /* Use the number of VPEs in core 0 for smp_num_siblings */ + if (!c) + smp_num_siblings = core_vpes; + for (v = 0; v < min_t(int, core_vpes, NR_CPUS - nvpes); v++) { cpu_data[nvpes + v].core = c; #ifdef CONFIG_MIPS_MT_SMP @@ -137,19 +76,14 @@ static void __init cps_smp_setup(void) __cpu_logical_map[v] = v; } + /* Set a coherent default CCA (CWB) */ + change_c0_config(CONF_CM_CMASK, 0x5); + /* Core 0 is powered up (we're running on it) */ bitmap_set(core_power, 0, 1); - /* Disable MT - we only want to run 1 TC per VPE */ - if (cpu_has_mipsmt) - dmt(); - /* Initialise core 0 */ - init_core(); - - /* Patch the start of mips_cps_core_entry to provide the CM base */ - entry_code = (u32 *)&mips_cps_core_entry; - UASM_i_LA(&entry_code, 3, (long)mips_cm_base); + mips_cps_core_init(); /* Make core 0 coherent with everything */ write_gcr_cl_coherence(0xff); @@ -157,15 +91,99 @@ static void __init cps_smp_setup(void) static void __init cps_prepare_cpus(unsigned int max_cpus) { + unsigned ncores, core_vpes, c, cca; + bool cca_unsuitable; + u32 *entry_code; + mips_mt_set_cpuoptions(); + + /* Detect whether the CCA is unsuited to multi-core SMP */ + cca = read_c0_config() & CONF_CM_CMASK; + switch (cca) { + case 0x4: /* CWBE */ + case 0x5: /* CWB */ + /* The CCA is coherent, multi-core is fine */ + cca_unsuitable = false; + break; + + default: + /* CCA is not coherent, multi-core is not usable */ + cca_unsuitable = true; + } + + /* Warn the user if the CCA prevents multi-core */ + ncores = mips_cm_numcores(); + if (cca_unsuitable && ncores > 1) { + pr_warn("Using only one core due to unsuitable CCA 0x%x\n", + cca); + + for_each_present_cpu(c) { + if (cpu_data[c].core) + set_cpu_present(c, false); + } + } + + /* + * Patch the start of mips_cps_core_entry to provide: + * + * v0 = CM base address + * s0 = kseg0 CCA + */ + entry_code = (u32 *)&mips_cps_core_entry; + UASM_i_LA(&entry_code, 3, (long)mips_cm_base); + uasm_i_addiu(&entry_code, 16, 0, cca); + dma_cache_wback_inv((unsigned long)&mips_cps_core_entry, + (void *)entry_code - (void *)&mips_cps_core_entry); + + /* Allocate core boot configuration structs */ + mips_cps_core_bootcfg = kcalloc(ncores, sizeof(*mips_cps_core_bootcfg), + GFP_KERNEL); + if (!mips_cps_core_bootcfg) { + pr_err("Failed to allocate boot config for %u cores\n", ncores); + goto err_out; + } + + /* Allocate VPE boot configuration structs */ + for (c = 0; c < ncores; c++) { + core_vpes = core_vpe_count(c); + mips_cps_core_bootcfg[c].vpe_config = kcalloc(core_vpes, + sizeof(*mips_cps_core_bootcfg[c].vpe_config), + GFP_KERNEL); + if (!mips_cps_core_bootcfg[c].vpe_config) { + pr_err("Failed to allocate %u VPE boot configs\n", + core_vpes); + goto err_out; + } + } + + /* Mark this CPU as booted */ + atomic_set(&mips_cps_core_bootcfg[current_cpu_data.core].vpe_mask, + 1 << cpu_vpe_id(¤t_cpu_data)); + + return; +err_out: + /* Clean up allocations */ + if (mips_cps_core_bootcfg) { + for (c = 0; c < ncores; c++) + kfree(mips_cps_core_bootcfg[c].vpe_config); + kfree(mips_cps_core_bootcfg); + mips_cps_core_bootcfg = NULL; + } + + /* Effectively disable SMP by declaring CPUs not present */ + for_each_possible_cpu(c) { + if (c == 0) + continue; + set_cpu_present(c, false); + } } -static void boot_core(struct boot_config *cfg) +static void boot_core(unsigned core) { u32 access; /* Select the appropriate core */ - write_gcr_cl_other(cfg->core << CM_GCR_Cx_OTHER_CORENUM_SHF); + write_gcr_cl_other(core << CM_GCR_Cx_OTHER_CORENUM_SHF); /* Set its reset vector */ write_gcr_co_reset_base(CKSEG1ADDR((unsigned long)mips_cps_core_entry)); @@ -175,104 +193,74 @@ static void boot_core(struct boot_config *cfg) /* Ensure the core can access the GCRs */ access = read_gcr_access(); - access |= 1 << (CM_GCR_ACCESS_ACCESSEN_SHF + cfg->core); + access |= 1 << (CM_GCR_ACCESS_ACCESSEN_SHF + core); write_gcr_access(access); - /* Copy cfg */ - mips_cps_bootcfg = *cfg; - if (mips_cpc_present()) { - /* Select the appropriate core */ - write_cpc_cl_other(cfg->core << CPC_Cx_OTHER_CORENUM_SHF); - /* Reset the core */ + mips_cpc_lock_other(core); write_cpc_co_cmd(CPC_Cx_CMD_RESET); + mips_cpc_unlock_other(); } else { /* Take the core out of reset */ write_gcr_co_reset_release(0); } /* The core is now powered up */ - bitmap_set(core_power, cfg->core, 1); + bitmap_set(core_power, core, 1); } -static void boot_vpe(void *info) +static void remote_vpe_boot(void *dummy) { - struct boot_config *cfg = info; - u32 tcstatus, vpeconf0; - - /* Enter VPE configuration state */ - dvpe(); - set_c0_mvpcontrol(MVPCONTROL_VPC); - - settc(cfg->vpe); - - /* Set the TC restart PC */ - write_tc_c0_tcrestart((unsigned long)&smp_bootstrap); - - /* Activate the TC, allow interrupts */ - tcstatus = read_tc_c0_tcstatus(); - tcstatus &= ~TCSTATUS_IXMT; - tcstatus |= TCSTATUS_A; - write_tc_c0_tcstatus(tcstatus); - - /* Clear the TC halt bit */ - write_tc_c0_tchalt(0); - - /* Activate the VPE */ - vpeconf0 = read_vpe_c0_vpeconf0(); - vpeconf0 |= VPECONF0_VPA; - write_vpe_c0_vpeconf0(vpeconf0); - - /* Set the stack & global pointer registers */ - write_tc_gpr_sp(cfg->sp); - write_tc_gpr_gp(cfg->gp); - - /* Leave VPE configuration state */ - clear_c0_mvpcontrol(MVPCONTROL_VPC); - - /* Enable other VPEs to execute */ - evpe(EVPE_ENABLE); + mips_cps_boot_vpes(); } static void cps_boot_secondary(int cpu, struct task_struct *idle) { - struct boot_config cfg; + unsigned core = cpu_data[cpu].core; + unsigned vpe_id = cpu_vpe_id(&cpu_data[cpu]); + struct core_boot_config *core_cfg = &mips_cps_core_bootcfg[core]; + struct vpe_boot_config *vpe_cfg = &core_cfg->vpe_config[vpe_id]; unsigned int remote; int err; - cfg.core = cpu_data[cpu].core; - cfg.vpe = cpu_vpe_id(&cpu_data[cpu]); - cfg.pc = (unsigned long)&smp_bootstrap; - cfg.sp = __KSTK_TOS(idle); - cfg.gp = (unsigned long)task_thread_info(idle); + vpe_cfg->pc = (unsigned long)&smp_bootstrap; + vpe_cfg->sp = __KSTK_TOS(idle); + vpe_cfg->gp = (unsigned long)task_thread_info(idle); + + atomic_or(1 << cpu_vpe_id(&cpu_data[cpu]), &core_cfg->vpe_mask); - if (!test_bit(cfg.core, core_power)) { + preempt_disable(); + + if (!test_bit(core, core_power)) { /* Boot a VPE on a powered down core */ - boot_core(&cfg); - return; + boot_core(core); + goto out; } - if (cfg.core != current_cpu_data.core) { + if (core != current_cpu_data.core) { /* Boot a VPE on another powered up core */ for (remote = 0; remote < NR_CPUS; remote++) { - if (cpu_data[remote].core != cfg.core) + if (cpu_data[remote].core != core) continue; if (cpu_online(remote)) break; } BUG_ON(remote >= NR_CPUS); - err = smp_call_function_single(remote, boot_vpe, &cfg, 1); + err = smp_call_function_single(remote, remote_vpe_boot, + NULL, 1); if (err) panic("Failed to call remote CPU\n"); - return; + goto out; } BUG_ON(!cpu_has_mipsmt); /* Boot a VPE on this core */ - boot_vpe(&cfg); + mips_cps_boot_vpes(); +out: + preempt_enable(); } static void cps_init_secondary(void) @@ -281,10 +269,6 @@ static void cps_init_secondary(void) if (cpu_has_mipsmt) dmt(); - /* TODO: revisit this assumption once hotplug is implemented */ - if (cpu_vpe_id(¤t_cpu_data) == 0) - init_core(); - change_c0_status(ST0_IM, STATUSF_IP3 | STATUSF_IP4 | STATUSF_IP6 | STATUSF_IP7); } @@ -302,10 +286,148 @@ static void cps_smp_finish(void) local_irq_enable(); } -static void cps_cpus_done(void) +#ifdef CONFIG_HOTPLUG_CPU + +static int cps_cpu_disable(void) +{ + unsigned cpu = smp_processor_id(); + struct core_boot_config *core_cfg; + + if (!cpu) + return -EBUSY; + + if (!cps_pm_support_state(CPS_PM_POWER_GATED)) + return -EINVAL; + + core_cfg = &mips_cps_core_bootcfg[current_cpu_data.core]; + atomic_sub(1 << cpu_vpe_id(¤t_cpu_data), &core_cfg->vpe_mask); + smp_mb__after_atomic_dec(); + set_cpu_online(cpu, false); + cpu_clear(cpu, cpu_callin_map); + + return 0; +} + +static DECLARE_COMPLETION(cpu_death_chosen); +static unsigned cpu_death_sibling; +static enum { + CPU_DEATH_HALT, + CPU_DEATH_POWER, +} cpu_death; + +void play_dead(void) +{ + unsigned cpu, core; + + local_irq_disable(); + idle_task_exit(); + cpu = smp_processor_id(); + cpu_death = CPU_DEATH_POWER; + + if (cpu_has_mipsmt) { + core = cpu_data[cpu].core; + + /* Look for another online VPE within the core */ + for_each_online_cpu(cpu_death_sibling) { + if (cpu_data[cpu_death_sibling].core != core) + continue; + + /* + * There is an online VPE within the core. Just halt + * this TC and leave the core alone. + */ + cpu_death = CPU_DEATH_HALT; + break; + } + } + + /* This CPU has chosen its way out */ + complete(&cpu_death_chosen); + + if (cpu_death == CPU_DEATH_HALT) { + /* Halt this TC */ + write_c0_tchalt(TCHALT_H); + instruction_hazard(); + } else { + /* Power down the core */ + cps_pm_enter_state(CPS_PM_POWER_GATED); + } + + /* This should never be reached */ + panic("Failed to offline CPU %u", cpu); +} + +static void wait_for_sibling_halt(void *ptr_cpu) { + unsigned cpu = (unsigned)ptr_cpu; + unsigned vpe_id = cpu_data[cpu].vpe_id; + unsigned halted; + unsigned long flags; + + do { + local_irq_save(flags); + settc(vpe_id); + halted = read_tc_c0_tchalt(); + local_irq_restore(flags); + } while (!(halted & TCHALT_H)); +} + +static void cps_cpu_die(unsigned int cpu) +{ + unsigned core = cpu_data[cpu].core; + unsigned stat; + int err; + + /* Wait for the cpu to choose its way out */ + if (!wait_for_completion_timeout(&cpu_death_chosen, + msecs_to_jiffies(5000))) { + pr_err("CPU%u: didn't offline\n", cpu); + return; + } + + /* + * Now wait for the CPU to actually offline. Without doing this that + * offlining may race with one or more of: + * + * - Onlining the CPU again. + * - Powering down the core if another VPE within it is offlined. + * - A sibling VPE entering a non-coherent state. + * + * In the non-MT halt case (ie. infinite loop) the CPU is doing nothing + * with which we could race, so do nothing. + */ + if (cpu_death == CPU_DEATH_POWER) { + /* + * Wait for the core to enter a powered down or clock gated + * state, the latter happening when a JTAG probe is connected + * in which case the CPC will refuse to power down the core. + */ + do { + mips_cpc_lock_other(core); + stat = read_cpc_co_stat_conf(); + stat &= CPC_Cx_STAT_CONF_SEQSTATE_MSK; + mips_cpc_unlock_other(); + } while (stat != CPC_Cx_STAT_CONF_SEQSTATE_D0 && + stat != CPC_Cx_STAT_CONF_SEQSTATE_D2 && + stat != CPC_Cx_STAT_CONF_SEQSTATE_U2); + + /* Indicate the core is powered off */ + bitmap_clear(core_power, core, 1); + } else if (cpu_has_mipsmt) { + /* + * Have a CPU with access to the offlined CPUs registers wait + * for its TC to halt. + */ + err = smp_call_function_single(cpu_death_sibling, + wait_for_sibling_halt, + (void *)cpu, 1); + if (err) + panic("Failed to call remote sibling CPU\n"); + } } +#endif /* CONFIG_HOTPLUG_CPU */ + static struct plat_smp_ops cps_smp_ops = { .smp_setup = cps_smp_setup, .prepare_cpus = cps_prepare_cpus, @@ -314,9 +436,18 @@ static struct plat_smp_ops cps_smp_ops = { .smp_finish = cps_smp_finish, .send_ipi_single = gic_send_ipi_single, .send_ipi_mask = gic_send_ipi_mask, - .cpus_done = cps_cpus_done, +#ifdef CONFIG_HOTPLUG_CPU + .cpu_disable = cps_cpu_disable, + .cpu_die = cps_cpu_die, +#endif }; +bool mips_cps_smp_in_use(void) +{ + extern struct plat_smp_ops *mp_ops; + return mp_ops == &cps_smp_ops; +} + int register_cps_smp_ops(void) { if (!mips_cm_present()) { diff --git a/arch/mips/kernel/smp-gic.c b/arch/mips/kernel/smp-gic.c index 3bb1f92ab52..3b21a96d1cc 100644 --- a/arch/mips/kernel/smp-gic.c +++ b/arch/mips/kernel/smp-gic.c @@ -15,12 +15,14 @@ #include <linux/printk.h> #include <asm/gic.h> +#include <asm/mips-cpc.h> #include <asm/smp-ops.h> void gic_send_ipi_single(int cpu, unsigned int action) { unsigned long flags; unsigned int intr; + unsigned int core = cpu_data[cpu].core; pr_debug("CPU%d: %s cpu %d action %u status %08x\n", smp_processor_id(), __func__, cpu, action, read_c0_status()); @@ -41,6 +43,15 @@ void gic_send_ipi_single(int cpu, unsigned int action) } gic_send_ipi(intr); + + if (mips_cpc_present() && (core != current_cpu_data.core)) { + while (!cpumask_test_cpu(cpu, &cpu_coherent_mask)) { + mips_cpc_lock_other(core); + write_cpc_co_cmd(CPC_Cx_CMD_PWRUP); + mips_cpc_unlock_other(); + } + } + local_irq_restore(flags); } diff --git a/arch/mips/kernel/smp-mt.c b/arch/mips/kernel/smp-mt.c index f8e13149604..3babf6e4f89 100644 --- a/arch/mips/kernel/smp-mt.c +++ b/arch/mips/kernel/smp-mt.c @@ -183,10 +183,6 @@ static void vsmp_smp_finish(void) local_irq_enable(); } -static void vsmp_cpus_done(void) -{ -} - /* * Setup the PC, SP, and GP of a secondary processor and start it * running! @@ -287,7 +283,6 @@ struct plat_smp_ops vsmp_smp_ops = { .send_ipi_mask = vsmp_send_ipi_mask, .init_secondary = vsmp_init_secondary, .smp_finish = vsmp_smp_finish, - .cpus_done = vsmp_cpus_done, .boot_secondary = vsmp_boot_secondary, .smp_setup = vsmp_smp_setup, .prepare_cpus = vsmp_prepare_cpus, diff --git a/arch/mips/kernel/smp-up.c b/arch/mips/kernel/smp-up.c index 7fde3e4d978..17878d71ef2 100644 --- a/arch/mips/kernel/smp-up.c +++ b/arch/mips/kernel/smp-up.c @@ -36,11 +36,6 @@ static void up_smp_finish(void) { } -/* Hook for after all CPUs are online */ -static void up_cpus_done(void) -{ -} - /* * Firmware CPU startup hook */ @@ -73,7 +68,6 @@ struct plat_smp_ops up_smp_ops = { .send_ipi_mask = up_send_ipi_mask, .init_secondary = up_init_secondary, .smp_finish = up_smp_finish, - .cpus_done = up_cpus_done, .boot_secondary = up_boot_secondary, .smp_setup = up_smp_setup, .prepare_cpus = up_prepare_cpus, diff --git a/arch/mips/kernel/smp.c b/arch/mips/kernel/smp.c index 0a022ee33b2..9bad52ede90 100644 --- a/arch/mips/kernel/smp.c +++ b/arch/mips/kernel/smp.c @@ -43,10 +43,6 @@ #include <asm/time.h> #include <asm/setup.h> -#ifdef CONFIG_MIPS_MT_SMTC -#include <asm/mipsmtregs.h> -#endif /* CONFIG_MIPS_MT_SMTC */ - volatile cpumask_t cpu_callin_map; /* Bitmask of started secondaries */ int __cpu_number_map[NR_CPUS]; /* Map physical to logical */ @@ -66,6 +62,8 @@ EXPORT_SYMBOL(cpu_sibling_map); /* representing cpus for which sibling maps can be computed */ static cpumask_t cpu_sibling_setup_map; +cpumask_t cpu_coherent_mask; + static inline void set_cpu_sibling_map(int cpu) { int i; @@ -102,12 +100,6 @@ asmlinkage void start_secondary(void) { unsigned int cpu; -#ifdef CONFIG_MIPS_MT_SMTC - /* Only do cpu_probe for first TC of CPU */ - if ((read_c0_tcbind() & TCBIND_CURTC) != 0) - __cpu_name[smp_processor_id()] = __cpu_name[0]; - else -#endif /* CONFIG_MIPS_MT_SMTC */ cpu_probe(); cpu_report(); per_cpu_trap_init(false); @@ -124,6 +116,7 @@ asmlinkage void start_secondary(void) cpu = smp_processor_id(); cpu_data[cpu].udelay_val = loops_per_jiffy; + cpu_set(cpu, cpu_coherent_mask); notify_cpu_starting(cpu); set_cpu_online(cpu, true); @@ -173,7 +166,6 @@ void smp_send_stop(void) void __init smp_cpus_done(unsigned int max_cpus) { - mp_ops->cpus_done(); } /* called from main before smp_init() */ @@ -186,6 +178,7 @@ void __init smp_prepare_cpus(unsigned int max_cpus) #ifndef CONFIG_HOTPLUG_CPU init_cpu_present(cpu_possible_mask); #endif + cpumask_copy(&cpu_coherent_mask, cpu_possible_mask); } /* preload SMP state for boot cpu */ @@ -238,13 +231,10 @@ static void flush_tlb_mm_ipi(void *mm) * o collapses to normal function call on UP kernels * o collapses to normal function call on systems with a single shared * primary cache. - * o CONFIG_MIPS_MT_SMTC currently implies there is only one physical core. */ static inline void smp_on_other_tlbs(void (*func) (void *info), void *info) { -#ifndef CONFIG_MIPS_MT_SMTC smp_call_function(func, info, 1); -#endif } static inline void smp_on_each_tlb(void (*func) (void *info), void *info) @@ -404,3 +394,46 @@ void dump_send_ipi(void (*dump_ipi_callback)(void *)) } EXPORT_SYMBOL(dump_send_ipi); #endif + +#ifdef CONFIG_GENERIC_CLOCKEVENTS_BROADCAST + +static DEFINE_PER_CPU(atomic_t, tick_broadcast_count); +static DEFINE_PER_CPU(struct call_single_data, tick_broadcast_csd); + +void tick_broadcast(const struct cpumask *mask) +{ + atomic_t *count; + struct call_single_data *csd; + int cpu; + + for_each_cpu(cpu, mask) { + count = &per_cpu(tick_broadcast_count, cpu); + csd = &per_cpu(tick_broadcast_csd, cpu); + + if (atomic_inc_return(count) == 1) + smp_call_function_single_async(cpu, csd); + } +} + +static void tick_broadcast_callee(void *info) +{ + int cpu = smp_processor_id(); + tick_receive_broadcast(); + atomic_set(&per_cpu(tick_broadcast_count, cpu), 0); +} + +static int __init tick_broadcast_init(void) +{ + struct call_single_data *csd; + int cpu; + + for (cpu = 0; cpu < NR_CPUS; cpu++) { + csd = &per_cpu(tick_broadcast_csd, cpu); + csd->func = tick_broadcast_callee; + } + + return 0; +} +early_initcall(tick_broadcast_init); + +#endif /* CONFIG_GENERIC_CLOCKEVENTS_BROADCAST */ diff --git a/arch/mips/kernel/smtc-asm.S b/arch/mips/kernel/smtc-asm.S deleted file mode 100644 index 2866863a39d..00000000000 --- a/arch/mips/kernel/smtc-asm.S +++ /dev/null @@ -1,133 +0,0 @@ -/* - * Assembly Language Functions for MIPS MT SMTC support - */ - -/* - * This file should be built into the kernel only if CONFIG_MIPS_MT_SMTC is set. */ - -#include <asm/regdef.h> -#include <asm/asmmacro.h> -#include <asm/stackframe.h> -#include <asm/irqflags.h> - -/* - * "Software Interrupt" linkage. - * - * This is invoked when an "Interrupt" is sent from one TC to another, - * where the TC to be interrupted is halted, has it's Restart address - * and Status values saved by the "remote control" thread, then modified - * to cause execution to begin here, in kenel mode. This code then - * disguises the TC state as that of an exception and transfers - * control to the general exception or vectored interrupt handler. - */ - .set noreorder - -/* -The __smtc_ipi_vector would use k0 and k1 as temporaries and -1) Set EXL (this is per-VPE, so this can't be done by proxy!) -2) Restore the K/CU and IXMT bits to the pre "exception" state - (EXL means no interrupts and access to the kernel map). -3) Set EPC to be the saved value of TCRestart. -4) Jump to the exception handler entry point passed by the sender. - -CAN WE PROVE THAT WE WON'T DO THIS IF INTS DISABLED?? -*/ - -/* - * Reviled and slandered vision: Set EXL and restore K/CU/IXMT - * state of pre-halt thread, then save everything and call - * thought some function pointer to imaginary_exception, which - * will parse a register value or memory message queue to - * deliver things like interprocessor interrupts. On return - * from that function, jump to the global ret_from_irq code - * to invoke the scheduler and return as appropriate. - */ - -#define PT_PADSLOT4 (PT_R0-8) -#define PT_PADSLOT5 (PT_R0-4) - - .text - .align 5 -FEXPORT(__smtc_ipi_vector) -#ifdef CONFIG_CPU_MICROMIPS - nop -#endif - .set noat - /* Disable thread scheduling to make Status update atomic */ - DMT 27 # dmt k1 - _ehb - /* Set EXL */ - mfc0 k0,CP0_STATUS - ori k0,k0,ST0_EXL - mtc0 k0,CP0_STATUS - _ehb - /* Thread scheduling now inhibited by EXL. Restore TE state. */ - andi k1,k1,VPECONTROL_TE - beqz k1,1f - emt -1: - /* - * The IPI sender has put some information on the anticipated - * kernel stack frame. If we were in user mode, this will be - * built above the saved kernel SP. If we were already in the - * kernel, it will be built above the current CPU SP. - * - * Were we in kernel mode, as indicated by CU0? - */ - sll k1,k0,3 - .set noreorder - bltz k1,2f - move k1,sp - .set reorder - /* - * If previously in user mode, set CU0 and use kernel stack. - */ - li k1,ST0_CU0 - or k1,k1,k0 - mtc0 k1,CP0_STATUS - _ehb - get_saved_sp - /* Interrupting TC will have pre-set values in slots in the new frame */ -2: subu k1,k1,PT_SIZE - /* Load TCStatus Value */ - lw k0,PT_TCSTATUS(k1) - /* Write it to TCStatus to restore CU/KSU/IXMT state */ - mtc0 k0,$2,1 - _ehb - lw k0,PT_EPC(k1) - mtc0 k0,CP0_EPC - /* Save all will redundantly recompute the SP, but use it for now */ - SAVE_ALL - CLI - TRACE_IRQS_OFF - /* Function to be invoked passed stack pad slot 5 */ - lw t0,PT_PADSLOT5(sp) - /* Argument from sender passed in stack pad slot 4 */ - lw a0,PT_PADSLOT4(sp) - LONG_L s0, TI_REGS($28) - LONG_S sp, TI_REGS($28) - PTR_LA ra, ret_from_irq - jr t0 - -/* - * Called from idle loop to provoke processing of queued IPIs - * First IPI message in queue passed as argument. - */ - -LEAF(self_ipi) - /* Before anything else, block interrupts */ - mfc0 t0,CP0_TCSTATUS - ori t1,t0,TCSTATUS_IXMT - mtc0 t1,CP0_TCSTATUS - _ehb - /* We know we're in kernel mode, so prepare stack frame */ - subu t1,sp,PT_SIZE - sw ra,PT_EPC(t1) - sw a0,PT_PADSLOT4(t1) - la t2,ipi_decode - sw t2,PT_PADSLOT5(t1) - /* Save pre-disable value of TCStatus */ - sw t0,PT_TCSTATUS(t1) - j __smtc_ipi_vector - nop -END(self_ipi) diff --git a/arch/mips/kernel/smtc-proc.c b/arch/mips/kernel/smtc-proc.c deleted file mode 100644 index 38635a996cb..00000000000 --- a/arch/mips/kernel/smtc-proc.c +++ /dev/null @@ -1,102 +0,0 @@ -/* - * /proc hooks for SMTC kernel - * Copyright (C) 2005 Mips Technologies, Inc - */ - -#include <linux/kernel.h> -#include <linux/sched.h> -#include <linux/cpumask.h> -#include <linux/interrupt.h> - -#include <asm/cpu.h> -#include <asm/processor.h> -#include <linux/atomic.h> -#include <asm/hardirq.h> -#include <asm/mmu_context.h> -#include <asm/mipsregs.h> -#include <asm/cacheflush.h> -#include <linux/proc_fs.h> -#include <linux/seq_file.h> - -#include <asm/smtc_proc.h> - -/* - * /proc diagnostic and statistics hooks - */ - -/* - * Statistics gathered - */ -unsigned long selfipis[NR_CPUS]; - -struct smtc_cpu_proc smtc_cpu_stats[NR_CPUS]; - -atomic_t smtc_fpu_recoveries; - -static int smtc_proc_show(struct seq_file *m, void *v) -{ - int i; - extern unsigned long ebase; - - seq_printf(m, "SMTC Status Word: 0x%08x\n", smtc_status); - seq_printf(m, "Config7: 0x%08x\n", read_c0_config7()); - seq_printf(m, "EBASE: 0x%08lx\n", ebase); - seq_printf(m, "Counter Interrupts taken per CPU (TC)\n"); - for (i=0; i < NR_CPUS; i++) - seq_printf(m, "%d: %ld\n", i, smtc_cpu_stats[i].timerints); - seq_printf(m, "Self-IPIs by CPU:\n"); - for(i = 0; i < NR_CPUS; i++) - seq_printf(m, "%d: %ld\n", i, smtc_cpu_stats[i].selfipis); - seq_printf(m, "%d Recoveries of \"stolen\" FPU\n", - atomic_read(&smtc_fpu_recoveries)); - return 0; -} - -static int smtc_proc_open(struct inode *inode, struct file *file) -{ - return single_open(file, smtc_proc_show, NULL); -} - -static const struct file_operations smtc_proc_fops = { - .open = smtc_proc_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; - -void init_smtc_stats(void) -{ - int i; - - for (i=0; i<NR_CPUS; i++) { - smtc_cpu_stats[i].timerints = 0; - smtc_cpu_stats[i].selfipis = 0; - } - - atomic_set(&smtc_fpu_recoveries, 0); - - proc_create("smtc", 0444, NULL, &smtc_proc_fops); -} - -static int proc_cpuinfo_chain_call(struct notifier_block *nfb, - unsigned long action_unused, void *data) -{ - struct proc_cpuinfo_notifier_args *pcn = data; - struct seq_file *m = pcn->m; - unsigned long n = pcn->n; - - if (!cpu_has_mipsmt) - return NOTIFY_OK; - - seq_printf(m, "VPE\t\t\t: %d\n", cpu_data[n].vpe_id); - seq_printf(m, "TC\t\t\t: %d\n", cpu_data[n].tc_id); - - return NOTIFY_OK; -} - -static int __init proc_cpuinfo_notifier_init(void) -{ - return proc_cpuinfo_notifier(proc_cpuinfo_chain_call, 0); -} - -subsys_initcall(proc_cpuinfo_notifier_init); diff --git a/arch/mips/kernel/smtc.c b/arch/mips/kernel/smtc.c deleted file mode 100644 index c1681d65dd5..00000000000 --- a/arch/mips/kernel/smtc.c +++ /dev/null @@ -1,1528 +0,0 @@ -/* - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version 2 - * of the License, or (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. - * - * Copyright (C) 2004 Mips Technologies, Inc - * Copyright (C) 2008 Kevin D. Kissell - */ - -#include <linux/clockchips.h> -#include <linux/kernel.h> -#include <linux/sched.h> -#include <linux/smp.h> -#include <linux/cpumask.h> -#include <linux/interrupt.h> -#include <linux/kernel_stat.h> -#include <linux/module.h> -#include <linux/ftrace.h> -#include <linux/slab.h> - -#include <asm/cpu.h> -#include <asm/processor.h> -#include <linux/atomic.h> -#include <asm/hardirq.h> -#include <asm/hazards.h> -#include <asm/irq.h> -#include <asm/idle.h> -#include <asm/mmu_context.h> -#include <asm/mipsregs.h> -#include <asm/cacheflush.h> -#include <asm/time.h> -#include <asm/addrspace.h> -#include <asm/smtc.h> -#include <asm/smtc_proc.h> -#include <asm/setup.h> - -/* - * SMTC Kernel needs to manipulate low-level CPU interrupt mask - * in do_IRQ. These are passed in setup_irq_smtc() and stored - * in this table. - */ -unsigned long irq_hwmask[NR_IRQS]; - -#define LOCK_MT_PRA() \ - local_irq_save(flags); \ - mtflags = dmt() - -#define UNLOCK_MT_PRA() \ - emt(mtflags); \ - local_irq_restore(flags) - -#define LOCK_CORE_PRA() \ - local_irq_save(flags); \ - mtflags = dvpe() - -#define UNLOCK_CORE_PRA() \ - evpe(mtflags); \ - local_irq_restore(flags) - -/* - * Data structures purely associated with SMTC parallelism - */ - - -/* - * Table for tracking ASIDs whose lifetime is prolonged. - */ - -asiduse smtc_live_asid[MAX_SMTC_TLBS][MAX_SMTC_ASIDS]; - -/* - * Number of InterProcessor Interrupt (IPI) message buffers to allocate - */ - -#define IPIBUF_PER_CPU 4 - -struct smtc_ipi_q IPIQ[NR_CPUS]; -static struct smtc_ipi_q freeIPIq; - - -/* - * Number of FPU contexts for each VPE - */ - -static int smtc_nconf1[MAX_SMTC_VPES]; - - -/* Forward declarations */ - -void ipi_decode(struct smtc_ipi *); -static void post_direct_ipi(int cpu, struct smtc_ipi *pipi); -static void setup_cross_vpe_interrupts(unsigned int nvpe); -void init_smtc_stats(void); - -/* Global SMTC Status */ - -unsigned int smtc_status; - -/* Boot command line configuration overrides */ - -static int vpe0limit; -static int ipibuffers; -static int nostlb; -static int asidmask; -unsigned long smtc_asid_mask = 0xff; - -static int __init vpe0tcs(char *str) -{ - get_option(&str, &vpe0limit); - - return 1; -} - -static int __init ipibufs(char *str) -{ - get_option(&str, &ipibuffers); - return 1; -} - -static int __init stlb_disable(char *s) -{ - nostlb = 1; - return 1; -} - -static int __init asidmask_set(char *str) -{ - get_option(&str, &asidmask); - switch (asidmask) { - case 0x1: - case 0x3: - case 0x7: - case 0xf: - case 0x1f: - case 0x3f: - case 0x7f: - case 0xff: - smtc_asid_mask = (unsigned long)asidmask; - break; - default: - printk("ILLEGAL ASID mask 0x%x from command line\n", asidmask); - } - return 1; -} - -__setup("vpe0tcs=", vpe0tcs); -__setup("ipibufs=", ipibufs); -__setup("nostlb", stlb_disable); -__setup("asidmask=", asidmask_set); - -#ifdef CONFIG_SMTC_IDLE_HOOK_DEBUG - -static int hang_trig; - -static int __init hangtrig_enable(char *s) -{ - hang_trig = 1; - return 1; -} - - -__setup("hangtrig", hangtrig_enable); - -#define DEFAULT_BLOCKED_IPI_LIMIT 32 - -static int timerq_limit = DEFAULT_BLOCKED_IPI_LIMIT; - -static int __init tintq(char *str) -{ - get_option(&str, &timerq_limit); - return 1; -} - -__setup("tintq=", tintq); - -static int imstuckcount[MAX_SMTC_VPES][8]; -/* vpemask represents IM/IE bits of per-VPE Status registers, low-to-high */ -static int vpemask[MAX_SMTC_VPES][8] = { - {0, 0, 1, 0, 0, 0, 0, 1}, - {0, 0, 0, 0, 0, 0, 0, 1} -}; -int tcnoprog[NR_CPUS]; -static atomic_t idle_hook_initialized = ATOMIC_INIT(0); -static int clock_hang_reported[NR_CPUS]; - -#endif /* CONFIG_SMTC_IDLE_HOOK_DEBUG */ - -/* - * Configure shared TLB - VPC configuration bit must be set by caller - */ - -static void smtc_configure_tlb(void) -{ - int i, tlbsiz, vpes; - unsigned long mvpconf0; - unsigned long config1val; - - /* Set up ASID preservation table */ - for (vpes=0; vpes<MAX_SMTC_TLBS; vpes++) { - for(i = 0; i < MAX_SMTC_ASIDS; i++) { - smtc_live_asid[vpes][i] = 0; - } - } - mvpconf0 = read_c0_mvpconf0(); - - if ((vpes = ((mvpconf0 & MVPCONF0_PVPE) - >> MVPCONF0_PVPE_SHIFT) + 1) > 1) { - /* If we have multiple VPEs, try to share the TLB */ - if ((mvpconf0 & MVPCONF0_TLBS) && !nostlb) { - /* - * If TLB sizing is programmable, shared TLB - * size is the total available complement. - * Otherwise, we have to take the sum of all - * static VPE TLB entries. - */ - if ((tlbsiz = ((mvpconf0 & MVPCONF0_PTLBE) - >> MVPCONF0_PTLBE_SHIFT)) == 0) { - /* - * If there's more than one VPE, there had better - * be more than one TC, because we need one to bind - * to each VPE in turn to be able to read - * its configuration state! - */ - settc(1); - /* Stop the TC from doing anything foolish */ - write_tc_c0_tchalt(TCHALT_H); - mips_ihb(); - /* No need to un-Halt - that happens later anyway */ - for (i=0; i < vpes; i++) { - write_tc_c0_tcbind(i); - /* - * To be 100% sure we're really getting the right - * information, we exit the configuration state - * and do an IHB after each rebinding. - */ - write_c0_mvpcontrol( - read_c0_mvpcontrol() & ~ MVPCONTROL_VPC ); - mips_ihb(); - /* - * Only count if the MMU Type indicated is TLB - */ - if (((read_vpe_c0_config() & MIPS_CONF_MT) >> 7) == 1) { - config1val = read_vpe_c0_config1(); - tlbsiz += ((config1val >> 25) & 0x3f) + 1; - } - - /* Put core back in configuration state */ - write_c0_mvpcontrol( - read_c0_mvpcontrol() | MVPCONTROL_VPC ); - mips_ihb(); - } - } - write_c0_mvpcontrol(read_c0_mvpcontrol() | MVPCONTROL_STLB); - ehb(); - - /* - * Setup kernel data structures to use software total, - * rather than read the per-VPE Config1 value. The values - * for "CPU 0" gets copied to all the other CPUs as part - * of their initialization in smtc_cpu_setup(). - */ - - /* MIPS32 limits TLB indices to 64 */ - if (tlbsiz > 64) - tlbsiz = 64; - cpu_data[0].tlbsize = current_cpu_data.tlbsize = tlbsiz; - smtc_status |= SMTC_TLB_SHARED; - local_flush_tlb_all(); - - printk("TLB of %d entry pairs shared by %d VPEs\n", - tlbsiz, vpes); - } else { - printk("WARNING: TLB Not Sharable on SMTC Boot!\n"); - } - } -} - - -/* - * Incrementally build the CPU map out of constituent MIPS MT cores, - * using the specified available VPEs and TCs. Plaform code needs - * to ensure that each MIPS MT core invokes this routine on reset, - * one at a time(!). - * - * This version of the build_cpu_map and prepare_cpus routines assumes - * that *all* TCs of a MIPS MT core will be used for Linux, and that - * they will be spread across *all* available VPEs (to minimise the - * loss of efficiency due to exception service serialization). - * An improved version would pick up configuration information and - * possibly leave some TCs/VPEs as "slave" processors. - * - * Use c0_MVPConf0 to find out how many TCs are available, setting up - * cpu_possible_mask and the logical/physical mappings. - */ - -int __init smtc_build_cpu_map(int start_cpu_slot) -{ - int i, ntcs; - - /* - * The CPU map isn't actually used for anything at this point, - * so it's not clear what else we should do apart from set - * everything up so that "logical" = "physical". - */ - ntcs = ((read_c0_mvpconf0() & MVPCONF0_PTC) >> MVPCONF0_PTC_SHIFT) + 1; - for (i=start_cpu_slot; i<NR_CPUS && i<ntcs; i++) { - set_cpu_possible(i, true); - __cpu_number_map[i] = i; - __cpu_logical_map[i] = i; - } -#ifdef CONFIG_MIPS_MT_FPAFF - /* Initialize map of CPUs with FPUs */ - cpus_clear(mt_fpu_cpumask); -#endif - - /* One of those TC's is the one booting, and not a secondary... */ - printk("%i available secondary CPU TC(s)\n", i - 1); - - return i; -} - -/* - * Common setup before any secondaries are started - * Make sure all CPUs are in a sensible state before we boot any of the - * secondaries. - * - * For MIPS MT "SMTC" operation, we set up all TCs, spread as evenly - * as possible across the available VPEs. - */ - -static void smtc_tc_setup(int vpe, int tc, int cpu) -{ - static int cp1contexts[MAX_SMTC_VPES]; - - /* - * Make a local copy of the available FPU contexts in order - * to keep track of TCs that can have one. - */ - if (tc == 1) - { - /* - * FIXME: Multi-core SMTC hasn't been tested and the - * maximum number of VPEs may change. - */ - cp1contexts[0] = smtc_nconf1[0] - 1; - cp1contexts[1] = smtc_nconf1[1]; - } - - settc(tc); - write_tc_c0_tchalt(TCHALT_H); - mips_ihb(); - write_tc_c0_tcstatus((read_tc_c0_tcstatus() - & ~(TCSTATUS_TKSU | TCSTATUS_DA | TCSTATUS_IXMT)) - | TCSTATUS_A); - /* - * TCContext gets an offset from the base of the IPIQ array - * to be used in low-level code to detect the presence of - * an active IPI queue. - */ - write_tc_c0_tccontext((sizeof(struct smtc_ipi_q) * cpu) << 16); - - /* Bind TC to VPE. */ - write_tc_c0_tcbind(vpe); - - /* In general, all TCs should have the same cpu_data indications. */ - memcpy(&cpu_data[cpu], &cpu_data[0], sizeof(struct cpuinfo_mips)); - - /* Check to see if there is a FPU context available for this TC. */ - if (!cp1contexts[vpe]) - cpu_data[cpu].options &= ~MIPS_CPU_FPU; - else - cp1contexts[vpe]--; - - /* Store the TC and VPE into the cpu_data structure. */ - cpu_data[cpu].vpe_id = vpe; - cpu_data[cpu].tc_id = tc; - - /* FIXME: Multi-core SMTC hasn't been tested, but be prepared. */ - cpu_data[cpu].core = (read_vpe_c0_ebase() >> 1) & 0xff; -} - -/* - * Tweak to get Count registers synced as closely as possible. The - * value seems good for 34K-class cores. - */ - -#define CP0_SKEW 8 - -void smtc_prepare_cpus(int cpus) -{ - int i, vpe, tc, ntc, nvpe, tcpervpe[NR_CPUS], slop, cpu; - unsigned long flags; - unsigned long val; - int nipi; - struct smtc_ipi *pipi; - - /* disable interrupts so we can disable MT */ - local_irq_save(flags); - /* disable MT so we can configure */ - dvpe(); - dmt(); - - spin_lock_init(&freeIPIq.lock); - - /* - * We probably don't have as many VPEs as we do SMP "CPUs", - * but it's possible - and in any case we'll never use more! - */ - for (i=0; i<NR_CPUS; i++) { - IPIQ[i].head = IPIQ[i].tail = NULL; - spin_lock_init(&IPIQ[i].lock); - IPIQ[i].depth = 0; - IPIQ[i].resched_flag = 0; /* No reschedules queued initially */ - } - - /* cpu_data index starts at zero */ - cpu = 0; - cpu_data[cpu].vpe_id = 0; - cpu_data[cpu].tc_id = 0; - cpu_data[cpu].core = (read_c0_ebase() >> 1) & 0xff; - cpu++; - - /* Report on boot-time options */ - mips_mt_set_cpuoptions(); - if (vpelimit > 0) - printk("Limit of %d VPEs set\n", vpelimit); - if (tclimit > 0) - printk("Limit of %d TCs set\n", tclimit); - if (nostlb) { - printk("Shared TLB Use Inhibited - UNSAFE for Multi-VPE Operation\n"); - } - if (asidmask) - printk("ASID mask value override to 0x%x\n", asidmask); - - /* Temporary */ -#ifdef CONFIG_SMTC_IDLE_HOOK_DEBUG - if (hang_trig) - printk("Logic Analyser Trigger on suspected TC hang\n"); -#endif /* CONFIG_SMTC_IDLE_HOOK_DEBUG */ - - /* Put MVPE's into 'configuration state' */ - write_c0_mvpcontrol( read_c0_mvpcontrol() | MVPCONTROL_VPC ); - - val = read_c0_mvpconf0(); - nvpe = ((val & MVPCONF0_PVPE) >> MVPCONF0_PVPE_SHIFT) + 1; - if (vpelimit > 0 && nvpe > vpelimit) - nvpe = vpelimit; - ntc = ((val & MVPCONF0_PTC) >> MVPCONF0_PTC_SHIFT) + 1; - if (ntc > NR_CPUS) - ntc = NR_CPUS; - if (tclimit > 0 && ntc > tclimit) - ntc = tclimit; - slop = ntc % nvpe; - for (i = 0; i < nvpe; i++) { - tcpervpe[i] = ntc / nvpe; - if (slop) { - if((slop - i) > 0) tcpervpe[i]++; - } - } - /* Handle command line override for VPE0 */ - if (vpe0limit > ntc) vpe0limit = ntc; - if (vpe0limit > 0) { - int slopslop; - if (vpe0limit < tcpervpe[0]) { - /* Reducing TC count - distribute to others */ - slop = tcpervpe[0] - vpe0limit; - slopslop = slop % (nvpe - 1); - tcpervpe[0] = vpe0limit; - for (i = 1; i < nvpe; i++) { - tcpervpe[i] += slop / (nvpe - 1); - if(slopslop && ((slopslop - (i - 1) > 0))) - tcpervpe[i]++; - } - } else if (vpe0limit > tcpervpe[0]) { - /* Increasing TC count - steal from others */ - slop = vpe0limit - tcpervpe[0]; - slopslop = slop % (nvpe - 1); - tcpervpe[0] = vpe0limit; - for (i = 1; i < nvpe; i++) { - tcpervpe[i] -= slop / (nvpe - 1); - if(slopslop && ((slopslop - (i - 1) > 0))) - tcpervpe[i]--; - } - } - } - - /* Set up shared TLB */ - smtc_configure_tlb(); - - for (tc = 0, vpe = 0 ; (vpe < nvpe) && (tc < ntc) ; vpe++) { - /* Get number of CP1 contexts for each VPE. */ - if (tc == 0) - { - /* - * Do not call settc() for TC0 or the FPU context - * value will be incorrect. Besides, we know that - * we are TC0 anyway. - */ - smtc_nconf1[0] = ((read_vpe_c0_vpeconf1() & - VPECONF1_NCP1) >> VPECONF1_NCP1_SHIFT); - if (nvpe == 2) - { - settc(1); - smtc_nconf1[1] = ((read_vpe_c0_vpeconf1() & - VPECONF1_NCP1) >> VPECONF1_NCP1_SHIFT); - settc(0); - } - } - if (tcpervpe[vpe] == 0) - continue; - if (vpe != 0) - printk(", "); - printk("VPE %d: TC", vpe); - for (i = 0; i < tcpervpe[vpe]; i++) { - /* - * TC 0 is bound to VPE 0 at reset, - * and is presumably executing this - * code. Leave it alone! - */ - if (tc != 0) { - smtc_tc_setup(vpe, tc, cpu); - if (vpe != 0) { - /* - * Set MVP bit (possibly again). Do it - * here to catch CPUs that have no TCs - * bound to the VPE at reset. In that - * case, a TC must be bound to the VPE - * before we can set VPEControl[MVP] - */ - write_vpe_c0_vpeconf0( - read_vpe_c0_vpeconf0() | - VPECONF0_MVP); - } - cpu++; - } - printk(" %d", tc); - tc++; - } - if (vpe != 0) { - /* - * Allow this VPE to control others. - */ - write_vpe_c0_vpeconf0(read_vpe_c0_vpeconf0() | - VPECONF0_MVP); - - /* - * Clear any stale software interrupts from VPE's Cause - */ - write_vpe_c0_cause(0); - - /* - * Clear ERL/EXL of VPEs other than 0 - * and set restricted interrupt enable/mask. - */ - write_vpe_c0_status((read_vpe_c0_status() - & ~(ST0_BEV | ST0_ERL | ST0_EXL | ST0_IM)) - | (STATUSF_IP0 | STATUSF_IP1 | STATUSF_IP7 - | ST0_IE)); - /* - * set config to be the same as vpe0, - * particularly kseg0 coherency alg - */ - write_vpe_c0_config(read_c0_config()); - /* Clear any pending timer interrupt */ - write_vpe_c0_compare(0); - /* Propagate Config7 */ - write_vpe_c0_config7(read_c0_config7()); - write_vpe_c0_count(read_c0_count() + CP0_SKEW); - ehb(); - } - /* enable multi-threading within VPE */ - write_vpe_c0_vpecontrol(read_vpe_c0_vpecontrol() | VPECONTROL_TE); - /* enable the VPE */ - write_vpe_c0_vpeconf0(read_vpe_c0_vpeconf0() | VPECONF0_VPA); - } - - /* - * Pull any physically present but unused TCs out of circulation. - */ - while (tc < (((val & MVPCONF0_PTC) >> MVPCONF0_PTC_SHIFT) + 1)) { - set_cpu_possible(tc, false); - set_cpu_present(tc, false); - tc++; - } - - /* release config state */ - write_c0_mvpcontrol( read_c0_mvpcontrol() & ~ MVPCONTROL_VPC ); - - printk("\n"); - - /* Set up coprocessor affinity CPU mask(s) */ - -#ifdef CONFIG_MIPS_MT_FPAFF - for (tc = 0; tc < ntc; tc++) { - if (cpu_data[tc].options & MIPS_CPU_FPU) - cpu_set(tc, mt_fpu_cpumask); - } -#endif - - /* set up ipi interrupts... */ - - /* If we have multiple VPEs running, set up the cross-VPE interrupt */ - - setup_cross_vpe_interrupts(nvpe); - - /* Set up queue of free IPI "messages". */ - nipi = NR_CPUS * IPIBUF_PER_CPU; - if (ipibuffers > 0) - nipi = ipibuffers; - - pipi = kmalloc(nipi *sizeof(struct smtc_ipi), GFP_KERNEL); - if (pipi == NULL) - panic("kmalloc of IPI message buffers failed"); - else - printk("IPI buffer pool of %d buffers\n", nipi); - for (i = 0; i < nipi; i++) { - smtc_ipi_nq(&freeIPIq, pipi); - pipi++; - } - - /* Arm multithreading and enable other VPEs - but all TCs are Halted */ - emt(EMT_ENABLE); - evpe(EVPE_ENABLE); - local_irq_restore(flags); - /* Initialize SMTC /proc statistics/diagnostics */ - init_smtc_stats(); -} - - -/* - * Setup the PC, SP, and GP of a secondary processor and start it - * running! - * smp_bootstrap is the place to resume from - * __KSTK_TOS(idle) is apparently the stack pointer - * (unsigned long)idle->thread_info the gp - * - */ -void smtc_boot_secondary(int cpu, struct task_struct *idle) -{ - extern u32 kernelsp[NR_CPUS]; - unsigned long flags; - int mtflags; - - LOCK_MT_PRA(); - if (cpu_data[cpu].vpe_id != cpu_data[smp_processor_id()].vpe_id) { - dvpe(); - } - settc(cpu_data[cpu].tc_id); - - /* pc */ - write_tc_c0_tcrestart((unsigned long)&smp_bootstrap); - - /* stack pointer */ - kernelsp[cpu] = __KSTK_TOS(idle); - write_tc_gpr_sp(__KSTK_TOS(idle)); - - /* global pointer */ - write_tc_gpr_gp((unsigned long)task_thread_info(idle)); - - smtc_status |= SMTC_MTC_ACTIVE; - write_tc_c0_tchalt(0); - if (cpu_data[cpu].vpe_id != cpu_data[smp_processor_id()].vpe_id) { - evpe(EVPE_ENABLE); - } - UNLOCK_MT_PRA(); -} - -void smtc_init_secondary(void) -{ -} - -void smtc_smp_finish(void) -{ - int cpu = smp_processor_id(); - - /* - * Lowest-numbered CPU per VPE starts a clock tick. - * Like per_cpu_trap_init() hack, this assumes that - * SMTC init code assigns TCs consdecutively and - * in ascending order across available VPEs. - */ - if (cpu > 0 && (cpu_data[cpu].vpe_id != cpu_data[cpu - 1].vpe_id)) - write_c0_compare(read_c0_count() + mips_hpt_frequency/HZ); - - local_irq_enable(); - - printk("TC %d going on-line as CPU %d\n", - cpu_data[smp_processor_id()].tc_id, smp_processor_id()); -} - -void smtc_cpus_done(void) -{ -} - -/* - * Support for SMTC-optimized driver IRQ registration - */ - -/* - * SMTC Kernel needs to manipulate low-level CPU interrupt mask - * in do_IRQ. These are passed in setup_irq_smtc() and stored - * in this table. - */ - -int setup_irq_smtc(unsigned int irq, struct irqaction * new, - unsigned long hwmask) -{ -#ifdef CONFIG_SMTC_IDLE_HOOK_DEBUG - unsigned int vpe = current_cpu_data.vpe_id; - - vpemask[vpe][irq - MIPS_CPU_IRQ_BASE] = 1; -#endif - irq_hwmask[irq] = hwmask; - - return setup_irq(irq, new); -} - -#ifdef CONFIG_MIPS_MT_SMTC_IRQAFF -/* - * Support for IRQ affinity to TCs - */ - -void smtc_set_irq_affinity(unsigned int irq, cpumask_t affinity) -{ - /* - * If a "fast path" cache of quickly decodable affinity state - * is maintained, this is where it gets done, on a call up - * from the platform affinity code. - */ -} - -void smtc_forward_irq(struct irq_data *d) -{ - unsigned int irq = d->irq; - int target; - - /* - * OK wise guy, now figure out how to get the IRQ - * to be serviced on an authorized "CPU". - * - * Ideally, to handle the situation where an IRQ has multiple - * eligible CPUS, we would maintain state per IRQ that would - * allow a fair distribution of service requests. Since the - * expected use model is any-or-only-one, for simplicity - * and efficiency, we just pick the easiest one to find. - */ - - target = cpumask_first(d->affinity); - - /* - * We depend on the platform code to have correctly processed - * IRQ affinity change requests to ensure that the IRQ affinity - * mask has been purged of bits corresponding to nonexistent and - * offline "CPUs", and to TCs bound to VPEs other than the VPE - * connected to the physical interrupt input for the interrupt - * in question. Otherwise we have a nasty problem with interrupt - * mask management. This is best handled in non-performance-critical - * platform IRQ affinity setting code, to minimize interrupt-time - * checks. - */ - - /* If no one is eligible, service locally */ - if (target >= NR_CPUS) - do_IRQ_no_affinity(irq); - else - smtc_send_ipi(target, IRQ_AFFINITY_IPI, irq); -} - -#endif /* CONFIG_MIPS_MT_SMTC_IRQAFF */ - -/* - * IPI model for SMTC is tricky, because interrupts aren't TC-specific. - * Within a VPE one TC can interrupt another by different approaches. - * The easiest to get right would probably be to make all TCs except - * the target IXMT and set a software interrupt, but an IXMT-based - * scheme requires that a handler must run before a new IPI could - * be sent, which would break the "broadcast" loops in MIPS MT. - * A more gonzo approach within a VPE is to halt the TC, extract - * its Restart, Status, and a couple of GPRs, and program the Restart - * address to emulate an interrupt. - * - * Within a VPE, one can be confident that the target TC isn't in - * a critical EXL state when halted, since the write to the Halt - * register could not have issued on the writing thread if the - * halting thread had EXL set. So k0 and k1 of the target TC - * can be used by the injection code. Across VPEs, one can't - * be certain that the target TC isn't in a critical exception - * state. So we try a two-step process of sending a software - * interrupt to the target VPE, which either handles the event - * itself (if it was the target) or injects the event within - * the VPE. - */ - -static void smtc_ipi_qdump(void) -{ - int i; - struct smtc_ipi *temp; - - for (i = 0; i < NR_CPUS ;i++) { - pr_info("IPIQ[%d]: head = 0x%x, tail = 0x%x, depth = %d\n", - i, (unsigned)IPIQ[i].head, (unsigned)IPIQ[i].tail, - IPIQ[i].depth); - temp = IPIQ[i].head; - - while (temp != IPIQ[i].tail) { - pr_debug("%d %d %d: ", temp->type, temp->dest, - (int)temp->arg); -#ifdef SMTC_IPI_DEBUG - pr_debug("%u %lu\n", temp->sender, temp->stamp); -#else - pr_debug("\n"); -#endif - temp = temp->flink; - } - } -} - -/* - * The standard atomic.h primitives don't quite do what we want - * here: We need an atomic add-and-return-previous-value (which - * could be done with atomic_add_return and a decrement) and an - * atomic set/zero-and-return-previous-value (which can't really - * be done with the atomic.h primitives). And since this is - * MIPS MT, we can assume that we have LL/SC. - */ -static inline int atomic_postincrement(atomic_t *v) -{ - unsigned long result; - - unsigned long temp; - - __asm__ __volatile__( - "1: ll %0, %2 \n" - " addu %1, %0, 1 \n" - " sc %1, %2 \n" - " beqz %1, 1b \n" - __WEAK_LLSC_MB - : "=&r" (result), "=&r" (temp), "=m" (v->counter) - : "m" (v->counter) - : "memory"); - - return result; -} - -void smtc_send_ipi(int cpu, int type, unsigned int action) -{ - int tcstatus; - struct smtc_ipi *pipi; - unsigned long flags; - int mtflags; - unsigned long tcrestart; - int set_resched_flag = (type == LINUX_SMP_IPI && - action == SMP_RESCHEDULE_YOURSELF); - - if (cpu == smp_processor_id()) { - printk("Cannot Send IPI to self!\n"); - return; - } - if (set_resched_flag && IPIQ[cpu].resched_flag != 0) - return; /* There is a reschedule queued already */ - - /* Set up a descriptor, to be delivered either promptly or queued */ - pipi = smtc_ipi_dq(&freeIPIq); - if (pipi == NULL) { - bust_spinlocks(1); - mips_mt_regdump(dvpe()); - panic("IPI Msg. Buffers Depleted"); - } - pipi->type = type; - pipi->arg = (void *)action; - pipi->dest = cpu; - if (cpu_data[cpu].vpe_id != cpu_data[smp_processor_id()].vpe_id) { - /* If not on same VPE, enqueue and send cross-VPE interrupt */ - IPIQ[cpu].resched_flag |= set_resched_flag; - smtc_ipi_nq(&IPIQ[cpu], pipi); - LOCK_CORE_PRA(); - settc(cpu_data[cpu].tc_id); - write_vpe_c0_cause(read_vpe_c0_cause() | C_SW1); - UNLOCK_CORE_PRA(); - } else { - /* - * Not sufficient to do a LOCK_MT_PRA (dmt) here, - * since ASID shootdown on the other VPE may - * collide with this operation. - */ - LOCK_CORE_PRA(); - settc(cpu_data[cpu].tc_id); - /* Halt the targeted TC */ - write_tc_c0_tchalt(TCHALT_H); - mips_ihb(); - - /* - * Inspect TCStatus - if IXMT is set, we have to queue - * a message. Otherwise, we set up the "interrupt" - * of the other TC - */ - tcstatus = read_tc_c0_tcstatus(); - - if ((tcstatus & TCSTATUS_IXMT) != 0) { - /* - * If we're in the the irq-off version of the wait - * loop, we need to force exit from the wait and - * do a direct post of the IPI. - */ - if (cpu_wait == r4k_wait_irqoff) { - tcrestart = read_tc_c0_tcrestart(); - if (address_is_in_r4k_wait_irqoff(tcrestart)) { - write_tc_c0_tcrestart(__pastwait); - tcstatus &= ~TCSTATUS_IXMT; - write_tc_c0_tcstatus(tcstatus); - goto postdirect; - } - } - /* - * Otherwise we queue the message for the target TC - * to pick up when he does a local_irq_restore() - */ - write_tc_c0_tchalt(0); - UNLOCK_CORE_PRA(); - IPIQ[cpu].resched_flag |= set_resched_flag; - smtc_ipi_nq(&IPIQ[cpu], pipi); - } else { -postdirect: - post_direct_ipi(cpu, pipi); - write_tc_c0_tchalt(0); - UNLOCK_CORE_PRA(); - } - } -} - -/* - * Send IPI message to Halted TC, TargTC/TargVPE already having been set - */ -static void post_direct_ipi(int cpu, struct smtc_ipi *pipi) -{ - struct pt_regs *kstack; - unsigned long tcstatus; - unsigned long tcrestart; - extern u32 kernelsp[NR_CPUS]; - extern void __smtc_ipi_vector(void); -//printk("%s: on %d for %d\n", __func__, smp_processor_id(), cpu); - - /* Extract Status, EPC from halted TC */ - tcstatus = read_tc_c0_tcstatus(); - tcrestart = read_tc_c0_tcrestart(); - /* If TCRestart indicates a WAIT instruction, advance the PC */ - if ((tcrestart & 0x80000000) - && ((*(unsigned int *)tcrestart & 0xfe00003f) == 0x42000020)) { - tcrestart += 4; - } - /* - * Save on TC's future kernel stack - * - * CU bit of Status is indicator that TC was - * already running on a kernel stack... - */ - if (tcstatus & ST0_CU0) { - /* Note that this "- 1" is pointer arithmetic */ - kstack = ((struct pt_regs *)read_tc_gpr_sp()) - 1; - } else { - kstack = ((struct pt_regs *)kernelsp[cpu]) - 1; - } - - kstack->cp0_epc = (long)tcrestart; - /* Save TCStatus */ - kstack->cp0_tcstatus = tcstatus; - /* Pass token of operation to be performed kernel stack pad area */ - kstack->pad0[4] = (unsigned long)pipi; - /* Pass address of function to be called likewise */ - kstack->pad0[5] = (unsigned long)&ipi_decode; - /* Set interrupt exempt and kernel mode */ - tcstatus |= TCSTATUS_IXMT; - tcstatus &= ~TCSTATUS_TKSU; - write_tc_c0_tcstatus(tcstatus); - ehb(); - /* Set TC Restart address to be SMTC IPI vector */ - write_tc_c0_tcrestart(__smtc_ipi_vector); -} - -static void ipi_resched_interrupt(void) -{ - scheduler_ipi(); -} - -static void ipi_call_interrupt(void) -{ - /* Invoke generic function invocation code in smp.c */ - smp_call_function_interrupt(); -} - -DECLARE_PER_CPU(struct clock_event_device, mips_clockevent_device); - -static void __irq_entry smtc_clock_tick_interrupt(void) -{ - unsigned int cpu = smp_processor_id(); - struct clock_event_device *cd; - int irq = MIPS_CPU_IRQ_BASE + 1; - - irq_enter(); - kstat_incr_irq_this_cpu(irq); - cd = &per_cpu(mips_clockevent_device, cpu); - cd->event_handler(cd); - irq_exit(); -} - -void ipi_decode(struct smtc_ipi *pipi) -{ - void *arg_copy = pipi->arg; - int type_copy = pipi->type; - - smtc_ipi_nq(&freeIPIq, pipi); - - switch (type_copy) { - case SMTC_CLOCK_TICK: - smtc_clock_tick_interrupt(); - break; - - case LINUX_SMP_IPI: - switch ((int)arg_copy) { - case SMP_RESCHEDULE_YOURSELF: - ipi_resched_interrupt(); - break; - case SMP_CALL_FUNCTION: - ipi_call_interrupt(); - break; - default: - printk("Impossible SMTC IPI Argument %p\n", arg_copy); - break; - } - break; -#ifdef CONFIG_MIPS_MT_SMTC_IRQAFF - case IRQ_AFFINITY_IPI: - /* - * Accept a "forwarded" interrupt that was initially - * taken by a TC who doesn't have affinity for the IRQ. - */ - do_IRQ_no_affinity((int)arg_copy); - break; -#endif /* CONFIG_MIPS_MT_SMTC_IRQAFF */ - default: - printk("Impossible SMTC IPI Type 0x%x\n", type_copy); - break; - } -} - -/* - * Similar to smtc_ipi_replay(), but invoked from context restore, - * so it reuses the current exception frame rather than set up a - * new one with self_ipi. - */ - -void deferred_smtc_ipi(void) -{ - int cpu = smp_processor_id(); - - /* - * Test is not atomic, but much faster than a dequeue, - * and the vast majority of invocations will have a null queue. - * If irq_disabled when this was called, then any IPIs queued - * after we test last will be taken on the next irq_enable/restore. - * If interrupts were enabled, then any IPIs added after the - * last test will be taken directly. - */ - - while (IPIQ[cpu].head != NULL) { - struct smtc_ipi_q *q = &IPIQ[cpu]; - struct smtc_ipi *pipi; - unsigned long flags; - - /* - * It may be possible we'll come in with interrupts - * already enabled. - */ - local_irq_save(flags); - spin_lock(&q->lock); - pipi = __smtc_ipi_dq(q); - spin_unlock(&q->lock); - if (pipi != NULL) { - if (pipi->type == LINUX_SMP_IPI && - (int)pipi->arg == SMP_RESCHEDULE_YOURSELF) - IPIQ[cpu].resched_flag = 0; - ipi_decode(pipi); - } - /* - * The use of the __raw_local restore isn't - * as obviously necessary here as in smtc_ipi_replay(), - * but it's more efficient, given that we're already - * running down the IPI queue. - */ - __arch_local_irq_restore(flags); - } -} - -/* - * Cross-VPE interrupts in the SMTC prototype use "software interrupts" - * set via cross-VPE MTTR manipulation of the Cause register. It would be - * in some regards preferable to have external logic for "doorbell" hardware - * interrupts. - */ - -static int cpu_ipi_irq = MIPS_CPU_IRQ_BASE + MIPS_CPU_IPI_IRQ; - -static irqreturn_t ipi_interrupt(int irq, void *dev_idm) -{ - int my_vpe = cpu_data[smp_processor_id()].vpe_id; - int my_tc = cpu_data[smp_processor_id()].tc_id; - int cpu; - struct smtc_ipi *pipi; - unsigned long tcstatus; - int sent; - unsigned long flags; - unsigned int mtflags; - unsigned int vpflags; - - /* - * So long as cross-VPE interrupts are done via - * MFTR/MTTR read-modify-writes of Cause, we need - * to stop other VPEs whenever the local VPE does - * anything similar. - */ - local_irq_save(flags); - vpflags = dvpe(); - clear_c0_cause(0x100 << MIPS_CPU_IPI_IRQ); - set_c0_status(0x100 << MIPS_CPU_IPI_IRQ); - irq_enable_hazard(); - evpe(vpflags); - local_irq_restore(flags); - - /* - * Cross-VPE Interrupt handler: Try to directly deliver IPIs - * queued for TCs on this VPE other than the current one. - * Return-from-interrupt should cause us to drain the queue - * for the current TC, so we ought not to have to do it explicitly here. - */ - - for_each_online_cpu(cpu) { - if (cpu_data[cpu].vpe_id != my_vpe) - continue; - - pipi = smtc_ipi_dq(&IPIQ[cpu]); - if (pipi != NULL) { - if (cpu_data[cpu].tc_id != my_tc) { - sent = 0; - LOCK_MT_PRA(); - settc(cpu_data[cpu].tc_id); - write_tc_c0_tchalt(TCHALT_H); - mips_ihb(); - tcstatus = read_tc_c0_tcstatus(); - if ((tcstatus & TCSTATUS_IXMT) == 0) { - post_direct_ipi(cpu, pipi); - sent = 1; - } - write_tc_c0_tchalt(0); - UNLOCK_MT_PRA(); - if (!sent) { - smtc_ipi_req(&IPIQ[cpu], pipi); - } - } else { - /* - * ipi_decode() should be called - * with interrupts off - */ - local_irq_save(flags); - if (pipi->type == LINUX_SMP_IPI && - (int)pipi->arg == SMP_RESCHEDULE_YOURSELF) - IPIQ[cpu].resched_flag = 0; - ipi_decode(pipi); - local_irq_restore(flags); - } - } - } - - return IRQ_HANDLED; -} - -static void ipi_irq_dispatch(void) -{ - do_IRQ(cpu_ipi_irq); -} - -static struct irqaction irq_ipi = { - .handler = ipi_interrupt, - .flags = IRQF_PERCPU, - .name = "SMTC_IPI" -}; - -static void setup_cross_vpe_interrupts(unsigned int nvpe) -{ - if (nvpe < 1) - return; - - if (!cpu_has_vint) - panic("SMTC Kernel requires Vectored Interrupt support"); - - set_vi_handler(MIPS_CPU_IPI_IRQ, ipi_irq_dispatch); - - setup_irq_smtc(cpu_ipi_irq, &irq_ipi, (0x100 << MIPS_CPU_IPI_IRQ)); - - irq_set_handler(cpu_ipi_irq, handle_percpu_irq); -} - -/* - * SMTC-specific hacks invoked from elsewhere in the kernel. - */ - - /* - * smtc_ipi_replay is called from raw_local_irq_restore - */ - -void smtc_ipi_replay(void) -{ - unsigned int cpu = smp_processor_id(); - - /* - * To the extent that we've ever turned interrupts off, - * we may have accumulated deferred IPIs. This is subtle. - * we should be OK: If we pick up something and dispatch - * it here, that's great. If we see nothing, but concurrent - * with this operation, another TC sends us an IPI, IXMT - * is clear, and we'll handle it as a real pseudo-interrupt - * and not a pseudo-pseudo interrupt. The important thing - * is to do the last check for queued message *after* the - * re-enabling of interrupts. - */ - while (IPIQ[cpu].head != NULL) { - struct smtc_ipi_q *q = &IPIQ[cpu]; - struct smtc_ipi *pipi; - unsigned long flags; - - /* - * It's just possible we'll come in with interrupts - * already enabled. - */ - local_irq_save(flags); - - spin_lock(&q->lock); - pipi = __smtc_ipi_dq(q); - spin_unlock(&q->lock); - /* - ** But use a raw restore here to avoid recursion. - */ - __arch_local_irq_restore(flags); - - if (pipi) { - self_ipi(pipi); - smtc_cpu_stats[cpu].selfipis++; - } - } -} - -EXPORT_SYMBOL(smtc_ipi_replay); - -void smtc_idle_loop_hook(void) -{ -#ifdef CONFIG_SMTC_IDLE_HOOK_DEBUG - int im; - int flags; - int mtflags; - int bit; - int vpe; - int tc; - int hook_ntcs; - /* - * printk within DMT-protected regions can deadlock, - * so buffer diagnostic messages for later output. - */ - char *pdb_msg; - char id_ho_db_msg[768]; /* worst-case use should be less than 700 */ - - if (atomic_read(&idle_hook_initialized) == 0) { /* fast test */ - if (atomic_add_return(1, &idle_hook_initialized) == 1) { - int mvpconf0; - /* Tedious stuff to just do once */ - mvpconf0 = read_c0_mvpconf0(); - hook_ntcs = ((mvpconf0 & MVPCONF0_PTC) >> MVPCONF0_PTC_SHIFT) + 1; - if (hook_ntcs > NR_CPUS) - hook_ntcs = NR_CPUS; - for (tc = 0; tc < hook_ntcs; tc++) { - tcnoprog[tc] = 0; - clock_hang_reported[tc] = 0; - } - for (vpe = 0; vpe < 2; vpe++) - for (im = 0; im < 8; im++) - imstuckcount[vpe][im] = 0; - printk("Idle loop test hook initialized for %d TCs\n", hook_ntcs); - atomic_set(&idle_hook_initialized, 1000); - } else { - /* Someone else is initializing in parallel - let 'em finish */ - while (atomic_read(&idle_hook_initialized) < 1000) - ; - } - } - - /* Have we stupidly left IXMT set somewhere? */ - if (read_c0_tcstatus() & 0x400) { - write_c0_tcstatus(read_c0_tcstatus() & ~0x400); - ehb(); - printk("Dangling IXMT in cpu_idle()\n"); - } - - /* Have we stupidly left an IM bit turned off? */ -#define IM_LIMIT 2000 - local_irq_save(flags); - mtflags = dmt(); - pdb_msg = &id_ho_db_msg[0]; - im = read_c0_status(); - vpe = current_cpu_data.vpe_id; - for (bit = 0; bit < 8; bit++) { - /* - * In current prototype, I/O interrupts - * are masked for VPE > 0 - */ - if (vpemask[vpe][bit]) { - if (!(im & (0x100 << bit))) - imstuckcount[vpe][bit]++; - else - imstuckcount[vpe][bit] = 0; - if (imstuckcount[vpe][bit] > IM_LIMIT) { - set_c0_status(0x100 << bit); - ehb(); - imstuckcount[vpe][bit] = 0; - pdb_msg += sprintf(pdb_msg, - "Dangling IM %d fixed for VPE %d\n", bit, - vpe); - } - } - } - - emt(mtflags); - local_irq_restore(flags); - if (pdb_msg != &id_ho_db_msg[0]) - printk("CPU%d: %s", smp_processor_id(), id_ho_db_msg); -#endif /* CONFIG_SMTC_IDLE_HOOK_DEBUG */ - - smtc_ipi_replay(); -} - -void smtc_soft_dump(void) -{ - int i; - - printk("Counter Interrupts taken per CPU (TC)\n"); - for (i=0; i < NR_CPUS; i++) { - printk("%d: %ld\n", i, smtc_cpu_stats[i].timerints); - } - printk("Self-IPI invocations:\n"); - for (i=0; i < NR_CPUS; i++) { - printk("%d: %ld\n", i, smtc_cpu_stats[i].selfipis); - } - smtc_ipi_qdump(); - printk("%d Recoveries of \"stolen\" FPU\n", - atomic_read(&smtc_fpu_recoveries)); -} - - -/* - * TLB management routines special to SMTC - */ - -void smtc_get_new_mmu_context(struct mm_struct *mm, unsigned long cpu) -{ - unsigned long flags, mtflags, tcstat, prevhalt, asid; - int tlb, i; - - /* - * It would be nice to be able to use a spinlock here, - * but this is invoked from within TLB flush routines - * that protect themselves with DVPE, so if a lock is - * held by another TC, it'll never be freed. - * - * DVPE/DMT must not be done with interrupts enabled, - * so even so most callers will already have disabled - * them, let's be really careful... - */ - - local_irq_save(flags); - if (smtc_status & SMTC_TLB_SHARED) { - mtflags = dvpe(); - tlb = 0; - } else { - mtflags = dmt(); - tlb = cpu_data[cpu].vpe_id; - } - asid = asid_cache(cpu); - - do { - if (!((asid += ASID_INC) & ASID_MASK) ) { - if (cpu_has_vtag_icache) - flush_icache_all(); - /* Traverse all online CPUs (hack requires contiguous range) */ - for_each_online_cpu(i) { - /* - * We don't need to worry about our own CPU, nor those of - * CPUs who don't share our TLB. - */ - if ((i != smp_processor_id()) && - ((smtc_status & SMTC_TLB_SHARED) || - (cpu_data[i].vpe_id == cpu_data[cpu].vpe_id))) { - settc(cpu_data[i].tc_id); - prevhalt = read_tc_c0_tchalt() & TCHALT_H; - if (!prevhalt) { - write_tc_c0_tchalt(TCHALT_H); - mips_ihb(); - } - tcstat = read_tc_c0_tcstatus(); - smtc_live_asid[tlb][(tcstat & ASID_MASK)] |= (asiduse)(0x1 << i); - if (!prevhalt) - write_tc_c0_tchalt(0); - } - } - if (!asid) /* fix version if needed */ - asid = ASID_FIRST_VERSION; - local_flush_tlb_all(); /* start new asid cycle */ - } - } while (smtc_live_asid[tlb][(asid & ASID_MASK)]); - - /* - * SMTC shares the TLB within VPEs and possibly across all VPEs. - */ - for_each_online_cpu(i) { - if ((smtc_status & SMTC_TLB_SHARED) || - (cpu_data[i].vpe_id == cpu_data[cpu].vpe_id)) - cpu_context(i, mm) = asid_cache(i) = asid; - } - - if (smtc_status & SMTC_TLB_SHARED) - evpe(mtflags); - else - emt(mtflags); - local_irq_restore(flags); -} - -/* - * Invoked from macros defined in mmu_context.h - * which must already have disabled interrupts - * and done a DVPE or DMT as appropriate. - */ - -void smtc_flush_tlb_asid(unsigned long asid) -{ - int entry; - unsigned long ehi; - - entry = read_c0_wired(); - - /* Traverse all non-wired entries */ - while (entry < current_cpu_data.tlbsize) { - write_c0_index(entry); - ehb(); - tlb_read(); - ehb(); - ehi = read_c0_entryhi(); - if ((ehi & ASID_MASK) == asid) { - /* - * Invalidate only entries with specified ASID, - * makiing sure all entries differ. - */ - write_c0_entryhi(CKSEG0 + (entry << (PAGE_SHIFT + 1))); - write_c0_entrylo0(0); - write_c0_entrylo1(0); - mtc0_tlbw_hazard(); - tlb_write_indexed(); - } - entry++; - } - write_c0_index(PARKED_INDEX); - tlbw_use_hazard(); -} - -/* - * Support for single-threading cache flush operations. - */ - -static int halt_state_save[NR_CPUS]; - -/* - * To really, really be sure that nothing is being done - * by other TCs, halt them all. This code assumes that - * a DVPE has already been done, so while their Halted - * state is theoretically architecturally unstable, in - * practice, it's not going to change while we're looking - * at it. - */ - -void smtc_cflush_lockdown(void) -{ - int cpu; - - for_each_online_cpu(cpu) { - if (cpu != smp_processor_id()) { - settc(cpu_data[cpu].tc_id); - halt_state_save[cpu] = read_tc_c0_tchalt(); - write_tc_c0_tchalt(TCHALT_H); - } - } - mips_ihb(); -} - -/* It would be cheating to change the cpu_online states during a flush! */ - -void smtc_cflush_release(void) -{ - int cpu; - - /* - * Start with a hazard barrier to ensure - * that all CACHE ops have played through. - */ - mips_ihb(); - - for_each_online_cpu(cpu) { - if (cpu != smp_processor_id()) { - settc(cpu_data[cpu].tc_id); - write_tc_c0_tchalt(halt_state_save[cpu]); - } - } - mips_ihb(); -} diff --git a/arch/mips/kernel/sync-r4k.c b/arch/mips/kernel/sync-r4k.c index c24ad5f4b32..2242bdd4370 100644 --- a/arch/mips/kernel/sync-r4k.c +++ b/arch/mips/kernel/sync-r4k.c @@ -6,8 +6,6 @@ * not have done anything significant (but they may have had interrupts * enabled briefly - prom_smp_finish() should not be responsible for enabling * interrupts...) - * - * FIXME: broken for SMTC */ #include <linux/kernel.h> @@ -33,14 +31,6 @@ void synchronise_count_master(int cpu) unsigned long flags; unsigned int initcount; -#ifdef CONFIG_MIPS_MT_SMTC - /* - * SMTC needs to synchronise per VPE, not per CPU - * ignore for now - */ - return; -#endif - printk(KERN_INFO "Synchronize counters for CPU %u: ", cpu); local_irq_save(flags); @@ -110,14 +100,6 @@ void synchronise_count_slave(int cpu) int i; unsigned int initcount; -#ifdef CONFIG_MIPS_MT_SMTC - /* - * SMTC needs to synchronise per VPE, not per CPU - * ignore for now - */ - return; -#endif - /* * Not every cpu is online at the time this gets called, * so we first wait for the master to say everyone is ready diff --git a/arch/mips/kernel/time.c b/arch/mips/kernel/time.c index dcb8e5d3bb8..8d0170969e2 100644 --- a/arch/mips/kernel/time.c +++ b/arch/mips/kernel/time.c @@ -26,7 +26,6 @@ #include <asm/cpu-features.h> #include <asm/cpu-type.h> #include <asm/div64.h> -#include <asm/smtc_ipi.h> #include <asm/time.h> /* diff --git a/arch/mips/kernel/traps.c b/arch/mips/kernel/traps.c index 8119ac2fdfc..51706d6dd5b 100644 --- a/arch/mips/kernel/traps.c +++ b/arch/mips/kernel/traps.c @@ -15,6 +15,7 @@ #include <linux/bug.h> #include <linux/compiler.h> #include <linux/context_tracking.h> +#include <linux/cpu_pm.h> #include <linux/kexec.h> #include <linux/init.h> #include <linux/kernel.h> @@ -370,9 +371,6 @@ void __noreturn die(const char *str, struct pt_regs *regs) { static int die_counter; int sig = SIGSEGV; -#ifdef CONFIG_MIPS_MT_SMTC - unsigned long dvpret; -#endif /* CONFIG_MIPS_MT_SMTC */ oops_enter(); @@ -382,13 +380,7 @@ void __noreturn die(const char *str, struct pt_regs *regs) console_verbose(); raw_spin_lock_irq(&die_lock); -#ifdef CONFIG_MIPS_MT_SMTC - dvpret = dvpe(); -#endif /* CONFIG_MIPS_MT_SMTC */ bust_spinlocks(1); -#ifdef CONFIG_MIPS_MT_SMTC - mips_mt_regdump(dvpret); -#endif /* CONFIG_MIPS_MT_SMTC */ printk("%s[#%d]:\n", str, ++die_counter); show_registers(regs); @@ -712,10 +704,12 @@ int process_fpemu_return(int sig, void __user *fault_addr) si.si_addr = fault_addr; si.si_signo = sig; if (sig == SIGSEGV) { + down_read(¤t->mm->mmap_sem); if (find_vma(current->mm, (unsigned long)fault_addr)) si.si_code = SEGV_ACCERR; else si.si_code = SEGV_MAPERR; + up_read(¤t->mm->mmap_sem); } else { si.si_code = BUS_ADRERR; } @@ -1759,19 +1753,6 @@ static void *set_vi_srs_handler(int n, vi_handler_t addr, int srs) extern char rollback_except_vec_vi; char *vec_start = using_rollback_handler() ? &rollback_except_vec_vi : &except_vec_vi; -#ifdef CONFIG_MIPS_MT_SMTC - /* - * We need to provide the SMTC vectored interrupt handler - * not only with the address of the handler, but with the - * Status.IM bit to be masked before going there. - */ - extern char except_vec_vi_mori; -#if defined(CONFIG_CPU_MICROMIPS) || defined(CONFIG_CPU_BIG_ENDIAN) - const int mori_offset = &except_vec_vi_mori - vec_start + 2; -#else - const int mori_offset = &except_vec_vi_mori - vec_start; -#endif -#endif /* CONFIG_MIPS_MT_SMTC */ #if defined(CONFIG_CPU_MICROMIPS) || defined(CONFIG_CPU_BIG_ENDIAN) const int lui_offset = &except_vec_vi_lui - vec_start + 2; const int ori_offset = &except_vec_vi_ori - vec_start + 2; @@ -1795,12 +1776,6 @@ static void *set_vi_srs_handler(int n, vi_handler_t addr, int srs) #else handler_len); #endif -#ifdef CONFIG_MIPS_MT_SMTC - BUG_ON(n > 7); /* Vector index %d exceeds SMTC maximum. */ - - h = (u16 *)(b + mori_offset); - *h = (0x100 << n); -#endif /* CONFIG_MIPS_MT_SMTC */ h = (u16 *)(b + lui_offset); *h = (handler >> 16) & 0xffff; h = (u16 *)(b + ori_offset); @@ -1865,32 +1840,16 @@ static int __init ulri_disable(char *s) } __setup("noulri", ulri_disable); -void per_cpu_trap_init(bool is_boot_cpu) +/* configure STATUS register */ +static void configure_status(void) { - unsigned int cpu = smp_processor_id(); - unsigned int status_set = ST0_CU0; - unsigned int hwrena = cpu_hwrena_impl_bits; -#ifdef CONFIG_MIPS_MT_SMTC - int secondaryTC = 0; - int bootTC = (cpu == 0); - - /* - * Only do per_cpu_trap_init() for first TC of Each VPE. - * Note that this hack assumes that the SMTC init code - * assigns TCs consecutively and in ascending order. - */ - - if (((read_c0_tcbind() & TCBIND_CURTC) != 0) && - ((read_c0_tcbind() & TCBIND_CURVPE) == cpu_data[cpu - 1].vpe_id)) - secondaryTC = 1; -#endif /* CONFIG_MIPS_MT_SMTC */ - /* * Disable coprocessors and select 32-bit or 64-bit addressing * and the 16/32 or 32/32 FPR register model. Reset the BEV * flag that some firmware may have left set and the TS bit (for * IP27). Set XX for ISA IV code to work. */ + unsigned int status_set = ST0_CU0; #ifdef CONFIG_64BIT status_set |= ST0_FR|ST0_KX|ST0_SX|ST0_UX; #endif @@ -1901,6 +1860,12 @@ void per_cpu_trap_init(bool is_boot_cpu) change_c0_status(ST0_CU|ST0_MX|ST0_RE|ST0_FR|ST0_BEV|ST0_TS|ST0_KX|ST0_SX|ST0_UX, status_set); +} + +/* configure HWRENA register */ +static void configure_hwrena(void) +{ + unsigned int hwrena = cpu_hwrena_impl_bits; if (cpu_has_mips_r2) hwrena |= 0x0000000f; @@ -1910,11 +1875,10 @@ void per_cpu_trap_init(bool is_boot_cpu) if (hwrena) write_c0_hwrena(hwrena); +} -#ifdef CONFIG_MIPS_MT_SMTC - if (!secondaryTC) { -#endif /* CONFIG_MIPS_MT_SMTC */ - +static void configure_exception_vector(void) +{ if (cpu_has_veic || cpu_has_vint) { unsigned long sr = set_c0_status(ST0_BEV); write_c0_ebase(ebase); @@ -1930,6 +1894,16 @@ void per_cpu_trap_init(bool is_boot_cpu) } else set_c0_cause(CAUSEF_IV); } +} + +void per_cpu_trap_init(bool is_boot_cpu) +{ + unsigned int cpu = smp_processor_id(); + + configure_status(); + configure_hwrena(); + + configure_exception_vector(); /* * Before R2 both interrupt numbers were fixed to 7, so on R2 only: @@ -1949,10 +1923,6 @@ void per_cpu_trap_init(bool is_boot_cpu) cp0_perfcount_irq = -1; } -#ifdef CONFIG_MIPS_MT_SMTC - } -#endif /* CONFIG_MIPS_MT_SMTC */ - if (!cpu_data[cpu].asid_cache) cpu_data[cpu].asid_cache = ASID_FIRST_VERSION; @@ -1961,23 +1931,10 @@ void per_cpu_trap_init(bool is_boot_cpu) BUG_ON(current->mm); enter_lazy_tlb(&init_mm, current); -#ifdef CONFIG_MIPS_MT_SMTC - if (bootTC) { -#endif /* CONFIG_MIPS_MT_SMTC */ /* Boot CPU's cache setup in setup_arch(). */ if (!is_boot_cpu) cpu_cache_init(); tlb_init(); -#ifdef CONFIG_MIPS_MT_SMTC - } else if (!secondaryTC) { - /* - * First TC in non-boot VPE must do subset of tlb_init() - * for MMU countrol registers. - */ - write_c0_pagemask(PM_DEFAULT_MASK); - write_c0_wired(0); - } -#endif /* CONFIG_MIPS_MT_SMTC */ TLBMISS_HANDLER_SETUP(); } @@ -2185,3 +2142,32 @@ void __init trap_init(void) cu2_notifier(default_cu2_call, 0x80000000); /* Run last */ } + +static int trap_pm_notifier(struct notifier_block *self, unsigned long cmd, + void *v) +{ + switch (cmd) { + case CPU_PM_ENTER_FAILED: + case CPU_PM_EXIT: + configure_status(); + configure_hwrena(); + configure_exception_vector(); + + /* Restore register with CPU number for TLB handlers */ + TLBMISS_HANDLER_RESTORE(); + + break; + } + + return NOTIFY_OK; +} + +static struct notifier_block trap_pm_notifier_block = { + .notifier_call = trap_pm_notifier, +}; + +static int __init trap_pm_init(void) +{ + return cpu_pm_register_notifier(&trap_pm_notifier_block); +} +arch_initcall(trap_pm_init); diff --git a/arch/mips/kernel/vpe-mt.c b/arch/mips/kernel/vpe-mt.c index 949ae0e1701..2e003b11a09 100644 --- a/arch/mips/kernel/vpe-mt.c +++ b/arch/mips/kernel/vpe-mt.c @@ -127,9 +127,8 @@ int vpe_run(struct vpe *v) clear_c0_mvpcontrol(MVPCONTROL_VPC); /* - * SMTC/SMVP kernels manage VPE enable independently, - * but uniprocessor kernels need to turn it on, even - * if that wasn't the pre-dvpe() state. + * SMVP kernels manage VPE enable independently, but uniprocessor + * kernels need to turn it on, even if that wasn't the pre-dvpe() state. */ #ifdef CONFIG_SMP evpe(vpeflags); @@ -454,12 +453,11 @@ int __init vpe_module_init(void) settc(tc); - /* Any TC that is bound to VPE0 gets left as is - in - * case we are running SMTC on VPE0. A TC that is bound - * to any other VPE gets bound to VPE0, ideally I'd like - * to make it homeless but it doesn't appear to let me - * bind a TC to a non-existent VPE. Which is perfectly - * reasonable. + /* + * A TC that is bound to any other VPE gets bound to + * VPE0, ideally I'd like to make it homeless but it + * doesn't appear to let me bind a TC to a non-existent + * VPE. Which is perfectly reasonable. * * The (un)bound state is visible to an EJTAG probe so * may notify GDB... diff --git a/arch/mips/lantiq/irq.c b/arch/mips/lantiq/irq.c index 85685e1cdb8..030568a70ac 100644 --- a/arch/mips/lantiq/irq.c +++ b/arch/mips/lantiq/irq.c @@ -61,7 +61,7 @@ /* we have a cascade of 8 irqs */ #define MIPS_CPU_IRQ_CASCADE 8 -#if defined(CONFIG_MIPS_MT_SMP) || defined(CONFIG_MIPS_MT_SMTC) +#ifdef CONFIG_MIPS_MT_SMP int gic_present; #endif @@ -440,7 +440,7 @@ int __init icu_of_init(struct device_node *node, struct device_node *parent) arch_init_ipiirq(MIPS_CPU_IRQ_BASE + MIPS_CPU_IPI_CALL_IRQ, &irq_call); #endif -#if !defined(CONFIG_MIPS_MT_SMP) && !defined(CONFIG_MIPS_MT_SMTC) +#ifndef CONFIG_MIPS_MT_SMP set_c0_status(IE_IRQ0 | IE_IRQ1 | IE_IRQ2 | IE_IRQ3 | IE_IRQ4 | IE_IRQ5); #else diff --git a/arch/mips/lib/delay.c b/arch/mips/lib/delay.c index 705cfb7c1a7..21d27c6819a 100644 --- a/arch/mips/lib/delay.c +++ b/arch/mips/lib/delay.c @@ -11,7 +11,9 @@ #include <linux/module.h> #include <linux/param.h> #include <linux/smp.h> +#include <linux/stringify.h> +#include <asm/asm.h> #include <asm/compiler.h> #include <asm/war.h> @@ -27,11 +29,7 @@ void __delay(unsigned long loops) " .set noreorder \n" " .align 3 \n" "1: bnez %0, 1b \n" -#if BITS_PER_LONG == 32 - " subu %0, %1 \n" -#else - " dsubu %0, %1 \n" -#endif + " " __stringify(LONG_SUBU) " %0, %1 \n" " .set reorder \n" : "=r" (loops) : GCC_DADDI_IMM_ASM() (1), "0" (loops)); diff --git a/arch/mips/lib/mips-atomic.c b/arch/mips/lib/mips-atomic.c index 6807f7172ea..57bcdaf1f1c 100644 --- a/arch/mips/lib/mips-atomic.c +++ b/arch/mips/lib/mips-atomic.c @@ -15,7 +15,7 @@ #include <linux/export.h> #include <linux/stringify.h> -#if !defined(CONFIG_CPU_MIPSR2) || defined(CONFIG_MIPS_MT_SMTC) +#ifndef CONFIG_CPU_MIPSR2 /* * For cli() we have to insert nops to make sure that the new value @@ -42,12 +42,7 @@ notrace void arch_local_irq_disable(void) __asm__ __volatile__( " .set push \n" " .set noat \n" -#ifdef CONFIG_MIPS_MT_SMTC - " mfc0 $1, $2, 1 \n" - " ori $1, 0x400 \n" - " .set noreorder \n" - " mtc0 $1, $2, 1 \n" -#elif defined(CONFIG_CPU_MIPSR2) +#if defined(CONFIG_CPU_MIPSR2) /* see irqflags.h for inline function */ #else " mfc0 $1,$12 \n" @@ -77,13 +72,7 @@ notrace unsigned long arch_local_irq_save(void) " .set push \n" " .set reorder \n" " .set noat \n" -#ifdef CONFIG_MIPS_MT_SMTC - " mfc0 %[flags], $2, 1 \n" - " ori $1, %[flags], 0x400 \n" - " .set noreorder \n" - " mtc0 $1, $2, 1 \n" - " andi %[flags], %[flags], 0x400 \n" -#elif defined(CONFIG_CPU_MIPSR2) +#if defined(CONFIG_CPU_MIPSR2) /* see irqflags.h for inline function */ #else " mfc0 %[flags], $12 \n" @@ -108,29 +97,13 @@ notrace void arch_local_irq_restore(unsigned long flags) { unsigned long __tmp1; -#ifdef CONFIG_MIPS_MT_SMTC - /* - * SMTC kernel needs to do a software replay of queued - * IPIs, at the cost of branch and call overhead on each - * local_irq_restore() - */ - if (unlikely(!(flags & 0x0400))) - smtc_ipi_replay(); -#endif preempt_disable(); __asm__ __volatile__( " .set push \n" " .set noreorder \n" " .set noat \n" -#ifdef CONFIG_MIPS_MT_SMTC - " mfc0 $1, $2, 1 \n" - " andi %[flags], 0x400 \n" - " ori $1, 0x400 \n" - " xori $1, 0x400 \n" - " or %[flags], $1 \n" - " mtc0 %[flags], $2, 1 \n" -#elif defined(CONFIG_CPU_MIPSR2) && defined(CONFIG_IRQ_CPU) +#if defined(CONFIG_CPU_MIPSR2) && defined(CONFIG_IRQ_CPU) /* see irqflags.h for inline function */ #elif defined(CONFIG_CPU_MIPSR2) /* see irqflags.h for inline function */ @@ -163,14 +136,7 @@ notrace void __arch_local_irq_restore(unsigned long flags) " .set push \n" " .set noreorder \n" " .set noat \n" -#ifdef CONFIG_MIPS_MT_SMTC - " mfc0 $1, $2, 1 \n" - " andi %[flags], 0x400 \n" - " ori $1, 0x400 \n" - " xori $1, 0x400 \n" - " or %[flags], $1 \n" - " mtc0 %[flags], $2, 1 \n" -#elif defined(CONFIG_CPU_MIPSR2) && defined(CONFIG_IRQ_CPU) +#if defined(CONFIG_CPU_MIPSR2) && defined(CONFIG_IRQ_CPU) /* see irqflags.h for inline function */ #elif defined(CONFIG_CPU_MIPSR2) /* see irqflags.h for inline function */ @@ -192,4 +158,4 @@ notrace void __arch_local_irq_restore(unsigned long flags) } EXPORT_SYMBOL(__arch_local_irq_restore); -#endif /* !defined(CONFIG_CPU_MIPSR2) || defined(CONFIG_MIPS_MT_SMTC) */ +#endif /* !CONFIG_CPU_MIPSR2 */ diff --git a/arch/mips/loongson/Kconfig b/arch/mips/loongson/Kconfig index 603d79a95f4..e6a86ccc442 100644 --- a/arch/mips/loongson/Kconfig +++ b/arch/mips/loongson/Kconfig @@ -95,10 +95,11 @@ config CS5536 config CS5536_MFGPT bool "CS5536 MFGPT Timer" - depends on CS5536 + depends on CS5536 && !HIGH_RES_TIMERS select MIPS_EXTERNAL_TIMER help - This option enables the mfgpt0 timer of AMD CS5536. + This option enables the mfgpt0 timer of AMD CS5536. With this timer + switched on you can not use high resolution timers. If you want to enable the Loongson2 CPUFreq Driver, Please enable this option at first, otherwise, You will get wrong system time. diff --git a/arch/mips/loongson/loongson-3/smp.c b/arch/mips/loongson/loongson-3/smp.c index c665fe16d4c..1e8894020ea 100644 --- a/arch/mips/loongson/loongson-3/smp.c +++ b/arch/mips/loongson/loongson-3/smp.c @@ -279,13 +279,6 @@ static void loongson3_boot_secondary(int cpu, struct task_struct *idle) loongson3_ipi_write64(startargs[0], (void *)(ipi_mailbox_buf[cpu]+0x0)); } -/* - * Final cleanup after all secondaries booted - */ -static void __init loongson3_cpus_done(void) -{ -} - #ifdef CONFIG_HOTPLUG_CPU static int loongson3_cpu_disable(void) @@ -432,7 +425,6 @@ struct plat_smp_ops loongson3_smp_ops = { .send_ipi_mask = loongson3_send_ipi_mask, .init_secondary = loongson3_init_secondary, .smp_finish = loongson3_smp_finish, - .cpus_done = loongson3_cpus_done, .boot_secondary = loongson3_boot_secondary, .smp_setup = loongson3_smp_setup, .prepare_cpus = loongson3_prepare_cpus, diff --git a/arch/mips/loongson1/Kconfig b/arch/mips/loongson1/Kconfig index fbf75f63579..e23c25d0996 100644 --- a/arch/mips/loongson1/Kconfig +++ b/arch/mips/loongson1/Kconfig @@ -14,6 +14,7 @@ config LOONGSON1_LS1B select SYS_SUPPORTS_32BIT_KERNEL select SYS_SUPPORTS_LITTLE_ENDIAN select SYS_SUPPORTS_HIGHMEM + select SYS_SUPPORTS_MIPS16 select SYS_HAS_EARLY_PRINTK select COMMON_CLK diff --git a/arch/mips/math-emu/Makefile b/arch/mips/math-emu/Makefile index 121a848a359..619cfc1a244 100644 --- a/arch/mips/math-emu/Makefile +++ b/arch/mips/math-emu/Makefile @@ -2,10 +2,12 @@ # Makefile for the Linux/MIPS kernel FPU emulation. # -obj-y := cp1emu.o ieee754m.o ieee754d.o ieee754dp.o ieee754sp.o ieee754.o \ - ieee754xcpt.o dp_frexp.o dp_modf.o dp_div.o dp_mul.o dp_sub.o \ - dp_add.o dp_fsp.o dp_cmp.o dp_logb.o dp_scalb.o dp_simple.o \ - dp_tint.o dp_fint.o dp_tlong.o dp_flong.o sp_frexp.o sp_modf.o \ - sp_div.o sp_mul.o sp_sub.o sp_add.o sp_fdp.o sp_cmp.o sp_logb.o \ - sp_scalb.o sp_simple.o sp_tint.o sp_fint.o sp_tlong.o sp_flong.o \ - dp_sqrt.o sp_sqrt.o kernel_linkage.o dsemul.o +obj-y += cp1emu.o ieee754dp.o ieee754sp.o ieee754.o dp_div.o dp_mul.o \ + dp_sub.o dp_add.o dp_fsp.o dp_cmp.o dp_simple.o dp_tint.o \ + dp_fint.o dp_tlong.o dp_flong.o sp_div.o sp_mul.o sp_sub.o \ + sp_add.o sp_fdp.o sp_cmp.o sp_simple.o sp_tint.o sp_fint.o \ + sp_tlong.o sp_flong.o dsemul.o + +lib-y += ieee754d.o dp_sqrt.o sp_sqrt.o + +obj-$(CONFIG_DEBUG_FS) += me-debugfs.o diff --git a/arch/mips/math-emu/cp1emu.c b/arch/mips/math-emu/cp1emu.c index 7b3c9acae68..736c17a226e 100644 --- a/arch/mips/math-emu/cp1emu.c +++ b/arch/mips/math-emu/cp1emu.c @@ -1,5 +1,5 @@ /* - * cp1emu.c: a MIPS coprocessor 1 (fpu) instruction emulator + * cp1emu.c: a MIPS coprocessor 1 (FPU) instruction emulator * * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. @@ -18,61 +18,46 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. * * A complete emulator for MIPS coprocessor 1 instructions. This is * required for #float(switch) or #float(trap), where it catches all * COP1 instructions via the "CoProcessor Unusable" exception. * * More surprisingly it is also required for #float(ieee), to help out - * the hardware fpu at the boundaries of the IEEE-754 representation + * the hardware FPU at the boundaries of the IEEE-754 representation * (denormalised values, infinities, underflow, etc). It is made * quite nasty because emulation of some non-COP1 instructions is * required, e.g. in branch delay slots. * - * Note if you know that you won't have an fpu, then you'll get much + * Note if you know that you won't have an FPU, then you'll get much * better performance by compiling with -msoft-float! */ #include <linux/sched.h> -#include <linux/module.h> #include <linux/debugfs.h> +#include <linux/kconfig.h> +#include <linux/percpu-defs.h> #include <linux/perf_event.h> +#include <asm/branch.h> #include <asm/inst.h> -#include <asm/bootinfo.h> -#include <asm/processor.h> #include <asm/ptrace.h> #include <asm/signal.h> -#include <asm/mipsregs.h> +#include <asm/uaccess.h> + +#include <asm/processor.h> #include <asm/fpu_emulator.h> #include <asm/fpu.h> -#include <asm/uaccess.h> -#include <asm/branch.h> #include "ieee754.h" -/* Strap kernel emulator for full MIPS IV emulation */ - -#ifdef __mips -#undef __mips -#endif -#define __mips 4 - /* Function which emulates a floating point instruction. */ static int fpu_emu(struct pt_regs *, struct mips_fpu_struct *, mips_instruction); -#if __mips >= 4 && __mips != 32 static int fpux_emu(struct pt_regs *, struct mips_fpu_struct *, mips_instruction, void *__user *); -#endif - -/* Further private data for which no space exists in mips_fpu_struct */ - -#ifdef CONFIG_DEBUG_FS -DEFINE_PER_CPU(struct mips_fpu_emulator_stats, fpuemustats); -#endif /* Control registers */ @@ -82,27 +67,6 @@ DEFINE_PER_CPU(struct mips_fpu_emulator_stats, fpuemustats); /* Determine rounding mode from the RM bits of the FCSR */ #define modeindex(v) ((v) & FPU_CSR_RM) -/* microMIPS bitfields */ -#define MM_POOL32A_MINOR_MASK 0x3f -#define MM_POOL32A_MINOR_SHIFT 0x6 -#define MM_MIPS32_COND_FC 0x30 - -/* Convert Mips rounding mode (0..3) to IEEE library modes. */ -static const unsigned char ieee_rm[4] = { - [FPU_CSR_RN] = IEEE754_RN, - [FPU_CSR_RZ] = IEEE754_RZ, - [FPU_CSR_RU] = IEEE754_RU, - [FPU_CSR_RD] = IEEE754_RD, -}; -/* Convert IEEE library modes to Mips rounding mode (0..3). */ -static const unsigned char mips_rm[4] = { - [IEEE754_RN] = FPU_CSR_RN, - [IEEE754_RZ] = FPU_CSR_RZ, - [IEEE754_RD] = FPU_CSR_RD, - [IEEE754_RU] = FPU_CSR_RU, -}; - -#if __mips >= 4 /* convert condition code register number to csr bit */ static const unsigned int fpucondbit[8] = { FPU_CSR_COND0, @@ -114,10 +78,6 @@ static const unsigned int fpucondbit[8] = { FPU_CSR_COND6, FPU_CSR_COND7 }; -#endif - -/* (microMIPS) Convert 16-bit register encoding to 32-bit register encoding. */ -static const unsigned int reg16to32map[8] = {16, 17, 2, 3, 4, 5, 6, 7}; /* (microMIPS) Convert certain microMIPS instructions to MIPS32 format. */ static const int sd_format[] = {16, 17, 0, 0, 0, 0, 0, 0}; @@ -466,199 +426,6 @@ static int microMIPS32_to_MIPS32(union mips_instruction *insn_ptr) return 0; } -int mm_isBranchInstr(struct pt_regs *regs, struct mm_decoded_insn dec_insn, - unsigned long *contpc) -{ - union mips_instruction insn = (union mips_instruction)dec_insn.insn; - int bc_false = 0; - unsigned int fcr31; - unsigned int bit; - - if (!cpu_has_mmips) - return 0; - - switch (insn.mm_i_format.opcode) { - case mm_pool32a_op: - if ((insn.mm_i_format.simmediate & MM_POOL32A_MINOR_MASK) == - mm_pool32axf_op) { - switch (insn.mm_i_format.simmediate >> - MM_POOL32A_MINOR_SHIFT) { - case mm_jalr_op: - case mm_jalrhb_op: - case mm_jalrs_op: - case mm_jalrshb_op: - if (insn.mm_i_format.rt != 0) /* Not mm_jr */ - regs->regs[insn.mm_i_format.rt] = - regs->cp0_epc + - dec_insn.pc_inc + - dec_insn.next_pc_inc; - *contpc = regs->regs[insn.mm_i_format.rs]; - return 1; - } - } - break; - case mm_pool32i_op: - switch (insn.mm_i_format.rt) { - case mm_bltzals_op: - case mm_bltzal_op: - regs->regs[31] = regs->cp0_epc + - dec_insn.pc_inc + - dec_insn.next_pc_inc; - /* Fall through */ - case mm_bltz_op: - if ((long)regs->regs[insn.mm_i_format.rs] < 0) - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - (insn.mm_i_format.simmediate << 1); - else - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - dec_insn.next_pc_inc; - return 1; - case mm_bgezals_op: - case mm_bgezal_op: - regs->regs[31] = regs->cp0_epc + - dec_insn.pc_inc + - dec_insn.next_pc_inc; - /* Fall through */ - case mm_bgez_op: - if ((long)regs->regs[insn.mm_i_format.rs] >= 0) - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - (insn.mm_i_format.simmediate << 1); - else - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - dec_insn.next_pc_inc; - return 1; - case mm_blez_op: - if ((long)regs->regs[insn.mm_i_format.rs] <= 0) - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - (insn.mm_i_format.simmediate << 1); - else - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - dec_insn.next_pc_inc; - return 1; - case mm_bgtz_op: - if ((long)regs->regs[insn.mm_i_format.rs] <= 0) - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - (insn.mm_i_format.simmediate << 1); - else - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - dec_insn.next_pc_inc; - return 1; - case mm_bc2f_op: - case mm_bc1f_op: - bc_false = 1; - /* Fall through */ - case mm_bc2t_op: - case mm_bc1t_op: - preempt_disable(); - if (is_fpu_owner()) - asm volatile("cfc1\t%0,$31" : "=r" (fcr31)); - else - fcr31 = current->thread.fpu.fcr31; - preempt_enable(); - - if (bc_false) - fcr31 = ~fcr31; - - bit = (insn.mm_i_format.rs >> 2); - bit += (bit != 0); - bit += 23; - if (fcr31 & (1 << bit)) - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - (insn.mm_i_format.simmediate << 1); - else - *contpc = regs->cp0_epc + - dec_insn.pc_inc + dec_insn.next_pc_inc; - return 1; - } - break; - case mm_pool16c_op: - switch (insn.mm_i_format.rt) { - case mm_jalr16_op: - case mm_jalrs16_op: - regs->regs[31] = regs->cp0_epc + - dec_insn.pc_inc + dec_insn.next_pc_inc; - /* Fall through */ - case mm_jr16_op: - *contpc = regs->regs[insn.mm_i_format.rs]; - return 1; - } - break; - case mm_beqz16_op: - if ((long)regs->regs[reg16to32map[insn.mm_b1_format.rs]] == 0) - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - (insn.mm_b1_format.simmediate << 1); - else - *contpc = regs->cp0_epc + - dec_insn.pc_inc + dec_insn.next_pc_inc; - return 1; - case mm_bnez16_op: - if ((long)regs->regs[reg16to32map[insn.mm_b1_format.rs]] != 0) - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - (insn.mm_b1_format.simmediate << 1); - else - *contpc = regs->cp0_epc + - dec_insn.pc_inc + dec_insn.next_pc_inc; - return 1; - case mm_b16_op: - *contpc = regs->cp0_epc + dec_insn.pc_inc + - (insn.mm_b0_format.simmediate << 1); - return 1; - case mm_beq32_op: - if (regs->regs[insn.mm_i_format.rs] == - regs->regs[insn.mm_i_format.rt]) - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - (insn.mm_i_format.simmediate << 1); - else - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - dec_insn.next_pc_inc; - return 1; - case mm_bne32_op: - if (regs->regs[insn.mm_i_format.rs] != - regs->regs[insn.mm_i_format.rt]) - *contpc = regs->cp0_epc + - dec_insn.pc_inc + - (insn.mm_i_format.simmediate << 1); - else - *contpc = regs->cp0_epc + - dec_insn.pc_inc + dec_insn.next_pc_inc; - return 1; - case mm_jalx32_op: - regs->regs[31] = regs->cp0_epc + - dec_insn.pc_inc + dec_insn.next_pc_inc; - *contpc = regs->cp0_epc + dec_insn.pc_inc; - *contpc >>= 28; - *contpc <<= 28; - *contpc |= (insn.j_format.target << 2); - return 1; - case mm_jals32_op: - case mm_jal32_op: - regs->regs[31] = regs->cp0_epc + - dec_insn.pc_inc + dec_insn.next_pc_inc; - /* Fall through */ - case mm_j32_op: - *contpc = regs->cp0_epc + dec_insn.pc_inc; - *contpc >>= 27; - *contpc <<= 27; - *contpc |= (insn.j_format.target << 1); - set_isa16_mode(*contpc); - return 1; - } - return 0; -} - /* * Redundant with logic already in kernel/branch.c, * embedded in compute_return_epc. At some point, @@ -817,7 +584,11 @@ static int isBranchInstr(struct pt_regs *regs, struct mm_decoded_insn dec_insn, if (insn.i_format.rs == bc_op) { preempt_disable(); if (is_fpu_owner()) - asm volatile("cfc1\t%0,$31" : "=r" (fcr31)); + asm volatile( + ".set push\n" + "\t.set mips1\n" + "\tcfc1\t%0,$31\n" + "\t.set pop" : "=r" (fcr31)); else fcr31 = current->thread.fpu.fcr31; preempt_enable(); @@ -867,23 +638,25 @@ static int isBranchInstr(struct pt_regs *regs, struct mm_decoded_insn dec_insn, */ static inline int cop1_64bit(struct pt_regs *xcp) { -#if defined(CONFIG_64BIT) && !defined(CONFIG_MIPS32_O32) - return 1; -#elif defined(CONFIG_32BIT) && !defined(CONFIG_MIPS_O32_FP64_SUPPORT) - return 0; -#else + if (config_enabled(CONFIG_64BIT) && !config_enabled(CONFIG_MIPS32_O32)) + return 1; + else if (config_enabled(CONFIG_32BIT) && + !config_enabled(CONFIG_MIPS_O32_FP64_SUPPORT)) + return 0; + return !test_thread_flag(TIF_32BIT_FPREGS); -#endif } -#define SIFROMREG(si, x) do { \ +#define SIFROMREG(si, x) \ +do { \ if (cop1_64bit(xcp)) \ (si) = get_fpr32(&ctx->fpr[x], 0); \ else \ (si) = get_fpr32(&ctx->fpr[(x) & ~1], (x) & 1); \ } while (0) -#define SITOREG(si, x) do { \ +#define SITOREG(si, x) \ +do { \ if (cop1_64bit(xcp)) { \ unsigned i; \ set_fpr32(&ctx->fpr[x], 0, si); \ @@ -896,17 +669,19 @@ static inline int cop1_64bit(struct pt_regs *xcp) #define SIFROMHREG(si, x) ((si) = get_fpr32(&ctx->fpr[x], 1)) -#define SITOHREG(si, x) do { \ +#define SITOHREG(si, x) \ +do { \ unsigned i; \ set_fpr32(&ctx->fpr[x], 1, si); \ for (i = 2; i < ARRAY_SIZE(ctx->fpr[x].val32); i++) \ set_fpr32(&ctx->fpr[x], i, 0); \ } while (0) -#define DIFROMREG(di, x) \ +#define DIFROMREG(di, x) \ ((di) = get_fpr64(&ctx->fpr[(x) & ~(cop1_64bit(xcp) == 0)], 0)) -#define DITOREG(di, x) do { \ +#define DITOREG(di, x) \ +do { \ unsigned fpr, i; \ fpr = (x) & ~(cop1_64bit(xcp) == 0); \ set_fpr64(&ctx->fpr[fpr], 0, di); \ @@ -927,23 +702,36 @@ static inline int cop1_64bit(struct pt_regs *xcp) static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, struct mm_decoded_insn dec_insn, void *__user *fault_addr) { - mips_instruction ir; unsigned long contpc = xcp->cp0_epc + dec_insn.pc_inc; - unsigned int cond; - int pc_inc; + unsigned int cond, cbit; + mips_instruction ir; + int likely, pc_inc; + u32 __user *wva; + u64 __user *dva; + u32 value; + u32 wval; + u64 dval; + int sig; + + /* + * These are giving gcc a gentle hint about what to expect in + * dec_inst in order to do better optimization. + */ + if (!cpu_has_mmips && dec_insn.micro_mips_mode) + unreachable(); /* XXX NEC Vr54xx bug workaround */ - if (xcp->cp0_cause & CAUSEF_BD) { + if (delay_slot(xcp)) { if (dec_insn.micro_mips_mode) { if (!mm_isBranchInstr(xcp, dec_insn, &contpc)) - xcp->cp0_cause &= ~CAUSEF_BD; + clear_delay_slot(xcp); } else { if (!isBranchInstr(xcp, dec_insn, &contpc)) - xcp->cp0_cause &= ~CAUSEF_BD; + clear_delay_slot(xcp); } } - if (xcp->cp0_cause & CAUSEF_BD) { + if (delay_slot(xcp)) { /* * The instruction to be emulated is in a branch delay slot * which means that we have to emulate the branch instruction @@ -985,96 +773,85 @@ static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, return SIGILL; } - emul: +emul: perf_sw_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, xcp, 0); MIPS_FPU_EMU_INC_STATS(emulated); switch (MIPSInst_OPCODE(ir)) { - case ldc1_op:{ - u64 __user *va = (u64 __user *) (xcp->regs[MIPSInst_RS(ir)] + - MIPSInst_SIMM(ir)); - u64 val; - + case ldc1_op: + dva = (u64 __user *) (xcp->regs[MIPSInst_RS(ir)] + + MIPSInst_SIMM(ir)); MIPS_FPU_EMU_INC_STATS(loads); - if (!access_ok(VERIFY_READ, va, sizeof(u64))) { + if (!access_ok(VERIFY_READ, dva, sizeof(u64))) { MIPS_FPU_EMU_INC_STATS(errors); - *fault_addr = va; + *fault_addr = dva; return SIGBUS; } - if (__get_user(val, va)) { + if (__get_user(dval, dva)) { MIPS_FPU_EMU_INC_STATS(errors); - *fault_addr = va; + *fault_addr = dva; return SIGSEGV; } - DITOREG(val, MIPSInst_RT(ir)); + DITOREG(dval, MIPSInst_RT(ir)); break; - } - - case sdc1_op:{ - u64 __user *va = (u64 __user *) (xcp->regs[MIPSInst_RS(ir)] + - MIPSInst_SIMM(ir)); - u64 val; + case sdc1_op: + dva = (u64 __user *) (xcp->regs[MIPSInst_RS(ir)] + + MIPSInst_SIMM(ir)); MIPS_FPU_EMU_INC_STATS(stores); - DIFROMREG(val, MIPSInst_RT(ir)); - if (!access_ok(VERIFY_WRITE, va, sizeof(u64))) { + DIFROMREG(dval, MIPSInst_RT(ir)); + if (!access_ok(VERIFY_WRITE, dva, sizeof(u64))) { MIPS_FPU_EMU_INC_STATS(errors); - *fault_addr = va; + *fault_addr = dva; return SIGBUS; } - if (__put_user(val, va)) { + if (__put_user(dval, dva)) { MIPS_FPU_EMU_INC_STATS(errors); - *fault_addr = va; + *fault_addr = dva; return SIGSEGV; } break; - } - - case lwc1_op:{ - u32 __user *va = (u32 __user *) (xcp->regs[MIPSInst_RS(ir)] + - MIPSInst_SIMM(ir)); - u32 val; + case lwc1_op: + wva = (u32 __user *) (xcp->regs[MIPSInst_RS(ir)] + + MIPSInst_SIMM(ir)); MIPS_FPU_EMU_INC_STATS(loads); - if (!access_ok(VERIFY_READ, va, sizeof(u32))) { + if (!access_ok(VERIFY_READ, wva, sizeof(u32))) { MIPS_FPU_EMU_INC_STATS(errors); - *fault_addr = va; + *fault_addr = wva; return SIGBUS; } - if (__get_user(val, va)) { + if (__get_user(wval, wva)) { MIPS_FPU_EMU_INC_STATS(errors); - *fault_addr = va; + *fault_addr = wva; return SIGSEGV; } - SITOREG(val, MIPSInst_RT(ir)); + SITOREG(wval, MIPSInst_RT(ir)); break; - } - - case swc1_op:{ - u32 __user *va = (u32 __user *) (xcp->regs[MIPSInst_RS(ir)] + - MIPSInst_SIMM(ir)); - u32 val; + case swc1_op: + wva = (u32 __user *) (xcp->regs[MIPSInst_RS(ir)] + + MIPSInst_SIMM(ir)); MIPS_FPU_EMU_INC_STATS(stores); - SIFROMREG(val, MIPSInst_RT(ir)); - if (!access_ok(VERIFY_WRITE, va, sizeof(u32))) { + SIFROMREG(wval, MIPSInst_RT(ir)); + if (!access_ok(VERIFY_WRITE, wva, sizeof(u32))) { MIPS_FPU_EMU_INC_STATS(errors); - *fault_addr = va; + *fault_addr = wva; return SIGBUS; } - if (__put_user(val, va)) { + if (__put_user(wval, wva)) { MIPS_FPU_EMU_INC_STATS(errors); - *fault_addr = va; + *fault_addr = wva; return SIGSEGV; } break; - } case cop1_op: switch (MIPSInst_RS(ir)) { - -#if defined(__mips64) case dmfc_op: + if (!cpu_has_mips_3_4_5 && !cpu_has_mips64) + return SIGILL; + /* copregister fs -> gpr[rt] */ if (MIPSInst_RT(ir) != 0) { DIFROMREG(xcp->regs[MIPSInst_RT(ir)], @@ -1083,10 +860,12 @@ static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, break; case dmtc_op: + if (!cpu_has_mips_3_4_5 && !cpu_has_mips64) + return SIGILL; + /* copregister fs <- rt */ DITOREG(xcp->regs[MIPSInst_RT(ir)], MIPSInst_RD(ir)); break; -#endif case mfhc_op: if (!cpu_has_mips_r2) @@ -1120,19 +899,14 @@ static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, SITOREG(xcp->regs[MIPSInst_RT(ir)], MIPSInst_RD(ir)); break; - case cfc_op:{ + case cfc_op: /* cop control register rd -> gpr[rt] */ - u32 value; - if (MIPSInst_RD(ir) == FPCREG_CSR) { value = ctx->fcr31; - value = (value & ~FPU_CSR_RM) | - mips_rm[modeindex(value)]; -#ifdef CSRTRACE - printk("%p gpr[%d]<-csr=%08x\n", - (void *) (xcp->cp0_epc), - MIPSInst_RT(ir), value); -#endif + value = (value & ~FPU_CSR_RM) | modeindex(value); + pr_debug("%p gpr[%d]<-csr=%08x\n", + (void *) (xcp->cp0_epc), + MIPSInst_RT(ir), value); } else if (MIPSInst_RD(ir) == FPCREG_RID) value = 0; @@ -1141,12 +915,9 @@ static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, if (MIPSInst_RT(ir)) xcp->regs[MIPSInst_RT(ir)] = value; break; - } - case ctc_op:{ + case ctc_op: /* copregister rd <- rt */ - u32 value; - if (MIPSInst_RT(ir) == 0) value = 0; else @@ -1155,37 +926,33 @@ static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, /* we only have one writable control reg */ if (MIPSInst_RD(ir) == FPCREG_CSR) { -#ifdef CSRTRACE - printk("%p gpr[%d]->csr=%08x\n", - (void *) (xcp->cp0_epc), - MIPSInst_RT(ir), value); -#endif + pr_debug("%p gpr[%d]->csr=%08x\n", + (void *) (xcp->cp0_epc), + MIPSInst_RT(ir), value); /* * Don't write reserved bits, * and convert to ieee library modes */ - ctx->fcr31 = (value & - ~(FPU_CSR_RSVD | FPU_CSR_RM)) | - ieee_rm[modeindex(value)]; + ctx->fcr31 = (value & ~(FPU_CSR_RSVD | FPU_CSR_RM)) | + modeindex(value); } if ((ctx->fcr31 >> 5) & ctx->fcr31 & FPU_CSR_ALL_E) { return SIGFPE; } break; - } - case bc_op:{ - int likely = 0; - - if (xcp->cp0_cause & CAUSEF_BD) + case bc_op: + if (delay_slot(xcp)) return SIGILL; -#if __mips >= 4 - cond = ctx->fcr31 & fpucondbit[MIPSInst_RT(ir) >> 2]; -#else - cond = ctx->fcr31 & FPU_CSR_COND; -#endif + if (cpu_has_mips_4_5_r) + cbit = fpucondbit[MIPSInst_RT(ir) >> 2]; + else + cbit = FPU_CSR_COND; + cond = ctx->fcr31 & cbit; + + likely = 0; switch (MIPSInst_RT(ir) & 3) { case bcfl_op: likely = 1; @@ -1201,10 +968,10 @@ static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, return SIGILL; } - xcp->cp0_cause |= CAUSEF_BD; + set_delay_slot(xcp); if (cond) { - /* branch taken: emulate dslot - * instruction + /* + * Branch taken: emulate dslot instruction */ xcp->cp0_epc += dec_insn.pc_inc; @@ -1238,23 +1005,37 @@ static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, switch (MIPSInst_OPCODE(ir)) { case lwc1_op: + goto emul; + case swc1_op: -#if (__mips >= 2 || defined(__mips64)) + goto emul; + case ldc1_op: case sdc1_op: -#endif + if (cpu_has_mips_2_3_4_5 || + cpu_has_mips64) + goto emul; + + return SIGILL; + goto emul; + case cop1_op: -#if __mips >= 4 && __mips != 32 - case cop1x_op: -#endif - /* its one of ours */ goto emul; -#if __mips >= 4 + + case cop1x_op: + if (cpu_has_mips_4_5 || cpu_has_mips64) + /* its one of ours */ + goto emul; + + return SIGILL; + case spec_op: + if (!cpu_has_mips_4_5_r) + return SIGILL; + if (MIPSInst_FUNC(ir) == movc_op) goto emul; break; -#endif } /* @@ -1262,10 +1043,7 @@ static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, * instruction in the dslot */ return mips_dsemul(xcp, ir, contpc); - } - else { - /* branch not taken */ - if (likely) { + } else if (likely) { /* branch not taken */ /* * branch likely nullifies * dslot if not taken @@ -1277,34 +1055,31 @@ static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, * dslot as normal insn */ } - } break; - } default: if (!(MIPSInst_RS(ir) & 0x10)) return SIGILL; - { - int sig; - /* a real fpu computation instruction */ - if ((sig = fpu_emu(xcp, ctx, ir))) - return sig; - } + /* a real fpu computation instruction */ + if ((sig = fpu_emu(xcp, ctx, ir))) + return sig; } break; -#if __mips >= 4 && __mips != 32 - case cop1x_op:{ - int sig = fpux_emu(xcp, ctx, ir, fault_addr); + case cop1x_op: + if (!cpu_has_mips_4_5 && !cpu_has_mips64) + return SIGILL; + + sig = fpux_emu(xcp, ctx, ir, fault_addr); if (sig) return sig; break; - } -#endif -#if __mips >= 4 case spec_op: + if (!cpu_has_mips_4_5_r) + return SIGILL; + if (MIPSInst_FUNC(ir) != movc_op) return SIGILL; cond = fpucondbit[MIPSInst_RT(ir) >> 2]; @@ -1312,8 +1087,6 @@ static int cop1Emulate(struct pt_regs *xcp, struct mips_fpu_struct *ctx, xcp->regs[MIPSInst_RD(ir)] = xcp->regs[MIPSInst_RS(ir)]; break; -#endif - default: sigill: return SIGILL; @@ -1321,7 +1094,7 @@ sigill: /* we did it !! */ xcp->cp0_epc = contpc; - xcp->cp0_cause &= ~CAUSEF_BD; + clear_delay_slot(xcp); return 0; } @@ -1342,44 +1115,42 @@ static const unsigned char cmptab[8] = { }; -#if __mips >= 4 && __mips != 32 - /* * Additional MIPS4 instructions */ -#define DEF3OP(name, p, f1, f2, f3) \ -static ieee754##p fpemu_##p##_##name(ieee754##p r, ieee754##p s, \ - ieee754##p t) \ -{ \ - struct _ieee754_csr ieee754_csr_save; \ - s = f1(s, t); \ - ieee754_csr_save = ieee754_csr; \ - s = f2(s, r); \ - ieee754_csr_save.cx |= ieee754_csr.cx; \ - ieee754_csr_save.sx |= ieee754_csr.sx; \ - s = f3(s); \ - ieee754_csr.cx |= ieee754_csr_save.cx; \ - ieee754_csr.sx |= ieee754_csr_save.sx; \ - return s; \ +#define DEF3OP(name, p, f1, f2, f3) \ +static union ieee754##p fpemu_##p##_##name(union ieee754##p r, \ + union ieee754##p s, union ieee754##p t) \ +{ \ + struct _ieee754_csr ieee754_csr_save; \ + s = f1(s, t); \ + ieee754_csr_save = ieee754_csr; \ + s = f2(s, r); \ + ieee754_csr_save.cx |= ieee754_csr.cx; \ + ieee754_csr_save.sx |= ieee754_csr.sx; \ + s = f3(s); \ + ieee754_csr.cx |= ieee754_csr_save.cx; \ + ieee754_csr.sx |= ieee754_csr_save.sx; \ + return s; \ } -static ieee754dp fpemu_dp_recip(ieee754dp d) +static union ieee754dp fpemu_dp_recip(union ieee754dp d) { return ieee754dp_div(ieee754dp_one(0), d); } -static ieee754dp fpemu_dp_rsqrt(ieee754dp d) +static union ieee754dp fpemu_dp_rsqrt(union ieee754dp d) { return ieee754dp_div(ieee754dp_one(0), ieee754dp_sqrt(d)); } -static ieee754sp fpemu_sp_recip(ieee754sp s) +static union ieee754sp fpemu_sp_recip(union ieee754sp s) { return ieee754sp_div(ieee754sp_one(0), s); } -static ieee754sp fpemu_sp_rsqrt(ieee754sp s) +static union ieee754sp fpemu_sp_rsqrt(union ieee754sp s) { return ieee754sp_div(ieee754sp_one(0), ieee754sp_sqrt(s)); } @@ -1403,8 +1174,8 @@ static int fpux_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, switch (MIPSInst_FMA_FFMT(ir)) { case s_fmt:{ /* 0 */ - ieee754sp(*handler) (ieee754sp, ieee754sp, ieee754sp); - ieee754sp fd, fr, fs, ft; + union ieee754sp(*handler) (union ieee754sp, union ieee754sp, union ieee754sp); + union ieee754sp fd, fr, fs, ft; u32 __user *va; u32 val; @@ -1467,18 +1238,26 @@ static int fpux_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, SPTOREG(fd, MIPSInst_FD(ir)); copcsr: - if (ieee754_cxtest(IEEE754_INEXACT)) + if (ieee754_cxtest(IEEE754_INEXACT)) { + MIPS_FPU_EMU_INC_STATS(ieee754_inexact); rcsr |= FPU_CSR_INE_X | FPU_CSR_INE_S; - if (ieee754_cxtest(IEEE754_UNDERFLOW)) + } + if (ieee754_cxtest(IEEE754_UNDERFLOW)) { + MIPS_FPU_EMU_INC_STATS(ieee754_underflow); rcsr |= FPU_CSR_UDF_X | FPU_CSR_UDF_S; - if (ieee754_cxtest(IEEE754_OVERFLOW)) + } + if (ieee754_cxtest(IEEE754_OVERFLOW)) { + MIPS_FPU_EMU_INC_STATS(ieee754_overflow); rcsr |= FPU_CSR_OVF_X | FPU_CSR_OVF_S; - if (ieee754_cxtest(IEEE754_INVALID_OPERATION)) + } + if (ieee754_cxtest(IEEE754_INVALID_OPERATION)) { + MIPS_FPU_EMU_INC_STATS(ieee754_invalidop); rcsr |= FPU_CSR_INV_X | FPU_CSR_INV_S; + } ctx->fcr31 = (ctx->fcr31 & ~FPU_CSR_ALL_X) | rcsr; if ((ctx->fcr31 >> 5) & ctx->fcr31 & FPU_CSR_ALL_E) { - /*printk ("SIGFPE: fpu csr = %08x\n", + /*printk ("SIGFPE: FPU csr = %08x\n", ctx->fcr31); */ return SIGFPE; } @@ -1492,8 +1271,8 @@ static int fpux_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, } case d_fmt:{ /* 1 */ - ieee754dp(*handler) (ieee754dp, ieee754dp, ieee754dp); - ieee754dp fd, fr, fs, ft; + union ieee754dp(*handler) (union ieee754dp, union ieee754dp, union ieee754dp); + union ieee754dp fd, fr, fs, ft; u64 __user *va; u64 val; @@ -1574,7 +1353,6 @@ static int fpux_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, return 0; } -#endif @@ -1586,23 +1364,25 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, { int rfmt; /* resulting format */ unsigned rcsr = 0; /* resulting csr */ + unsigned int oldrm; + unsigned int cbit; unsigned cond; union { - ieee754dp d; - ieee754sp s; + union ieee754dp d; + union ieee754sp s; int w; -#ifdef __mips64 s64 l; -#endif } rv; /* resulting value */ + u64 bits; MIPS_FPU_EMU_INC_STATS(cp1ops); switch (rfmt = (MIPSInst_FFMT(ir) & 0xf)) { - case s_fmt:{ /* 0 */ + case s_fmt: { /* 0 */ union { - ieee754sp(*b) (ieee754sp, ieee754sp); - ieee754sp(*u) (ieee754sp); + union ieee754sp(*b) (union ieee754sp, union ieee754sp); + union ieee754sp(*u) (union ieee754sp); } handler; + union ieee754sp fs, ft; switch (MIPSInst_FUNC(ir)) { /* binary ops */ @@ -1620,148 +1400,167 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, goto scopbop; /* unary ops */ -#if __mips >= 2 || defined(__mips64) case fsqrt_op: + if (!cpu_has_mips_4_5_r) + return SIGILL; + handler.u = ieee754sp_sqrt; goto scopuop; -#endif -#if __mips >= 4 && __mips != 32 + + /* + * Note that on some MIPS IV implementations such as the + * R5000 and R8000 the FSQRT and FRECIP instructions do not + * achieve full IEEE-754 accuracy - however this emulator does. + */ case frsqrt_op: + if (!cpu_has_mips_4_5_r2) + return SIGILL; + handler.u = fpemu_sp_rsqrt; goto scopuop; + case frecip_op: + if (!cpu_has_mips_4_5_r2) + return SIGILL; + handler.u = fpemu_sp_recip; goto scopuop; -#endif -#if __mips >= 4 + case fmovc_op: + if (!cpu_has_mips_4_5_r) + return SIGILL; + cond = fpucondbit[MIPSInst_FT(ir) >> 2]; if (((ctx->fcr31 & cond) != 0) != ((MIPSInst_FT(ir) & 1) != 0)) return 0; SPFROMREG(rv.s, MIPSInst_FS(ir)); break; + case fmovz_op: + if (!cpu_has_mips_4_5_r) + return SIGILL; + if (xcp->regs[MIPSInst_FT(ir)] != 0) return 0; SPFROMREG(rv.s, MIPSInst_FS(ir)); break; + case fmovn_op: + if (!cpu_has_mips_4_5_r) + return SIGILL; + if (xcp->regs[MIPSInst_FT(ir)] == 0) return 0; SPFROMREG(rv.s, MIPSInst_FS(ir)); break; -#endif + case fabs_op: handler.u = ieee754sp_abs; goto scopuop; + case fneg_op: handler.u = ieee754sp_neg; goto scopuop; + case fmov_op: /* an easy one */ SPFROMREG(rv.s, MIPSInst_FS(ir)); goto copcsr; /* binary op on handler */ - scopbop: - { - ieee754sp fs, ft; - - SPFROMREG(fs, MIPSInst_FS(ir)); - SPFROMREG(ft, MIPSInst_FT(ir)); - - rv.s = (*handler.b) (fs, ft); - goto copcsr; - } - scopuop: - { - ieee754sp fs; +scopbop: + SPFROMREG(fs, MIPSInst_FS(ir)); + SPFROMREG(ft, MIPSInst_FT(ir)); - SPFROMREG(fs, MIPSInst_FS(ir)); - rv.s = (*handler.u) (fs); - goto copcsr; - } - copcsr: - if (ieee754_cxtest(IEEE754_INEXACT)) + rv.s = (*handler.b) (fs, ft); + goto copcsr; +scopuop: + SPFROMREG(fs, MIPSInst_FS(ir)); + rv.s = (*handler.u) (fs); + goto copcsr; +copcsr: + if (ieee754_cxtest(IEEE754_INEXACT)) { + MIPS_FPU_EMU_INC_STATS(ieee754_inexact); rcsr |= FPU_CSR_INE_X | FPU_CSR_INE_S; - if (ieee754_cxtest(IEEE754_UNDERFLOW)) + } + if (ieee754_cxtest(IEEE754_UNDERFLOW)) { + MIPS_FPU_EMU_INC_STATS(ieee754_underflow); rcsr |= FPU_CSR_UDF_X | FPU_CSR_UDF_S; - if (ieee754_cxtest(IEEE754_OVERFLOW)) + } + if (ieee754_cxtest(IEEE754_OVERFLOW)) { + MIPS_FPU_EMU_INC_STATS(ieee754_overflow); rcsr |= FPU_CSR_OVF_X | FPU_CSR_OVF_S; - if (ieee754_cxtest(IEEE754_ZERO_DIVIDE)) + } + if (ieee754_cxtest(IEEE754_ZERO_DIVIDE)) { + MIPS_FPU_EMU_INC_STATS(ieee754_zerodiv); rcsr |= FPU_CSR_DIV_X | FPU_CSR_DIV_S; - if (ieee754_cxtest(IEEE754_INVALID_OPERATION)) + } + if (ieee754_cxtest(IEEE754_INVALID_OPERATION)) { + MIPS_FPU_EMU_INC_STATS(ieee754_invalidop); rcsr |= FPU_CSR_INV_X | FPU_CSR_INV_S; + } break; /* unary conv ops */ case fcvts_op: return SIGILL; /* not defined */ - case fcvtd_op:{ - ieee754sp fs; + case fcvtd_op: SPFROMREG(fs, MIPSInst_FS(ir)); rv.d = ieee754dp_fsp(fs); rfmt = d_fmt; goto copcsr; - } - case fcvtw_op:{ - ieee754sp fs; + case fcvtw_op: SPFROMREG(fs, MIPSInst_FS(ir)); rv.w = ieee754sp_tint(fs); rfmt = w_fmt; goto copcsr; - } -#if __mips >= 2 || defined(__mips64) case fround_op: case ftrunc_op: case fceil_op: - case ffloor_op:{ - unsigned int oldrm = ieee754_csr.rm; - ieee754sp fs; + case ffloor_op: + if (!cpu_has_mips_2_3_4_5 && !cpu_has_mips64) + return SIGILL; + oldrm = ieee754_csr.rm; SPFROMREG(fs, MIPSInst_FS(ir)); - ieee754_csr.rm = ieee_rm[modeindex(MIPSInst_FUNC(ir))]; + ieee754_csr.rm = modeindex(MIPSInst_FUNC(ir)); rv.w = ieee754sp_tint(fs); ieee754_csr.rm = oldrm; rfmt = w_fmt; goto copcsr; - } -#endif /* __mips >= 2 */ -#if defined(__mips64) - case fcvtl_op:{ - ieee754sp fs; + case fcvtl_op: + if (!cpu_has_mips_3_4_5 && !cpu_has_mips64) + return SIGILL; SPFROMREG(fs, MIPSInst_FS(ir)); rv.l = ieee754sp_tlong(fs); rfmt = l_fmt; goto copcsr; - } case froundl_op: case ftruncl_op: case fceill_op: - case ffloorl_op:{ - unsigned int oldrm = ieee754_csr.rm; - ieee754sp fs; + case ffloorl_op: + if (!cpu_has_mips_3_4_5 && !cpu_has_mips64) + return SIGILL; + oldrm = ieee754_csr.rm; SPFROMREG(fs, MIPSInst_FS(ir)); - ieee754_csr.rm = ieee_rm[modeindex(MIPSInst_FUNC(ir))]; + ieee754_csr.rm = modeindex(MIPSInst_FUNC(ir)); rv.l = ieee754sp_tlong(fs); ieee754_csr.rm = oldrm; rfmt = l_fmt; goto copcsr; - } -#endif /* defined(__mips64) */ default: if (MIPSInst_FUNC(ir) >= fcmp_op) { unsigned cmpop = MIPSInst_FUNC(ir) - fcmp_op; - ieee754sp fs, ft; + union ieee754sp fs, ft; SPFROMREG(fs, MIPSInst_FS(ir)); SPFROMREG(ft, MIPSInst_FT(ir)); @@ -1774,19 +1573,18 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, else goto copcsr; - } - else { + } else return SIGILL; - } break; } break; } - case d_fmt:{ + case d_fmt: { + union ieee754dp fs, ft; union { - ieee754dp(*b) (ieee754dp, ieee754dp); - ieee754dp(*u) (ieee754dp); + union ieee754dp(*b) (union ieee754dp, union ieee754dp); + union ieee754dp(*u) (union ieee754dp); } handler; switch (MIPSInst_FUNC(ir)) { @@ -1805,21 +1603,33 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, goto dcopbop; /* unary ops */ -#if __mips >= 2 || defined(__mips64) case fsqrt_op: + if (!cpu_has_mips_2_3_4_5_r) + return SIGILL; + handler.u = ieee754dp_sqrt; goto dcopuop; -#endif -#if __mips >= 4 && __mips != 32 + /* + * Note that on some MIPS IV implementations such as the + * R5000 and R8000 the FSQRT and FRECIP instructions do not + * achieve full IEEE-754 accuracy - however this emulator does. + */ case frsqrt_op: + if (!cpu_has_mips_4_5_r2) + return SIGILL; + handler.u = fpemu_dp_rsqrt; goto dcopuop; case frecip_op: + if (!cpu_has_mips_4_5_r2) + return SIGILL; + handler.u = fpemu_dp_recip; goto dcopuop; -#endif -#if __mips >= 4 case fmovc_op: + if (!cpu_has_mips_4_5_r) + return SIGILL; + cond = fpucondbit[MIPSInst_FT(ir) >> 2]; if (((ctx->fcr31 & cond) != 0) != ((MIPSInst_FT(ir) & 1) != 0)) @@ -1827,16 +1637,21 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, DPFROMREG(rv.d, MIPSInst_FS(ir)); break; case fmovz_op: + if (!cpu_has_mips_4_5_r) + return SIGILL; + if (xcp->regs[MIPSInst_FT(ir)] != 0) return 0; DPFROMREG(rv.d, MIPSInst_FS(ir)); break; case fmovn_op: + if (!cpu_has_mips_4_5_r) + return SIGILL; + if (xcp->regs[MIPSInst_FT(ir)] == 0) return 0; DPFROMREG(rv.d, MIPSInst_FS(ir)); break; -#endif case fabs_op: handler.u = ieee754dp_abs; goto dcopuop; @@ -1851,91 +1666,78 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, goto copcsr; /* binary op on handler */ - dcopbop:{ - ieee754dp fs, ft; - - DPFROMREG(fs, MIPSInst_FS(ir)); - DPFROMREG(ft, MIPSInst_FT(ir)); - - rv.d = (*handler.b) (fs, ft); - goto copcsr; - } - dcopuop:{ - ieee754dp fs; - - DPFROMREG(fs, MIPSInst_FS(ir)); - rv.d = (*handler.u) (fs); - goto copcsr; - } +dcopbop: + DPFROMREG(fs, MIPSInst_FS(ir)); + DPFROMREG(ft, MIPSInst_FT(ir)); - /* unary conv ops */ - case fcvts_op:{ - ieee754dp fs; + rv.d = (*handler.b) (fs, ft); + goto copcsr; +dcopuop: + DPFROMREG(fs, MIPSInst_FS(ir)); + rv.d = (*handler.u) (fs); + goto copcsr; + /* + * unary conv ops + */ + case fcvts_op: DPFROMREG(fs, MIPSInst_FS(ir)); rv.s = ieee754sp_fdp(fs); rfmt = s_fmt; goto copcsr; - } + case fcvtd_op: return SIGILL; /* not defined */ - case fcvtw_op:{ - ieee754dp fs; - + case fcvtw_op: DPFROMREG(fs, MIPSInst_FS(ir)); rv.w = ieee754dp_tint(fs); /* wrong */ rfmt = w_fmt; goto copcsr; - } -#if __mips >= 2 || defined(__mips64) case fround_op: case ftrunc_op: case fceil_op: - case ffloor_op:{ - unsigned int oldrm = ieee754_csr.rm; - ieee754dp fs; + case ffloor_op: + if (!cpu_has_mips_2_3_4_5_r) + return SIGILL; + oldrm = ieee754_csr.rm; DPFROMREG(fs, MIPSInst_FS(ir)); - ieee754_csr.rm = ieee_rm[modeindex(MIPSInst_FUNC(ir))]; + ieee754_csr.rm = modeindex(MIPSInst_FUNC(ir)); rv.w = ieee754dp_tint(fs); ieee754_csr.rm = oldrm; rfmt = w_fmt; goto copcsr; - } -#endif -#if defined(__mips64) - case fcvtl_op:{ - ieee754dp fs; + case fcvtl_op: + if (!cpu_has_mips_3_4_5 && !cpu_has_mips64) + return SIGILL; DPFROMREG(fs, MIPSInst_FS(ir)); rv.l = ieee754dp_tlong(fs); rfmt = l_fmt; goto copcsr; - } case froundl_op: case ftruncl_op: case fceill_op: - case ffloorl_op:{ - unsigned int oldrm = ieee754_csr.rm; - ieee754dp fs; + case ffloorl_op: + if (!cpu_has_mips_3_4_5 && !cpu_has_mips64) + return SIGILL; + oldrm = ieee754_csr.rm; DPFROMREG(fs, MIPSInst_FS(ir)); - ieee754_csr.rm = ieee_rm[modeindex(MIPSInst_FUNC(ir))]; + ieee754_csr.rm = modeindex(MIPSInst_FUNC(ir)); rv.l = ieee754dp_tlong(fs); ieee754_csr.rm = oldrm; rfmt = l_fmt; goto copcsr; - } -#endif /* __mips >= 3 */ default: if (MIPSInst_FUNC(ir) >= fcmp_op) { unsigned cmpop = MIPSInst_FUNC(ir) - fcmp_op; - ieee754dp fs, ft; + union ieee754dp fs, ft; DPFROMREG(fs, MIPSInst_FS(ir)); DPFROMREG(ft, MIPSInst_FT(ir)); @@ -1957,11 +1759,8 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, break; } break; - } - - case w_fmt:{ - ieee754sp fs; + case w_fmt: switch (MIPSInst_FUNC(ir)) { case fcvts_op: /* convert word to single precision real */ @@ -1981,9 +1780,11 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, break; } -#if defined(__mips64) - case l_fmt:{ - u64 bits; + case l_fmt: + + if (!cpu_has_mips_3_4_5 && !cpu_has_mips64) + return SIGILL; + DIFROMREG(bits, MIPSInst_FS(ir)); switch (MIPSInst_FUNC(ir)) { @@ -2001,8 +1802,6 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, return SIGILL; } break; - } -#endif default: return SIGILL; @@ -2017,7 +1816,7 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, */ ctx->fcr31 = (ctx->fcr31 & ~FPU_CSR_ALL_X) | rcsr; if ((ctx->fcr31 >> 5) & ctx->fcr31 & FPU_CSR_ALL_E) { - /*printk ("SIGFPE: fpu csr = %08x\n",ctx->fcr31); */ + /*printk ("SIGFPE: FPU csr = %08x\n",ctx->fcr31); */ return SIGFPE; } @@ -2025,18 +1824,18 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, * Now we can safely write the result back to the register file. */ switch (rfmt) { - case -1:{ -#if __mips >= 4 - cond = fpucondbit[MIPSInst_FD(ir) >> 2]; -#else - cond = FPU_CSR_COND; -#endif + case -1: + + if (cpu_has_mips_4_5_r) + cbit = fpucondbit[MIPSInst_RT(ir) >> 2]; + else + cbit = FPU_CSR_COND; if (rv.w) - ctx->fcr31 |= cond; + ctx->fcr31 |= cbit; else - ctx->fcr31 &= ~cond; + ctx->fcr31 &= ~cbit; break; - } + case d_fmt: DPTOREG(rv.d, MIPSInst_FD(ir)); break; @@ -2046,11 +1845,12 @@ static int fpu_emu(struct pt_regs *xcp, struct mips_fpu_struct *ctx, case w_fmt: SITOREG(rv.w, MIPSInst_FD(ir)); break; -#if defined(__mips64) case l_fmt: + if (!cpu_has_mips_3_4_5 && !cpu_has_mips64) + return SIGILL; + DITOREG(rv.l, MIPSInst_FD(ir)); break; -#endif default: return SIGILL; } @@ -2138,11 +1938,7 @@ int fpu_emulator_cop1Handler(struct pt_regs *xcp, struct mips_fpu_struct *ctx, * ieee754_csr. But ieee754_csr.rm is ieee * library modes. (not mips rounding mode) */ - /* convert to ieee library modes */ - ieee754_csr.rm = ieee_rm[ieee754_csr.rm]; sig = cop1Emulate(xcp, ctx, dec_insn, fault_addr); - /* revert to mips rounding mode */ - ieee754_csr.rm = mips_rm[ieee754_csr.rm]; } if (has_fpu) @@ -2155,58 +1951,8 @@ int fpu_emulator_cop1Handler(struct pt_regs *xcp, struct mips_fpu_struct *ctx, /* SIGILL indicates a non-fpu instruction */ if (sig == SIGILL && xcp->cp0_epc != oldepc) - /* but if epc has advanced, then ignore it */ + /* but if EPC has advanced, then ignore it */ sig = 0; return sig; } - -#ifdef CONFIG_DEBUG_FS - -static int fpuemu_stat_get(void *data, u64 *val) -{ - int cpu; - unsigned long sum = 0; - for_each_online_cpu(cpu) { - struct mips_fpu_emulator_stats *ps; - local_t *pv; - ps = &per_cpu(fpuemustats, cpu); - pv = (void *)ps + (unsigned long)data; - sum += local_read(pv); - } - *val = sum; - return 0; -} -DEFINE_SIMPLE_ATTRIBUTE(fops_fpuemu_stat, fpuemu_stat_get, NULL, "%llu\n"); - -extern struct dentry *mips_debugfs_dir; -static int __init debugfs_fpuemu(void) -{ - struct dentry *d, *dir; - - if (!mips_debugfs_dir) - return -ENODEV; - dir = debugfs_create_dir("fpuemustats", mips_debugfs_dir); - if (!dir) - return -ENOMEM; - -#define FPU_STAT_CREATE(M) \ - do { \ - d = debugfs_create_file(#M , S_IRUGO, dir, \ - (void *)offsetof(struct mips_fpu_emulator_stats, M), \ - &fops_fpuemu_stat); \ - if (!d) \ - return -ENOMEM; \ - } while (0) - - FPU_STAT_CREATE(emulated); - FPU_STAT_CREATE(loads); - FPU_STAT_CREATE(stores); - FPU_STAT_CREATE(cp1ops); - FPU_STAT_CREATE(cp1xops); - FPU_STAT_CREATE(errors); - - return 0; -} -__initcall(debugfs_fpuemu); -#endif diff --git a/arch/mips/math-emu/dp_add.c b/arch/mips/math-emu/dp_add.c index c57c8adc42c..7f64577df98 100644 --- a/arch/mips/math-emu/dp_add.c +++ b/arch/mips/math-emu/dp_add.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,24 +16,22 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - * + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754dp.h" -ieee754dp ieee754dp_add(ieee754dp x, ieee754dp y) +union ieee754dp ieee754dp_add(union ieee754dp x, union ieee754dp y) { + int s; + COMPXDP; COMPYDP; EXPLODEXDP; EXPLODEYDP; - CLEARCX; + ieee754_clearcx(); FLUSHXDP; FLUSHYDP; @@ -52,8 +48,8 @@ ieee754dp ieee754dp_add(ieee754dp x, ieee754dp y) case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_nanxcpt(ieee754dp_indef(), "add", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_nanxcpt(ieee754dp_indef()); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN): case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN): @@ -69,14 +65,14 @@ ieee754dp ieee754dp_add(ieee754dp x, ieee754dp y) return x; - /* Infinity handling - */ - + /* + * Infinity handling + */ case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF): if (xs == ys) return x; - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_xcpt(ieee754dp_indef(), "add", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_indef(); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF): case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF): @@ -88,15 +84,14 @@ ieee754dp ieee754dp_add(ieee754dp x, ieee754dp y) case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM): return x; - /* Zero handling - */ - + /* + * Zero handling + */ case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO): if (xs == ys) return x; else - return ieee754dp_zero(ieee754_csr.rm == - IEEE754_RD); + return ieee754dp_zero(ieee754_csr.rm == FPU_CSR_RD); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO): case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO): @@ -125,20 +120,24 @@ ieee754dp ieee754dp_add(ieee754dp x, ieee754dp y) assert(xm & DP_HIDDEN_BIT); assert(ym & DP_HIDDEN_BIT); - /* provide guard,round and stick bit space */ + /* + * Provide guard,round and stick bit space. + */ xm <<= 3; ym <<= 3; if (xe > ye) { - /* have to shift y fraction right to align + /* + * Have to shift y fraction right to align. */ - int s = xe - ye; + s = xe - ye; ym = XDPSRS(ym, s); ye += s; } else if (ye > xe) { - /* have to shift x fraction right to align + /* + * Have to shift x fraction right to align. */ - int s = ye - xe; + s = ye - xe; xm = XDPSRS(xm, s); xe += s; } @@ -146,14 +145,15 @@ ieee754dp ieee754dp_add(ieee754dp x, ieee754dp y) assert(xe <= DP_EMAX); if (xs == ys) { - /* generate 28 bit result of adding two 27 bit numbers - * leaving result in xm,xs,xe + /* + * Generate 28 bit result of adding two 27 bit numbers + * leaving result in xm, xs and xe. */ xm = xm + ym; xe = xe; xs = xs; - if (xm >> (DP_MBITS + 1 + 3)) { /* carry out */ + if (xm >> (DP_FBITS + 1 + 3)) { /* carry out */ xm = XDPSRS1(xm); xe++; } @@ -168,15 +168,16 @@ ieee754dp ieee754dp_add(ieee754dp x, ieee754dp y) xs = ys; } if (xm == 0) - return ieee754dp_zero(ieee754_csr.rm == - IEEE754_RD); + return ieee754dp_zero(ieee754_csr.rm == FPU_CSR_RD); - /* normalize to rounding precision */ - while ((xm >> (DP_MBITS + 3)) == 0) { + /* + * Normalize to rounding precision. + */ + while ((xm >> (DP_FBITS + 3)) == 0) { xm <<= 1; xe--; } - } - DPNORMRET2(xs, xe, xm, "add", x, y); + + return ieee754dp_format(xs, xe, xm); } diff --git a/arch/mips/math-emu/dp_cmp.c b/arch/mips/math-emu/dp_cmp.c index 0f32486b0ed..30f95f6e9ac 100644 --- a/arch/mips/math-emu/dp_cmp.c +++ b/arch/mips/math-emu/dp_cmp.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,16 +16,16 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754dp.h" -int ieee754dp_cmp(ieee754dp x, ieee754dp y, int cmp, int sig) +int ieee754dp_cmp(union ieee754dp x, union ieee754dp y, int cmp, int sig) { + s64 vx; + s64 vy; + COMPXDP; COMPYDP; @@ -35,21 +33,21 @@ int ieee754dp_cmp(ieee754dp x, ieee754dp y, int cmp, int sig) EXPLODEYDP; FLUSHXDP; FLUSHYDP; - CLEARCX; /* Even clear inexact flag here */ + ieee754_clearcx(); /* Even clear inexact flag here */ if (ieee754dp_isnan(x) || ieee754dp_isnan(y)) { if (sig || xc == IEEE754_CLASS_SNAN || yc == IEEE754_CLASS_SNAN) - SETCX(IEEE754_INVALID_OPERATION); + ieee754_setcx(IEEE754_INVALID_OPERATION); if (cmp & IEEE754_CUN) return 1; if (cmp & (IEEE754_CLT | IEEE754_CGT)) { - if (sig && SETANDTESTCX(IEEE754_INVALID_OPERATION)) - return ieee754si_xcpt(0, "fcmpf", x); + if (sig && ieee754_setandtestcx(IEEE754_INVALID_OPERATION)) + return 0; } return 0; } else { - s64 vx = x.bits; - s64 vy = y.bits; + vx = x.bits; + vy = y.bits; if (vx < 0) vx = -vx ^ DP_SIGN_BIT; diff --git a/arch/mips/math-emu/dp_div.c b/arch/mips/math-emu/dp_div.c index a1bce1b7c09..bef0e55e593 100644 --- a/arch/mips/math-emu/dp_div.c +++ b/arch/mips/math-emu/dp_div.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,23 +16,24 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754dp.h" -ieee754dp ieee754dp_div(ieee754dp x, ieee754dp y) +union ieee754dp ieee754dp_div(union ieee754dp x, union ieee754dp y) { + u64 rm; + int re; + u64 bm; + COMPXDP; COMPYDP; EXPLODEXDP; EXPLODEYDP; - CLEARCX; + ieee754_clearcx(); FLUSHXDP; FLUSHYDP; @@ -51,8 +50,8 @@ ieee754dp ieee754dp_div(ieee754dp x, ieee754dp y) case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_nanxcpt(ieee754dp_indef(), "div", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_nanxcpt(ieee754dp_indef()); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN): case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN): @@ -68,12 +67,12 @@ ieee754dp ieee754dp_div(ieee754dp x, ieee754dp y) return x; - /* Infinity handling - */ - + /* + * Infinity handling + */ case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_xcpt(ieee754dp_indef(), "div", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_indef(); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF): case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF): @@ -85,17 +84,17 @@ ieee754dp ieee754dp_div(ieee754dp x, ieee754dp y) case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM): return ieee754dp_inf(xs ^ ys); - /* Zero handling - */ - + /* + * Zero handling + */ case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_xcpt(ieee754dp_indef(), "div", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_indef(); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO): case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO): - SETCX(IEEE754_ZERO_DIVIDE); - return ieee754dp_xcpt(ieee754dp_inf(xs ^ ys), "div", x, y); + ieee754_setcx(IEEE754_ZERO_DIVIDE); + return ieee754dp_inf(xs ^ ys); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM): @@ -122,35 +121,34 @@ ieee754dp ieee754dp_div(ieee754dp x, ieee754dp y) xm <<= 3; ym <<= 3; - { - /* now the dirty work */ - - u64 rm = 0; - int re = xe - ye; - u64 bm; - - for (bm = DP_MBIT(DP_MBITS + 2); bm; bm >>= 1) { - if (xm >= ym) { - xm -= ym; - rm |= bm; - if (xm == 0) - break; - } - xm <<= 1; - } - rm <<= 1; - if (xm) - rm |= 1; /* have remainder, set sticky */ + /* now the dirty work */ - assert(rm); + rm = 0; + re = xe - ye; - /* normalise rm to rounding precision ? - */ - while ((rm >> (DP_MBITS + 3)) == 0) { - rm <<= 1; - re--; + for (bm = DP_MBIT(DP_FBITS + 2); bm; bm >>= 1) { + if (xm >= ym) { + xm -= ym; + rm |= bm; + if (xm == 0) + break; } + xm <<= 1; + } + + rm <<= 1; + if (xm) + rm |= 1; /* have remainder, set sticky */ - DPNORMRET2(xs == ys ? 0 : 1, re, rm, "div", x, y); + assert(rm); + + /* + * Normalise rm to rounding precision ? + */ + while ((rm >> (DP_FBITS + 3)) == 0) { + rm <<= 1; + re--; } + + return ieee754dp_format(xs == ys ? 0 : 1, re, rm); } diff --git a/arch/mips/math-emu/dp_fint.c b/arch/mips/math-emu/dp_fint.c index 88571288c9e..10258f0afd6 100644 --- a/arch/mips/math-emu/dp_fint.c +++ b/arch/mips/math-emu/dp_fint.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,21 +16,18 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754dp.h" -ieee754dp ieee754dp_fint(int x) +union ieee754dp ieee754dp_fint(int x) { u64 xm; int xe; int xs; - CLEARCX; + ieee754_clearcx(); if (x == 0) return ieee754dp_zero(0); @@ -51,29 +46,11 @@ ieee754dp ieee754dp_fint(int x) xm = x; } -#if 1 /* normalize - result can never be inexact or overflow */ - xe = DP_MBITS; - while ((xm >> DP_MBITS) == 0) { + xe = DP_FBITS; + while ((xm >> DP_FBITS) == 0) { xm <<= 1; xe--; } return builddp(xs, xe + DP_EBIAS, xm & ~DP_HIDDEN_BIT); -#else - /* normalize */ - xe = DP_MBITS + 3; - while ((xm >> (DP_MBITS + 3)) == 0) { - xm <<= 1; - xe--; - } - DPNORMRET1(xs, xe, xm, "fint", x); -#endif -} - -ieee754dp ieee754dp_funs(unsigned int u) -{ - if ((int) u < 0) - return ieee754dp_add(ieee754dp_1e31(), - ieee754dp_fint(u & ~(1 << 31))); - return ieee754dp_fint(u); } diff --git a/arch/mips/math-emu/dp_flong.c b/arch/mips/math-emu/dp_flong.c index 14fc01ec742..a267c2e39d7 100644 --- a/arch/mips/math-emu/dp_flong.c +++ b/arch/mips/math-emu/dp_flong.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,21 +16,18 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754dp.h" -ieee754dp ieee754dp_flong(s64 x) +union ieee754dp ieee754dp_flong(s64 x) { u64 xm; int xe; int xs; - CLEARCX; + ieee754_clearcx(); if (x == 0) return ieee754dp_zero(0); @@ -52,26 +47,19 @@ ieee754dp ieee754dp_flong(s64 x) } /* normalize */ - xe = DP_MBITS + 3; - if (xm >> (DP_MBITS + 1 + 3)) { + xe = DP_FBITS + 3; + if (xm >> (DP_FBITS + 1 + 3)) { /* shunt out overflow bits */ - while (xm >> (DP_MBITS + 1 + 3)) { + while (xm >> (DP_FBITS + 1 + 3)) { XDPSRSX1(); } } else { /* normalize in grs extended double precision */ - while ((xm >> (DP_MBITS + 3)) == 0) { + while ((xm >> (DP_FBITS + 3)) == 0) { xm <<= 1; xe--; } } - DPNORMRET1(xs, xe, xm, "dp_flong", x); -} -ieee754dp ieee754dp_fulong(u64 u) -{ - if ((s64) u < 0) - return ieee754dp_add(ieee754dp_1e63(), - ieee754dp_flong(u & ~(1ULL << 63))); - return ieee754dp_flong(u); + return ieee754dp_format(xs, xe, xm); } diff --git a/arch/mips/math-emu/dp_frexp.c b/arch/mips/math-emu/dp_frexp.c deleted file mode 100644 index cb15a5eaecb..00000000000 --- a/arch/mips/math-emu/dp_frexp.c +++ /dev/null @@ -1,52 +0,0 @@ -/* IEEE754 floating point arithmetic - * double precision: common utilities - */ -/* - * MIPS floating point support - * Copyright (C) 1994-2000 Algorithmics Ltd. - * - * ######################################################################## - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - */ - - -#include "ieee754dp.h" - -/* close to ieeep754dp_logb -*/ -ieee754dp ieee754dp_frexp(ieee754dp x, int *eptr) -{ - COMPXDP; - CLEARCX; - EXPLODEXDP; - - switch (xc) { - case IEEE754_CLASS_SNAN: - case IEEE754_CLASS_QNAN: - case IEEE754_CLASS_INF: - case IEEE754_CLASS_ZERO: - *eptr = 0; - return x; - case IEEE754_CLASS_DNORM: - DPDNORMX; - break; - case IEEE754_CLASS_NORM: - break; - } - *eptr = xe + 1; - return builddp(xs, -1 + DP_EBIAS, xm & ~DP_HIDDEN_BIT); -} diff --git a/arch/mips/math-emu/dp_fsp.c b/arch/mips/math-emu/dp_fsp.c index daed6834dc1..ffb69c5830b 100644 --- a/arch/mips/math-emu/dp_fsp.c +++ b/arch/mips/math-emu/dp_fsp.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,56 +16,58 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - +#include "ieee754sp.h" #include "ieee754dp.h" -ieee754dp ieee754dp_fsp(ieee754sp x) +union ieee754dp ieee754dp_fsp(union ieee754sp x) { COMPXSP; EXPLODEXSP; - CLEARCX; + ieee754_clearcx(); FLUSHXSP; switch (xc) { case IEEE754_CLASS_SNAN: - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_nanxcpt(ieee754dp_indef(), "fsp"); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_nanxcpt(ieee754dp_indef()); + case IEEE754_CLASS_QNAN: return ieee754dp_nanxcpt(builddp(xs, DP_EMAX + 1 + DP_EBIAS, ((u64) xm - << (DP_MBITS - - SP_MBITS))), "fsp", - x); + << (DP_FBITS - + SP_FBITS)))); case IEEE754_CLASS_INF: return ieee754dp_inf(xs); + case IEEE754_CLASS_ZERO: return ieee754dp_zero(xs); + case IEEE754_CLASS_DNORM: /* normalize */ - while ((xm >> SP_MBITS) == 0) { + while ((xm >> SP_FBITS) == 0) { xm <<= 1; xe--; } break; + case IEEE754_CLASS_NORM: break; } - /* CAN'T possibly overflow,underflow, or need rounding + /* + * Can't possibly overflow,underflow, or need rounding */ /* drop the hidden bit */ xm &= ~SP_HIDDEN_BIT; return builddp(xs, xe + DP_EBIAS, - (u64) xm << (DP_MBITS - SP_MBITS)); + (u64) xm << (DP_FBITS - SP_FBITS)); } diff --git a/arch/mips/math-emu/dp_logb.c b/arch/mips/math-emu/dp_logb.c deleted file mode 100644 index 151127e59f5..00000000000 --- a/arch/mips/math-emu/dp_logb.c +++ /dev/null @@ -1,53 +0,0 @@ -/* IEEE754 floating point arithmetic - * double precision: common utilities - */ -/* - * MIPS floating point support - * Copyright (C) 1994-2000 Algorithmics Ltd. - * - * ######################################################################## - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - */ - - -#include "ieee754dp.h" - -ieee754dp ieee754dp_logb(ieee754dp x) -{ - COMPXDP; - - CLEARCX; - - EXPLODEXDP; - - switch (xc) { - case IEEE754_CLASS_SNAN: - return ieee754dp_nanxcpt(x, "logb", x); - case IEEE754_CLASS_QNAN: - return x; - case IEEE754_CLASS_INF: - return ieee754dp_inf(0); - case IEEE754_CLASS_ZERO: - return ieee754dp_inf(1); - case IEEE754_CLASS_DNORM: - DPDNORMX; - break; - case IEEE754_CLASS_NORM: - break; - } - return ieee754dp_fint(xe); -} diff --git a/arch/mips/math-emu/dp_modf.c b/arch/mips/math-emu/dp_modf.c deleted file mode 100644 index b01f9cf6d40..00000000000 --- a/arch/mips/math-emu/dp_modf.c +++ /dev/null @@ -1,79 +0,0 @@ -/* IEEE754 floating point arithmetic - * double precision: common utilities - */ -/* - * MIPS floating point support - * Copyright (C) 1994-2000 Algorithmics Ltd. - * - * ######################################################################## - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - */ - - -#include "ieee754dp.h" - -/* modf function is always exact for a finite number -*/ -ieee754dp ieee754dp_modf(ieee754dp x, ieee754dp *ip) -{ - COMPXDP; - - CLEARCX; - - EXPLODEXDP; - - switch (xc) { - case IEEE754_CLASS_SNAN: - case IEEE754_CLASS_QNAN: - case IEEE754_CLASS_INF: - case IEEE754_CLASS_ZERO: - *ip = x; - return x; - case IEEE754_CLASS_DNORM: - /* far to small */ - *ip = ieee754dp_zero(xs); - return x; - case IEEE754_CLASS_NORM: - break; - } - if (xe < 0) { - *ip = ieee754dp_zero(xs); - return x; - } - if (xe >= DP_MBITS) { - *ip = x; - return ieee754dp_zero(xs); - } - /* generate ipart mantissa by clearing bottom bits - */ - *ip = builddp(xs, xe + DP_EBIAS, - ((xm >> (DP_MBITS - xe)) << (DP_MBITS - xe)) & - ~DP_HIDDEN_BIT); - - /* generate fpart mantissa by clearing top bits - * and normalizing (must be able to normalize) - */ - xm = (xm << (64 - (DP_MBITS - xe))) >> (64 - (DP_MBITS - xe)); - if (xm == 0) - return ieee754dp_zero(xs); - - while ((xm >> DP_MBITS) == 0) { - xm <<= 1; - xe--; - } - return builddp(xs, xe + DP_EBIAS, xm & ~DP_HIDDEN_BIT); -} diff --git a/arch/mips/math-emu/dp_mul.c b/arch/mips/math-emu/dp_mul.c index 09175f46192..d3acdedb5b9 100644 --- a/arch/mips/math-emu/dp_mul.c +++ b/arch/mips/math-emu/dp_mul.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,23 +16,32 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754dp.h" -ieee754dp ieee754dp_mul(ieee754dp x, ieee754dp y) +union ieee754dp ieee754dp_mul(union ieee754dp x, union ieee754dp y) { + int re; + int rs; + u64 rm; + unsigned lxm; + unsigned hxm; + unsigned lym; + unsigned hym; + u64 lrm; + u64 hrm; + u64 t; + u64 at; + COMPXDP; COMPYDP; EXPLODEXDP; EXPLODEYDP; - CLEARCX; + ieee754_clearcx(); FLUSHXDP; FLUSHYDP; @@ -51,8 +58,8 @@ ieee754dp ieee754dp_mul(ieee754dp x, ieee754dp y) case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_nanxcpt(ieee754dp_indef(), "mul", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_nanxcpt(ieee754dp_indef()); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN): case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN): @@ -68,12 +75,13 @@ ieee754dp ieee754dp_mul(ieee754dp x, ieee754dp y) return x; - /* Infinity handling */ - + /* + * Infinity handling + */ case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO): case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_xcpt(ieee754dp_indef(), "mul", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_indef(); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF): case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF): @@ -107,70 +115,59 @@ ieee754dp ieee754dp_mul(ieee754dp x, ieee754dp y) /* rm = xm * ym, re = xe+ye basically */ assert(xm & DP_HIDDEN_BIT); assert(ym & DP_HIDDEN_BIT); - { - int re = xe + ye; - int rs = xs ^ ys; - u64 rm; - /* shunt to top of word */ - xm <<= 64 - (DP_MBITS + 1); - ym <<= 64 - (DP_MBITS + 1); + re = xe + ye; + rs = xs ^ ys; + + /* shunt to top of word */ + xm <<= 64 - (DP_FBITS + 1); + ym <<= 64 - (DP_FBITS + 1); - /* multiply 32bits xm,ym to give high 32bits rm with stickness - */ + /* + * Multiply 32 bits xm, ym to give high 32 bits rm with stickness. + */ - /* 32 * 32 => 64 */ + /* 32 * 32 => 64 */ #define DPXMULT(x, y) ((u64)(x) * (u64)y) - { - unsigned lxm = xm; - unsigned hxm = xm >> 32; - unsigned lym = ym; - unsigned hym = ym >> 32; - u64 lrm; - u64 hrm; - - lrm = DPXMULT(lxm, lym); - hrm = DPXMULT(hxm, hym); - - { - u64 t = DPXMULT(lxm, hym); - { - u64 at = - lrm + (t << 32); - hrm += at < lrm; - lrm = at; - } - hrm = hrm + (t >> 32); - } - - { - u64 t = DPXMULT(hxm, lym); - { - u64 at = - lrm + (t << 32); - hrm += at < lrm; - lrm = at; - } - hrm = hrm + (t >> 32); - } - rm = hrm | (lrm != 0); - } - - /* - * sticky shift down to normal rounding precision - */ - if ((s64) rm < 0) { - rm = - (rm >> (64 - (DP_MBITS + 1 + 3))) | - ((rm << (DP_MBITS + 1 + 3)) != 0); + lxm = xm; + hxm = xm >> 32; + lym = ym; + hym = ym >> 32; + + lrm = DPXMULT(lxm, lym); + hrm = DPXMULT(hxm, hym); + + t = DPXMULT(lxm, hym); + + at = lrm + (t << 32); + hrm += at < lrm; + lrm = at; + + hrm = hrm + (t >> 32); + + t = DPXMULT(hxm, lym); + + at = lrm + (t << 32); + hrm += at < lrm; + lrm = at; + + hrm = hrm + (t >> 32); + + rm = hrm | (lrm != 0); + + /* + * Sticky shift down to normal rounding precision. + */ + if ((s64) rm < 0) { + rm = (rm >> (64 - (DP_FBITS + 1 + 3))) | + ((rm << (DP_FBITS + 1 + 3)) != 0); re++; - } else { - rm = - (rm >> (64 - (DP_MBITS + 1 + 3 + 1))) | - ((rm << (DP_MBITS + 1 + 3 + 1)) != 0); - } - assert(rm & (DP_HIDDEN_BIT << 3)); - DPNORMRET2(rs, re, rm, "mul", x, y); + } else { + rm = (rm >> (64 - (DP_FBITS + 1 + 3 + 1))) | + ((rm << (DP_FBITS + 1 + 3 + 1)) != 0); } + assert(rm & (DP_HIDDEN_BIT << 3)); + + return ieee754dp_format(rs, re, rm); } diff --git a/arch/mips/math-emu/dp_scalb.c b/arch/mips/math-emu/dp_scalb.c deleted file mode 100644 index 6f5df438dda..00000000000 --- a/arch/mips/math-emu/dp_scalb.c +++ /dev/null @@ -1,57 +0,0 @@ -/* IEEE754 floating point arithmetic - * double precision: common utilities - */ -/* - * MIPS floating point support - * Copyright (C) 1994-2000 Algorithmics Ltd. - * - * ######################################################################## - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - */ - - -#include "ieee754dp.h" - -ieee754dp ieee754dp_scalb(ieee754dp x, int n) -{ - COMPXDP; - - CLEARCX; - - EXPLODEXDP; - - switch (xc) { - case IEEE754_CLASS_SNAN: - return ieee754dp_nanxcpt(x, "scalb", x, n); - case IEEE754_CLASS_QNAN: - case IEEE754_CLASS_INF: - case IEEE754_CLASS_ZERO: - return x; - case IEEE754_CLASS_DNORM: - DPDNORMX; - break; - case IEEE754_CLASS_NORM: - break; - } - DPNORMRET2(xs, xe + n, xm << 3, "scalb", x, n); -} - - -ieee754dp ieee754dp_ldexp(ieee754dp x, int n) -{ - return ieee754dp_scalb(x, n); -} diff --git a/arch/mips/math-emu/dp_simple.c b/arch/mips/math-emu/dp_simple.c index 79ce2673a71..bccbe90efce 100644 --- a/arch/mips/math-emu/dp_simple.c +++ b/arch/mips/math-emu/dp_simple.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,33 +16,17 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754dp.h" -int ieee754dp_finite(ieee754dp x) -{ - return DPBEXP(x) != DP_EMAX + 1 + DP_EBIAS; -} - -ieee754dp ieee754dp_copysign(ieee754dp x, ieee754dp y) -{ - CLEARCX; - DPSIGN(x) = DPSIGN(y); - return x; -} - - -ieee754dp ieee754dp_neg(ieee754dp x) +union ieee754dp ieee754dp_neg(union ieee754dp x) { COMPXDP; EXPLODEXDP; - CLEARCX; + ieee754_clearcx(); FLUSHXDP; /* @@ -55,30 +37,29 @@ ieee754dp ieee754dp_neg(ieee754dp x) DPSIGN(x) ^= 1; if (xc == IEEE754_CLASS_SNAN) { - ieee754dp y = ieee754dp_indef(); - SETCX(IEEE754_INVALID_OPERATION); + union ieee754dp y = ieee754dp_indef(); + ieee754_setcx(IEEE754_INVALID_OPERATION); DPSIGN(y) = DPSIGN(x); - return ieee754dp_nanxcpt(y, "neg"); + return ieee754dp_nanxcpt(y); } return x; } - -ieee754dp ieee754dp_abs(ieee754dp x) +union ieee754dp ieee754dp_abs(union ieee754dp x) { COMPXDP; EXPLODEXDP; - CLEARCX; + ieee754_clearcx(); FLUSHXDP; /* Clear sign ALWAYS, irrespective of NaN */ DPSIGN(x) = 0; if (xc == IEEE754_CLASS_SNAN) { - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_nanxcpt(ieee754dp_indef(), "abs"); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_nanxcpt(ieee754dp_indef()); } return x; diff --git a/arch/mips/math-emu/dp_sqrt.c b/arch/mips/math-emu/dp_sqrt.c index b874d60a942..041bbb6124b 100644 --- a/arch/mips/math-emu/dp_sqrt.c +++ b/arch/mips/math-emu/dp_sqrt.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,12 +16,9 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754dp.h" static const unsigned table[] = { @@ -34,44 +29,49 @@ static const unsigned table[] = { 1742, 661, 130 }; -ieee754dp ieee754dp_sqrt(ieee754dp x) +union ieee754dp ieee754dp_sqrt(union ieee754dp x) { struct _ieee754_csr oldcsr; - ieee754dp y, z, t; + union ieee754dp y, z, t; unsigned scalx, yh; COMPXDP; EXPLODEXDP; - CLEARCX; + ieee754_clearcx(); FLUSHXDP; /* x == INF or NAN? */ switch (xc) { case IEEE754_CLASS_QNAN: /* sqrt(Nan) = Nan */ - return ieee754dp_nanxcpt(x, "sqrt"); + return ieee754dp_nanxcpt(x); + case IEEE754_CLASS_SNAN: - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_nanxcpt(ieee754dp_indef(), "sqrt"); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_nanxcpt(ieee754dp_indef()); + case IEEE754_CLASS_ZERO: /* sqrt(0) = 0 */ return x; + case IEEE754_CLASS_INF: if (xs) { /* sqrt(-Inf) = Nan */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_nanxcpt(ieee754dp_indef(), "sqrt"); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_nanxcpt(ieee754dp_indef()); } /* sqrt(+Inf) = Inf */ return x; + case IEEE754_CLASS_DNORM: DPDNORMX; /* fall through */ + case IEEE754_CLASS_NORM: if (xs) { /* sqrt(-x) = Nan */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_nanxcpt(ieee754dp_indef(), "sqrt"); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_nanxcpt(ieee754dp_indef()); } break; } @@ -80,7 +80,7 @@ ieee754dp ieee754dp_sqrt(ieee754dp x) oldcsr = ieee754_csr; ieee754_csr.mx &= ~IEEE754_INEXACT; ieee754_csr.sx &= ~IEEE754_INEXACT; - ieee754_csr.rm = IEEE754_RN; + ieee754_csr.rm = FPU_CSR_RN; /* adjust exponent to prevent overflow */ scalx = 0; @@ -110,19 +110,19 @@ ieee754dp ieee754dp_sqrt(ieee754dp x) /* triple to almost 56 sig. bits: y ~= sqrt(x) to within 1 ulp */ /* t=y*y; z=t; pt[n0]+=0x00100000; t+=z; z=(x-z)*y; */ z = t = ieee754dp_mul(y, y); - t.parts.bexp += 0x001; + t.bexp += 0x001; t = ieee754dp_add(t, z); z = ieee754dp_mul(ieee754dp_sub(x, z), y); /* t=z/(t+x) ; pt[n0]+=0x00100000; y+=t; */ t = ieee754dp_div(z, ieee754dp_add(t, x)); - t.parts.bexp += 0x001; + t.bexp += 0x001; y = ieee754dp_add(y, t); /* twiddle last bit to force y correctly rounded */ /* set RZ, clear INEX flag */ - ieee754_csr.rm = IEEE754_RZ; + ieee754_csr.rm = FPU_CSR_RZ; ieee754_csr.sx &= ~IEEE754_INEXACT; /* t=x/y; ...chopped quotient, possibly inexact */ @@ -139,10 +139,10 @@ ieee754dp ieee754dp_sqrt(ieee754dp x) oldcsr.sx |= IEEE754_INEXACT; switch (oldcsr.rm) { - case IEEE754_RP: + case FPU_CSR_RU: y.bits += 1; /* drop through */ - case IEEE754_RN: + case FPU_CSR_RN: t.bits += 1; break; } @@ -155,7 +155,7 @@ ieee754dp ieee754dp_sqrt(ieee754dp x) } /* py[n0]=py[n0]+scalx; ...scale back y */ - y.parts.bexp += scalx; + y.bexp += scalx; /* restore rounding mode, possibly set inexact */ ieee754_csr = oldcsr; diff --git a/arch/mips/math-emu/dp_sub.c b/arch/mips/math-emu/dp_sub.c index 91e0a4b5cbc..7a174029043 100644 --- a/arch/mips/math-emu/dp_sub.c +++ b/arch/mips/math-emu/dp_sub.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,23 +16,22 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754dp.h" -ieee754dp ieee754dp_sub(ieee754dp x, ieee754dp y) +union ieee754dp ieee754dp_sub(union ieee754dp x, union ieee754dp y) { + int s; + COMPXDP; COMPYDP; EXPLODEXDP; EXPLODEYDP; - CLEARCX; + ieee754_clearcx(); FLUSHXDP; FLUSHYDP; @@ -51,8 +48,8 @@ ieee754dp ieee754dp_sub(ieee754dp x, ieee754dp y) case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_nanxcpt(ieee754dp_indef(), "sub", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_nanxcpt(ieee754dp_indef()); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN): case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN): @@ -68,14 +65,14 @@ ieee754dp ieee754dp_sub(ieee754dp x, ieee754dp y) return x; - /* Infinity handling - */ - + /* + * Infinity handling + */ case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF): if (xs != ys) return x; - SETCX(IEEE754_INVALID_OPERATION); - return ieee754dp_xcpt(ieee754dp_indef(), "sub", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754dp_indef(); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF): case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF): @@ -87,15 +84,14 @@ ieee754dp ieee754dp_sub(ieee754dp x, ieee754dp y) case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM): return x; - /* Zero handling - */ - + /* + * Zero handling + */ case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO): if (xs != ys) return x; else - return ieee754dp_zero(ieee754_csr.rm == - IEEE754_RD); + return ieee754dp_zero(ieee754_csr.rm == FPU_CSR_RD); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO): case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO): @@ -136,15 +132,17 @@ ieee754dp ieee754dp_sub(ieee754dp x, ieee754dp y) ym <<= 3; if (xe > ye) { - /* have to shift y fraction right to align + /* + * Have to shift y fraction right to align */ - int s = xe - ye; + s = xe - ye; ym = XDPSRS(ym, s); ye += s; } else if (ye > xe) { - /* have to shift x fraction right to align + /* + * Have to shift x fraction right to align */ - int s = ye - xe; + s = ye - xe; xm = XDPSRS(xm, s); xe += s; } @@ -158,7 +156,7 @@ ieee754dp ieee754dp_sub(ieee754dp x, ieee754dp y) xe = xe; xs = xs; - if (xm >> (DP_MBITS + 1 + 3)) { /* carry out */ + if (xm >> (DP_FBITS + 1 + 3)) { /* carry out */ xm = XDPSRS1(xm); /* shift preserving sticky */ xe++; } @@ -173,7 +171,7 @@ ieee754dp ieee754dp_sub(ieee754dp x, ieee754dp y) xs = ys; } if (xm == 0) { - if (ieee754_csr.rm == IEEE754_RD) + if (ieee754_csr.rm == FPU_CSR_RD) return ieee754dp_zero(1); /* round negative inf. => sign = -1 */ else return ieee754dp_zero(0); /* other round modes => sign = 1 */ @@ -181,10 +179,11 @@ ieee754dp ieee754dp_sub(ieee754dp x, ieee754dp y) /* normalize to rounding precision */ - while ((xm >> (DP_MBITS + 3)) == 0) { + while ((xm >> (DP_FBITS + 3)) == 0) { xm <<= 1; xe--; } } - DPNORMRET2(xs, xe, xm, "sub", x, y); + + return ieee754dp_format(xs, xe, xm); } diff --git a/arch/mips/math-emu/dp_tint.c b/arch/mips/math-emu/dp_tint.c index 0ebe8598b94..6ffc336c530 100644 --- a/arch/mips/math-emu/dp_tint.c +++ b/arch/mips/math-emu/dp_tint.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,20 +16,21 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - -#include <linux/kernel.h> #include "ieee754dp.h" -int ieee754dp_tint(ieee754dp x) +int ieee754dp_tint(union ieee754dp x) { + u64 residue; + int round; + int sticky; + int odd; + COMPXDP; - CLEARCX; + ieee754_clearcx(); EXPLODEXDP; FLUSHXDP; @@ -40,10 +39,12 @@ int ieee754dp_tint(ieee754dp x) case IEEE754_CLASS_SNAN: case IEEE754_CLASS_QNAN: case IEEE754_CLASS_INF: - SETCX(IEEE754_INVALID_OPERATION); - return ieee754si_xcpt(ieee754si_indef(), "dp_tint", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754si_indef(); + case IEEE754_CLASS_ZERO: return 0; + case IEEE754_CLASS_DNORM: case IEEE754_CLASS_NORM: break; @@ -51,44 +52,39 @@ int ieee754dp_tint(ieee754dp x) if (xe > 31) { /* Set invalid. We will only use overflow for floating point overflow */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754si_xcpt(ieee754si_indef(), "dp_tint", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754si_indef(); } /* oh gawd */ - if (xe > DP_MBITS) { - xm <<= xe - DP_MBITS; - } else if (xe < DP_MBITS) { - u64 residue; - int round; - int sticky; - int odd; - + if (xe > DP_FBITS) { + xm <<= xe - DP_FBITS; + } else if (xe < DP_FBITS) { if (xe < -1) { residue = xm; round = 0; sticky = residue != 0; xm = 0; } else { - residue = xm << (64 - DP_MBITS + xe); + residue = xm << (64 - DP_FBITS + xe); round = (residue >> 63) != 0; sticky = (residue << 1) != 0; - xm >>= DP_MBITS - xe; + xm >>= DP_FBITS - xe; } /* Note: At this point upper 32 bits of xm are guaranteed to be zero */ odd = (xm & 0x1) != 0x0; switch (ieee754_csr.rm) { - case IEEE754_RN: + case FPU_CSR_RN: if (round && (sticky || odd)) xm++; break; - case IEEE754_RZ: + case FPU_CSR_RZ: break; - case IEEE754_RU: /* toward +Infinity */ + case FPU_CSR_RU: /* toward +Infinity */ if ((round || sticky) && !xs) xm++; break; - case IEEE754_RD: /* toward -Infinity */ + case FPU_CSR_RD: /* toward -Infinity */ if ((round || sticky) && xs) xm++; break; @@ -96,27 +92,14 @@ int ieee754dp_tint(ieee754dp x) /* look for valid corner case 0x80000000 */ if ((xm >> 31) != 0 && (xs == 0 || xm != 0x80000000)) { /* This can happen after rounding */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754si_xcpt(ieee754si_indef(), "dp_tint", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754si_indef(); } if (round || sticky) - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_INEXACT); } if (xs) return -xm; else return xm; } - - -unsigned int ieee754dp_tuns(ieee754dp x) -{ - ieee754dp hb = ieee754dp_1e31(); - - /* what if x < 0 ?? */ - if (ieee754dp_lt(x, hb)) - return (unsigned) ieee754dp_tint(x); - - return (unsigned) ieee754dp_tint(ieee754dp_sub(x, hb)) | - ((unsigned) 1 << 31); -} diff --git a/arch/mips/math-emu/dp_tlong.c b/arch/mips/math-emu/dp_tlong.c index 133ce2ba001..9cdc145b75e 100644 --- a/arch/mips/math-emu/dp_tlong.c +++ b/arch/mips/math-emu/dp_tlong.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,19 +16,21 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754dp.h" -s64 ieee754dp_tlong(ieee754dp x) +s64 ieee754dp_tlong(union ieee754dp x) { + u64 residue; + int round; + int sticky; + int odd; + COMPXDP; - CLEARCX; + ieee754_clearcx(); EXPLODEXDP; FLUSHXDP; @@ -39,10 +39,12 @@ s64 ieee754dp_tlong(ieee754dp x) case IEEE754_CLASS_SNAN: case IEEE754_CLASS_QNAN: case IEEE754_CLASS_INF: - SETCX(IEEE754_INVALID_OPERATION); - return ieee754di_xcpt(ieee754di_indef(), "dp_tlong", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754di_indef(); + case IEEE754_CLASS_ZERO: return 0; + case IEEE754_CLASS_DNORM: case IEEE754_CLASS_NORM: break; @@ -53,18 +55,13 @@ s64 ieee754dp_tlong(ieee754dp x) return -0x8000000000000000LL; /* Set invalid. We will only use overflow for floating point overflow */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754di_xcpt(ieee754di_indef(), "dp_tlong", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754di_indef(); } /* oh gawd */ - if (xe > DP_MBITS) { - xm <<= xe - DP_MBITS; - } else if (xe < DP_MBITS) { - u64 residue; - int round; - int sticky; - int odd; - + if (xe > DP_FBITS) { + xm <<= xe - DP_FBITS; + } else if (xe < DP_FBITS) { if (xe < -1) { residue = xm; round = 0; @@ -75,51 +72,38 @@ s64 ieee754dp_tlong(ieee754dp x) * so we do it in two steps. Be aware that xe * may be -1 */ residue = xm << (xe + 1); - residue <<= 63 - DP_MBITS; + residue <<= 63 - DP_FBITS; round = (residue >> 63) != 0; sticky = (residue << 1) != 0; - xm >>= DP_MBITS - xe; + xm >>= DP_FBITS - xe; } odd = (xm & 0x1) != 0x0; switch (ieee754_csr.rm) { - case IEEE754_RN: + case FPU_CSR_RN: if (round && (sticky || odd)) xm++; break; - case IEEE754_RZ: + case FPU_CSR_RZ: break; - case IEEE754_RU: /* toward +Infinity */ + case FPU_CSR_RU: /* toward +Infinity */ if ((round || sticky) && !xs) xm++; break; - case IEEE754_RD: /* toward -Infinity */ + case FPU_CSR_RD: /* toward -Infinity */ if ((round || sticky) && xs) xm++; break; } if ((xm >> 63) != 0) { /* This can happen after rounding */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754di_xcpt(ieee754di_indef(), "dp_tlong", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754di_indef(); } if (round || sticky) - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_INEXACT); } if (xs) return -xm; else return xm; } - - -u64 ieee754dp_tulong(ieee754dp x) -{ - ieee754dp hb = ieee754dp_1e63(); - - /* what if x < 0 ?? */ - if (ieee754dp_lt(x, hb)) - return (u64) ieee754dp_tlong(x); - - return (u64) ieee754dp_tlong(ieee754dp_sub(x, hb)) | - (1ULL << 63); -} diff --git a/arch/mips/math-emu/dsemul.c b/arch/mips/math-emu/dsemul.c index 7ea622ab8da..4f514f3724c 100644 --- a/arch/mips/math-emu/dsemul.c +++ b/arch/mips/math-emu/dsemul.c @@ -1,30 +1,12 @@ -#include <linux/compiler.h> -#include <linux/mm.h> -#include <linux/signal.h> -#include <linux/smp.h> - -#include <asm/asm.h> -#include <asm/bootinfo.h> -#include <asm/byteorder.h> -#include <asm/cpu.h> -#include <asm/inst.h> -#include <asm/processor.h> -#include <asm/uaccess.h> #include <asm/branch.h> -#include <asm/mipsregs.h> #include <asm/cacheflush.h> - #include <asm/fpu_emulator.h> +#include <asm/inst.h> +#include <asm/mipsregs.h> +#include <asm/uaccess.h> #include "ieee754.h" -/* Strap kernel emulator for full MIPS IV emulation */ - -#ifdef __mips -#undef __mips -#endif -#define __mips 4 - /* * Emulate the arbritrary instruction ir at xcp->cp0_epc. Required when * we have to emulate the instruction in a COP1 branch delay slot. Do @@ -59,13 +41,11 @@ int mips_dsemul(struct pt_regs *regs, mips_instruction ir, unsigned long cpc) (ir == 0)) { /* NOP is easy */ regs->cp0_epc = cpc; - regs->cp0_cause &= ~CAUSEF_BD; + clear_delay_slot(regs); return 0; } -#ifdef DSEMUL_TRACE - printk("dsemul %lx %lx\n", regs->cp0_epc, cpc); -#endif + pr_debug("dsemul %lx %lx\n", regs->cp0_epc, cpc); /* * The strategy is to push the instruction onto the user stack @@ -167,9 +147,8 @@ int do_dsemulret(struct pt_regs *xcp) * emulating the branch delay instruction. */ -#ifdef DSEMUL_TRACE - printk("dsemulret\n"); -#endif + pr_debug("dsemulret\n"); + if (__get_user(epc, &fr->epc)) { /* Saved EPC */ /* This is not a good situation to be in */ force_sig(SIGBUS, current); diff --git a/arch/mips/math-emu/ieee754.c b/arch/mips/math-emu/ieee754.c index 0015cf1989d..53f1d228708 100644 --- a/arch/mips/math-emu/ieee754.c +++ b/arch/mips/math-emu/ieee754.c @@ -10,8 +10,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -23,105 +21,69 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ +#include <linux/compiler.h> -#include "ieee754int.h" +#include "ieee754.h" #include "ieee754sp.h" #include "ieee754dp.h" -#define DP_EBIAS 1023 -#define DP_EMIN (-1022) -#define DP_EMAX 1023 - -#define SP_EBIAS 127 -#define SP_EMIN (-126) -#define SP_EMAX 127 - -/* special constants -*/ - - -#if (defined(BYTE_ORDER) && BYTE_ORDER == LITTLE_ENDIAN) || defined(__MIPSEL__) -#define SPSTR(s, b, m) {m, b, s} -#define DPSTR(s, b, mh, ml) {ml, mh, b, s} -#endif - -#ifdef __MIPSEB__ -#define SPSTR(s, b, m) {s, b, m} -#define DPSTR(s, b, mh, ml) {s, b, mh, ml} -#endif +/* + * Special constants + */ -const struct ieee754dp_konst __ieee754dp_spcvals[] = { - DPSTR(0, DP_EMIN - 1 + DP_EBIAS, 0, 0), /* + zero */ - DPSTR(1, DP_EMIN - 1 + DP_EBIAS, 0, 0), /* - zero */ - DPSTR(0, DP_EBIAS, 0, 0), /* + 1.0 */ - DPSTR(1, DP_EBIAS, 0, 0), /* - 1.0 */ - DPSTR(0, 3 + DP_EBIAS, 0x40000, 0), /* + 10.0 */ - DPSTR(1, 3 + DP_EBIAS, 0x40000, 0), /* - 10.0 */ - DPSTR(0, DP_EMAX + 1 + DP_EBIAS, 0, 0), /* + infinity */ - DPSTR(1, DP_EMAX + 1 + DP_EBIAS, 0, 0), /* - infinity */ - DPSTR(0, DP_EMAX+1+DP_EBIAS, 0x7FFFF, 0xFFFFFFFF), /* + indef quiet Nan */ - DPSTR(0, DP_EMAX + DP_EBIAS, 0xFFFFF, 0xFFFFFFFF), /* + max */ - DPSTR(1, DP_EMAX + DP_EBIAS, 0xFFFFF, 0xFFFFFFFF), /* - max */ - DPSTR(0, DP_EMIN + DP_EBIAS, 0, 0), /* + min normal */ - DPSTR(1, DP_EMIN + DP_EBIAS, 0, 0), /* - min normal */ - DPSTR(0, DP_EMIN - 1 + DP_EBIAS, 0, 1), /* + min denormal */ - DPSTR(1, DP_EMIN - 1 + DP_EBIAS, 0, 1), /* - min denormal */ - DPSTR(0, 31 + DP_EBIAS, 0, 0), /* + 1.0e31 */ - DPSTR(0, 63 + DP_EBIAS, 0, 0), /* + 1.0e63 */ -}; +#define DPCNST(s, b, m) \ +{ \ + .sign = (s), \ + .bexp = (b) + DP_EBIAS, \ + .mant = (m) \ +} -const struct ieee754sp_konst __ieee754sp_spcvals[] = { - SPSTR(0, SP_EMIN - 1 + SP_EBIAS, 0), /* + zero */ - SPSTR(1, SP_EMIN - 1 + SP_EBIAS, 0), /* - zero */ - SPSTR(0, SP_EBIAS, 0), /* + 1.0 */ - SPSTR(1, SP_EBIAS, 0), /* - 1.0 */ - SPSTR(0, 3 + SP_EBIAS, 0x200000), /* + 10.0 */ - SPSTR(1, 3 + SP_EBIAS, 0x200000), /* - 10.0 */ - SPSTR(0, SP_EMAX + 1 + SP_EBIAS, 0), /* + infinity */ - SPSTR(1, SP_EMAX + 1 + SP_EBIAS, 0), /* - infinity */ - SPSTR(0, SP_EMAX+1+SP_EBIAS, 0x3FFFFF), /* + indef quiet Nan */ - SPSTR(0, SP_EMAX + SP_EBIAS, 0x7FFFFF), /* + max normal */ - SPSTR(1, SP_EMAX + SP_EBIAS, 0x7FFFFF), /* - max normal */ - SPSTR(0, SP_EMIN + SP_EBIAS, 0), /* + min normal */ - SPSTR(1, SP_EMIN + SP_EBIAS, 0), /* - min normal */ - SPSTR(0, SP_EMIN - 1 + SP_EBIAS, 1), /* + min denormal */ - SPSTR(1, SP_EMIN - 1 + SP_EBIAS, 1), /* - min denormal */ - SPSTR(0, 31 + SP_EBIAS, 0), /* + 1.0e31 */ - SPSTR(0, 63 + SP_EBIAS, 0), /* + 1.0e63 */ +const union ieee754dp __ieee754dp_spcvals[] = { + DPCNST(0, DP_EMIN - 1, 0x0000000000000ULL), /* + zero */ + DPCNST(1, DP_EMIN - 1, 0x0000000000000ULL), /* - zero */ + DPCNST(0, 0, 0x0000000000000ULL), /* + 1.0 */ + DPCNST(1, 0, 0x0000000000000ULL), /* - 1.0 */ + DPCNST(0, 3, 0x4000000000000ULL), /* + 10.0 */ + DPCNST(1, 3, 0x4000000000000ULL), /* - 10.0 */ + DPCNST(0, DP_EMAX + 1, 0x0000000000000ULL), /* + infinity */ + DPCNST(1, DP_EMAX + 1, 0x0000000000000ULL), /* - infinity */ + DPCNST(0, DP_EMAX + 1, 0x7FFFFFFFFFFFFULL), /* + indef quiet Nan */ + DPCNST(0, DP_EMAX, 0xFFFFFFFFFFFFFULL), /* + max */ + DPCNST(1, DP_EMAX, 0xFFFFFFFFFFFFFULL), /* - max */ + DPCNST(0, DP_EMIN, 0x0000000000000ULL), /* + min normal */ + DPCNST(1, DP_EMIN, 0x0000000000000ULL), /* - min normal */ + DPCNST(0, DP_EMIN - 1, 0x0000000000001ULL), /* + min denormal */ + DPCNST(1, DP_EMIN - 1, 0x0000000000001ULL), /* - min denormal */ + DPCNST(0, 31, 0x0000000000000ULL), /* + 1.0e31 */ + DPCNST(0, 63, 0x0000000000000ULL), /* + 1.0e63 */ }; - -int ieee754si_xcpt(int r, const char *op, ...) -{ - struct ieee754xctx ax; - - if (!TSTX()) - return r; - ax.op = op; - ax.rt = IEEE754_RT_SI; - ax.rv.si = r; - va_start(ax.ap, op); - ieee754_xcpt(&ax); - va_end(ax.ap); - return ax.rv.si; +#define SPCNST(s, b, m) \ +{ \ + .sign = (s), \ + .bexp = (b) + SP_EBIAS, \ + .mant = (m) \ } -s64 ieee754di_xcpt(s64 r, const char *op, ...) -{ - struct ieee754xctx ax; - - if (!TSTX()) - return r; - ax.op = op; - ax.rt = IEEE754_RT_DI; - ax.rv.di = r; - va_start(ax.ap, op); - ieee754_xcpt(&ax); - va_end(ax.ap); - return ax.rv.di; -} +const union ieee754sp __ieee754sp_spcvals[] = { + SPCNST(0, SP_EMIN - 1, 0x000000), /* + zero */ + SPCNST(1, SP_EMIN - 1, 0x000000), /* - zero */ + SPCNST(0, 0, 0x000000), /* + 1.0 */ + SPCNST(1, 0, 0x000000), /* - 1.0 */ + SPCNST(0, 3, 0x200000), /* + 10.0 */ + SPCNST(1, 3, 0x200000), /* - 10.0 */ + SPCNST(0, SP_EMAX + 1, 0x000000), /* + infinity */ + SPCNST(1, SP_EMAX + 1, 0x000000), /* - infinity */ + SPCNST(0, SP_EMAX + 1, 0x3FFFFF), /* + indef quiet Nan */ + SPCNST(0, SP_EMAX, 0x7FFFFF), /* + max normal */ + SPCNST(1, SP_EMAX, 0x7FFFFF), /* - max normal */ + SPCNST(0, SP_EMIN, 0x000000), /* + min normal */ + SPCNST(1, SP_EMIN, 0x000000), /* - min normal */ + SPCNST(0, SP_EMIN - 1, 0x000001), /* + min denormal */ + SPCNST(1, SP_EMIN - 1, 0x000001), /* - min denormal */ + SPCNST(0, 31, 0x000000), /* + 1.0e31 */ + SPCNST(0, 63, 0x000000), /* + 1.0e63 */ +}; diff --git a/arch/mips/math-emu/ieee754.h b/arch/mips/math-emu/ieee754.h index 22796e01206..43c4fb522ac 100644 --- a/arch/mips/math-emu/ieee754.h +++ b/arch/mips/math-emu/ieee754.h @@ -13,7 +13,7 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. * * Nov 7, 2000 * Modification to allow integration with Linux kernel @@ -24,186 +24,93 @@ #ifndef __ARCH_MIPS_MATH_EMU_IEEE754_H #define __ARCH_MIPS_MATH_EMU_IEEE754_H +#include <linux/compiler.h> #include <asm/byteorder.h> +#include <linux/kernel.h> #include <linux/types.h> #include <linux/sched.h> +#include <asm/bitfield.h> -/* - * Not very pretty, but the Linux kernel's normal va_list definition - * does not allow it to be used as a structure element, as it is here. - */ -#ifndef _STDARG_H -#include <stdarg.h> -#endif - -#ifdef __LITTLE_ENDIAN -struct ieee754dp_konst { - unsigned mantlo:32; - unsigned manthi:20; - unsigned bexp:11; - unsigned sign:1; -}; -struct ieee754sp_konst { - unsigned mant:23; - unsigned bexp:8; - unsigned sign:1; -}; - -typedef union _ieee754dp { - struct ieee754dp_konst oparts; +union ieee754dp { struct { - u64 mant:52; - unsigned int bexp:11; - unsigned int sign:1; - } parts; + __BITFIELD_FIELD(unsigned int sign:1, + __BITFIELD_FIELD(unsigned int bexp:11, + __BITFIELD_FIELD(u64 mant:52, + ;))) + }; u64 bits; - double d; -} ieee754dp; - -typedef union _ieee754sp { - struct ieee754sp_konst parts; - float f; - u32 bits; -} ieee754sp; -#endif - -#ifdef __BIG_ENDIAN -struct ieee754dp_konst { - unsigned sign:1; - unsigned bexp:11; - unsigned manthi:20; - unsigned mantlo:32; }; -typedef union _ieee754dp { - struct ieee754dp_konst oparts; +union ieee754sp { struct { - unsigned int sign:1; - unsigned int bexp:11; - u64 mant:52; - } parts; - double d; - u64 bits; -} ieee754dp; - -struct ieee754sp_konst { - unsigned sign:1; - unsigned bexp:8; - unsigned mant:23; -}; - -typedef union _ieee754sp { - struct ieee754sp_konst parts; - float f; + __BITFIELD_FIELD(unsigned sign:1, + __BITFIELD_FIELD(unsigned bexp:8, + __BITFIELD_FIELD(unsigned mant:23, + ;))) + }; u32 bits; -} ieee754sp; -#endif +}; /* * single precision (often aka float) */ -int ieee754sp_finite(ieee754sp x); -int ieee754sp_class(ieee754sp x); - -ieee754sp ieee754sp_abs(ieee754sp x); -ieee754sp ieee754sp_neg(ieee754sp x); -ieee754sp ieee754sp_scalb(ieee754sp x, int); -ieee754sp ieee754sp_logb(ieee754sp x); - -/* x with sign of y */ -ieee754sp ieee754sp_copysign(ieee754sp x, ieee754sp y); - -ieee754sp ieee754sp_add(ieee754sp x, ieee754sp y); -ieee754sp ieee754sp_sub(ieee754sp x, ieee754sp y); -ieee754sp ieee754sp_mul(ieee754sp x, ieee754sp y); -ieee754sp ieee754sp_div(ieee754sp x, ieee754sp y); - -ieee754sp ieee754sp_fint(int x); -ieee754sp ieee754sp_funs(unsigned x); -ieee754sp ieee754sp_flong(s64 x); -ieee754sp ieee754sp_fulong(u64 x); -ieee754sp ieee754sp_fdp(ieee754dp x); - -int ieee754sp_tint(ieee754sp x); -unsigned int ieee754sp_tuns(ieee754sp x); -s64 ieee754sp_tlong(ieee754sp x); -u64 ieee754sp_tulong(ieee754sp x); - -int ieee754sp_cmp(ieee754sp x, ieee754sp y, int cop, int sig); -/* - * basic sp math - */ -ieee754sp ieee754sp_modf(ieee754sp x, ieee754sp * ip); -ieee754sp ieee754sp_frexp(ieee754sp x, int *exp); -ieee754sp ieee754sp_ldexp(ieee754sp x, int exp); +int ieee754sp_class(union ieee754sp x); -ieee754sp ieee754sp_ceil(ieee754sp x); -ieee754sp ieee754sp_floor(ieee754sp x); -ieee754sp ieee754sp_trunc(ieee754sp x); +union ieee754sp ieee754sp_abs(union ieee754sp x); +union ieee754sp ieee754sp_neg(union ieee754sp x); -ieee754sp ieee754sp_sqrt(ieee754sp x); +union ieee754sp ieee754sp_add(union ieee754sp x, union ieee754sp y); +union ieee754sp ieee754sp_sub(union ieee754sp x, union ieee754sp y); +union ieee754sp ieee754sp_mul(union ieee754sp x, union ieee754sp y); +union ieee754sp ieee754sp_div(union ieee754sp x, union ieee754sp y); -/* - * double precision (often aka double) -*/ -int ieee754dp_finite(ieee754dp x); -int ieee754dp_class(ieee754dp x); +union ieee754sp ieee754sp_fint(int x); +union ieee754sp ieee754sp_flong(s64 x); +union ieee754sp ieee754sp_fdp(union ieee754dp x); -/* x with sign of y */ -ieee754dp ieee754dp_copysign(ieee754dp x, ieee754dp y); +int ieee754sp_tint(union ieee754sp x); +s64 ieee754sp_tlong(union ieee754sp x); -ieee754dp ieee754dp_add(ieee754dp x, ieee754dp y); -ieee754dp ieee754dp_sub(ieee754dp x, ieee754dp y); -ieee754dp ieee754dp_mul(ieee754dp x, ieee754dp y); -ieee754dp ieee754dp_div(ieee754dp x, ieee754dp y); +int ieee754sp_cmp(union ieee754sp x, union ieee754sp y, int cop, int sig); -ieee754dp ieee754dp_abs(ieee754dp x); -ieee754dp ieee754dp_neg(ieee754dp x); -ieee754dp ieee754dp_scalb(ieee754dp x, int); +union ieee754sp ieee754sp_sqrt(union ieee754sp x); -/* return exponent as integer in floating point format - */ -ieee754dp ieee754dp_logb(ieee754dp x); +/* + * double precision (often aka double) +*/ +int ieee754dp_class(union ieee754dp x); -ieee754dp ieee754dp_fint(int x); -ieee754dp ieee754dp_funs(unsigned x); -ieee754dp ieee754dp_flong(s64 x); -ieee754dp ieee754dp_fulong(u64 x); -ieee754dp ieee754dp_fsp(ieee754sp x); +union ieee754dp ieee754dp_add(union ieee754dp x, union ieee754dp y); +union ieee754dp ieee754dp_sub(union ieee754dp x, union ieee754dp y); +union ieee754dp ieee754dp_mul(union ieee754dp x, union ieee754dp y); +union ieee754dp ieee754dp_div(union ieee754dp x, union ieee754dp y); -ieee754dp ieee754dp_ceil(ieee754dp x); -ieee754dp ieee754dp_floor(ieee754dp x); -ieee754dp ieee754dp_trunc(ieee754dp x); +union ieee754dp ieee754dp_abs(union ieee754dp x); +union ieee754dp ieee754dp_neg(union ieee754dp x); -int ieee754dp_tint(ieee754dp x); -unsigned int ieee754dp_tuns(ieee754dp x); -s64 ieee754dp_tlong(ieee754dp x); -u64 ieee754dp_tulong(ieee754dp x); +union ieee754dp ieee754dp_fint(int x); +union ieee754dp ieee754dp_flong(s64 x); +union ieee754dp ieee754dp_fsp(union ieee754sp x); -int ieee754dp_cmp(ieee754dp x, ieee754dp y, int cop, int sig); -/* - * basic sp math - */ -ieee754dp ieee754dp_modf(ieee754dp x, ieee754dp * ip); -ieee754dp ieee754dp_frexp(ieee754dp x, int *exp); -ieee754dp ieee754dp_ldexp(ieee754dp x, int exp); +int ieee754dp_tint(union ieee754dp x); +s64 ieee754dp_tlong(union ieee754dp x); -ieee754dp ieee754dp_ceil(ieee754dp x); -ieee754dp ieee754dp_floor(ieee754dp x); -ieee754dp ieee754dp_trunc(ieee754dp x); +int ieee754dp_cmp(union ieee754dp x, union ieee754dp y, int cop, int sig); -ieee754dp ieee754dp_sqrt(ieee754dp x); +union ieee754dp ieee754dp_sqrt(union ieee754dp x); /* 5 types of floating point number */ -#define IEEE754_CLASS_NORM 0x00 -#define IEEE754_CLASS_ZERO 0x01 -#define IEEE754_CLASS_DNORM 0x02 -#define IEEE754_CLASS_INF 0x03 -#define IEEE754_CLASS_SNAN 0x04 -#define IEEE754_CLASS_QNAN 0x05 +enum { + IEEE754_CLASS_NORM = 0x00, + IEEE754_CLASS_ZERO = 0x01, + IEEE754_CLASS_DNORM = 0x02, + IEEE754_CLASS_INF = 0x03, + IEEE754_CLASS_SNAN = 0x04, + IEEE754_CLASS_QNAN = 0x05, +}; /* exception numbers */ #define IEEE754_INEXACT 0x01 @@ -219,114 +126,84 @@ ieee754dp ieee754dp_sqrt(ieee754dp x); #define IEEE754_CGT 0x04 #define IEEE754_CUN 0x08 -/* rounding mode -*/ -#define IEEE754_RN 0 /* round to nearest */ -#define IEEE754_RZ 1 /* round toward zero */ -#define IEEE754_RD 2 /* round toward -Infinity */ -#define IEEE754_RU 3 /* round toward +Infinity */ - -/* other naming */ -#define IEEE754_RM IEEE754_RD -#define IEEE754_RP IEEE754_RU - /* "normal" comparisons */ -static inline int ieee754sp_eq(ieee754sp x, ieee754sp y) +static inline int ieee754sp_eq(union ieee754sp x, union ieee754sp y) { return ieee754sp_cmp(x, y, IEEE754_CEQ, 0); } -static inline int ieee754sp_ne(ieee754sp x, ieee754sp y) +static inline int ieee754sp_ne(union ieee754sp x, union ieee754sp y) { return ieee754sp_cmp(x, y, IEEE754_CLT | IEEE754_CGT | IEEE754_CUN, 0); } -static inline int ieee754sp_lt(ieee754sp x, ieee754sp y) +static inline int ieee754sp_lt(union ieee754sp x, union ieee754sp y) { return ieee754sp_cmp(x, y, IEEE754_CLT, 0); } -static inline int ieee754sp_le(ieee754sp x, ieee754sp y) +static inline int ieee754sp_le(union ieee754sp x, union ieee754sp y) { return ieee754sp_cmp(x, y, IEEE754_CLT | IEEE754_CEQ, 0); } -static inline int ieee754sp_gt(ieee754sp x, ieee754sp y) +static inline int ieee754sp_gt(union ieee754sp x, union ieee754sp y) { return ieee754sp_cmp(x, y, IEEE754_CGT, 0); } -static inline int ieee754sp_ge(ieee754sp x, ieee754sp y) +static inline int ieee754sp_ge(union ieee754sp x, union ieee754sp y) { return ieee754sp_cmp(x, y, IEEE754_CGT | IEEE754_CEQ, 0); } -static inline int ieee754dp_eq(ieee754dp x, ieee754dp y) +static inline int ieee754dp_eq(union ieee754dp x, union ieee754dp y) { return ieee754dp_cmp(x, y, IEEE754_CEQ, 0); } -static inline int ieee754dp_ne(ieee754dp x, ieee754dp y) +static inline int ieee754dp_ne(union ieee754dp x, union ieee754dp y) { return ieee754dp_cmp(x, y, IEEE754_CLT | IEEE754_CGT | IEEE754_CUN, 0); } -static inline int ieee754dp_lt(ieee754dp x, ieee754dp y) +static inline int ieee754dp_lt(union ieee754dp x, union ieee754dp y) { return ieee754dp_cmp(x, y, IEEE754_CLT, 0); } -static inline int ieee754dp_le(ieee754dp x, ieee754dp y) +static inline int ieee754dp_le(union ieee754dp x, union ieee754dp y) { return ieee754dp_cmp(x, y, IEEE754_CLT | IEEE754_CEQ, 0); } -static inline int ieee754dp_gt(ieee754dp x, ieee754dp y) +static inline int ieee754dp_gt(union ieee754dp x, union ieee754dp y) { return ieee754dp_cmp(x, y, IEEE754_CGT, 0); } -static inline int ieee754dp_ge(ieee754dp x, ieee754dp y) +static inline int ieee754dp_ge(union ieee754dp x, union ieee754dp y) { return ieee754dp_cmp(x, y, IEEE754_CGT | IEEE754_CEQ, 0); } - -/* - * Like strtod - */ -ieee754dp ieee754dp_fstr(const char *s, char **endp); -char *ieee754dp_tstr(ieee754dp x, int prec, int fmt, int af); - - /* * The control status register */ struct _ieee754_csr { -#ifdef __BIG_ENDIAN - unsigned pad0:7; - unsigned nod:1; /* set 1 for no denormalised numbers */ - unsigned c:1; /* condition */ - unsigned pad1:5; - unsigned cx:6; /* exceptions this operation */ - unsigned mx:5; /* exception enable mask */ - unsigned sx:5; /* exceptions total */ - unsigned rm:2; /* current rounding mode */ -#endif -#ifdef __LITTLE_ENDIAN - unsigned rm:2; /* current rounding mode */ - unsigned sx:5; /* exceptions total */ - unsigned mx:5; /* exception enable mask */ - unsigned cx:6; /* exceptions this operation */ - unsigned pad1:5; - unsigned c:1; /* condition */ - unsigned nod:1; /* set 1 for no denormalised numbers */ - unsigned pad0:7; -#endif + __BITFIELD_FIELD(unsigned pad0:7, + __BITFIELD_FIELD(unsigned nod:1, /* set 1 for no denormalised numbers */ + __BITFIELD_FIELD(unsigned c:1, /* condition */ + __BITFIELD_FIELD(unsigned pad1:5, + __BITFIELD_FIELD(unsigned cx:6, /* exceptions this operation */ + __BITFIELD_FIELD(unsigned mx:5, /* exception enable mask */ + __BITFIELD_FIELD(unsigned sx:5, /* exceptions total */ + __BITFIELD_FIELD(unsigned rm:2, /* current rounding mode */ + ;)))))))) }; #define ieee754_csr (*(struct _ieee754_csr *)(¤t->thread.fpu.fcr31)) @@ -377,8 +254,8 @@ static inline int ieee754_sxtest(unsigned n) } /* debugging */ -ieee754sp ieee754sp_dump(char *s, ieee754sp x); -ieee754dp ieee754dp_dump(char *s, ieee754dp x); +union ieee754sp ieee754sp_dump(char *s, union ieee754sp x); +union ieee754dp ieee754dp_dump(char *s, union ieee754dp x); #define IEEE754_SPCVAL_PZERO 0 #define IEEE754_SPCVAL_NZERO 1 @@ -398,10 +275,10 @@ ieee754dp ieee754dp_dump(char *s, ieee754dp x); #define IEEE754_SPCVAL_P1E31 15 /* + 1.0e31 */ #define IEEE754_SPCVAL_P1E63 16 /* + 1.0e63 */ -extern const struct ieee754dp_konst __ieee754dp_spcvals[]; -extern const struct ieee754sp_konst __ieee754sp_spcvals[]; -#define ieee754dp_spcvals ((const ieee754dp *)__ieee754dp_spcvals) -#define ieee754sp_spcvals ((const ieee754sp *)__ieee754sp_spcvals) +extern const union ieee754dp __ieee754dp_spcvals[]; +extern const union ieee754sp __ieee754sp_spcvals[]; +#define ieee754dp_spcvals ((const union ieee754dp *)__ieee754dp_spcvals) +#define ieee754sp_spcvals ((const union ieee754sp *)__ieee754sp_spcvals) /* * Return infinity with given sign @@ -431,28 +308,15 @@ extern const struct ieee754sp_konst __ieee754sp_spcvals[]; /* * Indefinite integer value */ -#define ieee754si_indef() INT_MAX -#ifdef LONG_LONG_MAX -#define ieee754di_indef() LONG_LONG_MAX -#else -#define ieee754di_indef() ((s64)(~0ULL>>1)) -#endif - -/* IEEE exception context, passed to handler */ -struct ieee754xctx { - const char *op; /* operation name */ - int rt; /* result type */ - union { - ieee754sp sp; /* single precision */ - ieee754dp dp; /* double precision */ -#ifdef IEEE854_XP - ieee754xp xp; /* extended precision */ -#endif - int si; /* standard signed integer (32bits) */ - s64 di; /* extended signed integer (64bits) */ - } rv; /* default result format implied by op */ - va_list ap; -}; +static inline int ieee754si_indef(void) +{ + return INT_MAX; +} + +static inline s64 ieee754di_indef(void) +{ + return S64_MAX; +} /* result types for xctx.rt */ #define IEEE754_RT_SP 0 @@ -461,8 +325,6 @@ struct ieee754xctx { #define IEEE754_RT_SI 3 #define IEEE754_RT_DI 4 -extern void ieee754_xcpt(struct ieee754xctx *xcp); - /* compat */ #define ieee754dp_fix(x) ieee754dp_tint(x) #define ieee754sp_fix(x) ieee754sp_tint(x) diff --git a/arch/mips/math-emu/ieee754d.c b/arch/mips/math-emu/ieee754d.c index 9599bdd3258..a04e8a7e5ac 100644 --- a/arch/mips/math-emu/ieee754d.c +++ b/arch/mips/math-emu/ieee754d.c @@ -16,7 +16,7 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. * * Nov 7, 2000 * Modified to build and operate in Linux kernel environment. @@ -25,38 +25,13 @@ * Copyright (C) 2000 MIPS Technologies, Inc. All rights reserved. */ -#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/printk.h> #include "ieee754.h" +#include "ieee754sp.h" +#include "ieee754dp.h" -#define DP_EBIAS 1023 -#define DP_EMIN (-1022) -#define DP_EMAX 1023 -#define DP_FBITS 52 - -#define SP_EBIAS 127 -#define SP_EMIN (-126) -#define SP_EMAX 127 -#define SP_FBITS 23 - -#define DP_MBIT(x) ((u64)1 << (x)) -#define DP_HIDDEN_BIT DP_MBIT(DP_FBITS) -#define DP_SIGN_BIT DP_MBIT(63) - - -#define SP_MBIT(x) ((u32)1 << (x)) -#define SP_HIDDEN_BIT SP_MBIT(SP_FBITS) -#define SP_SIGN_BIT SP_MBIT(31) - - -#define SPSIGN(sp) (sp.parts.sign) -#define SPBEXP(sp) (sp.parts.bexp) -#define SPMANT(sp) (sp.parts.mant) - -#define DPSIGN(dp) (dp.parts.sign) -#define DPBEXP(dp) (dp.parts.bexp) -#define DPMANT(dp) (dp.parts.mant) - -ieee754dp ieee754dp_dump(char *m, ieee754dp x) +union ieee754dp ieee754dp_dump(char *m, union ieee754dp x) { int i; @@ -96,7 +71,7 @@ ieee754dp ieee754dp_dump(char *m, ieee754dp x) return x; } -ieee754sp ieee754sp_dump(char *m, ieee754sp x) +union ieee754sp ieee754sp_dump(char *m, union ieee754sp x) { int i; diff --git a/arch/mips/math-emu/ieee754dp.c b/arch/mips/math-emu/ieee754dp.c index 068e56be8de..fd134675fc2 100644 --- a/arch/mips/math-emu/ieee754dp.c +++ b/arch/mips/math-emu/ieee754dp.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,104 +16,68 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ +#include <linux/compiler.h> #include "ieee754dp.h" -int ieee754dp_class(ieee754dp x) +int ieee754dp_class(union ieee754dp x) { COMPXDP; EXPLODEXDP; return xc; } -int ieee754dp_isnan(ieee754dp x) +int ieee754dp_isnan(union ieee754dp x) { return ieee754dp_class(x) >= IEEE754_CLASS_SNAN; } -int ieee754dp_issnan(ieee754dp x) +static inline int ieee754dp_issnan(union ieee754dp x) { assert(ieee754dp_isnan(x)); - return ((DPMANT(x) & DP_MBIT(DP_MBITS-1)) == DP_MBIT(DP_MBITS-1)); + return ((DPMANT(x) & DP_MBIT(DP_FBITS-1)) == DP_MBIT(DP_FBITS-1)); } -ieee754dp ieee754dp_xcpt(ieee754dp r, const char *op, ...) -{ - struct ieee754xctx ax; - if (!TSTX()) - return r; - - ax.op = op; - ax.rt = IEEE754_RT_DP; - ax.rv.dp = r; - va_start(ax.ap, op); - ieee754_xcpt(&ax); - va_end(ax.ap); - return ax.rv.dp; -} - -ieee754dp ieee754dp_nanxcpt(ieee754dp r, const char *op, ...) +union ieee754dp __cold ieee754dp_nanxcpt(union ieee754dp r) { - struct ieee754xctx ax; - assert(ieee754dp_isnan(r)); if (!ieee754dp_issnan(r)) /* QNAN does not cause invalid op !! */ return r; - if (!SETANDTESTCX(IEEE754_INVALID_OPERATION)) { + if (!ieee754_setandtestcx(IEEE754_INVALID_OPERATION)) { /* not enabled convert to a quiet NaN */ - DPMANT(r) &= (~DP_MBIT(DP_MBITS-1)); + DPMANT(r) &= (~DP_MBIT(DP_FBITS-1)); if (ieee754dp_isnan(r)) return r; else return ieee754dp_indef(); } - ax.op = op; - ax.rt = 0; - ax.rv.dp = r; - va_start(ax.ap, op); - ieee754_xcpt(&ax); - va_end(ax.ap); - return ax.rv.dp; + return r; } -ieee754dp ieee754dp_bestnan(ieee754dp x, ieee754dp y) -{ - assert(ieee754dp_isnan(x)); - assert(ieee754dp_isnan(y)); - - if (DPMANT(x) > DPMANT(y)) - return x; - else - return y; -} - - -static u64 get_rounding(int sn, u64 xm) +static u64 ieee754dp_get_rounding(int sn, u64 xm) { /* inexact must round of 3 bits */ if (xm & (DP_MBIT(3) - 1)) { switch (ieee754_csr.rm) { - case IEEE754_RZ: + case FPU_CSR_RZ: break; - case IEEE754_RN: + case FPU_CSR_RN: xm += 0x3 + ((xm >> 3) & 1); /* xm += (xm&0x8)?0x4:0x3 */ break; - case IEEE754_RU: /* toward +Infinity */ + case FPU_CSR_RU: /* toward +Infinity */ if (!sn) /* ?? */ xm += 0x8; break; - case IEEE754_RD: /* toward -Infinity */ + case FPU_CSR_RD: /* toward -Infinity */ if (sn) /* ?? */ xm += 0x8; break; @@ -130,11 +92,11 @@ static u64 get_rounding(int sn, u64 xm) * xe is an unbiased exponent * xm is 3bit extended precision value. */ -ieee754dp ieee754dp_format(int sn, int xe, u64 xm) +union ieee754dp ieee754dp_format(int sn, int xe, u64 xm) { assert(xm); /* we don't gen exact zeros (probably should) */ - assert((xm >> (DP_MBITS + 1 + 3)) == 0); /* no execess */ + assert((xm >> (DP_FBITS + 1 + 3)) == 0); /* no execess */ assert(xm & (DP_HIDDEN_BIT << 3)); if (xe < DP_EMIN) { @@ -142,32 +104,32 @@ ieee754dp ieee754dp_format(int sn, int xe, u64 xm) int es = DP_EMIN - xe; if (ieee754_csr.nod) { - SETCX(IEEE754_UNDERFLOW); - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_UNDERFLOW); + ieee754_setcx(IEEE754_INEXACT); switch(ieee754_csr.rm) { - case IEEE754_RN: - case IEEE754_RZ: + case FPU_CSR_RN: + case FPU_CSR_RZ: return ieee754dp_zero(sn); - case IEEE754_RU: /* toward +Infinity */ - if(sn == 0) + case FPU_CSR_RU: /* toward +Infinity */ + if (sn == 0) return ieee754dp_min(0); else return ieee754dp_zero(1); - case IEEE754_RD: /* toward -Infinity */ - if(sn == 0) + case FPU_CSR_RD: /* toward -Infinity */ + if (sn == 0) return ieee754dp_zero(0); else return ieee754dp_min(1); } } - if (xe == DP_EMIN - 1 - && get_rounding(sn, xm) >> (DP_MBITS + 1 + 3)) + if (xe == DP_EMIN - 1 && + ieee754dp_get_rounding(sn, xm) >> (DP_FBITS + 1 + 3)) { /* Not tiny after rounding */ - SETCX(IEEE754_INEXACT); - xm = get_rounding(sn, xm); + ieee754_setcx(IEEE754_INEXACT); + xm = ieee754dp_get_rounding(sn, xm); xm >>= 1; /* Clear grs bits */ xm &= ~(DP_MBIT(3) - 1); @@ -183,17 +145,17 @@ ieee754dp ieee754dp_format(int sn, int xe, u64 xm) } } if (xm & (DP_MBIT(3) - 1)) { - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_INEXACT); if ((xm & (DP_HIDDEN_BIT << 3)) == 0) { - SETCX(IEEE754_UNDERFLOW); + ieee754_setcx(IEEE754_UNDERFLOW); } /* inexact must round of 3 bits */ - xm = get_rounding(sn, xm); + xm = ieee754dp_get_rounding(sn, xm); /* adjust exponent for rounding add overflowing */ - if (xm >> (DP_MBITS + 3 + 1)) { + if (xm >> (DP_FBITS + 3 + 1)) { /* add causes mantissa overflow */ xm >>= 1; xe++; @@ -202,24 +164,24 @@ ieee754dp ieee754dp_format(int sn, int xe, u64 xm) /* strip grs bits */ xm >>= 3; - assert((xm >> (DP_MBITS + 1)) == 0); /* no execess */ + assert((xm >> (DP_FBITS + 1)) == 0); /* no execess */ assert(xe >= DP_EMIN); if (xe > DP_EMAX) { - SETCX(IEEE754_OVERFLOW); - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_OVERFLOW); + ieee754_setcx(IEEE754_INEXACT); /* -O can be table indexed by (rm,sn) */ switch (ieee754_csr.rm) { - case IEEE754_RN: + case FPU_CSR_RN: return ieee754dp_inf(sn); - case IEEE754_RZ: + case FPU_CSR_RZ: return ieee754dp_max(sn); - case IEEE754_RU: /* toward +Infinity */ + case FPU_CSR_RU: /* toward +Infinity */ if (sn == 0) return ieee754dp_inf(0); else return ieee754dp_max(1); - case IEEE754_RD: /* toward -Infinity */ + case FPU_CSR_RD: /* toward -Infinity */ if (sn == 0) return ieee754dp_max(0); else @@ -232,10 +194,10 @@ ieee754dp ieee754dp_format(int sn, int xe, u64 xm) /* we underflow (tiny/zero) */ assert(xe == DP_EMIN); if (ieee754_csr.mx & IEEE754_UNDERFLOW) - SETCX(IEEE754_UNDERFLOW); + ieee754_setcx(IEEE754_UNDERFLOW); return builddp(sn, DP_EMIN - 1 + DP_EBIAS, xm); } else { - assert((xm >> (DP_MBITS + 1)) == 0); /* no execess */ + assert((xm >> (DP_FBITS + 1)) == 0); /* no execess */ assert(xm & DP_HIDDEN_BIT); return builddp(sn, xe + DP_EBIAS, xm & ~DP_HIDDEN_BIT); diff --git a/arch/mips/math-emu/ieee754dp.h b/arch/mips/math-emu/ieee754dp.h index f139c724c59..61fd6fd3135 100644 --- a/arch/mips/math-emu/ieee754dp.h +++ b/arch/mips/math-emu/ieee754dp.h @@ -6,8 +6,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -19,64 +17,66 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ +#include <linux/compiler.h> #include "ieee754int.h" #define assert(expr) ((void)0) +#define DP_EBIAS 1023 +#define DP_EMIN (-1022) +#define DP_EMAX 1023 +#define DP_FBITS 52 +#define DP_MBITS 52 + +#define DP_MBIT(x) ((u64)1 << (x)) +#define DP_HIDDEN_BIT DP_MBIT(DP_FBITS) +#define DP_SIGN_BIT DP_MBIT(63) + +#define DPSIGN(dp) (dp.sign) +#define DPBEXP(dp) (dp.bexp) +#define DPMANT(dp) (dp.mant) + +static inline int ieee754dp_finite(union ieee754dp x) +{ + return DPBEXP(x) != DP_EMAX + 1 + DP_EBIAS; +} + /* 3bit extended double precision sticky right shift */ #define XDPSRS(v,rs) \ - ((rs > (DP_MBITS+3))?1:((v) >> (rs)) | ((v) << (64-(rs)) != 0)) + ((rs > (DP_FBITS+3))?1:((v) >> (rs)) | ((v) << (64-(rs)) != 0)) #define XDPSRSX1() \ - (xe++, (xm = (xm >> 1) | (xm & 1))) + (xe++, (xm = (xm >> 1) | (xm & 1))) #define XDPSRS1(v) \ - (((v) >> 1) | ((v) & 1)) + (((v) >> 1) | ((v) & 1)) /* convert denormal to normalized with extended exponent */ #define DPDNORMx(m,e) \ - while( (m >> DP_MBITS) == 0) { m <<= 1; e--; } + while ((m >> DP_FBITS) == 0) { m <<= 1; e--; } #define DPDNORMX DPDNORMx(xm, xe) #define DPDNORMY DPDNORMx(ym, ye) -static inline ieee754dp builddp(int s, int bx, u64 m) +static inline union ieee754dp builddp(int s, int bx, u64 m) { - ieee754dp r; + union ieee754dp r; assert((s) == 0 || (s) == 1); assert((bx) >= DP_EMIN - 1 + DP_EBIAS && (bx) <= DP_EMAX + 1 + DP_EBIAS); - assert(((m) >> DP_MBITS) == 0); + assert(((m) >> DP_FBITS) == 0); - r.parts.sign = s; - r.parts.bexp = bx; - r.parts.mant = m; - return r; -} + r.sign = s; + r.bexp = bx; + r.mant = m; -extern int ieee754dp_isnan(ieee754dp); -extern int ieee754dp_issnan(ieee754dp); -extern int ieee754si_xcpt(int, const char *, ...); -extern s64 ieee754di_xcpt(s64, const char *, ...); -extern ieee754dp ieee754dp_xcpt(ieee754dp, const char *, ...); -extern ieee754dp ieee754dp_nanxcpt(ieee754dp, const char *, ...); -extern ieee754dp ieee754dp_bestnan(ieee754dp, ieee754dp); -extern ieee754dp ieee754dp_format(int, int, u64); - - -#define DPNORMRET2(s, e, m, name, a0, a1) \ -{ \ - ieee754dp V = ieee754dp_format(s, e, m); \ - if(TSTX()) \ - return ieee754dp_xcpt(V, name, a0, a1); \ - else \ - return V; \ + return r; } -#define DPNORMRET1(s, e, m, name, a0) DPNORMRET2(s, e, m, name, a0, a0) +extern int ieee754dp_isnan(union ieee754dp); +extern union ieee754dp __cold ieee754dp_nanxcpt(union ieee754dp); +extern union ieee754dp ieee754dp_format(int, int, u64); diff --git a/arch/mips/math-emu/ieee754int.h b/arch/mips/math-emu/ieee754int.h index 4b6c6fb3530..f0365bb8674 100644 --- a/arch/mips/math-emu/ieee754int.h +++ b/arch/mips/math-emu/ieee754int.h @@ -6,8 +6,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -19,146 +17,125 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - +#ifndef __IEEE754INT_H +#define __IEEE754INT_H #include "ieee754.h" -#define DP_EBIAS 1023 -#define DP_EMIN (-1022) -#define DP_EMAX 1023 -#define DP_MBITS 52 - -#define SP_EBIAS 127 -#define SP_EMIN (-126) -#define SP_EMAX 127 -#define SP_MBITS 23 - -#define DP_MBIT(x) ((u64)1 << (x)) -#define DP_HIDDEN_BIT DP_MBIT(DP_MBITS) -#define DP_SIGN_BIT DP_MBIT(63) - -#define SP_MBIT(x) ((u32)1 << (x)) -#define SP_HIDDEN_BIT SP_MBIT(SP_MBITS) -#define SP_SIGN_BIT SP_MBIT(31) - - -#define SPSIGN(sp) (sp.parts.sign) -#define SPBEXP(sp) (sp.parts.bexp) -#define SPMANT(sp) (sp.parts.mant) - -#define DPSIGN(dp) (dp.parts.sign) -#define DPBEXP(dp) (dp.parts.bexp) -#define DPMANT(dp) (dp.parts.mant) - #define CLPAIR(x, y) ((x)*6+(y)) -#define CLEARCX \ - (ieee754_csr.cx = 0) - -#define SETCX(x) \ - (ieee754_csr.cx |= (x), ieee754_csr.sx |= (x)) +static inline void ieee754_clearcx(void) +{ + ieee754_csr.cx = 0; +} -#define SETANDTESTCX(x) \ - (SETCX(x), ieee754_csr.mx & (x)) +static inline void ieee754_setcx(const unsigned int flags) +{ + ieee754_csr.cx |= flags; + ieee754_csr.sx |= flags; +} -#define TSTX() \ - (ieee754_csr.cx & ieee754_csr.mx) +static inline int ieee754_setandtestcx(const unsigned int x) +{ + ieee754_setcx(x); + return ieee754_csr.mx & x; +} #define COMPXSP \ - unsigned xm; int xe; int xs __maybe_unused; int xc + unsigned xm; int xe; int xs __maybe_unused; int xc #define COMPYSP \ - unsigned ym; int ye; int ys; int yc - -#define EXPLODESP(v, vc, vs, ve, vm) \ -{\ - vs = SPSIGN(v);\ - ve = SPBEXP(v);\ - vm = SPMANT(v);\ - if(ve == SP_EMAX+1+SP_EBIAS){\ - if(vm == 0)\ - vc = IEEE754_CLASS_INF;\ - else if(vm & SP_MBIT(SP_MBITS-1)) \ - vc = IEEE754_CLASS_SNAN;\ - else \ - vc = IEEE754_CLASS_QNAN;\ - } else if(ve == SP_EMIN-1+SP_EBIAS) {\ - if(vm) {\ - ve = SP_EMIN;\ - vc = IEEE754_CLASS_DNORM;\ - } else\ - vc = IEEE754_CLASS_ZERO;\ - } else {\ - ve -= SP_EBIAS;\ - vm |= SP_HIDDEN_BIT;\ - vc = IEEE754_CLASS_NORM;\ - }\ + unsigned ym; int ye; int ys; int yc + +#define EXPLODESP(v, vc, vs, ve, vm) \ +{ \ + vs = SPSIGN(v); \ + ve = SPBEXP(v); \ + vm = SPMANT(v); \ + if (ve == SP_EMAX+1+SP_EBIAS) { \ + if (vm == 0) \ + vc = IEEE754_CLASS_INF; \ + else if (vm & SP_MBIT(SP_FBITS-1)) \ + vc = IEEE754_CLASS_SNAN; \ + else \ + vc = IEEE754_CLASS_QNAN; \ + } else if (ve == SP_EMIN-1+SP_EBIAS) { \ + if (vm) { \ + ve = SP_EMIN; \ + vc = IEEE754_CLASS_DNORM; \ + } else \ + vc = IEEE754_CLASS_ZERO; \ + } else { \ + ve -= SP_EBIAS; \ + vm |= SP_HIDDEN_BIT; \ + vc = IEEE754_CLASS_NORM; \ + } \ } #define EXPLODEXSP EXPLODESP(x, xc, xs, xe, xm) #define EXPLODEYSP EXPLODESP(y, yc, ys, ye, ym) #define COMPXDP \ -u64 xm; int xe; int xs __maybe_unused; int xc + u64 xm; int xe; int xs __maybe_unused; int xc #define COMPYDP \ -u64 ym; int ye; int ys; int yc - -#define EXPLODEDP(v, vc, vs, ve, vm) \ -{\ - vm = DPMANT(v);\ - vs = DPSIGN(v);\ - ve = DPBEXP(v);\ - if(ve == DP_EMAX+1+DP_EBIAS){\ - if(vm == 0)\ - vc = IEEE754_CLASS_INF;\ - else if(vm & DP_MBIT(DP_MBITS-1)) \ - vc = IEEE754_CLASS_SNAN;\ - else \ - vc = IEEE754_CLASS_QNAN;\ - } else if(ve == DP_EMIN-1+DP_EBIAS) {\ - if(vm) {\ - ve = DP_EMIN;\ - vc = IEEE754_CLASS_DNORM;\ - } else\ - vc = IEEE754_CLASS_ZERO;\ - } else {\ - ve -= DP_EBIAS;\ - vm |= DP_HIDDEN_BIT;\ - vc = IEEE754_CLASS_NORM;\ - }\ + u64 ym; int ye; int ys; int yc + +#define EXPLODEDP(v, vc, vs, ve, vm) \ +{ \ + vm = DPMANT(v); \ + vs = DPSIGN(v); \ + ve = DPBEXP(v); \ + if (ve == DP_EMAX+1+DP_EBIAS) { \ + if (vm == 0) \ + vc = IEEE754_CLASS_INF; \ + else if (vm & DP_MBIT(DP_FBITS-1)) \ + vc = IEEE754_CLASS_SNAN; \ + else \ + vc = IEEE754_CLASS_QNAN; \ + } else if (ve == DP_EMIN-1+DP_EBIAS) { \ + if (vm) { \ + ve = DP_EMIN; \ + vc = IEEE754_CLASS_DNORM; \ + } else \ + vc = IEEE754_CLASS_ZERO; \ + } else { \ + ve -= DP_EBIAS; \ + vm |= DP_HIDDEN_BIT; \ + vc = IEEE754_CLASS_NORM; \ + } \ } #define EXPLODEXDP EXPLODEDP(x, xc, xs, xe, xm) #define EXPLODEYDP EXPLODEDP(y, yc, ys, ye, ym) -#define FLUSHDP(v, vc, vs, ve, vm) \ - if(vc==IEEE754_CLASS_DNORM) {\ - if(ieee754_csr.nod) {\ - SETCX(IEEE754_INEXACT);\ - vc = IEEE754_CLASS_ZERO;\ - ve = DP_EMIN-1+DP_EBIAS;\ - vm = 0;\ - v = ieee754dp_zero(vs);\ - }\ +#define FLUSHDP(v, vc, vs, ve, vm) \ + if (vc==IEEE754_CLASS_DNORM) { \ + if (ieee754_csr.nod) { \ + ieee754_setcx(IEEE754_INEXACT); \ + vc = IEEE754_CLASS_ZERO; \ + ve = DP_EMIN-1+DP_EBIAS; \ + vm = 0; \ + v = ieee754dp_zero(vs); \ + } \ } -#define FLUSHSP(v, vc, vs, ve, vm) \ - if(vc==IEEE754_CLASS_DNORM) {\ - if(ieee754_csr.nod) {\ - SETCX(IEEE754_INEXACT);\ - vc = IEEE754_CLASS_ZERO;\ - ve = SP_EMIN-1+SP_EBIAS;\ - vm = 0;\ - v = ieee754sp_zero(vs);\ - }\ +#define FLUSHSP(v, vc, vs, ve, vm) \ + if (vc==IEEE754_CLASS_DNORM) { \ + if (ieee754_csr.nod) { \ + ieee754_setcx(IEEE754_INEXACT); \ + vc = IEEE754_CLASS_ZERO; \ + ve = SP_EMIN-1+SP_EBIAS; \ + vm = 0; \ + v = ieee754sp_zero(vs); \ + } \ } #define FLUSHXDP FLUSHDP(x, xc, xs, xe, xm) #define FLUSHYDP FLUSHDP(y, yc, ys, ye, ym) #define FLUSHXSP FLUSHSP(x, xc, xs, xe, xm) #define FLUSHYSP FLUSHSP(y, yc, ys, ye, ym) + +#endif /* __IEEE754INT_H */ diff --git a/arch/mips/math-emu/ieee754m.c b/arch/mips/math-emu/ieee754m.c deleted file mode 100644 index 24190f3c9dd..00000000000 --- a/arch/mips/math-emu/ieee754m.c +++ /dev/null @@ -1,55 +0,0 @@ -/* - * floor, trunc, ceil - */ -/* - * MIPS floating point support - * Copyright (C) 1994-2000 Algorithmics Ltd. - * - * ######################################################################## - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - */ - - -#include "ieee754.h" - -ieee754dp ieee754dp_floor(ieee754dp x) -{ - ieee754dp i; - - if (ieee754dp_lt(ieee754dp_modf(x, &i), ieee754dp_zero(0))) - return ieee754dp_sub(i, ieee754dp_one(0)); - else - return i; -} - -ieee754dp ieee754dp_ceil(ieee754dp x) -{ - ieee754dp i; - - if (ieee754dp_gt(ieee754dp_modf(x, &i), ieee754dp_zero(0))) - return ieee754dp_add(i, ieee754dp_one(0)); - else - return i; -} - -ieee754dp ieee754dp_trunc(ieee754dp x) -{ - ieee754dp i; - - (void) ieee754dp_modf(x, &i); - return i; -} diff --git a/arch/mips/math-emu/ieee754sp.c b/arch/mips/math-emu/ieee754sp.c index 15d1e36cfe6..d348efe9144 100644 --- a/arch/mips/math-emu/ieee754sp.c +++ b/arch/mips/math-emu/ieee754sp.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,105 +16,68 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ +#include <linux/compiler.h> #include "ieee754sp.h" -int ieee754sp_class(ieee754sp x) +int ieee754sp_class(union ieee754sp x) { COMPXSP; EXPLODEXSP; return xc; } -int ieee754sp_isnan(ieee754sp x) +int ieee754sp_isnan(union ieee754sp x) { return ieee754sp_class(x) >= IEEE754_CLASS_SNAN; } -int ieee754sp_issnan(ieee754sp x) +static inline int ieee754sp_issnan(union ieee754sp x) { assert(ieee754sp_isnan(x)); - return (SPMANT(x) & SP_MBIT(SP_MBITS-1)); + return (SPMANT(x) & SP_MBIT(SP_FBITS-1)); } -ieee754sp ieee754sp_xcpt(ieee754sp r, const char *op, ...) -{ - struct ieee754xctx ax; - - if (!TSTX()) - return r; - - ax.op = op; - ax.rt = IEEE754_RT_SP; - ax.rv.sp = r; - va_start(ax.ap, op); - ieee754_xcpt(&ax); - va_end(ax.ap); - return ax.rv.sp; -} - -ieee754sp ieee754sp_nanxcpt(ieee754sp r, const char *op, ...) +union ieee754sp __cold ieee754sp_nanxcpt(union ieee754sp r) { - struct ieee754xctx ax; - assert(ieee754sp_isnan(r)); if (!ieee754sp_issnan(r)) /* QNAN does not cause invalid op !! */ return r; - if (!SETANDTESTCX(IEEE754_INVALID_OPERATION)) { + if (!ieee754_setandtestcx(IEEE754_INVALID_OPERATION)) { /* not enabled convert to a quiet NaN */ - SPMANT(r) &= (~SP_MBIT(SP_MBITS-1)); + SPMANT(r) &= (~SP_MBIT(SP_FBITS-1)); if (ieee754sp_isnan(r)) return r; else return ieee754sp_indef(); } - ax.op = op; - ax.rt = 0; - ax.rv.sp = r; - va_start(ax.ap, op); - ieee754_xcpt(&ax); - va_end(ax.ap); - return ax.rv.sp; -} - -ieee754sp ieee754sp_bestnan(ieee754sp x, ieee754sp y) -{ - assert(ieee754sp_isnan(x)); - assert(ieee754sp_isnan(y)); - - if (SPMANT(x) > SPMANT(y)) - return x; - else - return y; + return r; } - -static unsigned get_rounding(int sn, unsigned xm) +static unsigned ieee754sp_get_rounding(int sn, unsigned xm) { /* inexact must round of 3 bits */ if (xm & (SP_MBIT(3) - 1)) { switch (ieee754_csr.rm) { - case IEEE754_RZ: + case FPU_CSR_RZ: break; - case IEEE754_RN: + case FPU_CSR_RN: xm += 0x3 + ((xm >> 3) & 1); /* xm += (xm&0x8)?0x4:0x3 */ break; - case IEEE754_RU: /* toward +Infinity */ + case FPU_CSR_RU: /* toward +Infinity */ if (!sn) /* ?? */ xm += 0x8; break; - case IEEE754_RD: /* toward -Infinity */ + case FPU_CSR_RD: /* toward -Infinity */ if (sn) /* ?? */ xm += 0x8; break; @@ -131,11 +92,11 @@ static unsigned get_rounding(int sn, unsigned xm) * xe is an unbiased exponent * xm is 3bit extended precision value. */ -ieee754sp ieee754sp_format(int sn, int xe, unsigned xm) +union ieee754sp ieee754sp_format(int sn, int xe, unsigned xm) { assert(xm); /* we don't gen exact zeros (probably should) */ - assert((xm >> (SP_MBITS + 1 + 3)) == 0); /* no execess */ + assert((xm >> (SP_FBITS + 1 + 3)) == 0); /* no execess */ assert(xm & (SP_HIDDEN_BIT << 3)); if (xe < SP_EMIN) { @@ -143,38 +104,37 @@ ieee754sp ieee754sp_format(int sn, int xe, unsigned xm) int es = SP_EMIN - xe; if (ieee754_csr.nod) { - SETCX(IEEE754_UNDERFLOW); - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_UNDERFLOW); + ieee754_setcx(IEEE754_INEXACT); switch(ieee754_csr.rm) { - case IEEE754_RN: - case IEEE754_RZ: + case FPU_CSR_RN: + case FPU_CSR_RZ: return ieee754sp_zero(sn); - case IEEE754_RU: /* toward +Infinity */ - if(sn == 0) + case FPU_CSR_RU: /* toward +Infinity */ + if (sn == 0) return ieee754sp_min(0); else return ieee754sp_zero(1); - case IEEE754_RD: /* toward -Infinity */ - if(sn == 0) + case FPU_CSR_RD: /* toward -Infinity */ + if (sn == 0) return ieee754sp_zero(0); else return ieee754sp_min(1); } } - if (xe == SP_EMIN - 1 - && get_rounding(sn, xm) >> (SP_MBITS + 1 + 3)) + if (xe == SP_EMIN - 1 && + ieee754sp_get_rounding(sn, xm) >> (SP_FBITS + 1 + 3)) { /* Not tiny after rounding */ - SETCX(IEEE754_INEXACT); - xm = get_rounding(sn, xm); + ieee754_setcx(IEEE754_INEXACT); + xm = ieee754sp_get_rounding(sn, xm); xm >>= 1; /* Clear grs bits */ xm &= ~(SP_MBIT(3) - 1); xe++; - } - else { + } else { /* sticky right shift es bits */ SPXSRSXn(es); @@ -183,17 +143,17 @@ ieee754sp ieee754sp_format(int sn, int xe, unsigned xm) } } if (xm & (SP_MBIT(3) - 1)) { - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_INEXACT); if ((xm & (SP_HIDDEN_BIT << 3)) == 0) { - SETCX(IEEE754_UNDERFLOW); + ieee754_setcx(IEEE754_UNDERFLOW); } /* inexact must round of 3 bits */ - xm = get_rounding(sn, xm); + xm = ieee754sp_get_rounding(sn, xm); /* adjust exponent for rounding add overflowing */ - if (xm >> (SP_MBITS + 1 + 3)) { + if (xm >> (SP_FBITS + 1 + 3)) { /* add causes mantissa overflow */ xm >>= 1; xe++; @@ -202,24 +162,24 @@ ieee754sp ieee754sp_format(int sn, int xe, unsigned xm) /* strip grs bits */ xm >>= 3; - assert((xm >> (SP_MBITS + 1)) == 0); /* no execess */ + assert((xm >> (SP_FBITS + 1)) == 0); /* no execess */ assert(xe >= SP_EMIN); if (xe > SP_EMAX) { - SETCX(IEEE754_OVERFLOW); - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_OVERFLOW); + ieee754_setcx(IEEE754_INEXACT); /* -O can be table indexed by (rm,sn) */ switch (ieee754_csr.rm) { - case IEEE754_RN: + case FPU_CSR_RN: return ieee754sp_inf(sn); - case IEEE754_RZ: + case FPU_CSR_RZ: return ieee754sp_max(sn); - case IEEE754_RU: /* toward +Infinity */ + case FPU_CSR_RU: /* toward +Infinity */ if (sn == 0) return ieee754sp_inf(0); else return ieee754sp_max(1); - case IEEE754_RD: /* toward -Infinity */ + case FPU_CSR_RD: /* toward -Infinity */ if (sn == 0) return ieee754sp_max(0); else @@ -232,10 +192,10 @@ ieee754sp ieee754sp_format(int sn, int xe, unsigned xm) /* we underflow (tiny/zero) */ assert(xe == SP_EMIN); if (ieee754_csr.mx & IEEE754_UNDERFLOW) - SETCX(IEEE754_UNDERFLOW); + ieee754_setcx(IEEE754_UNDERFLOW); return buildsp(sn, SP_EMIN - 1 + SP_EBIAS, xm); } else { - assert((xm >> (SP_MBITS + 1)) == 0); /* no execess */ + assert((xm >> (SP_FBITS + 1)) == 0); /* no execess */ assert(xm & SP_HIDDEN_BIT); return buildsp(sn, xe + SP_EBIAS, xm & ~SP_HIDDEN_BIT); diff --git a/arch/mips/math-emu/ieee754sp.h b/arch/mips/math-emu/ieee754sp.h index 754fd54649b..ad268e33231 100644 --- a/arch/mips/math-emu/ieee754sp.h +++ b/arch/mips/math-emu/ieee754sp.h @@ -6,8 +6,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -19,70 +17,71 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ +#include <linux/compiler.h> #include "ieee754int.h" #define assert(expr) ((void)0) +#define SP_EBIAS 127 +#define SP_EMIN (-126) +#define SP_EMAX 127 +#define SP_FBITS 23 +#define SP_MBITS 23 + +#define SP_MBIT(x) ((u32)1 << (x)) +#define SP_HIDDEN_BIT SP_MBIT(SP_FBITS) +#define SP_SIGN_BIT SP_MBIT(31) + +#define SPSIGN(sp) (sp.sign) +#define SPBEXP(sp) (sp.bexp) +#define SPMANT(sp) (sp.mant) + +static inline int ieee754sp_finite(union ieee754sp x) +{ + return SPBEXP(x) != SP_EMAX + 1 + SP_EBIAS; +} + /* 3bit extended single precision sticky right shift */ -#define SPXSRSXn(rs) \ - (xe += rs, \ - xm = (rs > (SP_MBITS+3))?1:((xm) >> (rs)) | ((xm) << (32-(rs)) != 0)) +#define SPXSRSXn(rs) \ + (xe += rs, \ + xm = (rs > (SP_FBITS+3))?1:((xm) >> (rs)) | ((xm) << (32-(rs)) != 0)) #define SPXSRSX1() \ - (xe++, (xm = (xm >> 1) | (xm & 1))) + (xe++, (xm = (xm >> 1) | (xm & 1))) -#define SPXSRSYn(rs) \ - (ye+=rs, \ - ym = (rs > (SP_MBITS+3))?1:((ym) >> (rs)) | ((ym) << (32-(rs)) != 0)) +#define SPXSRSYn(rs) \ + (ye+=rs, \ + ym = (rs > (SP_FBITS+3))?1:((ym) >> (rs)) | ((ym) << (32-(rs)) != 0)) #define SPXSRSY1() \ - (ye++, (ym = (ym >> 1) | (ym & 1))) + (ye++, (ym = (ym >> 1) | (ym & 1))) /* convert denormal to normalized with extended exponent */ #define SPDNORMx(m,e) \ - while( (m >> SP_MBITS) == 0) { m <<= 1; e--; } + while ((m >> SP_FBITS) == 0) { m <<= 1; e--; } #define SPDNORMX SPDNORMx(xm, xe) #define SPDNORMY SPDNORMx(ym, ye) -static inline ieee754sp buildsp(int s, int bx, unsigned m) +static inline union ieee754sp buildsp(int s, int bx, unsigned m) { - ieee754sp r; + union ieee754sp r; assert((s) == 0 || (s) == 1); assert((bx) >= SP_EMIN - 1 + SP_EBIAS && (bx) <= SP_EMAX + 1 + SP_EBIAS); - assert(((m) >> SP_MBITS) == 0); + assert(((m) >> SP_FBITS) == 0); - r.parts.sign = s; - r.parts.bexp = bx; - r.parts.mant = m; + r.sign = s; + r.bexp = bx; + r.mant = m; return r; } -extern int ieee754sp_isnan(ieee754sp); -extern int ieee754sp_issnan(ieee754sp); -extern int ieee754si_xcpt(int, const char *, ...); -extern s64 ieee754di_xcpt(s64, const char *, ...); -extern ieee754sp ieee754sp_xcpt(ieee754sp, const char *, ...); -extern ieee754sp ieee754sp_nanxcpt(ieee754sp, const char *, ...); -extern ieee754sp ieee754sp_bestnan(ieee754sp, ieee754sp); -extern ieee754sp ieee754sp_format(int, int, unsigned); - - -#define SPNORMRET2(s, e, m, name, a0, a1) \ -{ \ - ieee754sp V = ieee754sp_format(s, e, m); \ - if(TSTX()) \ - return ieee754sp_xcpt(V, name, a0, a1); \ - else \ - return V; \ -} - -#define SPNORMRET1(s, e, m, name, a0) SPNORMRET2(s, e, m, name, a0, a0) +extern int ieee754sp_isnan(union ieee754sp); +extern union ieee754sp __cold ieee754sp_nanxcpt(union ieee754sp); +extern union ieee754sp ieee754sp_format(int, int, unsigned); diff --git a/arch/mips/math-emu/ieee754xcpt.c b/arch/mips/math-emu/ieee754xcpt.c deleted file mode 100644 index 967167116ae..00000000000 --- a/arch/mips/math-emu/ieee754xcpt.c +++ /dev/null @@ -1,47 +0,0 @@ -/* - * MIPS floating point support - * Copyright (C) 1994-2000 Algorithmics Ltd. - * - * ######################################################################## - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - */ - -/************************************************************************** - * Nov 7, 2000 - * Added preprocessor hacks to map to Linux kernel diagnostics. - * - * Kevin D. Kissell, kevink@mips.com and Carsten Langgaard, carstenl@mips.com - * Copyright (C) 2000 MIPS Technologies, Inc. All rights reserved. - *************************************************************************/ - -#include <linux/kernel.h> -#include "ieee754.h" - -/* - * Very naff exception handler (you can plug in your own and - * override this). - */ - -static const char *const rtnames[] = { - "sp", "dp", "xp", "si", "di" -}; - -void ieee754_xcpt(struct ieee754xctx *xcp) -{ - printk(KERN_DEBUG "floating point exception in \"%s\", type=%s\n", - xcp->op, rtnames[xcp->rt]); -} diff --git a/arch/mips/math-emu/kernel_linkage.c b/arch/mips/math-emu/kernel_linkage.c deleted file mode 100644 index eb58a85b315..00000000000 --- a/arch/mips/math-emu/kernel_linkage.c +++ /dev/null @@ -1,45 +0,0 @@ -/* - * Kevin D. Kissell, kevink@mips and Carsten Langgaard, carstenl@mips.com - * Copyright (C) 2000 MIPS Technologies, Inc. All rights reserved. - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * Routines corresponding to Linux kernel FP context - * manipulation primitives for the Algorithmics MIPS - * FPU Emulator - */ -#include <linux/sched.h> -#include <asm/processor.h> -#include <asm/signal.h> -#include <asm/uaccess.h> - -#include <asm/fpu.h> -#include <asm/fpu_emulator.h> - -#define SIGNALLING_NAN 0x7ff800007ff80000LL - -void fpu_emulator_init_fpu(void) -{ - static int first = 1; - int i; - - if (first) { - first = 0; - printk("Algorithmics/MIPS FPU Emulator v1.5\n"); - } - - current->thread.fpu.fcr31 = 0; - for (i = 0; i < 32; i++) - set_fpr64(¤t->thread.fpu.fpr[i], 0, SIGNALLING_NAN); -} diff --git a/arch/mips/math-emu/me-debugfs.c b/arch/mips/math-emu/me-debugfs.c new file mode 100644 index 00000000000..becdd63e14a --- /dev/null +++ b/arch/mips/math-emu/me-debugfs.c @@ -0,0 +1,67 @@ +#include <linux/cpumask.h> +#include <linux/debugfs.h> +#include <linux/fs.h> +#include <linux/init.h> +#include <linux/percpu.h> +#include <linux/types.h> +#include <asm/fpu_emulator.h> +#include <asm/local.h> + +DEFINE_PER_CPU(struct mips_fpu_emulator_stats, fpuemustats); + +static int fpuemu_stat_get(void *data, u64 *val) +{ + int cpu; + unsigned long sum = 0; + + for_each_online_cpu(cpu) { + struct mips_fpu_emulator_stats *ps; + local_t *pv; + + ps = &per_cpu(fpuemustats, cpu); + pv = (void *)ps + (unsigned long)data; + sum += local_read(pv); + } + *val = sum; + return 0; +} +DEFINE_SIMPLE_ATTRIBUTE(fops_fpuemu_stat, fpuemu_stat_get, NULL, "%llu\n"); + +extern struct dentry *mips_debugfs_dir; +static int __init debugfs_fpuemu(void) +{ + struct dentry *d, *dir; + + if (!mips_debugfs_dir) + return -ENODEV; + dir = debugfs_create_dir("fpuemustats", mips_debugfs_dir); + if (!dir) + return -ENOMEM; + +#define FPU_EMU_STAT_OFFSET(m) \ + offsetof(struct mips_fpu_emulator_stats, m) + +#define FPU_STAT_CREATE(m) \ +do { \ + d = debugfs_create_file(#m , S_IRUGO, dir, \ + (void *)FPU_EMU_STAT_OFFSET(m), \ + &fops_fpuemu_stat); \ + if (!d) \ + return -ENOMEM; \ +} while (0) + + FPU_STAT_CREATE(emulated); + FPU_STAT_CREATE(loads); + FPU_STAT_CREATE(stores); + FPU_STAT_CREATE(cp1ops); + FPU_STAT_CREATE(cp1xops); + FPU_STAT_CREATE(errors); + FPU_STAT_CREATE(ieee754_inexact); + FPU_STAT_CREATE(ieee754_underflow); + FPU_STAT_CREATE(ieee754_overflow); + FPU_STAT_CREATE(ieee754_zerodiv); + FPU_STAT_CREATE(ieee754_invalidop); + + return 0; +} +__initcall(debugfs_fpuemu); diff --git a/arch/mips/math-emu/sp_add.c b/arch/mips/math-emu/sp_add.c index c446e64637e..2d84d460cb6 100644 --- a/arch/mips/math-emu/sp_add.c +++ b/arch/mips/math-emu/sp_add.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,23 +16,22 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" -ieee754sp ieee754sp_add(ieee754sp x, ieee754sp y) +union ieee754sp ieee754sp_add(union ieee754sp x, union ieee754sp y) { + int s; + COMPXSP; COMPYSP; EXPLODEXSP; EXPLODEYSP; - CLEARCX; + ieee754_clearcx(); FLUSHXSP; FLUSHYSP; @@ -51,8 +48,8 @@ ieee754sp ieee754sp_add(ieee754sp x, ieee754sp y) case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_nanxcpt(ieee754sp_indef(), "add", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_nanxcpt(ieee754sp_indef()); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN): case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN): @@ -68,14 +65,14 @@ ieee754sp ieee754sp_add(ieee754sp x, ieee754sp y) return x; - /* Infinity handling - */ - + /* + * Infinity handling + */ case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF): if (xs == ys) return x; - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_xcpt(ieee754sp_indef(), "add", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_indef(); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF): case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF): @@ -87,15 +84,14 @@ ieee754sp ieee754sp_add(ieee754sp x, ieee754sp y) case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM): return x; - /* Zero handling - */ - + /* + * Zero handling + */ case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO): if (xs == ys) return x; else - return ieee754sp_zero(ieee754_csr.rm == - IEEE754_RD); + return ieee754sp_zero(ieee754_csr.rm == FPU_CSR_RD); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO): case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO): @@ -108,6 +104,8 @@ ieee754sp ieee754sp_add(ieee754sp x, ieee754sp y) case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM): SPDNORMX; + /* FALL THROUGH */ + case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_DNORM): SPDNORMY; break; @@ -122,33 +120,38 @@ ieee754sp ieee754sp_add(ieee754sp x, ieee754sp y) assert(xm & SP_HIDDEN_BIT); assert(ym & SP_HIDDEN_BIT); - /* provide guard,round and stick bit space */ + /* + * Provide guard, round and stick bit space. + */ xm <<= 3; ym <<= 3; if (xe > ye) { - /* have to shift y fraction right to align + /* + * Have to shift y fraction right to align. */ - int s = xe - ye; + s = xe - ye; SPXSRSYn(s); } else if (ye > xe) { - /* have to shift x fraction right to align + /* + * Have to shift x fraction right to align. */ - int s = ye - xe; + s = ye - xe; SPXSRSXn(s); } assert(xe == ye); assert(xe <= SP_EMAX); if (xs == ys) { - /* generate 28 bit result of adding two 27 bit numbers - * leaving result in xm,xs,xe + /* + * Generate 28 bit result of adding two 27 bit numbers + * leaving result in xm, xs and xe. */ xm = xm + ym; xe = xe; xs = xs; - if (xm >> (SP_MBITS + 1 + 3)) { /* carry out */ + if (xm >> (SP_FBITS + 1 + 3)) { /* carry out */ SPXSRSX1(); } } else { @@ -162,15 +165,16 @@ ieee754sp ieee754sp_add(ieee754sp x, ieee754sp y) xs = ys; } if (xm == 0) - return ieee754sp_zero(ieee754_csr.rm == - IEEE754_RD); + return ieee754sp_zero(ieee754_csr.rm == FPU_CSR_RD); - /* normalize in extended single precision */ - while ((xm >> (SP_MBITS + 3)) == 0) { + /* + * Normalize in extended single precision + */ + while ((xm >> (SP_FBITS + 3)) == 0) { xm <<= 1; xe--; } - } - SPNORMRET2(xs, xe, xm, "add", x, y); + + return ieee754sp_format(xs, xe, xm); } diff --git a/arch/mips/math-emu/sp_cmp.c b/arch/mips/math-emu/sp_cmp.c index 716cf37e246..addbccb2f55 100644 --- a/arch/mips/math-emu/sp_cmp.c +++ b/arch/mips/math-emu/sp_cmp.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,16 +16,16 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" -int ieee754sp_cmp(ieee754sp x, ieee754sp y, int cmp, int sig) +int ieee754sp_cmp(union ieee754sp x, union ieee754sp y, int cmp, int sig) { + int vx; + int vy; + COMPXSP; COMPYSP; @@ -35,21 +33,21 @@ int ieee754sp_cmp(ieee754sp x, ieee754sp y, int cmp, int sig) EXPLODEYSP; FLUSHXSP; FLUSHYSP; - CLEARCX; /* Even clear inexact flag here */ + ieee754_clearcx(); /* Even clear inexact flag here */ if (ieee754sp_isnan(x) || ieee754sp_isnan(y)) { if (sig || xc == IEEE754_CLASS_SNAN || yc == IEEE754_CLASS_SNAN) - SETCX(IEEE754_INVALID_OPERATION); + ieee754_setcx(IEEE754_INVALID_OPERATION); if (cmp & IEEE754_CUN) return 1; if (cmp & (IEEE754_CLT | IEEE754_CGT)) { - if (sig && SETANDTESTCX(IEEE754_INVALID_OPERATION)) - return ieee754si_xcpt(0, "fcmpf", x); + if (sig && ieee754_setandtestcx(IEEE754_INVALID_OPERATION)) + return 0; } return 0; } else { - int vx = x.bits; - int vy = y.bits; + vx = x.bits; + vy = y.bits; if (vx < 0) vx = -vx ^ SP_SIGN_BIT; diff --git a/arch/mips/math-emu/sp_div.c b/arch/mips/math-emu/sp_div.c index d7747928c95..721f317aa87 100644 --- a/arch/mips/math-emu/sp_div.c +++ b/arch/mips/math-emu/sp_div.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,23 +16,24 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" -ieee754sp ieee754sp_div(ieee754sp x, ieee754sp y) +union ieee754sp ieee754sp_div(union ieee754sp x, union ieee754sp y) { + unsigned rm; + int re; + unsigned bm; + COMPXSP; COMPYSP; EXPLODEXSP; EXPLODEYSP; - CLEARCX; + ieee754_clearcx(); FLUSHXSP; FLUSHYSP; @@ -51,8 +50,8 @@ ieee754sp ieee754sp_div(ieee754sp x, ieee754sp y) case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_nanxcpt(ieee754sp_indef(), "div", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_nanxcpt(ieee754sp_indef()); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN): case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN): @@ -68,12 +67,12 @@ ieee754sp ieee754sp_div(ieee754sp x, ieee754sp y) return x; - /* Infinity handling - */ - + /* + * Infinity handling + */ case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_xcpt(ieee754sp_indef(), "div", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_indef(); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF): case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF): @@ -85,17 +84,17 @@ ieee754sp ieee754sp_div(ieee754sp x, ieee754sp y) case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM): return ieee754sp_inf(xs ^ ys); - /* Zero handling - */ - + /* + * Zero handling + */ case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_xcpt(ieee754sp_indef(), "div", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_indef(); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO): case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO): - SETCX(IEEE754_ZERO_DIVIDE); - return ieee754sp_xcpt(ieee754sp_inf(xs ^ ys), "div", x, y); + ieee754_setcx(IEEE754_ZERO_DIVIDE); + return ieee754sp_inf(xs ^ ys); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM): @@ -122,35 +121,33 @@ ieee754sp ieee754sp_div(ieee754sp x, ieee754sp y) xm <<= 3; ym <<= 3; - { - /* now the dirty work */ - - unsigned rm = 0; - int re = xe - ye; - unsigned bm; - - for (bm = SP_MBIT(SP_MBITS + 2); bm; bm >>= 1) { - if (xm >= ym) { - xm -= ym; - rm |= bm; - if (xm == 0) - break; - } - xm <<= 1; - } - rm <<= 1; - if (xm) - rm |= 1; /* have remainder, set sticky */ + /* now the dirty work */ - assert(rm); + rm = 0; + re = xe - ye; - /* normalise rm to rounding precision ? - */ - while ((rm >> (SP_MBITS + 3)) == 0) { - rm <<= 1; - re--; + for (bm = SP_MBIT(SP_FBITS + 2); bm; bm >>= 1) { + if (xm >= ym) { + xm -= ym; + rm |= bm; + if (xm == 0) + break; } + xm <<= 1; + } + + rm <<= 1; + if (xm) + rm |= 1; /* have remainder, set sticky */ - SPNORMRET2(xs == ys ? 0 : 1, re, rm, "div", x, y); + assert(rm); + + /* normalise rm to rounding precision ? + */ + while ((rm >> (SP_FBITS + 3)) == 0) { + rm <<= 1; + re--; } + + return ieee754sp_format(xs == ys ? 0 : 1, re, rm); } diff --git a/arch/mips/math-emu/sp_fdp.c b/arch/mips/math-emu/sp_fdp.c index e1515aae016..1b266fb1697 100644 --- a/arch/mips/math-emu/sp_fdp.c +++ b/arch/mips/math-emu/sp_fdp.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,59 +16,61 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" +#include "ieee754dp.h" -ieee754sp ieee754sp_fdp(ieee754dp x) +union ieee754sp ieee754sp_fdp(union ieee754dp x) { + u32 rm; + COMPXDP; - ieee754sp nan; + union ieee754sp nan; EXPLODEXDP; - CLEARCX; + ieee754_clearcx(); FLUSHXDP; switch (xc) { case IEEE754_CLASS_SNAN: - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_nanxcpt(ieee754sp_indef(), "fdp"); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_nanxcpt(ieee754sp_indef()); + case IEEE754_CLASS_QNAN: nan = buildsp(xs, SP_EMAX + 1 + SP_EBIAS, (u32) - (xm >> (DP_MBITS - SP_MBITS))); + (xm >> (DP_FBITS - SP_FBITS))); if (!ieee754sp_isnan(nan)) nan = ieee754sp_indef(); - return ieee754sp_nanxcpt(nan, "fdp", x); + return ieee754sp_nanxcpt(nan); + case IEEE754_CLASS_INF: return ieee754sp_inf(xs); + case IEEE754_CLASS_ZERO: return ieee754sp_zero(xs); + case IEEE754_CLASS_DNORM: /* can't possibly be sp representable */ - SETCX(IEEE754_UNDERFLOW); - SETCX(IEEE754_INEXACT); - if ((ieee754_csr.rm == IEEE754_RU && !xs) || - (ieee754_csr.rm == IEEE754_RD && xs)) - return ieee754sp_xcpt(ieee754sp_mind(xs), "fdp", x); - return ieee754sp_xcpt(ieee754sp_zero(xs), "fdp", x); + ieee754_setcx(IEEE754_UNDERFLOW); + ieee754_setcx(IEEE754_INEXACT); + if ((ieee754_csr.rm == FPU_CSR_RU && !xs) || + (ieee754_csr.rm == FPU_CSR_RD && xs)) + return ieee754sp_mind(xs); + return ieee754sp_zero(xs); + case IEEE754_CLASS_NORM: break; } - { - u32 rm; - - /* convert from DP_MBITS to SP_MBITS+3 with sticky right shift - */ - rm = (xm >> (DP_MBITS - (SP_MBITS + 3))) | - ((xm << (64 - (DP_MBITS - (SP_MBITS + 3)))) != 0); + /* + * Convert from DP_FBITS to SP_FBITS+3 with sticky right shift. + */ + rm = (xm >> (DP_FBITS - (SP_FBITS + 3))) | + ((xm << (64 - (DP_FBITS - (SP_FBITS + 3)))) != 0); - SPNORMRET1(xs, xe, rm, "fdp", x); - } + return ieee754sp_format(xs, xe, rm); } diff --git a/arch/mips/math-emu/sp_fint.c b/arch/mips/math-emu/sp_fint.c index 9694d6c016c..d5d8495b2cc 100644 --- a/arch/mips/math-emu/sp_fint.c +++ b/arch/mips/math-emu/sp_fint.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,21 +16,18 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" -ieee754sp ieee754sp_fint(int x) +union ieee754sp ieee754sp_fint(int x) { unsigned xm; int xe; int xs; - CLEARCX; + ieee754_clearcx(); if (x == 0) return ieee754sp_zero(0); @@ -50,30 +45,21 @@ ieee754sp ieee754sp_fint(int x) } else { xm = x; } - xe = SP_MBITS + 3; + xe = SP_FBITS + 3; - if (xm >> (SP_MBITS + 1 + 3)) { + if (xm >> (SP_FBITS + 1 + 3)) { /* shunt out overflow bits */ - while (xm >> (SP_MBITS + 1 + 3)) { + while (xm >> (SP_FBITS + 1 + 3)) { SPXSRSX1(); } } else { /* normalize in grs extended single precision */ - while ((xm >> (SP_MBITS + 3)) == 0) { + while ((xm >> (SP_FBITS + 3)) == 0) { xm <<= 1; xe--; } } - SPNORMRET1(xs, xe, xm, "fint", x); -} - - -ieee754sp ieee754sp_funs(unsigned int u) -{ - if ((int) u < 0) - return ieee754sp_add(ieee754sp_1e31(), - ieee754sp_fint(u & ~(1 << 31))); - return ieee754sp_fint(u); + return ieee754sp_format(xs, xe, xm); } diff --git a/arch/mips/math-emu/sp_flong.c b/arch/mips/math-emu/sp_flong.c index 16a651f2986..012e30ce758 100644 --- a/arch/mips/math-emu/sp_flong.c +++ b/arch/mips/math-emu/sp_flong.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,21 +16,18 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" -ieee754sp ieee754sp_flong(s64 x) +union ieee754sp ieee754sp_flong(s64 x) { u64 xm; /* <--- need 64-bit mantissa temp */ int xe; int xs; - CLEARCX; + ieee754_clearcx(); if (x == 0) return ieee754sp_zero(0); @@ -50,29 +45,20 @@ ieee754sp ieee754sp_flong(s64 x) } else { xm = x; } - xe = SP_MBITS + 3; + xe = SP_FBITS + 3; - if (xm >> (SP_MBITS + 1 + 3)) { + if (xm >> (SP_FBITS + 1 + 3)) { /* shunt out overflow bits */ - while (xm >> (SP_MBITS + 1 + 3)) { + while (xm >> (SP_FBITS + 1 + 3)) { SPXSRSX1(); } } else { /* normalize in grs extended single precision */ - while ((xm >> (SP_MBITS + 3)) == 0) { + while ((xm >> (SP_FBITS + 3)) == 0) { xm <<= 1; xe--; } } - SPNORMRET1(xs, xe, xm, "sp_flong", x); -} - - -ieee754sp ieee754sp_fulong(u64 u) -{ - if ((s64) u < 0) - return ieee754sp_add(ieee754sp_1e63(), - ieee754sp_flong(u & ~(1ULL << 63))); - return ieee754sp_flong(u); + return ieee754sp_format(xs, xe, xm); } diff --git a/arch/mips/math-emu/sp_frexp.c b/arch/mips/math-emu/sp_frexp.c deleted file mode 100644 index 5bc993c3004..00000000000 --- a/arch/mips/math-emu/sp_frexp.c +++ /dev/null @@ -1,52 +0,0 @@ -/* IEEE754 floating point arithmetic - * single precision - */ -/* - * MIPS floating point support - * Copyright (C) 1994-2000 Algorithmics Ltd. - * - * ######################################################################## - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - */ - - -#include "ieee754sp.h" - -/* close to ieeep754sp_logb -*/ -ieee754sp ieee754sp_frexp(ieee754sp x, int *eptr) -{ - COMPXSP; - CLEARCX; - EXPLODEXSP; - - switch (xc) { - case IEEE754_CLASS_SNAN: - case IEEE754_CLASS_QNAN: - case IEEE754_CLASS_INF: - case IEEE754_CLASS_ZERO: - *eptr = 0; - return x; - case IEEE754_CLASS_DNORM: - SPDNORMX; - break; - case IEEE754_CLASS_NORM: - break; - } - *eptr = xe + 1; - return buildsp(xs, -1 + SP_EBIAS, xm & ~SP_HIDDEN_BIT); -} diff --git a/arch/mips/math-emu/sp_logb.c b/arch/mips/math-emu/sp_logb.c deleted file mode 100644 index 9c14e0c75bd..00000000000 --- a/arch/mips/math-emu/sp_logb.c +++ /dev/null @@ -1,53 +0,0 @@ -/* IEEE754 floating point arithmetic - * single precision - */ -/* - * MIPS floating point support - * Copyright (C) 1994-2000 Algorithmics Ltd. - * - * ######################################################################## - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - */ - - -#include "ieee754sp.h" - -ieee754sp ieee754sp_logb(ieee754sp x) -{ - COMPXSP; - - CLEARCX; - - EXPLODEXSP; - - switch (xc) { - case IEEE754_CLASS_SNAN: - return ieee754sp_nanxcpt(x, "logb", x); - case IEEE754_CLASS_QNAN: - return x; - case IEEE754_CLASS_INF: - return ieee754sp_inf(0); - case IEEE754_CLASS_ZERO: - return ieee754sp_inf(1); - case IEEE754_CLASS_DNORM: - SPDNORMX; - break; - case IEEE754_CLASS_NORM: - break; - } - return ieee754sp_fint(xe); -} diff --git a/arch/mips/math-emu/sp_modf.c b/arch/mips/math-emu/sp_modf.c deleted file mode 100644 index 25a0fbaa055..00000000000 --- a/arch/mips/math-emu/sp_modf.c +++ /dev/null @@ -1,79 +0,0 @@ -/* IEEE754 floating point arithmetic - * single precision - */ -/* - * MIPS floating point support - * Copyright (C) 1994-2000 Algorithmics Ltd. - * - * ######################################################################## - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - */ - - -#include "ieee754sp.h" - -/* modf function is always exact for a finite number -*/ -ieee754sp ieee754sp_modf(ieee754sp x, ieee754sp *ip) -{ - COMPXSP; - - CLEARCX; - - EXPLODEXSP; - - switch (xc) { - case IEEE754_CLASS_SNAN: - case IEEE754_CLASS_QNAN: - case IEEE754_CLASS_INF: - case IEEE754_CLASS_ZERO: - *ip = x; - return x; - case IEEE754_CLASS_DNORM: - /* far to small */ - *ip = ieee754sp_zero(xs); - return x; - case IEEE754_CLASS_NORM: - break; - } - if (xe < 0) { - *ip = ieee754sp_zero(xs); - return x; - } - if (xe >= SP_MBITS) { - *ip = x; - return ieee754sp_zero(xs); - } - /* generate ipart mantissa by clearing bottom bits - */ - *ip = buildsp(xs, xe + SP_EBIAS, - ((xm >> (SP_MBITS - xe)) << (SP_MBITS - xe)) & - ~SP_HIDDEN_BIT); - - /* generate fpart mantissa by clearing top bits - * and normalizing (must be able to normalize) - */ - xm = (xm << (32 - (SP_MBITS - xe))) >> (32 - (SP_MBITS - xe)); - if (xm == 0) - return ieee754sp_zero(xs); - - while ((xm >> SP_MBITS) == 0) { - xm <<= 1; - xe--; - } - return buildsp(xs, xe + SP_EBIAS, xm & ~SP_HIDDEN_BIT); -} diff --git a/arch/mips/math-emu/sp_mul.c b/arch/mips/math-emu/sp_mul.c index fa4675cf2aa..890c13a2965 100644 --- a/arch/mips/math-emu/sp_mul.c +++ b/arch/mips/math-emu/sp_mul.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,23 +16,32 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" -ieee754sp ieee754sp_mul(ieee754sp x, ieee754sp y) +union ieee754sp ieee754sp_mul(union ieee754sp x, union ieee754sp y) { + int re; + int rs; + unsigned rm; + unsigned short lxm; + unsigned short hxm; + unsigned short lym; + unsigned short hym; + unsigned lrm; + unsigned hrm; + unsigned t; + unsigned at; + COMPXSP; COMPYSP; EXPLODEXSP; EXPLODEYSP; - CLEARCX; + ieee754_clearcx(); FLUSHXSP; FLUSHYSP; @@ -51,8 +58,8 @@ ieee754sp ieee754sp_mul(ieee754sp x, ieee754sp y) case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_nanxcpt(ieee754sp_indef(), "mul", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_nanxcpt(ieee754sp_indef()); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN): case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN): @@ -68,12 +75,13 @@ ieee754sp ieee754sp_mul(ieee754sp x, ieee754sp y) return x; - /* Infinity handling */ - + /* + * Infinity handling + */ case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_ZERO): case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_xcpt(ieee754sp_indef(), "mul", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_indef(); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_INF): case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF): @@ -108,63 +116,50 @@ ieee754sp ieee754sp_mul(ieee754sp x, ieee754sp y) assert(xm & SP_HIDDEN_BIT); assert(ym & SP_HIDDEN_BIT); - { - int re = xe + ye; - int rs = xs ^ ys; - unsigned rm; - - /* shunt to top of word */ - xm <<= 32 - (SP_MBITS + 1); - ym <<= 32 - (SP_MBITS + 1); - - /* multiply 32bits xm,ym to give high 32bits rm with stickness - */ - { - unsigned short lxm = xm & 0xffff; - unsigned short hxm = xm >> 16; - unsigned short lym = ym & 0xffff; - unsigned short hym = ym >> 16; - unsigned lrm; - unsigned hrm; - - lrm = lxm * lym; /* 16 * 16 => 32 */ - hrm = hxm * hym; /* 16 * 16 => 32 */ - - { - unsigned t = lxm * hym; /* 16 * 16 => 32 */ - { - unsigned at = lrm + (t << 16); - hrm += at < lrm; - lrm = at; - } - hrm = hrm + (t >> 16); - } - - { - unsigned t = hxm * lym; /* 16 * 16 => 32 */ - { - unsigned at = lrm + (t << 16); - hrm += at < lrm; - lrm = at; - } - hrm = hrm + (t >> 16); - } - rm = hrm | (lrm != 0); - } - - /* - * sticky shift down to normal rounding precision - */ - if ((int) rm < 0) { - rm = (rm >> (32 - (SP_MBITS + 1 + 3))) | - ((rm << (SP_MBITS + 1 + 3)) != 0); - re++; - } else { - rm = (rm >> (32 - (SP_MBITS + 1 + 3 + 1))) | - ((rm << (SP_MBITS + 1 + 3 + 1)) != 0); - } - assert(rm & (SP_HIDDEN_BIT << 3)); - - SPNORMRET2(rs, re, rm, "mul", x, y); + re = xe + ye; + rs = xs ^ ys; + + /* shunt to top of word */ + xm <<= 32 - (SP_FBITS + 1); + ym <<= 32 - (SP_FBITS + 1); + + /* + * Multiply 32 bits xm, ym to give high 32 bits rm with stickness. + */ + lxm = xm & 0xffff; + hxm = xm >> 16; + lym = ym & 0xffff; + hym = ym >> 16; + + lrm = lxm * lym; /* 16 * 16 => 32 */ + hrm = hxm * hym; /* 16 * 16 => 32 */ + + t = lxm * hym; /* 16 * 16 => 32 */ + at = lrm + (t << 16); + hrm += at < lrm; + lrm = at; + hrm = hrm + (t >> 16); + + t = hxm * lym; /* 16 * 16 => 32 */ + at = lrm + (t << 16); + hrm += at < lrm; + lrm = at; + hrm = hrm + (t >> 16); + + rm = hrm | (lrm != 0); + + /* + * Sticky shift down to normal rounding precision. + */ + if ((int) rm < 0) { + rm = (rm >> (32 - (SP_FBITS + 1 + 3))) | + ((rm << (SP_FBITS + 1 + 3)) != 0); + re++; + } else { + rm = (rm >> (32 - (SP_FBITS + 1 + 3 + 1))) | + ((rm << (SP_FBITS + 1 + 3 + 1)) != 0); } + assert(rm & (SP_HIDDEN_BIT << 3)); + + return ieee754sp_format(rs, re, rm); } diff --git a/arch/mips/math-emu/sp_scalb.c b/arch/mips/math-emu/sp_scalb.c deleted file mode 100644 index dd76196984c..00000000000 --- a/arch/mips/math-emu/sp_scalb.c +++ /dev/null @@ -1,57 +0,0 @@ -/* IEEE754 floating point arithmetic - * single precision - */ -/* - * MIPS floating point support - * Copyright (C) 1994-2000 Algorithmics Ltd. - * - * ######################################################################## - * - * This program is free software; you can distribute it and/or modify it - * under the terms of the GNU General Public License (Version 2) as - * published by the Free Software Foundation. - * - * This program is distributed in the hope it will be useful, but WITHOUT - * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or - * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License - * for more details. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## - */ - - -#include "ieee754sp.h" - -ieee754sp ieee754sp_scalb(ieee754sp x, int n) -{ - COMPXSP; - - CLEARCX; - - EXPLODEXSP; - - switch (xc) { - case IEEE754_CLASS_SNAN: - return ieee754sp_nanxcpt(x, "scalb", x, n); - case IEEE754_CLASS_QNAN: - case IEEE754_CLASS_INF: - case IEEE754_CLASS_ZERO: - return x; - case IEEE754_CLASS_DNORM: - SPDNORMX; - break; - case IEEE754_CLASS_NORM: - break; - } - SPNORMRET2(xs, xe + n, xm << 3, "scalb", x, n); -} - - -ieee754sp ieee754sp_ldexp(ieee754sp x, int n) -{ - return ieee754sp_scalb(x, n); -} diff --git a/arch/mips/math-emu/sp_simple.c b/arch/mips/math-emu/sp_simple.c index ae4fcfafd85..f1ffaa9a17e 100644 --- a/arch/mips/math-emu/sp_simple.c +++ b/arch/mips/math-emu/sp_simple.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,33 +16,17 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" -int ieee754sp_finite(ieee754sp x) -{ - return SPBEXP(x) != SP_EMAX + 1 + SP_EBIAS; -} - -ieee754sp ieee754sp_copysign(ieee754sp x, ieee754sp y) -{ - CLEARCX; - SPSIGN(x) = SPSIGN(y); - return x; -} - - -ieee754sp ieee754sp_neg(ieee754sp x) +union ieee754sp ieee754sp_neg(union ieee754sp x) { COMPXSP; EXPLODEXSP; - CLEARCX; + ieee754_clearcx(); FLUSHXSP; /* @@ -55,30 +37,29 @@ ieee754sp ieee754sp_neg(ieee754sp x) SPSIGN(x) ^= 1; if (xc == IEEE754_CLASS_SNAN) { - ieee754sp y = ieee754sp_indef(); - SETCX(IEEE754_INVALID_OPERATION); + union ieee754sp y = ieee754sp_indef(); + ieee754_setcx(IEEE754_INVALID_OPERATION); SPSIGN(y) = SPSIGN(x); - return ieee754sp_nanxcpt(y, "neg"); + return ieee754sp_nanxcpt(y); } return x; } - -ieee754sp ieee754sp_abs(ieee754sp x) +union ieee754sp ieee754sp_abs(union ieee754sp x) { COMPXSP; EXPLODEXSP; - CLEARCX; + ieee754_clearcx(); FLUSHXSP; /* Clear sign ALWAYS, irrespective of NaN */ SPSIGN(x) = 0; if (xc == IEEE754_CLASS_SNAN) { - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_nanxcpt(ieee754sp_indef(), "abs"); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_nanxcpt(ieee754sp_indef()); } return x; diff --git a/arch/mips/math-emu/sp_sqrt.c b/arch/mips/math-emu/sp_sqrt.c index fed20175f5f..b7c098a86f9 100644 --- a/arch/mips/math-emu/sp_sqrt.c +++ b/arch/mips/math-emu/sp_sqrt.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,15 +16,12 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" -ieee754sp ieee754sp_sqrt(ieee754sp x) +union ieee754sp ieee754sp_sqrt(union ieee754sp x) { int ix, s, q, m, t, i; unsigned int r; @@ -35,34 +30,38 @@ ieee754sp ieee754sp_sqrt(ieee754sp x) /* take care of Inf and NaN */ EXPLODEXSP; - CLEARCX; + ieee754_clearcx(); FLUSHXSP; /* x == INF or NAN? */ switch (xc) { case IEEE754_CLASS_QNAN: /* sqrt(Nan) = Nan */ - return ieee754sp_nanxcpt(x, "sqrt"); + return ieee754sp_nanxcpt(x); + case IEEE754_CLASS_SNAN: - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_nanxcpt(ieee754sp_indef(), "sqrt"); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_nanxcpt(ieee754sp_indef()); + case IEEE754_CLASS_ZERO: /* sqrt(0) = 0 */ return x; + case IEEE754_CLASS_INF: if (xs) { /* sqrt(-Inf) = Nan */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_nanxcpt(ieee754sp_indef(), "sqrt"); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_nanxcpt(ieee754sp_indef()); } /* sqrt(+Inf) = Inf */ return x; + case IEEE754_CLASS_DNORM: case IEEE754_CLASS_NORM: if (xs) { /* sqrt(-x) = Nan */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_nanxcpt(ieee754sp_indef(), "sqrt"); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_nanxcpt(ieee754sp_indef()); } break; } @@ -99,12 +98,12 @@ ieee754sp ieee754sp_sqrt(ieee754sp x) } if (ix != 0) { - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_INEXACT); switch (ieee754_csr.rm) { - case IEEE754_RP: + case FPU_CSR_RU: q += 2; break; - case IEEE754_RN: + case FPU_CSR_RN: q += (q & 1); break; } diff --git a/arch/mips/math-emu/sp_sub.c b/arch/mips/math-emu/sp_sub.c index e595c6f3d0b..8592e49032b 100644 --- a/arch/mips/math-emu/sp_sub.c +++ b/arch/mips/math-emu/sp_sub.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,23 +16,22 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" -ieee754sp ieee754sp_sub(ieee754sp x, ieee754sp y) +union ieee754sp ieee754sp_sub(union ieee754sp x, union ieee754sp y) { + int s; + COMPXSP; COMPYSP; EXPLODEXSP; EXPLODEYSP; - CLEARCX; + ieee754_clearcx(); FLUSHXSP; FLUSHYSP; @@ -51,8 +48,8 @@ ieee754sp ieee754sp_sub(ieee754sp x, ieee754sp y) case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_DNORM): case CLPAIR(IEEE754_CLASS_SNAN, IEEE754_CLASS_INF): - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_nanxcpt(ieee754sp_indef(), "sub", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_nanxcpt(ieee754sp_indef()); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_QNAN): case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_QNAN): @@ -68,14 +65,14 @@ ieee754sp ieee754sp_sub(ieee754sp x, ieee754sp y) return x; - /* Infinity handling - */ - + /* + * Infinity handling + */ case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_INF): if (xs != ys) return x; - SETCX(IEEE754_INVALID_OPERATION); - return ieee754sp_xcpt(ieee754sp_indef(), "sub", x, y); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754sp_indef(); case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_INF): case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_INF): @@ -87,15 +84,14 @@ ieee754sp ieee754sp_sub(ieee754sp x, ieee754sp y) case CLPAIR(IEEE754_CLASS_INF, IEEE754_CLASS_DNORM): return x; - /* Zero handling - */ - + /* + * Zero handling + */ case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_ZERO): if (xs != ys) return x; else - return ieee754sp_zero(ieee754_csr.rm == - IEEE754_RD); + return ieee754sp_zero(ieee754_csr.rm == FPU_CSR_RD); case CLPAIR(IEEE754_CLASS_NORM, IEEE754_CLASS_ZERO): case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_ZERO): @@ -104,7 +100,7 @@ ieee754sp ieee754sp_sub(ieee754sp x, ieee754sp y) case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_NORM): case CLPAIR(IEEE754_CLASS_ZERO, IEEE754_CLASS_DNORM): /* quick fix up */ - DPSIGN(y) ^= 1; + SPSIGN(y) ^= 1; return y; case CLPAIR(IEEE754_CLASS_DNORM, IEEE754_CLASS_DNORM): @@ -133,14 +129,16 @@ ieee754sp ieee754sp_sub(ieee754sp x, ieee754sp y) ym <<= 3; if (xe > ye) { - /* have to shift y fraction right to align + /* + * have to shift y fraction right to align */ - int s = xe - ye; + s = xe - ye; SPXSRSYn(s); } else if (ye > xe) { - /* have to shift x fraction right to align + /* + * have to shift x fraction right to align */ - int s = ye - xe; + s = ye - xe; SPXSRSXn(s); } assert(xe == ye); @@ -153,7 +151,7 @@ ieee754sp ieee754sp_sub(ieee754sp x, ieee754sp y) xe = xe; xs = xs; - if (xm >> (SP_MBITS + 1 + 3)) { /* carry out */ + if (xm >> (SP_FBITS + 1 + 3)) { /* carry out */ SPXSRSX1(); /* shift preserving sticky */ } } else { @@ -167,17 +165,18 @@ ieee754sp ieee754sp_sub(ieee754sp x, ieee754sp y) xs = ys; } if (xm == 0) { - if (ieee754_csr.rm == IEEE754_RD) + if (ieee754_csr.rm == FPU_CSR_RD) return ieee754sp_zero(1); /* round negative inf. => sign = -1 */ else return ieee754sp_zero(0); /* other round modes => sign = 1 */ } /* normalize to rounding precision */ - while ((xm >> (SP_MBITS + 3)) == 0) { + while ((xm >> (SP_FBITS + 3)) == 0) { xm <<= 1; xe--; } } - SPNORMRET2(xs, xe, xm, "sub", x, y); + + return ieee754sp_format(xs, xe, xm); } diff --git a/arch/mips/math-emu/sp_tint.c b/arch/mips/math-emu/sp_tint.c index 0fe9acc7716..091299a3179 100644 --- a/arch/mips/math-emu/sp_tint.c +++ b/arch/mips/math-emu/sp_tint.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,20 +16,21 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - -#include <linux/kernel.h> #include "ieee754sp.h" -int ieee754sp_tint(ieee754sp x) +int ieee754sp_tint(union ieee754sp x) { + u32 residue; + int round; + int sticky; + int odd; + COMPXSP; - CLEARCX; + ieee754_clearcx(); EXPLODEXSP; FLUSHXSP; @@ -40,10 +39,12 @@ int ieee754sp_tint(ieee754sp x) case IEEE754_CLASS_SNAN: case IEEE754_CLASS_QNAN: case IEEE754_CLASS_INF: - SETCX(IEEE754_INVALID_OPERATION); - return ieee754si_xcpt(ieee754si_indef(), "sp_tint", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754si_indef(); + case IEEE754_CLASS_ZERO: return 0; + case IEEE754_CLASS_DNORM: case IEEE754_CLASS_NORM: break; @@ -54,18 +55,13 @@ int ieee754sp_tint(ieee754sp x) return -0x80000000; /* Set invalid. We will only use overflow for floating point overflow */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754si_xcpt(ieee754si_indef(), "sp_tint", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754si_indef(); } /* oh gawd */ - if (xe > SP_MBITS) { - xm <<= xe - SP_MBITS; + if (xe > SP_FBITS) { + xm <<= xe - SP_FBITS; } else { - u32 residue; - int round; - int sticky; - int odd; - if (xe < -1) { residue = xm; round = 0; @@ -76,51 +72,38 @@ int ieee754sp_tint(ieee754sp x) * so we do it in two steps. Be aware that xe * may be -1 */ residue = xm << (xe + 1); - residue <<= 31 - SP_MBITS; + residue <<= 31 - SP_FBITS; round = (residue >> 31) != 0; sticky = (residue << 1) != 0; - xm >>= SP_MBITS - xe; + xm >>= SP_FBITS - xe; } odd = (xm & 0x1) != 0x0; switch (ieee754_csr.rm) { - case IEEE754_RN: + case FPU_CSR_RN: if (round && (sticky || odd)) xm++; break; - case IEEE754_RZ: + case FPU_CSR_RZ: break; - case IEEE754_RU: /* toward +Infinity */ + case FPU_CSR_RU: /* toward +Infinity */ if ((round || sticky) && !xs) xm++; break; - case IEEE754_RD: /* toward -Infinity */ + case FPU_CSR_RD: /* toward -Infinity */ if ((round || sticky) && xs) xm++; break; } if ((xm >> 31) != 0) { /* This can happen after rounding */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754si_xcpt(ieee754si_indef(), "sp_tint", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754si_indef(); } if (round || sticky) - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_INEXACT); } if (xs) return -xm; else return xm; } - - -unsigned int ieee754sp_tuns(ieee754sp x) -{ - ieee754sp hb = ieee754sp_1e31(); - - /* what if x < 0 ?? */ - if (ieee754sp_lt(x, hb)) - return (unsigned) ieee754sp_tint(x); - - return (unsigned) ieee754sp_tint(ieee754sp_sub(x, hb)) | - ((unsigned) 1 << 31); -} diff --git a/arch/mips/math-emu/sp_tlong.c b/arch/mips/math-emu/sp_tlong.c index d0ca6e22be2..9f3c742c1ce 100644 --- a/arch/mips/math-emu/sp_tlong.c +++ b/arch/mips/math-emu/sp_tlong.c @@ -5,8 +5,6 @@ * MIPS floating point support * Copyright (C) 1994-2000 Algorithmics Ltd. * - * ######################################################################## - * * This program is free software; you can distribute it and/or modify it * under the terms of the GNU General Public License (Version 2) as * published by the Free Software Foundation. @@ -18,19 +16,22 @@ * * You should have received a copy of the GNU General Public License along * with this program; if not, write to the Free Software Foundation, Inc., - * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * ######################################################################## + * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. */ - #include "ieee754sp.h" +#include "ieee754dp.h" -s64 ieee754sp_tlong(ieee754sp x) +s64 ieee754sp_tlong(union ieee754sp x) { + u32 residue; + int round; + int sticky; + int odd; + COMPXDP; /* <-- need 64-bit mantissa tmp */ - CLEARCX; + ieee754_clearcx(); EXPLODEXSP; FLUSHXSP; @@ -39,10 +40,12 @@ s64 ieee754sp_tlong(ieee754sp x) case IEEE754_CLASS_SNAN: case IEEE754_CLASS_QNAN: case IEEE754_CLASS_INF: - SETCX(IEEE754_INVALID_OPERATION); - return ieee754di_xcpt(ieee754di_indef(), "sp_tlong", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754di_indef(); + case IEEE754_CLASS_ZERO: return 0; + case IEEE754_CLASS_DNORM: case IEEE754_CLASS_NORM: break; @@ -53,69 +56,51 @@ s64 ieee754sp_tlong(ieee754sp x) return -0x8000000000000000LL; /* Set invalid. We will only use overflow for floating point overflow */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754di_xcpt(ieee754di_indef(), "sp_tlong", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754di_indef(); } /* oh gawd */ - if (xe > SP_MBITS) { - xm <<= xe - SP_MBITS; - } else if (xe < SP_MBITS) { - u32 residue; - int round; - int sticky; - int odd; - + if (xe > SP_FBITS) { + xm <<= xe - SP_FBITS; + } else if (xe < SP_FBITS) { if (xe < -1) { residue = xm; round = 0; sticky = residue != 0; xm = 0; } else { - residue = xm << (32 - SP_MBITS + xe); + residue = xm << (32 - SP_FBITS + xe); round = (residue >> 31) != 0; sticky = (residue << 1) != 0; - xm >>= SP_MBITS - xe; + xm >>= SP_FBITS - xe; } odd = (xm & 0x1) != 0x0; switch (ieee754_csr.rm) { - case IEEE754_RN: + case FPU_CSR_RN: if (round && (sticky || odd)) xm++; break; - case IEEE754_RZ: + case FPU_CSR_RZ: break; - case IEEE754_RU: /* toward +Infinity */ + case FPU_CSR_RU: /* toward +Infinity */ if ((round || sticky) && !xs) xm++; break; - case IEEE754_RD: /* toward -Infinity */ + case FPU_CSR_RD: /* toward -Infinity */ if ((round || sticky) && xs) xm++; break; } if ((xm >> 63) != 0) { /* This can happen after rounding */ - SETCX(IEEE754_INVALID_OPERATION); - return ieee754di_xcpt(ieee754di_indef(), "sp_tlong", x); + ieee754_setcx(IEEE754_INVALID_OPERATION); + return ieee754di_indef(); } if (round || sticky) - SETCX(IEEE754_INEXACT); + ieee754_setcx(IEEE754_INEXACT); } if (xs) return -xm; else return xm; } - - -u64 ieee754sp_tulong(ieee754sp x) -{ - ieee754sp hb = ieee754sp_1e63(); - - /* what if x < 0 ?? */ - if (ieee754sp_lt(x, hb)) - return (u64) ieee754sp_tlong(x); - - return (u64) ieee754sp_tlong(ieee754sp_sub(x, hb)) | - (1ULL << 63); -} diff --git a/arch/mips/mm/c-octeon.c b/arch/mips/mm/c-octeon.c index f41a5c5b086..05b1d7cf951 100644 --- a/arch/mips/mm/c-octeon.c +++ b/arch/mips/mm/c-octeon.c @@ -137,8 +137,10 @@ static void octeon_flush_cache_sigtramp(unsigned long addr) { struct vm_area_struct *vma; + down_read(¤t->mm->mmap_sem); vma = find_vma(current->mm, addr); octeon_flush_icache_all_cores(vma); + up_read(¤t->mm->mmap_sem); } diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c index 1c74a6ad072..f2e8302fa70 100644 --- a/arch/mips/mm/c-r4k.c +++ b/arch/mips/mm/c-r4k.c @@ -7,6 +7,7 @@ * Copyright (C) 1997, 1998, 1999, 2000, 2001, 2002 Ralf Baechle (ralf@gnu.org) * Copyright (C) 1999, 2000 Silicon Graphics, Inc. */ +#include <linux/cpu_pm.h> #include <linux/hardirq.h> #include <linux/init.h> #include <linux/highmem.h> @@ -50,7 +51,7 @@ static inline void r4k_on_each_cpu(void (*func) (void *info), void *info) { preempt_disable(); -#if !defined(CONFIG_MIPS_MT_SMP) && !defined(CONFIG_MIPS_MT_SMTC) +#ifndef CONFIG_MIPS_MT_SMP smp_call_function(func, info, 1); #endif func(info); @@ -105,22 +106,37 @@ static inline void r4k_blast_dcache_page_dc32(unsigned long addr) static inline void r4k_blast_dcache_page_dc64(unsigned long addr) { - R4600_HIT_CACHEOP_WAR_IMPL; blast_dcache64_page(addr); } +static inline void r4k_blast_dcache_page_dc128(unsigned long addr) +{ + blast_dcache128_page(addr); +} + static void r4k_blast_dcache_page_setup(void) { unsigned long dc_lsize = cpu_dcache_line_size(); - if (dc_lsize == 0) + switch (dc_lsize) { + case 0: r4k_blast_dcache_page = (void *)cache_noop; - else if (dc_lsize == 16) + break; + case 16: r4k_blast_dcache_page = blast_dcache16_page; - else if (dc_lsize == 32) + break; + case 32: r4k_blast_dcache_page = r4k_blast_dcache_page_dc32; - else if (dc_lsize == 64) + break; + case 64: r4k_blast_dcache_page = r4k_blast_dcache_page_dc64; + break; + case 128: + r4k_blast_dcache_page = r4k_blast_dcache_page_dc128; + break; + default: + break; + } } #ifndef CONFIG_EVA @@ -159,6 +175,8 @@ static void r4k_blast_dcache_page_indexed_setup(void) r4k_blast_dcache_page_indexed = blast_dcache32_page_indexed; else if (dc_lsize == 64) r4k_blast_dcache_page_indexed = blast_dcache64_page_indexed; + else if (dc_lsize == 128) + r4k_blast_dcache_page_indexed = blast_dcache128_page_indexed; } void (* r4k_blast_dcache)(void); @@ -176,6 +194,8 @@ static void r4k_blast_dcache_setup(void) r4k_blast_dcache = blast_dcache32; else if (dc_lsize == 64) r4k_blast_dcache = blast_dcache64; + else if (dc_lsize == 128) + r4k_blast_dcache = blast_dcache128; } /* force code alignment (used for TX49XX_ICACHE_INDEX_INV_WAR) */ @@ -265,6 +285,8 @@ static void r4k_blast_icache_page_setup(void) r4k_blast_icache_page = blast_icache32_page; else if (ic_lsize == 64) r4k_blast_icache_page = blast_icache64_page; + else if (ic_lsize == 128) + r4k_blast_icache_page = blast_icache128_page; } #ifndef CONFIG_EVA @@ -338,6 +360,8 @@ static void r4k_blast_icache_setup(void) r4k_blast_icache = blast_icache32; } else if (ic_lsize == 64) r4k_blast_icache = blast_icache64; + else if (ic_lsize == 128) + r4k_blast_icache = blast_icache128; } static void (* r4k_blast_scache_page)(unsigned long addr); @@ -428,7 +452,7 @@ static void r4k___flush_cache_all(void) static inline int has_valid_asid(const struct mm_struct *mm) { -#if defined(CONFIG_MIPS_MT_SMP) || defined(CONFIG_MIPS_MT_SMTC) +#ifdef CONFIG_MIPS_MT_SMP int i; for_each_online_cpu(i) @@ -1094,6 +1118,21 @@ static void probe_pcache(void) c->dcache.waybit = 0; break; + case CPU_CAVIUM_OCTEON3: + /* For now lie about the number of ways. */ + c->icache.linesz = 128; + c->icache.sets = 16; + c->icache.ways = 8; + c->icache.flags |= MIPS_CACHE_VTAG; + icache_size = c->icache.sets * c->icache.ways * c->icache.linesz; + + c->dcache.linesz = 128; + c->dcache.ways = 8; + c->dcache.sets = 8; + dcache_size = c->dcache.sets * c->dcache.ways * c->dcache.linesz; + c->options |= MIPS_CPU_PREFETCH; + break; + default: if (!(config & MIPS_CONF_M)) panic("Don't know how to probe P-caches on this cpu."); @@ -1414,6 +1453,7 @@ static void setup_scache(void) loongson3_sc_init(); return; + case CPU_CAVIUM_OCTEON3: case CPU_XLP: /* don't need to worry about L2, fully coherent */ return; @@ -1644,3 +1684,26 @@ void r4k_cache_init(void) coherency_setup(); board_cache_error_setup = r4k_cache_error_setup; } + +static int r4k_cache_pm_notifier(struct notifier_block *self, unsigned long cmd, + void *v) +{ + switch (cmd) { + case CPU_PM_ENTER_FAILED: + case CPU_PM_EXIT: + coherency_setup(); + break; + } + + return NOTIFY_OK; +} + +static struct notifier_block r4k_cache_pm_notifier_block = { + .notifier_call = r4k_cache_pm_notifier, +}; + +int __init r4k_cache_init_pm(void) +{ + return cpu_pm_register_notifier(&r4k_cache_pm_notifier_block); +} +arch_initcall(r4k_cache_init_pm); diff --git a/arch/mips/mm/init.c b/arch/mips/mm/init.c index 4fc74c78265..6e4413330e3 100644 --- a/arch/mips/mm/init.c +++ b/arch/mips/mm/init.c @@ -44,27 +44,6 @@ #include <asm/tlb.h> #include <asm/fixmap.h> -/* Atomicity and interruptability */ -#ifdef CONFIG_MIPS_MT_SMTC - -#include <asm/mipsmtregs.h> - -#define ENTER_CRITICAL(flags) \ - { \ - unsigned int mvpflags; \ - local_irq_save(flags);\ - mvpflags = dvpe() -#define EXIT_CRITICAL(flags) \ - evpe(mvpflags); \ - local_irq_restore(flags); \ - } -#else - -#define ENTER_CRITICAL(flags) local_irq_save(flags) -#define EXIT_CRITICAL(flags) local_irq_restore(flags) - -#endif /* CONFIG_MIPS_MT_SMTC */ - /* * We have up to 8 empty zeroed pages so we can map one of the right colour * when needed. This is necessary only on R4000 / R4400 SC and MC versions @@ -100,21 +79,7 @@ void setup_zero_pages(void) zero_page_mask = ((PAGE_SIZE << order) - 1) & PAGE_MASK; } -#ifdef CONFIG_MIPS_MT_SMTC -static pte_t *kmap_coherent_pte; -static void __init kmap_coherent_init(void) -{ - unsigned long vaddr; - - /* cache the first coherent kmap pte */ - vaddr = __fix_to_virt(FIX_CMAP_BEGIN); - kmap_coherent_pte = kmap_get_fixmap_pte(vaddr); -} -#else -static inline void kmap_coherent_init(void) {} -#endif - -void *kmap_coherent(struct page *page, unsigned long addr) +static void *__kmap_pgprot(struct page *page, unsigned long addr, pgprot_t prot) { enum fixed_addresses idx; unsigned long vaddr, flags, entrylo; @@ -126,58 +91,48 @@ void *kmap_coherent(struct page *page, unsigned long addr) pagefault_disable(); idx = (addr >> PAGE_SHIFT) & (FIX_N_COLOURS - 1); -#ifdef CONFIG_MIPS_MT_SMTC - idx += FIX_N_COLOURS * smp_processor_id() + - (in_interrupt() ? (FIX_N_COLOURS * NR_CPUS) : 0); -#else idx += in_interrupt() ? FIX_N_COLOURS : 0; -#endif vaddr = __fix_to_virt(FIX_CMAP_END - idx); - pte = mk_pte(page, PAGE_KERNEL); + pte = mk_pte(page, prot); #if defined(CONFIG_64BIT_PHYS_ADDR) && defined(CONFIG_CPU_MIPS32) entrylo = pte.pte_high; #else entrylo = pte_to_entrylo(pte_val(pte)); #endif - ENTER_CRITICAL(flags); + local_irq_save(flags); old_ctx = read_c0_entryhi(); write_c0_entryhi(vaddr & (PAGE_MASK << 1)); write_c0_entrylo0(entrylo); write_c0_entrylo1(entrylo); -#ifdef CONFIG_MIPS_MT_SMTC - set_pte(kmap_coherent_pte - (FIX_CMAP_END - idx), pte); - /* preload TLB instead of local_flush_tlb_one() */ - mtc0_tlbw_hazard(); - tlb_probe(); - tlb_probe_hazard(); - tlbidx = read_c0_index(); - mtc0_tlbw_hazard(); - if (tlbidx < 0) - tlb_write_random(); - else - tlb_write_indexed(); -#else tlbidx = read_c0_wired(); write_c0_wired(tlbidx + 1); write_c0_index(tlbidx); mtc0_tlbw_hazard(); tlb_write_indexed(); -#endif tlbw_use_hazard(); write_c0_entryhi(old_ctx); - EXIT_CRITICAL(flags); + local_irq_restore(flags); return (void*) vaddr; } +void *kmap_coherent(struct page *page, unsigned long addr) +{ + return __kmap_pgprot(page, addr, PAGE_KERNEL); +} + +void *kmap_noncoherent(struct page *page, unsigned long addr) +{ + return __kmap_pgprot(page, addr, PAGE_KERNEL_NC); +} + void kunmap_coherent(void) { -#ifndef CONFIG_MIPS_MT_SMTC unsigned int wired; unsigned long flags, old_ctx; - ENTER_CRITICAL(flags); + local_irq_save(flags); old_ctx = read_c0_entryhi(); wired = read_c0_wired() - 1; write_c0_wired(wired); @@ -189,8 +144,7 @@ void kunmap_coherent(void) tlb_write_indexed(); tlbw_use_hazard(); write_c0_entryhi(old_ctx); - EXIT_CRITICAL(flags); -#endif + local_irq_restore(flags); pagefault_enable(); } @@ -256,7 +210,7 @@ EXPORT_SYMBOL_GPL(copy_from_user_page); void __init fixrange_init(unsigned long start, unsigned long end, pgd_t *pgd_base) { -#if defined(CONFIG_HIGHMEM) || defined(CONFIG_MIPS_MT_SMTC) +#ifdef CONFIG_HIGHMEM pgd_t *pgd; pud_t *pud; pmd_t *pmd; @@ -327,8 +281,6 @@ void __init paging_init(void) #ifdef CONFIG_HIGHMEM kmap_init(); #endif - kmap_coherent_init(); - #ifdef CONFIG_ZONE_DMA max_zone_pfns[ZONE_DMA] = MAX_DMA_PFN; #endif diff --git a/arch/mips/mm/tlb-r4k.c b/arch/mips/mm/tlb-r4k.c index eeaf50f5df2..3914e27456f 100644 --- a/arch/mips/mm/tlb-r4k.c +++ b/arch/mips/mm/tlb-r4k.c @@ -8,6 +8,7 @@ * Carsten Langgaard, carstenl@mips.com * Copyright (C) 2002 MIPS Technologies, Inc. All rights reserved. */ +#include <linux/cpu_pm.h> #include <linux/init.h> #include <linux/sched.h> #include <linux/smp.h> @@ -25,28 +26,6 @@ extern void build_tlb_refill_handler(void); -/* Atomicity and interruptability */ -#ifdef CONFIG_MIPS_MT_SMTC - -#include <asm/smtc.h> -#include <asm/mipsmtregs.h> - -#define ENTER_CRITICAL(flags) \ - { \ - unsigned int mvpflags; \ - local_irq_save(flags);\ - mvpflags = dvpe() -#define EXIT_CRITICAL(flags) \ - evpe(mvpflags); \ - local_irq_restore(flags); \ - } -#else - -#define ENTER_CRITICAL(flags) local_irq_save(flags) -#define EXIT_CRITICAL(flags) local_irq_restore(flags) - -#endif /* CONFIG_MIPS_MT_SMTC */ - /* * LOONGSON2/3 has a 4 entry itlb which is a subset of dtlb, * unfortunately, itlb is not totally transparent to software. @@ -75,7 +54,7 @@ void local_flush_tlb_all(void) unsigned long old_ctx; int entry, ftlbhighset; - ENTER_CRITICAL(flags); + local_irq_save(flags); /* Save old context and create impossible VPN2 value */ old_ctx = read_c0_entryhi(); write_c0_entrylo0(0); @@ -112,7 +91,7 @@ void local_flush_tlb_all(void) tlbw_use_hazard(); write_c0_entryhi(old_ctx); flush_itlb(); - EXIT_CRITICAL(flags); + local_irq_restore(flags); } EXPORT_SYMBOL(local_flush_tlb_all); @@ -142,7 +121,7 @@ void local_flush_tlb_range(struct vm_area_struct *vma, unsigned long start, if (cpu_context(cpu, mm) != 0) { unsigned long size, flags; - ENTER_CRITICAL(flags); + local_irq_save(flags); start = round_down(start, PAGE_SIZE << 1); end = round_up(end, PAGE_SIZE << 1); size = (end - start) >> (PAGE_SHIFT + 1); @@ -176,7 +155,7 @@ void local_flush_tlb_range(struct vm_area_struct *vma, unsigned long start, drop_mmu_context(mm, cpu); } flush_itlb(); - EXIT_CRITICAL(flags); + local_irq_restore(flags); } } @@ -184,7 +163,7 @@ void local_flush_tlb_kernel_range(unsigned long start, unsigned long end) { unsigned long size, flags; - ENTER_CRITICAL(flags); + local_irq_save(flags); size = (end - start + (PAGE_SIZE - 1)) >> PAGE_SHIFT; size = (size + 1) >> 1; if (size <= (current_cpu_data.tlbsizeftlbsets ? @@ -220,7 +199,7 @@ void local_flush_tlb_kernel_range(unsigned long start, unsigned long end) local_flush_tlb_all(); } flush_itlb(); - EXIT_CRITICAL(flags); + local_irq_restore(flags); } void local_flush_tlb_page(struct vm_area_struct *vma, unsigned long page) @@ -233,7 +212,7 @@ void local_flush_tlb_page(struct vm_area_struct *vma, unsigned long page) newpid = cpu_asid(cpu, vma->vm_mm); page &= (PAGE_MASK << 1); - ENTER_CRITICAL(flags); + local_irq_save(flags); oldpid = read_c0_entryhi(); write_c0_entryhi(page | newpid); mtc0_tlbw_hazard(); @@ -253,7 +232,7 @@ void local_flush_tlb_page(struct vm_area_struct *vma, unsigned long page) finish: write_c0_entryhi(oldpid); flush_itlb_vm(vma); - EXIT_CRITICAL(flags); + local_irq_restore(flags); } } @@ -266,7 +245,7 @@ void local_flush_tlb_one(unsigned long page) unsigned long flags; int oldpid, idx; - ENTER_CRITICAL(flags); + local_irq_save(flags); oldpid = read_c0_entryhi(); page &= (PAGE_MASK << 1); write_c0_entryhi(page); @@ -285,7 +264,7 @@ void local_flush_tlb_one(unsigned long page) } write_c0_entryhi(oldpid); flush_itlb(); - EXIT_CRITICAL(flags); + local_irq_restore(flags); } /* @@ -308,7 +287,7 @@ void __update_tlb(struct vm_area_struct * vma, unsigned long address, pte_t pte) if (current->active_mm != vma->vm_mm) return; - ENTER_CRITICAL(flags); + local_irq_save(flags); pid = read_c0_entryhi() & ASID_MASK; address &= (PAGE_MASK << 1); @@ -358,7 +337,7 @@ void __update_tlb(struct vm_area_struct * vma, unsigned long address, pte_t pte) } tlbw_use_hazard(); flush_itlb_vm(vma); - EXIT_CRITICAL(flags); + local_irq_restore(flags); } void add_wired_entry(unsigned long entrylo0, unsigned long entrylo1, @@ -369,7 +348,7 @@ void add_wired_entry(unsigned long entrylo0, unsigned long entrylo1, unsigned long old_pagemask; unsigned long old_ctx; - ENTER_CRITICAL(flags); + local_irq_save(flags); /* Save old context and create impossible VPN2 value */ old_ctx = read_c0_entryhi(); old_pagemask = read_c0_pagemask(); @@ -389,7 +368,7 @@ void add_wired_entry(unsigned long entrylo0, unsigned long entrylo1, tlbw_use_hazard(); /* What is the hazard here? */ write_c0_pagemask(old_pagemask); local_flush_tlb_all(); - EXIT_CRITICAL(flags); + local_irq_restore(flags); } #ifdef CONFIG_TRANSPARENT_HUGEPAGE @@ -399,13 +378,13 @@ int __init has_transparent_hugepage(void) unsigned int mask; unsigned long flags; - ENTER_CRITICAL(flags); + local_irq_save(flags); write_c0_pagemask(PM_HUGE_MASK); back_to_back_c0_hazard(); mask = read_c0_pagemask(); write_c0_pagemask(PM_DEFAULT_MASK); - EXIT_CRITICAL(flags); + local_irq_restore(flags); return mask == PM_HUGE_MASK; } @@ -421,7 +400,10 @@ static int __init set_ntlb(char *str) __setup("ntlb=", set_ntlb); -void tlb_init(void) +/* + * Configure TLB (for init or after a CPU has been powered off). + */ +static void r4k_tlb_configure(void) { /* * You should never change this register: @@ -453,6 +435,11 @@ void tlb_init(void) local_flush_tlb_all(); /* Did I tell you that ARC SUCKS? */ +} + +void tlb_init(void) +{ + r4k_tlb_configure(); if (ntlb) { if (ntlb > 1 && ntlb <= current_cpu_data.tlbsize) { @@ -466,3 +453,26 @@ void tlb_init(void) build_tlb_refill_handler(); } + +static int r4k_tlb_pm_notifier(struct notifier_block *self, unsigned long cmd, + void *v) +{ + switch (cmd) { + case CPU_PM_ENTER_FAILED: + case CPU_PM_EXIT: + r4k_tlb_configure(); + break; + } + + return NOTIFY_OK; +} + +static struct notifier_block r4k_tlb_pm_notifier_block = { + .notifier_call = r4k_tlb_pm_notifier, +}; + +static int __init r4k_tlb_init_pm(void) +{ + return cpu_pm_register_notifier(&r4k_tlb_pm_notifier_block); +} +arch_initcall(r4k_tlb_init_pm); diff --git a/arch/mips/mm/tlbex.c b/arch/mips/mm/tlbex.c index f99ec587b15..e80e10bafc8 100644 --- a/arch/mips/mm/tlbex.c +++ b/arch/mips/mm/tlbex.c @@ -1256,7 +1256,7 @@ static void build_r4000_tlb_refill_handler(void) memset(relocs, 0, sizeof(relocs)); memset(final_handler, 0, sizeof(final_handler)); - if ((scratch_reg >= 0 || scratchpad_available()) && use_bbit_insns()) { + if (IS_ENABLED(CONFIG_64BIT) && (scratch_reg >= 0 || scratchpad_available()) && use_bbit_insns()) { htlb_info = build_fast_tlb_refill_handler(&p, &l, &r, K0, K1, scratch_reg); vmalloc_mode = refill_scratch; diff --git a/arch/mips/mm/uasm-micromips.c b/arch/mips/mm/uasm-micromips.c index b8d580ca02e..775c2800cba 100644 --- a/arch/mips/mm/uasm-micromips.c +++ b/arch/mips/mm/uasm-micromips.c @@ -63,6 +63,7 @@ static struct insn insn_table_MM[] = { { insn_cache, M(mm_pool32b_op, 0, 0, mm_cache_func, 0, 0), RT | RS | SIMM }, { insn_daddu, 0, 0 }, { insn_daddiu, 0, 0 }, + { insn_divu, M(mm_pool32a_op, 0, 0, 0, mm_divu_op, mm_pool32axf_op), RT | RS }, { insn_dmfc0, 0, 0 }, { insn_dmtc0, 0, 0 }, { insn_dsll, 0, 0 }, @@ -78,14 +79,20 @@ static struct insn insn_table_MM[] = { { insn_ext, M(mm_pool32a_op, 0, 0, 0, 0, mm_ext_op), RT | RS | RD | RE }, { insn_j, M(mm_j32_op, 0, 0, 0, 0, 0), JIMM }, { insn_jal, M(mm_jal32_op, 0, 0, 0, 0, 0), JIMM }, + { insn_jalr, M(mm_pool32a_op, 0, 0, 0, mm_jalr_op, mm_pool32axf_op), RT | RS }, { insn_jr, M(mm_pool32a_op, 0, 0, 0, mm_jalr_op, mm_pool32axf_op), RS }, + { insn_lb, M(mm_lb32_op, 0, 0, 0, 0, 0), RT | RS | SIMM }, { insn_ld, 0, 0 }, + { insn_lh, M(mm_lh32_op, 0, 0, 0, 0, 0), RS | RS | SIMM }, { insn_ll, M(mm_pool32c_op, 0, 0, (mm_ll_func << 1), 0, 0), RS | RT | SIMM }, { insn_lld, 0, 0 }, { insn_lui, M(mm_pool32i_op, mm_lui_op, 0, 0, 0, 0), RS | SIMM }, { insn_lw, M(mm_lw32_op, 0, 0, 0, 0, 0), RT | RS | SIMM }, { insn_mfc0, M(mm_pool32a_op, 0, 0, 0, mm_mfc0_op, mm_pool32axf_op), RT | RS | RD }, + { insn_mfhi, M(mm_pool32a_op, 0, 0, 0, mm_mfhi32_op, mm_pool32axf_op), RS }, + { insn_mflo, M(mm_pool32a_op, 0, 0, 0, mm_mflo32_op, mm_pool32axf_op), RS }, { insn_mtc0, M(mm_pool32a_op, 0, 0, 0, mm_mtc0_op, mm_pool32axf_op), RT | RS | RD }, + { insn_mul, M(mm_pool32a_op, 0, 0, 0, 0, mm_mul_op), RT | RS | RD }, { insn_or, M(mm_pool32a_op, 0, 0, 0, 0, mm_or32_op), RT | RS | RD }, { insn_ori, M(mm_ori32_op, 0, 0, 0, 0, 0), RT | RS | UIMM }, { insn_pref, M(mm_pool32c_op, 0, 0, (mm_pref_func << 1), 0, 0), RT | RS | SIMM }, @@ -94,15 +101,22 @@ static struct insn insn_table_MM[] = { { insn_scd, 0, 0 }, { insn_sd, 0, 0 }, { insn_sll, M(mm_pool32a_op, 0, 0, 0, 0, mm_sll32_op), RT | RS | RD }, + { insn_sllv, M(mm_pool32a_op, 0, 0, 0, 0, mm_sllv32_op), RT | RS | RD }, + { insn_sltiu, M(mm_sltiu32_op, 0, 0, 0, 0, 0), RT | RS | SIMM }, + { insn_sltu, M(mm_pool32a_op, 0, 0, 0, 0, mm_sltu_op), RT | RS | RD }, { insn_sra, M(mm_pool32a_op, 0, 0, 0, 0, mm_sra_op), RT | RS | RD }, { insn_srl, M(mm_pool32a_op, 0, 0, 0, 0, mm_srl32_op), RT | RS | RD }, + { insn_srlv, M(mm_pool32a_op, 0, 0, 0, 0, mm_srlv32_op), RT | RS | RD }, { insn_rotr, M(mm_pool32a_op, 0, 0, 0, 0, mm_rotr_op), RT | RS | RD }, { insn_subu, M(mm_pool32a_op, 0, 0, 0, 0, mm_subu32_op), RT | RS | RD }, { insn_sw, M(mm_sw32_op, 0, 0, 0, 0, 0), RT | RS | SIMM }, + { insn_sync, M(mm_pool32a_op, 0, 0, 0, mm_sync_op, mm_pool32axf_op), RS }, { insn_tlbp, M(mm_pool32a_op, 0, 0, 0, mm_tlbp_op, mm_pool32axf_op), 0 }, { insn_tlbr, M(mm_pool32a_op, 0, 0, 0, mm_tlbr_op, mm_pool32axf_op), 0 }, { insn_tlbwi, M(mm_pool32a_op, 0, 0, 0, mm_tlbwi_op, mm_pool32axf_op), 0 }, { insn_tlbwr, M(mm_pool32a_op, 0, 0, 0, mm_tlbwr_op, mm_pool32axf_op), 0 }, + { insn_wait, M(mm_pool32a_op, 0, 0, 0, mm_wait_op, mm_pool32axf_op), SCIMM }, + { insn_wsbh, M(mm_pool32a_op, 0, 0, 0, mm_wsbh_op, mm_pool32axf_op), RT | RS }, { insn_xor, M(mm_pool32a_op, 0, 0, 0, 0, mm_xor32_op), RT | RS | RD }, { insn_xori, M(mm_xori32_op, 0, 0, 0, 0, 0), RT | RS | UIMM }, { insn_dins, 0, 0 }, diff --git a/arch/mips/mm/uasm-mips.c b/arch/mips/mm/uasm-mips.c index 3abd609518c..38792c2364f 100644 --- a/arch/mips/mm/uasm-mips.c +++ b/arch/mips/mm/uasm-mips.c @@ -67,6 +67,7 @@ static struct insn insn_table[] = { { insn_daddu, M(spec_op, 0, 0, 0, 0, daddu_op), RS | RT | RD }, { insn_dinsm, M(spec3_op, 0, 0, 0, 0, dinsm_op), RS | RT | RD | RE }, { insn_dins, M(spec3_op, 0, 0, 0, 0, dins_op), RS | RT | RD | RE }, + { insn_divu, M(spec_op, 0, 0, 0, 0, divu_op), RS | RT }, { insn_dmfc0, M(cop0_op, dmfc_op, 0, 0, 0, 0), RT | RD | SET}, { insn_dmtc0, M(cop0_op, dmtc_op, 0, 0, 0, 0), RT | RD | SET}, { insn_drotr32, M(spec_op, 1, 0, 0, 0, dsrl32_op), RT | RD | RE }, @@ -82,17 +83,23 @@ static struct insn insn_table[] = { { insn_ins, M(spec3_op, 0, 0, 0, 0, ins_op), RS | RT | RD | RE }, { insn_j, M(j_op, 0, 0, 0, 0, 0), JIMM }, { insn_jal, M(jal_op, 0, 0, 0, 0, 0), JIMM }, + { insn_jalr, M(spec_op, 0, 0, 0, 0, jalr_op), RS | RD }, { insn_j, M(j_op, 0, 0, 0, 0, 0), JIMM }, { insn_jr, M(spec_op, 0, 0, 0, 0, jr_op), RS }, + { insn_lb, M(lb_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_ld, M(ld_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_ldx, M(spec3_op, 0, 0, 0, ldx_op, lx_op), RS | RT | RD }, + { insn_lh, M(lw_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_lld, M(lld_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_ll, M(ll_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_lui, M(lui_op, 0, 0, 0, 0, 0), RT | SIMM }, { insn_lw, M(lw_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_lwx, M(spec3_op, 0, 0, 0, lwx_op, lx_op), RS | RT | RD }, { insn_mfc0, M(cop0_op, mfc_op, 0, 0, 0, 0), RT | RD | SET}, + { insn_mfhi, M(spec_op, 0, 0, 0, 0, mfhi_op), RD }, + { insn_mflo, M(spec_op, 0, 0, 0, 0, mflo_op), RD }, { insn_mtc0, M(cop0_op, mtc_op, 0, 0, 0, 0), RT | RD | SET}, + { insn_mul, M(spec2_op, 0, 0, 0, 0, mul_op), RS | RT | RD}, { insn_ori, M(ori_op, 0, 0, 0, 0, 0), RS | RT | UIMM }, { insn_or, M(spec_op, 0, 0, 0, 0, or_op), RS | RT | RD }, { insn_pref, M(pref_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, @@ -102,17 +109,25 @@ static struct insn insn_table[] = { { insn_sc, M(sc_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_sd, M(sd_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_sll, M(spec_op, 0, 0, 0, 0, sll_op), RT | RD | RE }, + { insn_sllv, M(spec_op, 0, 0, 0, 0, sllv_op), RS | RT | RD }, + { insn_sltiu, M(sltiu_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, + { insn_sltu, M(spec_op, 0, 0, 0, 0, sltu_op), RS | RT | RD }, { insn_sra, M(spec_op, 0, 0, 0, 0, sra_op), RT | RD | RE }, { insn_srl, M(spec_op, 0, 0, 0, 0, srl_op), RT | RD | RE }, + { insn_srlv, M(spec_op, 0, 0, 0, 0, srlv_op), RS | RT | RD }, { insn_subu, M(spec_op, 0, 0, 0, 0, subu_op), RS | RT | RD }, { insn_sw, M(sw_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, + { insn_sync, M(spec_op, 0, 0, 0, 0, sync_op), RE }, { insn_syscall, M(spec_op, 0, 0, 0, 0, syscall_op), SCIMM}, { insn_tlbp, M(cop0_op, cop_op, 0, 0, 0, tlbp_op), 0 }, { insn_tlbr, M(cop0_op, cop_op, 0, 0, 0, tlbr_op), 0 }, { insn_tlbwi, M(cop0_op, cop_op, 0, 0, 0, tlbwi_op), 0 }, { insn_tlbwr, M(cop0_op, cop_op, 0, 0, 0, tlbwr_op), 0 }, + { insn_wait, M(cop0_op, cop_op, 0, 0, 0, wait_op), SCIMM }, + { insn_wsbh, M(spec3_op, 0, 0, 0, wsbh_op, bshfl_op), RT | RD }, { insn_xori, M(xori_op, 0, 0, 0, 0, 0), RS | RT | UIMM }, { insn_xor, M(spec_op, 0, 0, 0, 0, xor_op), RS | RT | RD }, + { insn_yield, M(spec3_op, 0, 0, 0, 0, yield_op), RS | RD }, { insn_invalid, 0, 0 } }; diff --git a/arch/mips/mm/uasm.c b/arch/mips/mm/uasm.c index b9d14b6c7f5..00515805fe4 100644 --- a/arch/mips/mm/uasm.c +++ b/arch/mips/mm/uasm.c @@ -47,14 +47,16 @@ enum opcode { insn_addiu, insn_addu, insn_and, insn_andi, insn_bbit0, insn_bbit1, insn_beq, insn_beql, insn_bgez, insn_bgezl, insn_bltz, insn_bltzl, insn_bne, insn_cache, insn_daddiu, insn_daddu, insn_dins, insn_dinsm, - insn_dmfc0, insn_dmtc0, insn_drotr, insn_drotr32, insn_dsll, + insn_divu, insn_dmfc0, insn_dmtc0, insn_drotr, insn_drotr32, insn_dsll, insn_dsll32, insn_dsra, insn_dsrl, insn_dsrl32, insn_dsubu, insn_eret, - insn_ext, insn_ins, insn_j, insn_jal, insn_jr, insn_ld, insn_ldx, - insn_ll, insn_lld, insn_lui, insn_lw, insn_lwx, insn_mfc0, insn_mtc0, + insn_ext, insn_ins, insn_j, insn_jal, insn_jalr, insn_jr, insn_lb, + insn_ld, insn_ldx, insn_lh, insn_ll, insn_lld, insn_lui, insn_lw, + insn_lwx, insn_mfc0, insn_mfhi, insn_mflo, insn_mtc0, insn_mul, insn_or, insn_ori, insn_pref, insn_rfe, insn_rotr, insn_sc, insn_scd, - insn_sd, insn_sll, insn_sra, insn_srl, insn_subu, insn_sw, - insn_syscall, insn_tlbp, insn_tlbr, insn_tlbwi, insn_tlbwr, insn_xor, - insn_xori, + insn_sd, insn_sll, insn_sllv, insn_sltiu, insn_sltu, insn_sra, + insn_srl, insn_srlv, insn_subu, insn_sw, insn_sync, insn_syscall, + insn_tlbp, insn_tlbr, insn_tlbwi, insn_tlbwr, insn_wait, insn_wsbh, + insn_xor, insn_xori, insn_yield, }; struct insn { @@ -144,6 +146,13 @@ Ip_u2u1u3(op) \ } \ UASM_EXPORT_SYMBOL(uasm_i##op); +#define I_u3u2u1(op) \ +Ip_u3u2u1(op) \ +{ \ + build_insn(buf, insn##op, c, b, a); \ +} \ +UASM_EXPORT_SYMBOL(uasm_i##op); + #define I_u3u1u2(op) \ Ip_u3u1u2(op) \ { \ @@ -200,6 +209,13 @@ Ip_u1u2(op) \ } \ UASM_EXPORT_SYMBOL(uasm_i##op); +#define I_u2u1(op) \ +Ip_u1u2(op) \ +{ \ + build_insn(buf, insn##op, b, a); \ +} \ +UASM_EXPORT_SYMBOL(uasm_i##op); + #define I_u1s2(op) \ Ip_u1s2(op) \ { \ @@ -237,6 +253,7 @@ I_u1u2u3(_dmfc0) I_u1u2u3(_dmtc0) I_u2u1s3(_daddiu) I_u3u1u2(_daddu) +I_u1u2(_divu) I_u2u1u3(_dsll) I_u2u1u3(_dsll32) I_u2u1u3(_dsra) @@ -250,14 +267,20 @@ I_u2u1msbdu3(_ext) I_u2u1msbu3(_ins) I_u1(_j) I_u1(_jal) +I_u2u1(_jalr) I_u1(_jr) +I_u2s3u1(_lb) I_u2s3u1(_ld) +I_u2s3u1(_lh) I_u2s3u1(_ll) I_u2s3u1(_lld) I_u1s2(_lui) I_u2s3u1(_lw) I_u1u2u3(_mfc0) +I_u1(_mfhi) +I_u1(_mflo) I_u1u2u3(_mtc0) +I_u3u1u2(_mul) I_u2u1u3(_ori) I_u3u1u2(_or) I_0(_rfe) @@ -265,17 +288,25 @@ I_u2s3u1(_sc) I_u2s3u1(_scd) I_u2s3u1(_sd) I_u2u1u3(_sll) +I_u3u2u1(_sllv) +I_u2u1s3(_sltiu) +I_u3u1u2(_sltu) I_u2u1u3(_sra) I_u2u1u3(_srl) +I_u3u2u1(_srlv) I_u2u1u3(_rotr) I_u3u1u2(_subu) I_u2s3u1(_sw) +I_u1(_sync) I_0(_tlbp) I_0(_tlbr) I_0(_tlbwi) I_0(_tlbwr) +I_u1(_wait); +I_u2u1(_wsbh) I_u3u1u2(_xor) I_u2u1u3(_xori) +I_u2u1(_yield) I_u2u1msbu3(_dins); I_u2u1msb32u3(_dinsm); I_u1(_syscall); @@ -469,6 +500,14 @@ void ISAFUNC(uasm_il_b)(u32 **p, struct uasm_reloc **r, int lid) } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_b)); +void ISAFUNC(uasm_il_beq)(u32 **p, struct uasm_reloc **r, unsigned int r1, + unsigned int r2, int lid) +{ + uasm_r_mips_pc16(r, *p, lid); + ISAFUNC(uasm_i_beq)(p, r1, r2, 0); +} +UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_beq)); + void ISAFUNC(uasm_il_beqz)(u32 **p, struct uasm_reloc **r, unsigned int reg, int lid) { diff --git a/arch/mips/mti-malta/Makefile b/arch/mips/mti-malta/Makefile index eae0ba3876d..b9510ea8db5 100644 --- a/arch/mips/mti-malta/Makefile +++ b/arch/mips/mti-malta/Makefile @@ -9,5 +9,4 @@ obj-y := malta-amon.o malta-display.o malta-init.o \ malta-int.o malta-memory.o malta-platform.o \ malta-reset.o malta-setup.o malta-time.o -# FIXME FIXME FIXME -obj-$(CONFIG_MIPS_MT_SMTC) += malta-smtc.o +obj-$(CONFIG_MIPS_MALTA_PM) += malta-pm.o diff --git a/arch/mips/mti-malta/malta-init.c b/arch/mips/mti-malta/malta-init.c index 4f9e44d358b..0f60256d378 100644 --- a/arch/mips/mti-malta/malta-init.c +++ b/arch/mips/mti-malta/malta-init.c @@ -116,8 +116,6 @@ phys_t mips_cpc_default_phys_base(void) return CPC_BASE_ADDR; } -extern struct plat_smp_ops msmtc_smp_ops; - void __init prom_init(void) { mips_display_message("LINUX"); @@ -304,8 +302,4 @@ mips_pci_controller: return; if (!register_vsmp_smp_ops()) return; - -#ifdef CONFIG_MIPS_MT_SMTC - register_smp_ops(&msmtc_smp_ops); -#endif } diff --git a/arch/mips/mti-malta/malta-int.c b/arch/mips/mti-malta/malta-int.c index b71ee809191..ecc2785f785 100644 --- a/arch/mips/mti-malta/malta-int.c +++ b/arch/mips/mti-malta/malta-int.c @@ -504,28 +504,9 @@ void __init arch_init_irq(void) } else if (cpu_has_vint) { set_vi_handler(MIPSCPU_INT_I8259A, malta_hw0_irqdispatch); set_vi_handler(MIPSCPU_INT_COREHI, corehi_irqdispatch); -#ifdef CONFIG_MIPS_MT_SMTC - setup_irq_smtc(MIPS_CPU_IRQ_BASE+MIPSCPU_INT_I8259A, &i8259irq, - (0x100 << MIPSCPU_INT_I8259A)); - setup_irq_smtc(MIPS_CPU_IRQ_BASE+MIPSCPU_INT_COREHI, - &corehi_irqaction, (0x100 << MIPSCPU_INT_COREHI)); - /* - * Temporary hack to ensure that the subsidiary device - * interrupts coing in via the i8259A, but associated - * with low IRQ numbers, will restore the Status.IM - * value associated with the i8259A. - */ - { - int i; - - for (i = 0; i < 16; i++) - irq_hwmask[i] = (0x100 << MIPSCPU_INT_I8259A); - } -#else /* Not SMTC */ setup_irq(MIPS_CPU_IRQ_BASE+MIPSCPU_INT_I8259A, &i8259irq); setup_irq(MIPS_CPU_IRQ_BASE+MIPSCPU_INT_COREHI, &corehi_irqaction); -#endif /* CONFIG_MIPS_MT_SMTC */ } else { setup_irq(MIPS_CPU_IRQ_BASE+MIPSCPU_INT_I8259A, &i8259irq); setup_irq(MIPS_CPU_IRQ_BASE+MIPSCPU_INT_COREHI, diff --git a/arch/mips/mti-malta/malta-memory.c b/arch/mips/mti-malta/malta-memory.c index f2364e41968..6d977309675 100644 --- a/arch/mips/mti-malta/malta-memory.c +++ b/arch/mips/mti-malta/malta-memory.c @@ -26,8 +26,8 @@ unsigned long physical_memsize = 0L; fw_memblock_t * __init fw_getmdesc(int eva) { - char *memsize_str, *ememsize_str __maybe_unused = NULL, *ptr; - unsigned long memsize = 0, ememsize __maybe_unused = 0; + char *memsize_str, *ememsize_str = NULL, *ptr; + unsigned long memsize = 0, ememsize = 0; static char cmdline[COMMAND_LINE_SIZE] __initdata; int tmp; diff --git a/arch/mips/mti-malta/malta-pm.c b/arch/mips/mti-malta/malta-pm.c new file mode 100644 index 00000000000..c1e456c01a4 --- /dev/null +++ b/arch/mips/mti-malta/malta-pm.c @@ -0,0 +1,96 @@ +/* + * Copyright (C) 2014 Imagination Technologies + * Author: Paul Burton <paul.burton@imgtec.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + */ + +#include <linux/delay.h> +#include <linux/init.h> +#include <linux/io.h> +#include <linux/pci.h> + +#include <asm/mach-malta/malta-pm.h> + +static struct pci_bus *pm_pci_bus; +static resource_size_t pm_io_offset; + +int mips_pm_suspend(unsigned state) +{ + int spec_devid; + u16 sts; + + if (!pm_pci_bus || !pm_io_offset) + return -ENODEV; + + /* Ensure the power button status is clear */ + while (1) { + sts = inw(pm_io_offset + PIIX4_FUNC3IO_PMSTS); + if (!(sts & PIIX4_FUNC3IO_PMSTS_PWRBTN_STS)) + break; + outw(sts, pm_io_offset + PIIX4_FUNC3IO_PMSTS); + } + + /* Enable entry to suspend */ + outw(state | PIIX4_FUNC3IO_PMCNTRL_SUS_EN, + pm_io_offset + PIIX4_FUNC3IO_PMCNTRL); + + /* If the special cycle occurs too soon this doesn't work... */ + mdelay(10); + + /* + * The PIIX4 will enter the suspend state only after seeing a special + * cycle with the correct magic data on the PCI bus. Generate that + * cycle now. + */ + spec_devid = PCI_DEVID(0, PCI_DEVFN(0x1f, 0x7)); + pci_bus_write_config_dword(pm_pci_bus, spec_devid, 0, + PIIX4_SUSPEND_MAGIC); + + /* Give the system some time to power down */ + mdelay(1000); + + return 0; +} + +static int __init malta_pm_setup(void) +{ + struct pci_dev *dev; + int res, io_region = PCI_BRIDGE_RESOURCES; + + /* Find a reference to the PCI bus */ + pm_pci_bus = pci_find_next_bus(NULL); + if (!pm_pci_bus) { + pr_warn("malta-pm: failed to find reference to PCI bus\n"); + return -ENODEV; + } + + /* Find the PIIX4 PM device */ + dev = pci_get_subsys(PCI_VENDOR_ID_INTEL, + PCI_DEVICE_ID_INTEL_82371AB_3, PCI_ANY_ID, + PCI_ANY_ID, NULL); + if (!dev) { + pr_warn("malta-pm: failed to find PIIX4 PM\n"); + return -ENODEV; + } + + /* Request access to the PIIX4 PM IO registers */ + res = pci_request_region(dev, io_region, "PIIX4 PM IO registers"); + if (res) { + pr_warn("malta-pm: failed to request PM IO registers (%d)\n", + res); + pci_dev_put(dev); + return -ENODEV; + } + + /* Find the offset to the PIIX4 PM IO registers */ + pm_io_offset = pci_resource_start(dev, io_region); + + pci_dev_put(dev); + return 0; +} + +late_initcall(malta_pm_setup); diff --git a/arch/mips/mti-malta/malta-reset.c b/arch/mips/mti-malta/malta-reset.c index d627d4b2b47..2fd2cc2c503 100644 --- a/arch/mips/mti-malta/malta-reset.c +++ b/arch/mips/mti-malta/malta-reset.c @@ -10,6 +10,7 @@ #include <linux/pm.h> #include <asm/reboot.h> +#include <asm/mach-malta/malta-pm.h> #define SOFTRES_REG 0x1f000500 #define GORESET 0x42 @@ -24,17 +25,22 @@ static void mips_machine_restart(char *command) static void mips_machine_halt(void) { - unsigned int __iomem *softres_reg = - ioremap(SOFTRES_REG, sizeof(unsigned int)); + while (true); +} - __raw_writel(GORESET, softres_reg); +static void mips_machine_power_off(void) +{ + mips_pm_suspend(PIIX4_FUNC3IO_PMCNTRL_SUS_TYP_SOFF); + + pr_info("Failed to power down, resetting\n"); + mips_machine_restart(NULL); } static int __init mips_reboot_setup(void) { _machine_restart = mips_machine_restart; _machine_halt = mips_machine_halt; - pm_power_off = mips_machine_halt; + pm_power_off = mips_machine_power_off; return 0; } diff --git a/arch/mips/mti-malta/malta-setup.c b/arch/mips/mti-malta/malta-setup.c index bf621516aff..db7c9e5826a 100644 --- a/arch/mips/mti-malta/malta-setup.c +++ b/arch/mips/mti-malta/malta-setup.c @@ -77,11 +77,7 @@ const char *get_system_type(void) return "MIPS Malta"; } -#if defined(CONFIG_MIPS_MT_SMTC) -const char display_string[] = " SMTC LINUX ON MALTA "; -#else const char display_string[] = " LINUX ON MALTA "; -#endif /* CONFIG_MIPS_MT_SMTC */ #ifdef CONFIG_BLK_DEV_FD static void __init fd_activate(void) diff --git a/arch/mips/mti-malta/malta-smtc.c b/arch/mips/mti-malta/malta-smtc.c deleted file mode 100644 index c4849904f01..00000000000 --- a/arch/mips/mti-malta/malta-smtc.c +++ /dev/null @@ -1,162 +0,0 @@ -/* - * Malta Platform-specific hooks for SMP operation - */ -#include <linux/irq.h> -#include <linux/init.h> - -#include <asm/mipsregs.h> -#include <asm/mipsmtregs.h> -#include <asm/smtc.h> -#include <asm/smtc_ipi.h> - -/* VPE/SMP Prototype implements platform interfaces directly */ - -/* - * Cause the specified action to be performed on a targeted "CPU" - */ - -static void msmtc_send_ipi_single(int cpu, unsigned int action) -{ - /* "CPU" may be TC of same VPE, VPE of same CPU, or different CPU */ - smtc_send_ipi(cpu, LINUX_SMP_IPI, action); -} - -static void msmtc_send_ipi_mask(const struct cpumask *mask, unsigned int action) -{ - unsigned int i; - - for_each_cpu(i, mask) - msmtc_send_ipi_single(i, action); -} - -/* - * Post-config but pre-boot cleanup entry point - */ -static void msmtc_init_secondary(void) -{ - int myvpe; - - /* Don't enable Malta I/O interrupts (IP2) for secondary VPEs */ - myvpe = read_c0_tcbind() & TCBIND_CURVPE; - if (myvpe != 0) { - /* Ideally, this should be done only once per VPE, but... */ - clear_c0_status(ST0_IM); - set_c0_status((0x100 << cp0_compare_irq) - | (0x100 << MIPS_CPU_IPI_IRQ)); - if (cp0_perfcount_irq >= 0) - set_c0_status(0x100 << cp0_perfcount_irq); - } - - smtc_init_secondary(); -} - -/* - * Platform "CPU" startup hook - */ -static void msmtc_boot_secondary(int cpu, struct task_struct *idle) -{ - smtc_boot_secondary(cpu, idle); -} - -/* - * SMP initialization finalization entry point - */ -static void msmtc_smp_finish(void) -{ - smtc_smp_finish(); -} - -/* - * Hook for after all CPUs are online - */ - -static void msmtc_cpus_done(void) -{ -} - -/* - * Platform SMP pre-initialization - * - * As noted above, we can assume a single CPU for now - * but it may be multithreaded. - */ - -static void __init msmtc_smp_setup(void) -{ - /* - * we won't get the definitive value until - * we've run smtc_prepare_cpus later, but - * we would appear to need an upper bound now. - */ - smp_num_siblings = smtc_build_cpu_map(0); -} - -static void __init msmtc_prepare_cpus(unsigned int max_cpus) -{ - smtc_prepare_cpus(max_cpus); -} - -struct plat_smp_ops msmtc_smp_ops = { - .send_ipi_single = msmtc_send_ipi_single, - .send_ipi_mask = msmtc_send_ipi_mask, - .init_secondary = msmtc_init_secondary, - .smp_finish = msmtc_smp_finish, - .cpus_done = msmtc_cpus_done, - .boot_secondary = msmtc_boot_secondary, - .smp_setup = msmtc_smp_setup, - .prepare_cpus = msmtc_prepare_cpus, -}; - -#ifdef CONFIG_MIPS_MT_SMTC_IRQAFF -/* - * IRQ affinity hook - */ - - -int plat_set_irq_affinity(struct irq_data *d, const struct cpumask *affinity, - bool force) -{ - cpumask_t tmask; - int cpu = 0; - void smtc_set_irq_affinity(unsigned int irq, cpumask_t aff); - - /* - * On the legacy Malta development board, all I/O interrupts - * are routed through the 8259 and combined in a single signal - * to the CPU daughterboard, and on the CoreFPGA2/3 34K models, - * that signal is brought to IP2 of both VPEs. To avoid racing - * concurrent interrupt service events, IP2 is enabled only on - * one VPE, by convention VPE0. So long as no bits are ever - * cleared in the affinity mask, there will never be any - * interrupt forwarding. But as soon as a program or operator - * sets affinity for one of the related IRQs, we need to make - * sure that we don't ever try to forward across the VPE boundary, - * at least not until we engineer a system where the interrupt - * _ack() or _end() function can somehow know that it corresponds - * to an interrupt taken on another VPE, and perform the appropriate - * restoration of Status.IM state using MFTR/MTTR instead of the - * normal local behavior. We also ensure that no attempt will - * be made to forward to an offline "CPU". - */ - - cpumask_copy(&tmask, affinity); - for_each_cpu(cpu, affinity) { - if ((cpu_data[cpu].vpe_id != 0) || !cpu_online(cpu)) - cpu_clear(cpu, tmask); - } - cpumask_copy(d->affinity, &tmask); - - if (cpus_empty(tmask)) - /* - * We could restore a default mask here, but the - * runtime code can anyway deal with the null set - */ - printk(KERN_WARNING - "IRQ affinity leaves no legal CPU for IRQ %d\n", d->irq); - - /* Do any generic SMTC IRQ affinity setup */ - smtc_set_irq_affinity(d->irq, tmask); - - return IRQ_SET_MASK_OK_NOCOPY; -} -#endif /* CONFIG_MIPS_MT_SMTC_IRQAFF */ diff --git a/arch/mips/mti-sead3/sead3-pic32-i2c-drv.c b/arch/mips/mti-sead3/sead3-pic32-i2c-drv.c index b921e5ec507..80fe194cfa5 100644 --- a/arch/mips/mti-sead3/sead3-pic32-i2c-drv.c +++ b/arch/mips/mti-sead3/sead3-pic32-i2c-drv.c @@ -312,16 +312,13 @@ static int i2c_platform_probe(struct platform_device *pdev) pr_debug("i2c_platform_probe\n"); r = platform_get_resource(pdev, IORESOURCE_MEM, 0); - if (!r) { - ret = -ENODEV; - goto out; - } + if (!r) + return -ENODEV; - priv = kzalloc(sizeof(struct i2c_platform_data), GFP_KERNEL); - if (!priv) { - ret = -ENOMEM; - goto out; - } + priv = devm_kzalloc(&pdev->dev, sizeof(struct i2c_platform_data), + GFP_KERNEL); + if (!priv) + return -ENOMEM; /* FIXME: need to allocate resource in PIC32 space */ #if 0 @@ -330,10 +327,8 @@ static int i2c_platform_probe(struct platform_device *pdev) #else priv->base = r->start; #endif - if (!priv->base) { - ret = -EBUSY; - goto out_mem; - } + if (!priv->base) + return -EBUSY; priv->xfer_timeout = 200; priv->ack_timeout = 200; @@ -348,17 +343,13 @@ static int i2c_platform_probe(struct platform_device *pdev) i2c_platform_setup(priv); ret = i2c_add_numbered_adapter(&priv->adap); - if (ret == 0) { - platform_set_drvdata(pdev, priv); - return 0; + if (ret) { + i2c_platform_disable(priv); + return ret; } - i2c_platform_disable(priv); - -out_mem: - kfree(priv); -out: - return ret; + platform_set_drvdata(pdev, priv); + return 0; } static int i2c_platform_remove(struct platform_device *pdev) @@ -369,7 +360,6 @@ static int i2c_platform_remove(struct platform_device *pdev) platform_set_drvdata(pdev, NULL); i2c_del_adapter(&priv->adap); i2c_platform_disable(priv); - kfree(priv); return 0; } diff --git a/arch/mips/net/Makefile b/arch/mips/net/Makefile new file mode 100644 index 00000000000..ae74b3a91f5 --- /dev/null +++ b/arch/mips/net/Makefile @@ -0,0 +1,3 @@ +# MIPS networking code + +obj-$(CONFIG_BPF_JIT) += bpf_jit.o diff --git a/arch/mips/net/bpf_jit.c b/arch/mips/net/bpf_jit.c new file mode 100644 index 00000000000..a67b9753330 --- /dev/null +++ b/arch/mips/net/bpf_jit.c @@ -0,0 +1,1399 @@ +/* + * Just-In-Time compiler for BPF filters on MIPS + * + * Copyright (c) 2014 Imagination Technologies Ltd. + * Author: Markos Chandras <markos.chandras@imgtec.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; version 2 of the License. + */ + +#include <linux/bitops.h> +#include <linux/compiler.h> +#include <linux/errno.h> +#include <linux/filter.h> +#include <linux/if_vlan.h> +#include <linux/kconfig.h> +#include <linux/moduleloader.h> +#include <linux/netdevice.h> +#include <linux/string.h> +#include <linux/slab.h> +#include <linux/types.h> +#include <asm/bitops.h> +#include <asm/cacheflush.h> +#include <asm/cpu-features.h> +#include <asm/uasm.h> + +#include "bpf_jit.h" + +/* ABI + * + * s0 1st scratch register + * s1 2nd scratch register + * s2 offset register + * s3 BPF register A + * s4 BPF register X + * s5 *skb + * s6 *scratch memory + * + * On entry (*bpf_func)(*skb, *filter) + * a0 = MIPS_R_A0 = skb; + * a1 = MIPS_R_A1 = filter; + * + * Stack + * ... + * M[15] + * M[14] + * M[13] + * ... + * M[0] <-- r_M + * saved reg k-1 + * saved reg k-2 + * ... + * saved reg 0 <-- r_sp + * <no argument area> + * + * Packet layout + * + * <--------------------- len ------------------------> + * <--skb-len(r_skb_hl)-->< ----- skb->data_len ------> + * ---------------------------------------------------- + * | skb->data | + * ---------------------------------------------------- + */ + +#define RSIZE (sizeof(unsigned long)) +#define ptr typeof(unsigned long) + +/* ABI specific return values */ +#ifdef CONFIG_32BIT /* O32 */ +#ifdef CONFIG_CPU_LITTLE_ENDIAN +#define r_err MIPS_R_V1 +#define r_val MIPS_R_V0 +#else /* CONFIG_CPU_LITTLE_ENDIAN */ +#define r_err MIPS_R_V0 +#define r_val MIPS_R_V1 +#endif +#else /* N64 */ +#define r_err MIPS_R_V0 +#define r_val MIPS_R_V0 +#endif + +#define r_ret MIPS_R_V0 + +/* + * Use 2 scratch registers to avoid pipeline interlocks. + * There is no overhead during epilogue and prologue since + * any of the $s0-$s6 registers will only be preserved if + * they are going to actually be used. + */ +#define r_s0 MIPS_R_S0 /* scratch reg 1 */ +#define r_s1 MIPS_R_S1 /* scratch reg 2 */ +#define r_off MIPS_R_S2 +#define r_A MIPS_R_S3 +#define r_X MIPS_R_S4 +#define r_skb MIPS_R_S5 +#define r_M MIPS_R_S6 +#define r_tmp_imm MIPS_R_T6 /* No need to preserve this */ +#define r_tmp MIPS_R_T7 /* No need to preserve this */ +#define r_zero MIPS_R_ZERO +#define r_sp MIPS_R_SP +#define r_ra MIPS_R_RA + +#define SCRATCH_OFF(k) (4 * (k)) + +/* JIT flags */ +#define SEEN_CALL (1 << BPF_MEMWORDS) +#define SEEN_SREG_SFT (BPF_MEMWORDS + 1) +#define SEEN_SREG_BASE (1 << SEEN_SREG_SFT) +#define SEEN_SREG(x) (SEEN_SREG_BASE << (x)) +#define SEEN_S0 SEEN_SREG(0) +#define SEEN_S1 SEEN_SREG(1) +#define SEEN_OFF SEEN_SREG(2) +#define SEEN_A SEEN_SREG(3) +#define SEEN_X SEEN_SREG(4) +#define SEEN_SKB SEEN_SREG(5) +#define SEEN_MEM SEEN_SREG(6) + +/* Arguments used by JIT */ +#define ARGS_USED_BY_JIT 2 /* only applicable to 64-bit */ + +#define FLAG_NEED_X_RESET (1 << 0) + +#define SBIT(x) (1 << (x)) /* Signed version of BIT() */ + +/** + * struct jit_ctx - JIT context + * @skf: The sk_filter + * @prologue_bytes: Number of bytes for prologue + * @idx: Instruction index + * @flags: JIT flags + * @offsets: Instruction offsets + * @target: Memory location for the compiled filter + */ +struct jit_ctx { + const struct sk_filter *skf; + unsigned int prologue_bytes; + u32 idx; + u32 flags; + u32 *offsets; + u32 *target; +}; + + +static inline int optimize_div(u32 *k) +{ + /* power of 2 divides can be implemented with right shift */ + if (!(*k & (*k-1))) { + *k = ilog2(*k); + return 1; + } + + return 0; +} + +/* Simply emit the instruction if the JIT memory space has been allocated */ +#define emit_instr(ctx, func, ...) \ +do { \ + if ((ctx)->target != NULL) { \ + u32 *p = &(ctx)->target[ctx->idx]; \ + uasm_i_##func(&p, ##__VA_ARGS__); \ + } \ + (ctx)->idx++; \ +} while (0) + +/* Determine if immediate is within the 16-bit signed range */ +static inline bool is_range16(s32 imm) +{ + if (imm >= SBIT(15) || imm < -SBIT(15)) + return true; + return false; +} + +static inline void emit_addu(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, addu, dst, src1, src2); +} + +static inline void emit_nop(struct jit_ctx *ctx) +{ + emit_instr(ctx, nop); +} + +/* Load a u32 immediate to a register */ +static inline void emit_load_imm(unsigned int dst, u32 imm, struct jit_ctx *ctx) +{ + if (ctx->target != NULL) { + /* addiu can only handle s16 */ + if (is_range16(imm)) { + u32 *p = &ctx->target[ctx->idx]; + uasm_i_lui(&p, r_tmp_imm, (s32)imm >> 16); + p = &ctx->target[ctx->idx + 1]; + uasm_i_ori(&p, dst, r_tmp_imm, imm & 0xffff); + } else { + u32 *p = &ctx->target[ctx->idx]; + uasm_i_addiu(&p, dst, r_zero, imm); + } + } + ctx->idx++; + + if (is_range16(imm)) + ctx->idx++; +} + +static inline void emit_or(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, or, dst, src1, src2); +} + +static inline void emit_ori(unsigned int dst, unsigned src, u32 imm, + struct jit_ctx *ctx) +{ + if (imm >= BIT(16)) { + emit_load_imm(r_tmp, imm, ctx); + emit_or(dst, src, r_tmp, ctx); + } else { + emit_instr(ctx, ori, dst, src, imm); + } +} + + +static inline void emit_daddu(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, daddu, dst, src1, src2); +} + +static inline void emit_daddiu(unsigned int dst, unsigned int src, + int imm, struct jit_ctx *ctx) +{ + /* + * Only used for stack, so the imm is relatively small + * and it fits in 15-bits + */ + emit_instr(ctx, daddiu, dst, src, imm); +} + +static inline void emit_addiu(unsigned int dst, unsigned int src, + u32 imm, struct jit_ctx *ctx) +{ + if (is_range16(imm)) { + emit_load_imm(r_tmp, imm, ctx); + emit_addu(dst, r_tmp, src, ctx); + } else { + emit_instr(ctx, addiu, dst, src, imm); + } +} + +static inline void emit_and(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, and, dst, src1, src2); +} + +static inline void emit_andi(unsigned int dst, unsigned int src, + u32 imm, struct jit_ctx *ctx) +{ + /* If imm does not fit in u16 then load it to register */ + if (imm >= BIT(16)) { + emit_load_imm(r_tmp, imm, ctx); + emit_and(dst, src, r_tmp, ctx); + } else { + emit_instr(ctx, andi, dst, src, imm); + } +} + +static inline void emit_xor(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, xor, dst, src1, src2); +} + +static inline void emit_xori(ptr dst, ptr src, u32 imm, struct jit_ctx *ctx) +{ + /* If imm does not fit in u16 then load it to register */ + if (imm >= BIT(16)) { + emit_load_imm(r_tmp, imm, ctx); + emit_xor(dst, src, r_tmp, ctx); + } else { + emit_instr(ctx, xori, dst, src, imm); + } +} + +static inline void emit_stack_offset(int offset, struct jit_ctx *ctx) +{ + if (config_enabled(CONFIG_64BIT)) + emit_instr(ctx, daddiu, r_sp, r_sp, offset); + else + emit_instr(ctx, addiu, r_sp, r_sp, offset); + +} + +static inline void emit_subu(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, subu, dst, src1, src2); +} + +static inline void emit_neg(unsigned int reg, struct jit_ctx *ctx) +{ + emit_subu(reg, r_zero, reg, ctx); +} + +static inline void emit_sllv(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + emit_instr(ctx, sllv, dst, src, sa); +} + +static inline void emit_sll(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + /* sa is 5-bits long */ + BUG_ON(sa >= BIT(5)); + emit_instr(ctx, sll, dst, src, sa); +} + +static inline void emit_srlv(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + emit_instr(ctx, srlv, dst, src, sa); +} + +static inline void emit_srl(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + /* sa is 5-bits long */ + BUG_ON(sa >= BIT(5)); + emit_instr(ctx, srl, dst, src, sa); +} + +static inline void emit_sltu(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, sltu, dst, src1, src2); +} + +static inline void emit_sltiu(unsigned dst, unsigned int src, + unsigned int imm, struct jit_ctx *ctx) +{ + /* 16 bit immediate */ + if (is_range16((s32)imm)) { + emit_load_imm(r_tmp, imm, ctx); + emit_sltu(dst, src, r_tmp, ctx); + } else { + emit_instr(ctx, sltiu, dst, src, imm); + } + +} + +/* Store register on the stack */ +static inline void emit_store_stack_reg(ptr reg, ptr base, + unsigned int offset, + struct jit_ctx *ctx) +{ + if (config_enabled(CONFIG_64BIT)) + emit_instr(ctx, sd, reg, offset, base); + else + emit_instr(ctx, sw, reg, offset, base); +} + +static inline void emit_store(ptr reg, ptr base, unsigned int offset, + struct jit_ctx *ctx) +{ + emit_instr(ctx, sw, reg, offset, base); +} + +static inline void emit_load_stack_reg(ptr reg, ptr base, + unsigned int offset, + struct jit_ctx *ctx) +{ + if (config_enabled(CONFIG_64BIT)) + emit_instr(ctx, ld, reg, offset, base); + else + emit_instr(ctx, lw, reg, offset, base); +} + +static inline void emit_load(unsigned int reg, unsigned int base, + unsigned int offset, struct jit_ctx *ctx) +{ + emit_instr(ctx, lw, reg, offset, base); +} + +static inline void emit_load_byte(unsigned int reg, unsigned int base, + unsigned int offset, struct jit_ctx *ctx) +{ + emit_instr(ctx, lb, reg, offset, base); +} + +static inline void emit_half_load(unsigned int reg, unsigned int base, + unsigned int offset, struct jit_ctx *ctx) +{ + emit_instr(ctx, lh, reg, offset, base); +} + +static inline void emit_mul(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, mul, dst, src1, src2); +} + +static inline void emit_div(unsigned int dst, unsigned int src, + struct jit_ctx *ctx) +{ + if (ctx->target != NULL) { + u32 *p = &ctx->target[ctx->idx]; + uasm_i_divu(&p, dst, src); + p = &ctx->target[ctx->idx + 1]; + uasm_i_mfhi(&p, dst); + } + ctx->idx += 2; /* 2 insts */ +} + +static inline void emit_mod(unsigned int dst, unsigned int src, + struct jit_ctx *ctx) +{ + if (ctx->target != NULL) { + u32 *p = &ctx->target[ctx->idx]; + uasm_i_divu(&p, dst, src); + p = &ctx->target[ctx->idx + 1]; + uasm_i_mflo(&p, dst); + } + ctx->idx += 2; /* 2 insts */ +} + +static inline void emit_dsll(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + emit_instr(ctx, dsll, dst, src, sa); +} + +static inline void emit_dsrl32(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + emit_instr(ctx, dsrl32, dst, src, sa); +} + +static inline void emit_wsbh(unsigned int dst, unsigned int src, + struct jit_ctx *ctx) +{ + emit_instr(ctx, wsbh, dst, src); +} + +/* load a function pointer to register */ +static inline void emit_load_func(unsigned int reg, ptr imm, + struct jit_ctx *ctx) +{ + if (config_enabled(CONFIG_64BIT)) { + /* At this point imm is always 64-bit */ + emit_load_imm(r_tmp, (u64)imm >> 32, ctx); + emit_dsll(r_tmp_imm, r_tmp, 16, ctx); /* left shift by 16 */ + emit_ori(r_tmp, r_tmp_imm, (imm >> 16) & 0xffff, ctx); + emit_dsll(r_tmp_imm, r_tmp, 16, ctx); /* left shift by 16 */ + emit_ori(reg, r_tmp_imm, imm & 0xffff, ctx); + } else { + emit_load_imm(reg, imm, ctx); + } +} + +/* Move to real MIPS register */ +static inline void emit_reg_move(ptr dst, ptr src, struct jit_ctx *ctx) +{ + if (config_enabled(CONFIG_64BIT)) + emit_daddu(dst, src, r_zero, ctx); + else + emit_addu(dst, src, r_zero, ctx); +} + +/* Move to JIT (32-bit) register */ +static inline void emit_jit_reg_move(ptr dst, ptr src, struct jit_ctx *ctx) +{ + emit_addu(dst, src, r_zero, ctx); +} + +/* Compute the immediate value for PC-relative branches. */ +static inline u32 b_imm(unsigned int tgt, struct jit_ctx *ctx) +{ + if (ctx->target == NULL) + return 0; + + /* + * We want a pc-relative branch. We only do forward branches + * so tgt is always after pc. tgt is the instruction offset + * we want to jump to. + + * Branch on MIPS: + * I: target_offset <- sign_extend(offset) + * I+1: PC += target_offset (delay slot) + * + * ctx->idx currently points to the branch instruction + * but the offset is added to the delay slot so we need + * to subtract 4. + */ + return ctx->offsets[tgt] - + (ctx->idx * 4 - ctx->prologue_bytes) - 4; +} + +static inline void emit_bcond(int cond, unsigned int reg1, unsigned int reg2, + unsigned int imm, struct jit_ctx *ctx) +{ + if (ctx->target != NULL) { + u32 *p = &ctx->target[ctx->idx]; + + switch (cond) { + case MIPS_COND_EQ: + uasm_i_beq(&p, reg1, reg2, imm); + break; + case MIPS_COND_NE: + uasm_i_bne(&p, reg1, reg2, imm); + break; + case MIPS_COND_ALL: + uasm_i_b(&p, imm); + break; + default: + pr_warn("%s: Unhandled branch conditional: %d\n", + __func__, cond); + } + } + ctx->idx++; +} + +static inline void emit_b(unsigned int imm, struct jit_ctx *ctx) +{ + emit_bcond(MIPS_COND_ALL, r_zero, r_zero, imm, ctx); +} + +static inline void emit_jalr(unsigned int link, unsigned int reg, + struct jit_ctx *ctx) +{ + emit_instr(ctx, jalr, link, reg); +} + +static inline void emit_jr(unsigned int reg, struct jit_ctx *ctx) +{ + emit_instr(ctx, jr, reg); +} + +static inline u16 align_sp(unsigned int num) +{ + /* Double word alignment for 32-bit, quadword for 64-bit */ + unsigned int align = config_enabled(CONFIG_64BIT) ? 16 : 8; + num = (num + (align - 1)) & -align; + return num; +} + +static inline void update_on_xread(struct jit_ctx *ctx) +{ + if (!(ctx->flags & SEEN_X)) + ctx->flags |= FLAG_NEED_X_RESET; + + ctx->flags |= SEEN_X; +} + +static bool is_load_to_a(u16 inst) +{ + switch (inst) { + case BPF_S_LD_W_LEN: + case BPF_S_LD_W_ABS: + case BPF_S_LD_H_ABS: + case BPF_S_LD_B_ABS: + case BPF_S_ANC_CPU: + case BPF_S_ANC_IFINDEX: + case BPF_S_ANC_MARK: + case BPF_S_ANC_PROTOCOL: + case BPF_S_ANC_RXHASH: + case BPF_S_ANC_VLAN_TAG: + case BPF_S_ANC_VLAN_TAG_PRESENT: + case BPF_S_ANC_QUEUE: + return true; + default: + return false; + } +} + +static void save_bpf_jit_regs(struct jit_ctx *ctx, unsigned offset) +{ + int i = 0, real_off = 0; + u32 sflags, tmp_flags; + + /* Adjust the stack pointer */ + emit_stack_offset(-align_sp(offset), ctx); + + if (ctx->flags & SEEN_CALL) { + /* Argument save area */ + if (config_enabled(CONFIG_64BIT)) + /* Bottom of current frame */ + real_off = align_sp(offset) - RSIZE; + else + /* Top of previous frame */ + real_off = align_sp(offset) + RSIZE; + emit_store_stack_reg(MIPS_R_A0, r_sp, real_off, ctx); + emit_store_stack_reg(MIPS_R_A1, r_sp, real_off + RSIZE, ctx); + + real_off = 0; + } + + tmp_flags = sflags = ctx->flags >> SEEN_SREG_SFT; + /* sflags is essentially a bitmap */ + while (tmp_flags) { + if ((sflags >> i) & 0x1) { + emit_store_stack_reg(MIPS_R_S0 + i, r_sp, real_off, + ctx); + real_off += RSIZE; + } + i++; + tmp_flags >>= 1; + } + + /* save return address */ + if (ctx->flags & SEEN_CALL) { + emit_store_stack_reg(r_ra, r_sp, real_off, ctx); + real_off += RSIZE; + } + + /* Setup r_M leaving the alignment gap if necessary */ + if (ctx->flags & SEEN_MEM) { + if (real_off % (RSIZE * 2)) + real_off += RSIZE; + emit_addiu(r_M, r_sp, real_off, ctx); + } +} + +static void restore_bpf_jit_regs(struct jit_ctx *ctx, + unsigned int offset) +{ + int i, real_off = 0; + u32 sflags, tmp_flags; + + if (ctx->flags & SEEN_CALL) { + if (config_enabled(CONFIG_64BIT)) + /* Bottom of current frame */ + real_off = align_sp(offset) - RSIZE; + else + /* Top of previous frame */ + real_off = align_sp(offset) + RSIZE; + emit_load_stack_reg(MIPS_R_A0, r_sp, real_off, ctx); + emit_load_stack_reg(MIPS_R_A1, r_sp, real_off + RSIZE, ctx); + + real_off = 0; + } + + tmp_flags = sflags = ctx->flags >> SEEN_SREG_SFT; + /* sflags is a bitmap */ + i = 0; + while (tmp_flags) { + if ((sflags >> i) & 0x1) { + emit_load_stack_reg(MIPS_R_S0 + i, r_sp, real_off, + ctx); + real_off += RSIZE; + } + i++; + tmp_flags >>= 1; + } + + /* restore return address */ + if (ctx->flags & SEEN_CALL) + emit_load_stack_reg(r_ra, r_sp, real_off, ctx); + + /* Restore the sp and discard the scrach memory */ + emit_stack_offset(align_sp(offset), ctx); +} + +static unsigned int get_stack_depth(struct jit_ctx *ctx) +{ + int sp_off = 0; + + + /* How may s* regs do we need to preserved? */ + sp_off += hweight32(ctx->flags >> SEEN_SREG_SFT) * RSIZE; + + if (ctx->flags & SEEN_MEM) + sp_off += 4 * BPF_MEMWORDS; /* BPF_MEMWORDS are 32-bit */ + + if (ctx->flags & SEEN_CALL) + /* + * The JIT code make calls to external functions using 2 + * arguments. Therefore, for o32 we don't need to allocate + * space because we don't care if the argumetns are lost + * across calls. We do need however to preserve incoming + * arguments but the space is already allocated for us by + * the caller. On the other hand, for n64, we need to allocate + * this space ourselves. We need to preserve $ra as well. + */ + sp_off += config_enabled(CONFIG_64BIT) ? + (ARGS_USED_BY_JIT + 1) * RSIZE : RSIZE; + + /* + * Subtract the bytes for the last registers since we only care about + * the location on the stack pointer. + */ + return sp_off - RSIZE; +} + +static void build_prologue(struct jit_ctx *ctx) +{ + u16 first_inst = ctx->skf->insns[0].code; + int sp_off; + + /* Calculate the total offset for the stack pointer */ + sp_off = get_stack_depth(ctx); + save_bpf_jit_regs(ctx, sp_off); + + if (ctx->flags & SEEN_SKB) + emit_reg_move(r_skb, MIPS_R_A0, ctx); + + if (ctx->flags & FLAG_NEED_X_RESET) + emit_jit_reg_move(r_X, r_zero, ctx); + + /* Do not leak kernel data to userspace */ + if ((first_inst != BPF_S_RET_K) && !(is_load_to_a(first_inst))) + emit_jit_reg_move(r_A, r_zero, ctx); +} + +static void build_epilogue(struct jit_ctx *ctx) +{ + unsigned int sp_off; + + /* Calculate the total offset for the stack pointer */ + + sp_off = get_stack_depth(ctx); + restore_bpf_jit_regs(ctx, sp_off); + + /* Return */ + emit_jr(r_ra, ctx); + emit_nop(ctx); +} + +static u64 jit_get_skb_b(struct sk_buff *skb, unsigned offset) +{ + u8 ret; + int err; + + err = skb_copy_bits(skb, offset, &ret, 1); + + return (u64)err << 32 | ret; +} + +static u64 jit_get_skb_h(struct sk_buff *skb, unsigned offset) +{ + u16 ret; + int err; + + err = skb_copy_bits(skb, offset, &ret, 2); + + return (u64)err << 32 | ntohs(ret); +} + +static u64 jit_get_skb_w(struct sk_buff *skb, unsigned offset) +{ + u32 ret; + int err; + + err = skb_copy_bits(skb, offset, &ret, 4); + + return (u64)err << 32 | ntohl(ret); +} + +#define PKT_TYPE_MAX 7 +static int pkt_type_offset(void) +{ + struct sk_buff skb_probe = { + .pkt_type = ~0, + }; + char *ct = (char *)&skb_probe; + unsigned int off; + + for (off = 0; off < sizeof(struct sk_buff); off++) { + if (ct[off] == PKT_TYPE_MAX) + return off; + } + pr_err_once("Please fix pkt_type_offset(), as pkt_type couldn't be found\n"); + return -1; +} + +static int build_body(struct jit_ctx *ctx) +{ + void *load_func[] = {jit_get_skb_b, jit_get_skb_h, jit_get_skb_w}; + const struct sk_filter *prog = ctx->skf; + const struct sock_filter *inst; + unsigned int i, off, load_order, condt; + u32 k, b_off __maybe_unused; + + for (i = 0; i < prog->len; i++) { + inst = &(prog->insns[i]); + pr_debug("%s: code->0x%02x, jt->0x%x, jf->0x%x, k->0x%x\n", + __func__, inst->code, inst->jt, inst->jf, inst->k); + k = inst->k; + + if (ctx->target == NULL) + ctx->offsets[i] = ctx->idx * 4; + + switch (inst->code) { + case BPF_S_LD_IMM: + /* A <- k ==> li r_A, k */ + ctx->flags |= SEEN_A; + emit_load_imm(r_A, k, ctx); + break; + case BPF_S_LD_W_LEN: + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, len) != 4); + /* A <- len ==> lw r_A, offset(skb) */ + ctx->flags |= SEEN_SKB | SEEN_A; + off = offsetof(struct sk_buff, len); + emit_load(r_A, r_skb, off, ctx); + break; + case BPF_S_LD_MEM: + /* A <- M[k] ==> lw r_A, offset(M) */ + ctx->flags |= SEEN_MEM | SEEN_A; + emit_load(r_A, r_M, SCRATCH_OFF(k), ctx); + break; + case BPF_S_LD_W_ABS: + /* A <- P[k:4] */ + load_order = 2; + goto load; + case BPF_S_LD_H_ABS: + /* A <- P[k:2] */ + load_order = 1; + goto load; + case BPF_S_LD_B_ABS: + /* A <- P[k:1] */ + load_order = 0; +load: + emit_load_imm(r_off, k, ctx); +load_common: + ctx->flags |= SEEN_CALL | SEEN_OFF | SEEN_S0 | + SEEN_SKB | SEEN_A; + + emit_load_func(r_s0, (ptr)load_func[load_order], + ctx); + emit_reg_move(MIPS_R_A0, r_skb, ctx); + emit_jalr(MIPS_R_RA, r_s0, ctx); + /* Load second argument to delay slot */ + emit_reg_move(MIPS_R_A1, r_off, ctx); + /* Check the error value */ + if (config_enabled(CONFIG_64BIT)) { + /* Get error code from the top 32-bits */ + emit_dsrl32(r_s0, r_val, 0, ctx); + /* Branch to 3 instructions ahead */ + emit_bcond(MIPS_COND_NE, r_s0, r_zero, 3 << 2, + ctx); + } else { + /* Branch to 3 instructions ahead */ + emit_bcond(MIPS_COND_NE, r_err, r_zero, 3 << 2, + ctx); + } + emit_nop(ctx); + /* We are good */ + emit_b(b_imm(i + 1, ctx), ctx); + emit_jit_reg_move(r_A, r_val, ctx); + /* Return with error */ + emit_b(b_imm(prog->len, ctx), ctx); + emit_reg_move(r_ret, r_zero, ctx); + break; + case BPF_S_LD_W_IND: + /* A <- P[X + k:4] */ + load_order = 2; + goto load_ind; + case BPF_S_LD_H_IND: + /* A <- P[X + k:2] */ + load_order = 1; + goto load_ind; + case BPF_S_LD_B_IND: + /* A <- P[X + k:1] */ + load_order = 0; +load_ind: + update_on_xread(ctx); + ctx->flags |= SEEN_OFF | SEEN_X; + emit_addiu(r_off, r_X, k, ctx); + goto load_common; + case BPF_S_LDX_IMM: + /* X <- k */ + ctx->flags |= SEEN_X; + emit_load_imm(r_X, k, ctx); + break; + case BPF_S_LDX_MEM: + /* X <- M[k] */ + ctx->flags |= SEEN_X | SEEN_MEM; + emit_load(r_X, r_M, SCRATCH_OFF(k), ctx); + break; + case BPF_S_LDX_W_LEN: + /* X <- len */ + ctx->flags |= SEEN_X | SEEN_SKB; + off = offsetof(struct sk_buff, len); + emit_load(r_X, r_skb, off, ctx); + break; + case BPF_S_LDX_B_MSH: + /* X <- 4 * (P[k:1] & 0xf) */ + ctx->flags |= SEEN_X | SEEN_CALL | SEEN_S0 | SEEN_SKB; + /* Load offset to a1 */ + emit_load_func(r_s0, (ptr)jit_get_skb_b, ctx); + /* + * This may emit two instructions so it may not fit + * in the delay slot. So use a0 in the delay slot. + */ + emit_load_imm(MIPS_R_A1, k, ctx); + emit_jalr(MIPS_R_RA, r_s0, ctx); + emit_reg_move(MIPS_R_A0, r_skb, ctx); /* delay slot */ + /* Check the error value */ + if (config_enabled(CONFIG_64BIT)) { + /* Top 32-bits of $v0 on 64-bit */ + emit_dsrl32(r_s0, r_val, 0, ctx); + emit_bcond(MIPS_COND_NE, r_s0, r_zero, + 3 << 2, ctx); + } else { + emit_bcond(MIPS_COND_NE, r_err, r_zero, + 3 << 2, ctx); + } + /* No need for delay slot */ + /* We are good */ + /* X <- P[1:K] & 0xf */ + emit_andi(r_X, r_val, 0xf, ctx); + /* X << 2 */ + emit_b(b_imm(i + 1, ctx), ctx); + emit_sll(r_X, r_X, 2, ctx); /* delay slot */ + /* Return with error */ + emit_b(b_imm(prog->len, ctx), ctx); + emit_load_imm(r_ret, 0, ctx); /* delay slot */ + break; + case BPF_S_ST: + /* M[k] <- A */ + ctx->flags |= SEEN_MEM | SEEN_A; + emit_store(r_A, r_M, SCRATCH_OFF(k), ctx); + break; + case BPF_S_STX: + /* M[k] <- X */ + ctx->flags |= SEEN_MEM | SEEN_X; + emit_store(r_X, r_M, SCRATCH_OFF(k), ctx); + break; + case BPF_S_ALU_ADD_K: + /* A += K */ + ctx->flags |= SEEN_A; + emit_addiu(r_A, r_A, k, ctx); + break; + case BPF_S_ALU_ADD_X: + /* A += X */ + ctx->flags |= SEEN_A | SEEN_X; + emit_addu(r_A, r_A, r_X, ctx); + break; + case BPF_S_ALU_SUB_K: + /* A -= K */ + ctx->flags |= SEEN_A; + emit_addiu(r_A, r_A, -k, ctx); + break; + case BPF_S_ALU_SUB_X: + /* A -= X */ + ctx->flags |= SEEN_A | SEEN_X; + emit_subu(r_A, r_A, r_X, ctx); + break; + case BPF_S_ALU_MUL_K: + /* A *= K */ + /* Load K to scratch register before MUL */ + ctx->flags |= SEEN_A | SEEN_S0; + emit_load_imm(r_s0, k, ctx); + emit_mul(r_A, r_A, r_s0, ctx); + break; + case BPF_S_ALU_MUL_X: + /* A *= X */ + update_on_xread(ctx); + ctx->flags |= SEEN_A | SEEN_X; + emit_mul(r_A, r_A, r_X, ctx); + break; + case BPF_S_ALU_DIV_K: + /* A /= k */ + if (k == 1) + break; + if (optimize_div(&k)) { + ctx->flags |= SEEN_A; + emit_srl(r_A, r_A, k, ctx); + break; + } + ctx->flags |= SEEN_A | SEEN_S0; + emit_load_imm(r_s0, k, ctx); + emit_div(r_A, r_s0, ctx); + break; + case BPF_S_ALU_MOD_K: + /* A %= k */ + if (k == 1 || optimize_div(&k)) { + ctx->flags |= SEEN_A; + emit_jit_reg_move(r_A, r_zero, ctx); + } else { + ctx->flags |= SEEN_A | SEEN_S0; + emit_load_imm(r_s0, k, ctx); + emit_mod(r_A, r_s0, ctx); + } + break; + case BPF_S_ALU_DIV_X: + /* A /= X */ + update_on_xread(ctx); + ctx->flags |= SEEN_X | SEEN_A; + /* Check if r_X is zero */ + emit_bcond(MIPS_COND_EQ, r_X, r_zero, + b_imm(prog->len, ctx), ctx); + emit_load_imm(r_val, 0, ctx); /* delay slot */ + emit_div(r_A, r_X, ctx); + break; + case BPF_S_ALU_MOD_X: + /* A %= X */ + update_on_xread(ctx); + ctx->flags |= SEEN_X | SEEN_A; + /* Check if r_X is zero */ + emit_bcond(MIPS_COND_EQ, r_X, r_zero, + b_imm(prog->len, ctx), ctx); + emit_load_imm(r_val, 0, ctx); /* delay slot */ + emit_mod(r_A, r_X, ctx); + break; + case BPF_S_ALU_OR_K: + /* A |= K */ + ctx->flags |= SEEN_A; + emit_ori(r_A, r_A, k, ctx); + break; + case BPF_S_ALU_OR_X: + /* A |= X */ + update_on_xread(ctx); + ctx->flags |= SEEN_A; + emit_ori(r_A, r_A, r_X, ctx); + break; + case BPF_S_ALU_XOR_K: + /* A ^= k */ + ctx->flags |= SEEN_A; + emit_xori(r_A, r_A, k, ctx); + break; + case BPF_S_ANC_ALU_XOR_X: + case BPF_S_ALU_XOR_X: + /* A ^= X */ + update_on_xread(ctx); + ctx->flags |= SEEN_A; + emit_xor(r_A, r_A, r_X, ctx); + break; + case BPF_S_ALU_AND_K: + /* A &= K */ + ctx->flags |= SEEN_A; + emit_andi(r_A, r_A, k, ctx); + break; + case BPF_S_ALU_AND_X: + /* A &= X */ + update_on_xread(ctx); + ctx->flags |= SEEN_A | SEEN_X; + emit_and(r_A, r_A, r_X, ctx); + break; + case BPF_S_ALU_LSH_K: + /* A <<= K */ + ctx->flags |= SEEN_A; + emit_sll(r_A, r_A, k, ctx); + break; + case BPF_S_ALU_LSH_X: + /* A <<= X */ + ctx->flags |= SEEN_A | SEEN_X; + update_on_xread(ctx); + emit_sllv(r_A, r_A, r_X, ctx); + break; + case BPF_S_ALU_RSH_K: + /* A >>= K */ + ctx->flags |= SEEN_A; + emit_srl(r_A, r_A, k, ctx); + break; + case BPF_S_ALU_RSH_X: + ctx->flags |= SEEN_A | SEEN_X; + update_on_xread(ctx); + emit_srlv(r_A, r_A, r_X, ctx); + break; + case BPF_S_ALU_NEG: + /* A = -A */ + ctx->flags |= SEEN_A; + emit_neg(r_A, ctx); + break; + case BPF_S_JMP_JA: + /* pc += K */ + emit_b(b_imm(i + k + 1, ctx), ctx); + emit_nop(ctx); + break; + case BPF_S_JMP_JEQ_K: + /* pc += ( A == K ) ? pc->jt : pc->jf */ + condt = MIPS_COND_EQ | MIPS_COND_K; + goto jmp_cmp; + case BPF_S_JMP_JEQ_X: + ctx->flags |= SEEN_X; + /* pc += ( A == X ) ? pc->jt : pc->jf */ + condt = MIPS_COND_EQ | MIPS_COND_X; + goto jmp_cmp; + case BPF_S_JMP_JGE_K: + /* pc += ( A >= K ) ? pc->jt : pc->jf */ + condt = MIPS_COND_GE | MIPS_COND_K; + goto jmp_cmp; + case BPF_S_JMP_JGE_X: + ctx->flags |= SEEN_X; + /* pc += ( A >= X ) ? pc->jt : pc->jf */ + condt = MIPS_COND_GE | MIPS_COND_X; + goto jmp_cmp; + case BPF_S_JMP_JGT_K: + /* pc += ( A > K ) ? pc->jt : pc->jf */ + condt = MIPS_COND_GT | MIPS_COND_K; + goto jmp_cmp; + case BPF_S_JMP_JGT_X: + ctx->flags |= SEEN_X; + /* pc += ( A > X ) ? pc->jt : pc->jf */ + condt = MIPS_COND_GT | MIPS_COND_X; +jmp_cmp: + /* Greater or Equal */ + if ((condt & MIPS_COND_GE) || + (condt & MIPS_COND_GT)) { + if (condt & MIPS_COND_K) { /* K */ + ctx->flags |= SEEN_S0 | SEEN_A; + emit_sltiu(r_s0, r_A, k, ctx); + } else { /* X */ + ctx->flags |= SEEN_S0 | SEEN_A | + SEEN_X; + emit_sltu(r_s0, r_A, r_X, ctx); + } + /* A < (K|X) ? r_scrach = 1 */ + b_off = b_imm(i + inst->jf + 1, ctx); + emit_bcond(MIPS_COND_GT, r_s0, r_zero, b_off, + ctx); + emit_nop(ctx); + /* A > (K|X) ? scratch = 0 */ + if (condt & MIPS_COND_GT) { + /* Checking for equality */ + ctx->flags |= SEEN_S0 | SEEN_A | SEEN_X; + if (condt & MIPS_COND_K) + emit_load_imm(r_s0, k, ctx); + else + emit_jit_reg_move(r_s0, r_X, + ctx); + b_off = b_imm(i + inst->jf + 1, ctx); + emit_bcond(MIPS_COND_EQ, r_A, r_s0, + b_off, ctx); + emit_nop(ctx); + /* Finally, A > K|X */ + b_off = b_imm(i + inst->jt + 1, ctx); + emit_b(b_off, ctx); + emit_nop(ctx); + } else { + /* A >= (K|X) so jump */ + b_off = b_imm(i + inst->jt + 1, ctx); + emit_b(b_off, ctx); + emit_nop(ctx); + } + } else { + /* A == K|X */ + if (condt & MIPS_COND_K) { /* K */ + ctx->flags |= SEEN_S0 | SEEN_A; + emit_load_imm(r_s0, k, ctx); + /* jump true */ + b_off = b_imm(i + inst->jt + 1, ctx); + emit_bcond(MIPS_COND_EQ, r_A, r_s0, + b_off, ctx); + emit_nop(ctx); + /* jump false */ + b_off = b_imm(i + inst->jf + 1, + ctx); + emit_bcond(MIPS_COND_NE, r_A, r_s0, + b_off, ctx); + emit_nop(ctx); + } else { /* X */ + /* jump true */ + ctx->flags |= SEEN_A | SEEN_X; + b_off = b_imm(i + inst->jt + 1, + ctx); + emit_bcond(MIPS_COND_EQ, r_A, r_X, + b_off, ctx); + emit_nop(ctx); + /* jump false */ + b_off = b_imm(i + inst->jf + 1, ctx); + emit_bcond(MIPS_COND_NE, r_A, r_X, + b_off, ctx); + emit_nop(ctx); + } + } + break; + case BPF_S_JMP_JSET_K: + ctx->flags |= SEEN_S0 | SEEN_S1 | SEEN_A; + /* pc += (A & K) ? pc -> jt : pc -> jf */ + emit_load_imm(r_s1, k, ctx); + emit_and(r_s0, r_A, r_s1, ctx); + /* jump true */ + b_off = b_imm(i + inst->jt + 1, ctx); + emit_bcond(MIPS_COND_NE, r_s0, r_zero, b_off, ctx); + emit_nop(ctx); + /* jump false */ + b_off = b_imm(i + inst->jf + 1, ctx); + emit_b(b_off, ctx); + emit_nop(ctx); + break; + case BPF_S_JMP_JSET_X: + ctx->flags |= SEEN_S0 | SEEN_X | SEEN_A; + /* pc += (A & X) ? pc -> jt : pc -> jf */ + emit_and(r_s0, r_A, r_X, ctx); + /* jump true */ + b_off = b_imm(i + inst->jt + 1, ctx); + emit_bcond(MIPS_COND_NE, r_s0, r_zero, b_off, ctx); + emit_nop(ctx); + /* jump false */ + b_off = b_imm(i + inst->jf + 1, ctx); + emit_b(b_off, ctx); + emit_nop(ctx); + break; + case BPF_S_RET_A: + ctx->flags |= SEEN_A; + if (i != prog->len - 1) + /* + * If this is not the last instruction + * then jump to the epilogue + */ + emit_b(b_imm(prog->len, ctx), ctx); + emit_reg_move(r_ret, r_A, ctx); /* delay slot */ + break; + case BPF_S_RET_K: + /* + * It can emit two instructions so it does not fit on + * the delay slot. + */ + emit_load_imm(r_ret, k, ctx); + if (i != prog->len - 1) { + /* + * If this is not the last instruction + * then jump to the epilogue + */ + emit_b(b_imm(prog->len, ctx), ctx); + emit_nop(ctx); + } + break; + case BPF_S_MISC_TAX: + /* X = A */ + ctx->flags |= SEEN_X | SEEN_A; + emit_jit_reg_move(r_X, r_A, ctx); + break; + case BPF_S_MISC_TXA: + /* A = X */ + ctx->flags |= SEEN_A | SEEN_X; + update_on_xread(ctx); + emit_jit_reg_move(r_A, r_X, ctx); + break; + /* AUX */ + case BPF_S_ANC_PROTOCOL: + /* A = ntohs(skb->protocol */ + ctx->flags |= SEEN_SKB | SEEN_OFF | SEEN_A; + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, + protocol) != 2); + off = offsetof(struct sk_buff, protocol); + emit_half_load(r_A, r_skb, off, ctx); +#ifdef CONFIG_CPU_LITTLE_ENDIAN + /* This needs little endian fixup */ + if (cpu_has_mips_r2) { + /* R2 and later have the wsbh instruction */ + emit_wsbh(r_A, r_A, ctx); + } else { + /* Get first byte */ + emit_andi(r_tmp_imm, r_A, 0xff, ctx); + /* Shift it */ + emit_sll(r_tmp, r_tmp_imm, 8, ctx); + /* Get second byte */ + emit_srl(r_tmp_imm, r_A, 8, ctx); + emit_andi(r_tmp_imm, r_tmp_imm, 0xff, ctx); + /* Put everyting together in r_A */ + emit_or(r_A, r_tmp, r_tmp_imm, ctx); + } +#endif + break; + case BPF_S_ANC_CPU: + ctx->flags |= SEEN_A | SEEN_OFF; + /* A = current_thread_info()->cpu */ + BUILD_BUG_ON(FIELD_SIZEOF(struct thread_info, + cpu) != 4); + off = offsetof(struct thread_info, cpu); + /* $28/gp points to the thread_info struct */ + emit_load(r_A, 28, off, ctx); + break; + case BPF_S_ANC_IFINDEX: + /* A = skb->dev->ifindex */ + ctx->flags |= SEEN_SKB | SEEN_A | SEEN_S0; + off = offsetof(struct sk_buff, dev); + emit_load(r_s0, r_skb, off, ctx); + /* error (0) in the delay slot */ + emit_bcond(MIPS_COND_EQ, r_s0, r_zero, + b_imm(prog->len, ctx), ctx); + emit_reg_move(r_ret, r_zero, ctx); + BUILD_BUG_ON(FIELD_SIZEOF(struct net_device, + ifindex) != 4); + off = offsetof(struct net_device, ifindex); + emit_load(r_A, r_s0, off, ctx); + break; + case BPF_S_ANC_MARK: + ctx->flags |= SEEN_SKB | SEEN_A; + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, mark) != 4); + off = offsetof(struct sk_buff, mark); + emit_load(r_A, r_skb, off, ctx); + break; + case BPF_S_ANC_RXHASH: + ctx->flags |= SEEN_SKB | SEEN_A; + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, hash) != 4); + off = offsetof(struct sk_buff, hash); + emit_load(r_A, r_skb, off, ctx); + break; + case BPF_S_ANC_VLAN_TAG: + case BPF_S_ANC_VLAN_TAG_PRESENT: + ctx->flags |= SEEN_SKB | SEEN_S0 | SEEN_A; + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, + vlan_tci) != 2); + off = offsetof(struct sk_buff, vlan_tci); + emit_half_load(r_s0, r_skb, off, ctx); + if (inst->code == BPF_S_ANC_VLAN_TAG) + emit_and(r_A, r_s0, VLAN_VID_MASK, ctx); + else + emit_and(r_A, r_s0, VLAN_TAG_PRESENT, ctx); + break; + case BPF_S_ANC_PKTTYPE: + off = pkt_type_offset(); + + if (off < 0) + return -1; + emit_load_byte(r_tmp, r_skb, off, ctx); + /* Keep only the last 3 bits */ + emit_andi(r_A, r_tmp, PKT_TYPE_MAX, ctx); + break; + case BPF_S_ANC_QUEUE: + ctx->flags |= SEEN_SKB | SEEN_A; + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, + queue_mapping) != 2); + BUILD_BUG_ON(offsetof(struct sk_buff, + queue_mapping) > 0xff); + off = offsetof(struct sk_buff, queue_mapping); + emit_half_load(r_A, r_skb, off, ctx); + break; + default: + pr_warn("%s: Unhandled opcode: 0x%02x\n", __FILE__, + inst->code); + return -1; + } + } + + /* compute offsets only during the first pass */ + if (ctx->target == NULL) + ctx->offsets[i] = ctx->idx * 4; + + return 0; +} + +int bpf_jit_enable __read_mostly; + +void bpf_jit_compile(struct sk_filter *fp) +{ + struct jit_ctx ctx; + unsigned int alloc_size, tmp_idx; + + if (!bpf_jit_enable) + return; + + memset(&ctx, 0, sizeof(ctx)); + + ctx.offsets = kcalloc(fp->len, sizeof(*ctx.offsets), GFP_KERNEL); + if (ctx.offsets == NULL) + return; + + ctx.skf = fp; + + if (build_body(&ctx)) + goto out; + + tmp_idx = ctx.idx; + build_prologue(&ctx); + ctx.prologue_bytes = (ctx.idx - tmp_idx) * 4; + /* just to complete the ctx.idx count */ + build_epilogue(&ctx); + + alloc_size = 4 * ctx.idx; + ctx.target = module_alloc(alloc_size); + if (ctx.target == NULL) + goto out; + + /* Clean it */ + memset(ctx.target, 0, alloc_size); + + ctx.idx = 0; + + /* Generate the actual JIT code */ + build_prologue(&ctx); + build_body(&ctx); + build_epilogue(&ctx); + + /* Update the icache */ + flush_icache_range((ptr)ctx.target, (ptr)(ctx.target + ctx.idx)); + + if (bpf_jit_enable > 1) + /* Dump JIT code */ + bpf_jit_dump(fp->len, alloc_size, 2, ctx.target); + + fp->bpf_func = (void *)ctx.target; + fp->jited = 1; + +out: + kfree(ctx.offsets); +} + +void bpf_jit_free(struct sk_filter *fp) +{ + if (fp->jited) + module_free(NULL, fp->bpf_func); + kfree(fp); +} diff --git a/arch/mips/net/bpf_jit.h b/arch/mips/net/bpf_jit.h new file mode 100644 index 00000000000..3a5751b4335 --- /dev/null +++ b/arch/mips/net/bpf_jit.h @@ -0,0 +1,44 @@ +/* + * Just-In-Time compiler for BPF filters on MIPS + * + * Copyright (c) 2014 Imagination Technologies Ltd. + * Author: Markos Chandras <markos.chandras@imgtec.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; version 2 of the License. + */ + +#ifndef BPF_JIT_MIPS_OP_H +#define BPF_JIT_MIPS_OP_H + +/* Registers used by JIT */ +#define MIPS_R_ZERO 0 +#define MIPS_R_V0 2 +#define MIPS_R_V1 3 +#define MIPS_R_A0 4 +#define MIPS_R_A1 5 +#define MIPS_R_T6 14 +#define MIPS_R_T7 15 +#define MIPS_R_S0 16 +#define MIPS_R_S1 17 +#define MIPS_R_S2 18 +#define MIPS_R_S3 19 +#define MIPS_R_S4 20 +#define MIPS_R_S5 21 +#define MIPS_R_S6 22 +#define MIPS_R_S7 23 +#define MIPS_R_SP 29 +#define MIPS_R_RA 31 + +/* Conditional codes */ +#define MIPS_COND_EQ 0x1 +#define MIPS_COND_GE (0x1 << 1) +#define MIPS_COND_GT (0x1 << 2) +#define MIPS_COND_NE (0x1 << 3) +#define MIPS_COND_ALL (0x1 << 4) +/* Conditionals on X register or K immediate */ +#define MIPS_COND_X (0x1 << 5) +#define MIPS_COND_K (0x1 << 6) + +#endif /* BPF_JIT_MIPS_OP_H */ diff --git a/arch/mips/netlogic/common/irq.c b/arch/mips/netlogic/common/irq.c index 5afc4b7fce0..c100b9afa0a 100644 --- a/arch/mips/netlogic/common/irq.c +++ b/arch/mips/netlogic/common/irq.c @@ -203,6 +203,8 @@ void nlm_set_pic_extra_ack(int node, int irq, void (*xack)(struct irq_data *)) xirq = nlm_irq_to_xirq(node, irq); pic_data = irq_get_handler_data(xirq); + if (WARN_ON(!pic_data)) + return; pic_data->extra_ack = xack; } diff --git a/arch/mips/netlogic/common/reset.S b/arch/mips/netlogic/common/reset.S index b231fe1e7a0..701c4bcb9e4 100644 --- a/arch/mips/netlogic/common/reset.S +++ b/arch/mips/netlogic/common/reset.S @@ -35,6 +35,7 @@ #include <asm/asm.h> #include <asm/asm-offsets.h> +#include <asm/cpu.h> #include <asm/cacheops.h> #include <asm/regdef.h> #include <asm/mipsregs.h> @@ -74,13 +75,25 @@ .endm /* + * Allow access to physical mem >64G by enabling ELPA in PAGEGRAIN + * register. This is needed before going to C code since the SP can + * in this region. Called from all HW threads. + */ +.macro xlp_early_mmu_init + mfc0 t0, CP0_PAGEMASK, 1 + li t1, (1 << 29) /* ELPA bit */ + or t0, t1 + mtc0 t0, CP0_PAGEMASK, 1 +.endm + +/* * L1D cache has to be flushed before enabling threads in XLP. * On XLP8xx/XLP3xx, we do a low level flush using processor control * registers. On XLPII CPUs, usual cache instructions work. */ .macro xlp_flush_l1_dcache mfc0 t0, CP0_EBASE, 0 - andi t0, t0, 0xff00 + andi t0, t0, PRID_IMP_MASK slt t1, t0, 0x1200 beqz t1, 15f nop @@ -159,11 +172,15 @@ FEXPORT(nlm_reset_entry) 1: /* Entry point on core wakeup */ mfc0 t0, CP0_EBASE, 0 /* processor ID */ - andi t0, 0xff00 + andi t0, PRID_IMP_MASK li t1, 0x1500 /* XLP 9xx */ beq t0, t1, 2f /* does not need to set coherent */ nop + li t1, 0x1300 /* XLP 5xx */ + beq t0, t1, 2f /* does not need to set coherent */ + nop + /* set bit in SYS coherent register for the core */ mfc0 t0, CP0_EBASE, 1 mfc0 t1, CP0_EBASE, 1 @@ -197,6 +214,9 @@ FEXPORT(nlm_reset_entry) EXPORT(nlm_boot_siblings) /* core L1D flush before enable threads */ xlp_flush_l1_dcache + /* save ra and sp, will be used later (only for boot cpu) */ + dmtc0 ra, $22, 6 + dmtc0 sp, $22, 7 /* Enable hw threads by writing to MAP_THREADMODE of the core */ li t0, CKSEG1ADDR(RESET_DATA_PHYS) lw t1, BOOT_THREAD_MODE(t0) /* t1 <- thread mode */ @@ -225,6 +245,8 @@ EXPORT(nlm_boot_siblings) #endif mtc0 t1, CP0_STATUS + xlp_early_mmu_init + /* mark CPU ready */ li t3, CKSEG1ADDR(RESET_DATA_PHYS) ADDIU t1, t3, BOOT_CPU_READY @@ -238,14 +260,12 @@ EXPORT(nlm_boot_siblings) nop /* - * For the boot CPU, we have to restore registers and - * return + * For the boot CPU, we have to restore ra and sp and return, rest + * of the registers will be restored by the caller */ -4: dmfc0 t0, $4, 2 /* restore SP from UserLocal */ - li t1, 0xfadebeef - dmtc0 t1, $4, 2 /* restore SP from UserLocal */ - PTR_SUBU sp, t0, PT_SIZE - RESTORE_ALL +4: + dmfc0 ra, $22, 6 + dmfc0 sp, $22, 7 jr ra nop EXPORT(nlm_reset_entry_end) @@ -253,6 +273,7 @@ EXPORT(nlm_reset_entry_end) LEAF(nlm_init_boot_cpu) #ifdef CONFIG_CPU_XLP xlp_config_lsu + xlp_early_mmu_init #endif jr ra nop diff --git a/arch/mips/netlogic/common/smp.c b/arch/mips/netlogic/common/smp.c index 6baae15cc7b..4fde7ac76cc 100644 --- a/arch/mips/netlogic/common/smp.c +++ b/arch/mips/netlogic/common/smp.c @@ -135,10 +135,6 @@ void nlm_smp_finish(void) local_irq_enable(); } -void nlm_cpus_done(void) -{ -} - /* * Boot all other cpus in the system, initialize them, and bring them into * the boot function @@ -198,7 +194,7 @@ void __init nlm_smp_setup(void) cpumask_scnprintf(buf, ARRAY_SIZE(buf), cpu_possible_mask); pr_info("Possible CPU mask: %s\n", buf); - /* check with the cores we have worken up */ + /* check with the cores we have woken up */ for (ncore = 0, i = 0; i < NLM_NR_NODES; i++) ncore += hweight32(nlm_get_node(i)->coremask); @@ -213,6 +209,7 @@ static int nlm_parse_cpumask(cpumask_t *wakeup_mask) { uint32_t core0_thr_mask, core_thr_mask; int threadmode, i, j; + char buf[64]; core0_thr_mask = 0; for (i = 0; i < NLM_THREADS_PER_CORE; i++) @@ -247,8 +244,8 @@ static int nlm_parse_cpumask(cpumask_t *wakeup_mask) return threadmode; unsupp: - panic("Unsupported CPU mask %lx", - (unsigned long)cpumask_bits(wakeup_mask)[0]); + cpumask_scnprintf(buf, ARRAY_SIZE(buf), wakeup_mask); + panic("Unsupported CPU mask %s", buf); return 0; } @@ -277,7 +274,6 @@ struct plat_smp_ops nlm_smp_ops = { .send_ipi_mask = nlm_send_ipi_mask, .init_secondary = nlm_init_secondary, .smp_finish = nlm_smp_finish, - .cpus_done = nlm_cpus_done, .boot_secondary = nlm_boot_secondary, .smp_setup = nlm_smp_setup, .prepare_cpus = nlm_prepare_cpus, diff --git a/arch/mips/netlogic/common/smpboot.S b/arch/mips/netlogic/common/smpboot.S index 8597657c27f..805355b0bd0 100644 --- a/arch/mips/netlogic/common/smpboot.S +++ b/arch/mips/netlogic/common/smpboot.S @@ -54,8 +54,9 @@ .set noat .set arch=xlr /* for mfcr/mtcr, XLR is sufficient */ -FEXPORT(xlp_boot_core0_siblings) /* "Master" cpu starts from here */ - dmtc0 sp, $4, 2 /* SP saved in UserLocal */ +/* Called by the boot cpu to wake up its sibling threads */ +NESTED(xlp_boot_core0_siblings, PT_SIZE, sp) + /* CPU register contents lost when enabling threads, save them first */ SAVE_ALL sync /* find the location to which nlm_boot_siblings was relocated */ @@ -65,9 +66,12 @@ FEXPORT(xlp_boot_core0_siblings) /* "Master" cpu starts from here */ dsubu t2, t1 daddu t2, t0 /* call it */ - jr t2 + jalr t2 nop - /* not reached */ + RESTORE_ALL + jr ra + nop +END(xlp_boot_core0_siblings) NESTED(nlm_boot_secondary_cpus, 16, sp) /* Initialize CP0 Status */ diff --git a/arch/mips/netlogic/common/time.c b/arch/mips/netlogic/common/time.c index 13391b8a603..0c0a1a606f7 100644 --- a/arch/mips/netlogic/common/time.c +++ b/arch/mips/netlogic/common/time.c @@ -82,6 +82,7 @@ static struct clocksource csrc_pic = { static void nlm_init_pic_timer(void) { uint64_t picbase = nlm_get_node(0)->picbase; + u32 picfreq; nlm_pic_set_timer(picbase, PIC_CLOCK_TIMER, ~0ULL, 0, 0); if (current_cpu_data.cputype == CPU_XLR) { @@ -92,7 +93,9 @@ static void nlm_init_pic_timer(void) csrc_pic.read = nlm_get_pic_timer; } csrc_pic.rating = 1000; - clocksource_register_hz(&csrc_pic, pic_timer_freq()); + picfreq = pic_timer_freq(); + clocksource_register_hz(&csrc_pic, picfreq); + pr_info("PIC clock source added, frequency %d\n", picfreq); } void __init plat_time_init(void) diff --git a/arch/mips/netlogic/dts/xlp_gvp.dts b/arch/mips/netlogic/dts/xlp_gvp.dts index 047d27f5448..bb4ecd1d47f 100644 --- a/arch/mips/netlogic/dts/xlp_gvp.dts +++ b/arch/mips/netlogic/dts/xlp_gvp.dts @@ -26,11 +26,12 @@ interrupt-parent = <&pic>; interrupts = <17>; }; - pic: pic@4000 { - interrupt-controller; + pic: pic@110000 { + compatible = "netlogic,xlp-pic"; #address-cells = <0>; #interrupt-cells = <1>; reg = <0 0x110000 0x200>; + interrupt-controller; }; nor_flash@1,0 { diff --git a/arch/mips/netlogic/xlp/Makefile b/arch/mips/netlogic/xlp/Makefile index ed9a93c0465..be358a8050c 100644 --- a/arch/mips/netlogic/xlp/Makefile +++ b/arch/mips/netlogic/xlp/Makefile @@ -2,3 +2,5 @@ obj-y += setup.o nlm_hal.o cop2-ex.o dt.o obj-$(CONFIG_SMP) += wakeup.o obj-$(CONFIG_USB) += usb-init.o obj-$(CONFIG_USB) += usb-init-xlp2.o +obj-$(CONFIG_SATA_AHCI) += ahci-init.o +obj-$(CONFIG_SATA_AHCI) += ahci-init-xlp2.o diff --git a/arch/mips/netlogic/xlp/ahci-init-xlp2.c b/arch/mips/netlogic/xlp/ahci-init-xlp2.c new file mode 100644 index 00000000000..c83dbf3689e --- /dev/null +++ b/arch/mips/netlogic/xlp/ahci-init-xlp2.c @@ -0,0 +1,377 @@ +/* + * Copyright (c) 2003-2014 Broadcom Corporation + * All Rights Reserved + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the Broadcom + * license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY BROADCOM ``AS IS'' AND ANY EXPRESS OR + * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED + * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL BROADCOM OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include <linux/dma-mapping.h> +#include <linux/kernel.h> +#include <linux/delay.h> +#include <linux/init.h> +#include <linux/pci.h> +#include <linux/irq.h> +#include <linux/bitops.h> +#include <linux/pci_ids.h> +#include <linux/nodemask.h> + +#include <asm/cpu.h> +#include <asm/mipsregs.h> + +#include <asm/netlogic/common.h> +#include <asm/netlogic/haldefs.h> +#include <asm/netlogic/mips-extns.h> +#include <asm/netlogic/xlp-hal/xlp.h> +#include <asm/netlogic/xlp-hal/iomap.h> + +#define SATA_CTL 0x0 +#define SATA_STATUS 0x1 /* Status Reg */ +#define SATA_INT 0x2 /* Interrupt Reg */ +#define SATA_INT_MASK 0x3 /* Interrupt Mask Reg */ +#define SATA_BIU_TIMEOUT 0x4 +#define AXIWRSPERRLOG 0x5 +#define AXIRDSPERRLOG 0x6 +#define BiuTimeoutLow 0x7 +#define BiuTimeoutHi 0x8 +#define BiuSlvErLow 0x9 +#define BiuSlvErHi 0xa +#define IO_CONFIG_SWAP_DIS 0xb +#define CR_REG_TIMER 0xc +#define CORE_ID 0xd +#define AXI_SLAVE_OPT1 0xe +#define PHY_MEM_ACCESS 0xf +#define PHY0_CNTRL 0x10 +#define PHY0_STAT 0x11 +#define PHY0_RX_ALIGN 0x12 +#define PHY0_RX_EQ_LO 0x13 +#define PHY0_RX_EQ_HI 0x14 +#define PHY0_BIST_LOOP 0x15 +#define PHY1_CNTRL 0x16 +#define PHY1_STAT 0x17 +#define PHY1_RX_ALIGN 0x18 +#define PHY1_RX_EQ_LO 0x19 +#define PHY1_RX_EQ_HI 0x1a +#define PHY1_BIST_LOOP 0x1b +#define RdExBase 0x1c +#define RdExLimit 0x1d +#define CacheAllocBase 0x1e +#define CacheAllocLimit 0x1f +#define BiuSlaveCmdGstNum 0x20 + +/*SATA_CTL Bits */ +#define SATA_RST_N BIT(0) /* Active low reset sata_core phy */ +#define SataCtlReserve0 BIT(1) +#define M_CSYSREQ BIT(2) /* AXI master low power, not used */ +#define S_CSYSREQ BIT(3) /* AXI slave low power, not used */ +#define P0_CP_DET BIT(8) /* Reserved, bring in from pad */ +#define P0_MP_SW BIT(9) /* Mech Switch */ +#define P0_DISABLE BIT(10) /* disable p0 */ +#define P0_ACT_LED_EN BIT(11) /* Active LED enable */ +#define P0_IRST_HARD_SYNTH BIT(12) /* PHY hard synth reset */ +#define P0_IRST_HARD_TXRX BIT(13) /* PHY lane hard reset */ +#define P0_IRST_POR BIT(14) /* PHY power on reset*/ +#define P0_IPDTXL BIT(15) /* PHY Tx lane dis/power down */ +#define P0_IPDRXL BIT(16) /* PHY Rx lane dis/power down */ +#define P0_IPDIPDMSYNTH BIT(17) /* PHY synthesizer dis/porwer down */ +#define P0_CP_POD_EN BIT(18) /* CP_POD enable */ +#define P0_AT_BYPASS BIT(19) /* P0 address translation by pass */ +#define P1_CP_DET BIT(20) /* Reserved,Cold Detect */ +#define P1_MP_SW BIT(21) /* Mech Switch */ +#define P1_DISABLE BIT(22) /* disable p1 */ +#define P1_ACT_LED_EN BIT(23) /* Active LED enable */ +#define P1_IRST_HARD_SYNTH BIT(24) /* PHY hard synth reset */ +#define P1_IRST_HARD_TXRX BIT(25) /* PHY lane hard reset */ +#define P1_IRST_POR BIT(26) /* PHY power on reset*/ +#define P1_IPDTXL BIT(27) /* PHY Tx lane dis/porwer down */ +#define P1_IPDRXL BIT(28) /* PHY Rx lane dis/porwer down */ +#define P1_IPDIPDMSYNTH BIT(29) /* PHY synthesizer dis/porwer down */ +#define P1_CP_POD_EN BIT(30) +#define P1_AT_BYPASS BIT(31) /* P1 address translation by pass */ + +/* Status register */ +#define M_CACTIVE BIT(0) /* m_cactive, not used */ +#define S_CACTIVE BIT(1) /* s_cactive, not used */ +#define P0_PHY_READY BIT(8) /* phy is ready */ +#define P0_CP_POD BIT(9) /* Cold PowerOn */ +#define P0_SLUMBER BIT(10) /* power mode slumber */ +#define P0_PATIAL BIT(11) /* power mode patial */ +#define P0_PHY_SIG_DET BIT(12) /* phy dignal detect */ +#define P0_PHY_CALI BIT(13) /* phy calibration done */ +#define P1_PHY_READY BIT(16) /* phy is ready */ +#define P1_CP_POD BIT(17) /* Cold PowerOn */ +#define P1_SLUMBER BIT(18) /* power mode slumber */ +#define P1_PATIAL BIT(19) /* power mode patial */ +#define P1_PHY_SIG_DET BIT(20) /* phy dignal detect */ +#define P1_PHY_CALI BIT(21) /* phy calibration done */ + +/* SATA CR_REG_TIMER bits */ +#define CR_TIME_SCALE (0x1000 << 0) + +/* SATA PHY specific registers start and end address */ +#define RXCDRCALFOSC0 0x0065 +#define CALDUTY 0x006e +#define RXDPIF 0x8065 +#define PPMDRIFTMAX_HI 0x80A4 + +#define nlm_read_sata_reg(b, r) nlm_read_reg(b, r) +#define nlm_write_sata_reg(b, r, v) nlm_write_reg(b, r, v) +#define nlm_get_sata_pcibase(node) \ + nlm_pcicfg_base(XLP9XX_IO_SATA_OFFSET(node)) +#define nlm_get_sata_regbase(node) \ + (nlm_get_sata_pcibase(node) + 0x100) + +/* SATA PHY config for register block 1 0x0065 .. 0x006e */ +static const u8 sata_phy_config1[] = { + 0xC9, 0xC9, 0x07, 0x07, 0x18, 0x18, 0x01, 0x01, 0x22, 0x00 +}; + +/* SATA PHY config for register block 2 0x0x8065 .. 0x0x80A4 */ +static const u8 sata_phy_config2[] = { + 0xAA, 0x00, 0x4C, 0xC9, 0xC9, 0x07, 0x07, 0x18, + 0x18, 0x05, 0x0C, 0x10, 0x00, 0x10, 0x00, 0xFF, + 0xCF, 0xF7, 0xE1, 0xF5, 0xFD, 0xFD, 0xFF, 0xFF, + 0xFF, 0xFF, 0xE3, 0xE7, 0xDB, 0xF5, 0xFD, 0xFD, + 0xF5, 0xF5, 0xFF, 0xFF, 0xE3, 0xE7, 0xDB, 0xF5, + 0xFD, 0xFD, 0xF5, 0xF5, 0xFF, 0xFF, 0xFF, 0xF5, + 0x3F, 0x00, 0x32, 0x00, 0x03, 0x01, 0x05, 0x05, + 0x04, 0x00, 0x00, 0x08, 0x04, 0x00, 0x00, 0x04, +}; + +const int sata_phy_debug = 0; /* set to verify PHY writes */ + +static void sata_clear_glue_reg(u64 regbase, u32 off, u32 bit) +{ + u32 reg_val; + + reg_val = nlm_read_sata_reg(regbase, off); + nlm_write_sata_reg(regbase, off, (reg_val & ~bit)); +} + +static void sata_set_glue_reg(u64 regbase, u32 off, u32 bit) +{ + u32 reg_val; + + reg_val = nlm_read_sata_reg(regbase, off); + nlm_write_sata_reg(regbase, off, (reg_val | bit)); +} + +static void write_phy_reg(u64 regbase, u32 addr, u32 physel, u8 data) +{ + nlm_write_sata_reg(regbase, PHY_MEM_ACCESS, + (1u << 31) | (physel << 24) | (data << 16) | addr); + udelay(850); +} + +static u8 read_phy_reg(u64 regbase, u32 addr, u32 physel) +{ + u32 val; + + nlm_write_sata_reg(regbase, PHY_MEM_ACCESS, + (0 << 31) | (physel << 24) | (0 << 16) | addr); + udelay(850); + val = nlm_read_sata_reg(regbase, PHY_MEM_ACCESS); + return (val >> 16) & 0xff; +} + +static void config_sata_phy(u64 regbase) +{ + u32 port, i, reg; + + for (port = 0; port < 2; port++) { + for (i = 0, reg = RXCDRCALFOSC0; reg <= CALDUTY; reg++, i++) + write_phy_reg(regbase, reg, port, sata_phy_config1[i]); + + for (i = 0, reg = RXDPIF; reg <= PPMDRIFTMAX_HI; reg++, i++) + write_phy_reg(regbase, reg, port, sata_phy_config2[i]); + } +} + +static void check_phy_register(u64 regbase, u32 addr, u32 physel, u8 xdata) +{ + u8 data; + + data = read_phy_reg(regbase, addr, physel); + pr_info("PHY read addr = 0x%x physel = %d data = 0x%x %s\n", + addr, physel, data, data == xdata ? "TRUE" : "FALSE"); +} + +static void verify_sata_phy_config(u64 regbase) +{ + u32 port, i, reg; + + for (port = 0; port < 2; port++) { + for (i = 0, reg = RXCDRCALFOSC0; reg <= CALDUTY; reg++, i++) + check_phy_register(regbase, reg, port, + sata_phy_config1[i]); + + for (i = 0, reg = RXDPIF; reg <= PPMDRIFTMAX_HI; reg++, i++) + check_phy_register(regbase, reg, port, + sata_phy_config2[i]); + } +} + +static void nlm_sata_firmware_init(int node) +{ + u32 reg_val; + u64 regbase; + int n; + + pr_info("Initializing XLP9XX On-chip AHCI...\n"); + regbase = nlm_get_sata_regbase(node); + + /* Reset port0 */ + sata_clear_glue_reg(regbase, SATA_CTL, P0_IRST_POR); + sata_clear_glue_reg(regbase, SATA_CTL, P0_IRST_HARD_TXRX); + sata_clear_glue_reg(regbase, SATA_CTL, P0_IRST_HARD_SYNTH); + sata_clear_glue_reg(regbase, SATA_CTL, P0_IPDTXL); + sata_clear_glue_reg(regbase, SATA_CTL, P0_IPDRXL); + sata_clear_glue_reg(regbase, SATA_CTL, P0_IPDIPDMSYNTH); + + /* port1 */ + sata_clear_glue_reg(regbase, SATA_CTL, P1_IRST_POR); + sata_clear_glue_reg(regbase, SATA_CTL, P1_IRST_HARD_TXRX); + sata_clear_glue_reg(regbase, SATA_CTL, P1_IRST_HARD_SYNTH); + sata_clear_glue_reg(regbase, SATA_CTL, P1_IPDTXL); + sata_clear_glue_reg(regbase, SATA_CTL, P1_IPDRXL); + sata_clear_glue_reg(regbase, SATA_CTL, P1_IPDIPDMSYNTH); + udelay(300); + + /* Set PHY */ + sata_set_glue_reg(regbase, SATA_CTL, P0_IPDTXL); + sata_set_glue_reg(regbase, SATA_CTL, P0_IPDRXL); + sata_set_glue_reg(regbase, SATA_CTL, P0_IPDIPDMSYNTH); + sata_set_glue_reg(regbase, SATA_CTL, P1_IPDTXL); + sata_set_glue_reg(regbase, SATA_CTL, P1_IPDRXL); + sata_set_glue_reg(regbase, SATA_CTL, P1_IPDIPDMSYNTH); + + udelay(1000); + sata_set_glue_reg(regbase, SATA_CTL, P0_IRST_POR); + udelay(1000); + sata_set_glue_reg(regbase, SATA_CTL, P1_IRST_POR); + udelay(1000); + + /* setup PHY */ + config_sata_phy(regbase); + if (sata_phy_debug) + verify_sata_phy_config(regbase); + + udelay(1000); + sata_set_glue_reg(regbase, SATA_CTL, P0_IRST_HARD_TXRX); + sata_set_glue_reg(regbase, SATA_CTL, P0_IRST_HARD_SYNTH); + sata_set_glue_reg(regbase, SATA_CTL, P1_IRST_HARD_TXRX); + sata_set_glue_reg(regbase, SATA_CTL, P1_IRST_HARD_SYNTH); + udelay(300); + + /* Override reset in serial PHY mode */ + sata_set_glue_reg(regbase, CR_REG_TIMER, CR_TIME_SCALE); + /* Set reset SATA */ + sata_set_glue_reg(regbase, SATA_CTL, SATA_RST_N); + sata_set_glue_reg(regbase, SATA_CTL, M_CSYSREQ); + sata_set_glue_reg(regbase, SATA_CTL, S_CSYSREQ); + + pr_debug("Waiting for PHYs to come up.\n"); + n = 10000; + do { + reg_val = nlm_read_sata_reg(regbase, SATA_STATUS); + if ((reg_val & P1_PHY_READY) && (reg_val & P0_PHY_READY)) + break; + udelay(10); + } while (--n > 0); + + if (reg_val & P0_PHY_READY) + pr_info("PHY0 is up.\n"); + else + pr_info("PHY0 is down.\n"); + if (reg_val & P1_PHY_READY) + pr_info("PHY1 is up.\n"); + else + pr_info("PHY1 is down.\n"); + + pr_info("XLP AHCI Init Done.\n"); +} + +static int __init nlm_ahci_init(void) +{ + int node; + + if (!cpu_is_xlp9xx()) + return 0; + for (node = 0; node < NLM_NR_NODES; node++) + if (nlm_node_present(node)) + nlm_sata_firmware_init(node); + return 0; +} + +static void nlm_sata_intr_ack(struct irq_data *data) +{ + u64 regbase; + u32 val; + int node; + + node = data->irq / NLM_IRQS_PER_NODE; + regbase = nlm_get_sata_regbase(node); + val = nlm_read_sata_reg(regbase, SATA_INT); + sata_set_glue_reg(regbase, SATA_INT, val); +} + +static void nlm_sata_fixup_bar(struct pci_dev *dev) +{ + dev->resource[5] = dev->resource[0]; + memset(&dev->resource[0], 0, sizeof(dev->resource[0])); +} + +static void nlm_sata_fixup_final(struct pci_dev *dev) +{ + u32 val; + u64 regbase; + int node; + + /* Find end bridge function to find node */ + node = xlp_socdev_to_node(dev); + regbase = nlm_get_sata_regbase(node); + + /* clear pending interrupts and then enable them */ + val = nlm_read_sata_reg(regbase, SATA_INT); + sata_set_glue_reg(regbase, SATA_INT, val); + + /* Enable only the core interrupt */ + sata_set_glue_reg(regbase, SATA_INT_MASK, 0x1); + + dev->irq = nlm_irq_to_xirq(node, PIC_SATA_IRQ); + nlm_set_pic_extra_ack(node, PIC_SATA_IRQ, nlm_sata_intr_ack); +} + +arch_initcall(nlm_ahci_init); + +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_BROADCOM, PCI_DEVICE_ID_XLP9XX_SATA, + nlm_sata_fixup_bar); + +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_BROADCOM, PCI_DEVICE_ID_XLP9XX_SATA, + nlm_sata_fixup_final); diff --git a/arch/mips/netlogic/xlp/ahci-init.c b/arch/mips/netlogic/xlp/ahci-init.c new file mode 100644 index 00000000000..a9d0fae0210 --- /dev/null +++ b/arch/mips/netlogic/xlp/ahci-init.c @@ -0,0 +1,209 @@ +/* + * Copyright (c) 2003-2014 Broadcom Corporation + * All Rights Reserved + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the Broadcom + * license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * + * THIS SOFTWARE IS PROVIDED BY BROADCOM ``AS IS'' AND ANY EXPRESS OR + * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED + * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL BROADCOM OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include <linux/dma-mapping.h> +#include <linux/kernel.h> +#include <linux/delay.h> +#include <linux/init.h> +#include <linux/pci.h> +#include <linux/irq.h> +#include <linux/bitops.h> + +#include <asm/cpu.h> +#include <asm/mipsregs.h> + +#include <asm/netlogic/haldefs.h> +#include <asm/netlogic/xlp-hal/xlp.h> +#include <asm/netlogic/common.h> +#include <asm/netlogic/xlp-hal/iomap.h> +#include <asm/netlogic/mips-extns.h> + +#define SATA_CTL 0x0 +#define SATA_STATUS 0x1 /* Status Reg */ +#define SATA_INT 0x2 /* Interrupt Reg */ +#define SATA_INT_MASK 0x3 /* Interrupt Mask Reg */ +#define SATA_CR_REG_TIMER 0x4 /* PHY Conrol Timer Reg */ +#define SATA_CORE_ID 0x5 /* Core ID Reg */ +#define SATA_AXI_SLAVE_OPT1 0x6 /* AXI Slave Options Reg */ +#define SATA_PHY_LOS_LEV 0x7 /* PHY LOS Level Reg */ +#define SATA_PHY_MULTI 0x8 /* PHY Multiplier Reg */ +#define SATA_PHY_CLK_SEL 0x9 /* Clock Select Reg */ +#define SATA_PHY_AMP1_GEN1 0xa /* PHY Transmit Amplitude Reg 1 */ +#define SATA_PHY_AMP1_GEN2 0xb /* PHY Transmit Amplitude Reg 2 */ +#define SATA_PHY_AMP1_GEN3 0xc /* PHY Transmit Amplitude Reg 3 */ +#define SATA_PHY_PRE1 0xd /* PHY Transmit Preemphasis Reg 1 */ +#define SATA_PHY_PRE2 0xe /* PHY Transmit Preemphasis Reg 2 */ +#define SATA_PHY_PRE3 0xf /* PHY Transmit Preemphasis Reg 3 */ +#define SATA_SPDMODE 0x10 /* Speed Mode Reg */ +#define SATA_REFCLK 0x11 /* Reference Clock Control Reg */ +#define SATA_BYTE_SWAP_DIS 0x12 /* byte swap disable */ + +/*SATA_CTL Bits */ +#define SATA_RST_N BIT(0) +#define PHY0_RESET_N BIT(16) +#define PHY1_RESET_N BIT(17) +#define PHY2_RESET_N BIT(18) +#define PHY3_RESET_N BIT(19) +#define M_CSYSREQ BIT(2) +#define S_CSYSREQ BIT(3) + +/*SATA_STATUS Bits */ +#define P0_PHY_READY BIT(4) +#define P1_PHY_READY BIT(5) +#define P2_PHY_READY BIT(6) +#define P3_PHY_READY BIT(7) + +#define nlm_read_sata_reg(b, r) nlm_read_reg(b, r) +#define nlm_write_sata_reg(b, r, v) nlm_write_reg(b, r, v) +#define nlm_get_sata_pcibase(node) \ + nlm_pcicfg_base(XLP_IO_SATA_OFFSET(node)) +/* SATA device specific configuration registers are starts at 0x900 offset */ +#define nlm_get_sata_regbase(node) \ + (nlm_get_sata_pcibase(node) + 0x900) + +static void sata_clear_glue_reg(uint64_t regbase, uint32_t off, uint32_t bit) +{ + uint32_t reg_val; + + reg_val = nlm_read_sata_reg(regbase, off); + nlm_write_sata_reg(regbase, off, (reg_val & ~bit)); +} + +static void sata_set_glue_reg(uint64_t regbase, uint32_t off, uint32_t bit) +{ + uint32_t reg_val; + + reg_val = nlm_read_sata_reg(regbase, off); + nlm_write_sata_reg(regbase, off, (reg_val | bit)); +} + +static void nlm_sata_firmware_init(int node) +{ + uint32_t reg_val; + uint64_t regbase; + int i; + + pr_info("XLP AHCI Initialization started.\n"); + regbase = nlm_get_sata_regbase(node); + + /* Reset SATA */ + sata_clear_glue_reg(regbase, SATA_CTL, SATA_RST_N); + /* Reset PHY */ + sata_clear_glue_reg(regbase, SATA_CTL, + (PHY3_RESET_N | PHY2_RESET_N + | PHY1_RESET_N | PHY0_RESET_N)); + + /* Set SATA */ + sata_set_glue_reg(regbase, SATA_CTL, SATA_RST_N); + /* Set PHY */ + sata_set_glue_reg(regbase, SATA_CTL, + (PHY3_RESET_N | PHY2_RESET_N + | PHY1_RESET_N | PHY0_RESET_N)); + + pr_debug("Waiting for PHYs to come up.\n"); + i = 0; + do { + reg_val = nlm_read_sata_reg(regbase, SATA_STATUS); + i++; + } while (((reg_val & 0xF0) != 0xF0) && (i < 10000)); + + for (i = 0; i < 4; i++) { + if (reg_val & (P0_PHY_READY << i)) + pr_info("PHY%d is up.\n", i); + else + pr_info("PHY%d is down.\n", i); + } + + pr_info("XLP AHCI init done.\n"); +} + +static int __init nlm_ahci_init(void) +{ + int node = 0; + int chip = read_c0_prid() & PRID_REV_MASK; + + if (chip == PRID_IMP_NETLOGIC_XLP3XX) + nlm_sata_firmware_init(node); + return 0; +} + +static void nlm_sata_intr_ack(struct irq_data *data) +{ + uint32_t val = 0; + uint64_t regbase; + + regbase = nlm_get_sata_regbase(nlm_nodeid()); + val = nlm_read_sata_reg(regbase, SATA_INT); + sata_set_glue_reg(regbase, SATA_INT, val); +} + +static void nlm_sata_fixup_bar(struct pci_dev *dev) +{ + /* + * The AHCI resource is in BAR 0, move it to + * BAR 5, where it is expected + */ + dev->resource[5] = dev->resource[0]; + memset(&dev->resource[0], 0, sizeof(dev->resource[0])); +} + +static void nlm_sata_fixup_final(struct pci_dev *dev) +{ + uint32_t val; + uint64_t regbase; + int node = 0; /* XLP3XX does not support multi-node */ + + regbase = nlm_get_sata_regbase(node); + + /* clear pending interrupts and then enable them */ + val = nlm_read_sata_reg(regbase, SATA_INT); + sata_set_glue_reg(regbase, SATA_INT, val); + + /* Mask the core interrupt. If all the interrupts + * are enabled there are spurious interrupt flow + * happening, to avoid only enable core interrupt + * mask. + */ + sata_set_glue_reg(regbase, SATA_INT_MASK, 0x1); + + dev->irq = PIC_SATA_IRQ; + nlm_set_pic_extra_ack(node, PIC_SATA_IRQ, nlm_sata_intr_ack); +} + +arch_initcall(nlm_ahci_init); + +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_NETLOGIC, PCI_DEVICE_ID_NLM_SATA, + nlm_sata_fixup_bar); +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_NETLOGIC, PCI_DEVICE_ID_NLM_SATA, + nlm_sata_fixup_final); diff --git a/arch/mips/netlogic/xlp/dt.c b/arch/mips/netlogic/xlp/dt.c index bdde33147bc..7cc46032b28 100644 --- a/arch/mips/netlogic/xlp/dt.c +++ b/arch/mips/netlogic/xlp/dt.c @@ -48,9 +48,10 @@ static void *xlp_fdt_blob; void __init *xlp_dt_init(void *fdtp) { if (!fdtp) { - switch (current_cpu_data.processor_id & 0xff00) { + switch (current_cpu_data.processor_id & PRID_IMP_MASK) { #ifdef CONFIG_DT_XLP_GVP case PRID_IMP_NETLOGIC_XLP9XX: + case PRID_IMP_NETLOGIC_XLP5XX: fdtp = __dtb_xlp_gvp_begin; break; #endif diff --git a/arch/mips/netlogic/xlp/nlm_hal.c b/arch/mips/netlogic/xlp/nlm_hal.c index 997cd9ee10d..bc24beb3a42 100644 --- a/arch/mips/netlogic/xlp/nlm_hal.c +++ b/arch/mips/netlogic/xlp/nlm_hal.c @@ -54,6 +54,8 @@ void nlm_node_init(int node) struct nlm_soc_info *nodep; nodep = nlm_get_node(node); + if (node == 0) + nodep->coremask = 1; /* node 0, boot cpu */ nodep->sysbase = nlm_get_sys_regbase(node); nodep->picbase = nlm_get_pic_regbase(node); nodep->ebase = read_c0_ebase() & (~((1 << 12) - 1)); @@ -64,31 +66,39 @@ void nlm_node_init(int node) spin_lock_init(&nodep->piclock); } -int nlm_irq_to_irt(int irq) +static int xlp9xx_irq_to_irt(int irq) +{ + switch (irq) { + case PIC_GPIO_IRQ: + return 12; + case PIC_9XX_XHCI_0_IRQ: + return 114; + case PIC_9XX_XHCI_1_IRQ: + return 115; + case PIC_UART_0_IRQ: + return 133; + case PIC_UART_1_IRQ: + return 134; + case PIC_SATA_IRQ: + return 143; + case PIC_SPI_IRQ: + return 152; + case PIC_MMC_IRQ: + return 153; + case PIC_PCIE_LINK_LEGACY_IRQ(0): + case PIC_PCIE_LINK_LEGACY_IRQ(1): + case PIC_PCIE_LINK_LEGACY_IRQ(2): + case PIC_PCIE_LINK_LEGACY_IRQ(3): + return 191 + irq - PIC_PCIE_LINK_LEGACY_IRQ_BASE; + } + return -1; +} + +static int xlp_irq_to_irt(int irq) { uint64_t pcibase; int devoff, irt; - /* bypass for 9xx */ - if (cpu_is_xlp9xx()) { - switch (irq) { - case PIC_9XX_XHCI_0_IRQ: - return 114; - case PIC_9XX_XHCI_1_IRQ: - return 115; - case PIC_UART_0_IRQ: - return 133; - case PIC_UART_1_IRQ: - return 134; - case PIC_PCIE_LINK_LEGACY_IRQ(0): - case PIC_PCIE_LINK_LEGACY_IRQ(1): - case PIC_PCIE_LINK_LEGACY_IRQ(2): - case PIC_PCIE_LINK_LEGACY_IRQ(3): - return 191 + irq - PIC_PCIE_LINK_LEGACY_IRQ_BASE; - } - return -1; - } - devoff = 0; switch (irq) { case PIC_UART_0_IRQ: @@ -98,7 +108,7 @@ int nlm_irq_to_irt(int irq) devoff = XLP_IO_UART1_OFFSET(0); break; case PIC_MMC_IRQ: - devoff = XLP_IO_SD_OFFSET(0); + devoff = XLP_IO_MMC_OFFSET(0); break; case PIC_I2C_0_IRQ: /* I2C will be fixed up */ case PIC_I2C_1_IRQ: @@ -109,6 +119,18 @@ int nlm_irq_to_irt(int irq) else devoff = XLP_IO_I2C0_OFFSET(0); break; + case PIC_SATA_IRQ: + devoff = XLP_IO_SATA_OFFSET(0); + break; + case PIC_GPIO_IRQ: + devoff = XLP_IO_GPIO_OFFSET(0); + break; + case PIC_NAND_IRQ: + devoff = XLP_IO_NAND_OFFSET(0); + break; + case PIC_SPI_IRQ: + devoff = XLP_IO_SPI_OFFSET(0); + break; default: if (cpu_is_xlpii()) { switch (irq) { @@ -164,61 +186,123 @@ int nlm_irq_to_irt(int irq) /* HW bug, PCI IRT entries are bad on early silicon, fix */ irt = PIC_IRT_PCIE_LINK_INDEX(irq - PIC_PCIE_LINK_LEGACY_IRQ_BASE); - } else if (irq >= PIC_PCIE_LINK_MSI_IRQ(0) && - irq <= PIC_PCIE_LINK_MSI_IRQ(3)) { - irt = -2; - } else if (irq >= PIC_PCIE_MSIX_IRQ(0) && - irq <= PIC_PCIE_MSIX_IRQ(3)) { - irt = -2; } else { irt = -1; } return irt; } -unsigned int nlm_get_core_frequency(int node, int core) +int nlm_irq_to_irt(int irq) { - unsigned int pll_divf, pll_divr, dfs_div, ext_div; - unsigned int rstval, dfsval, denom; - uint64_t num, sysbase; + /* return -2 for irqs without 1-1 mapping */ + if (irq >= PIC_PCIE_LINK_MSI_IRQ(0) && irq <= PIC_PCIE_LINK_MSI_IRQ(3)) + return -2; + if (irq >= PIC_PCIE_MSIX_IRQ(0) && irq <= PIC_PCIE_MSIX_IRQ(3)) + return -2; - sysbase = nlm_get_node(node)->sysbase; if (cpu_is_xlp9xx()) - rstval = nlm_read_sys_reg(sysbase, SYS_9XX_POWER_ON_RESET_CFG); + return xlp9xx_irq_to_irt(irq); else - rstval = nlm_read_sys_reg(sysbase, SYS_POWER_ON_RESET_CFG); - if (cpu_is_xlpii()) { - num = 1000000ULL * (400 * 3 + 100 * (rstval >> 26)); - denom = 3; + return xlp_irq_to_irt(irq); +} + +static unsigned int nlm_xlp2_get_core_frequency(int node, int core) +{ + unsigned int pll_post_div, ctrl_val0, ctrl_val1, denom; + uint64_t num, sysbase, clockbase; + + if (cpu_is_xlp9xx()) { + clockbase = nlm_get_clock_regbase(node); + ctrl_val0 = nlm_read_sys_reg(clockbase, + SYS_9XX_CPU_PLL_CTRL0(core)); + ctrl_val1 = nlm_read_sys_reg(clockbase, + SYS_9XX_CPU_PLL_CTRL1(core)); } else { - dfsval = nlm_read_sys_reg(sysbase, SYS_CORE_DFS_DIV_VALUE); - pll_divf = ((rstval >> 10) & 0x7f) + 1; - pll_divr = ((rstval >> 8) & 0x3) + 1; - ext_div = ((rstval >> 30) & 0x3) + 1; - dfs_div = ((dfsval >> (core * 4)) & 0xf) + 1; - - num = 800000000ULL * pll_divf; - denom = 3 * pll_divr * ext_div * dfs_div; + sysbase = nlm_get_node(node)->sysbase; + ctrl_val0 = nlm_read_sys_reg(sysbase, + SYS_CPU_PLL_CTRL0(core)); + ctrl_val1 = nlm_read_sys_reg(sysbase, + SYS_CPU_PLL_CTRL1(core)); + } + + /* Find PLL post divider value */ + switch ((ctrl_val0 >> 24) & 0x7) { + case 1: + pll_post_div = 2; + break; + case 3: + pll_post_div = 4; + break; + case 7: + pll_post_div = 8; + break; + case 6: + pll_post_div = 16; + break; + case 0: + default: + pll_post_div = 1; + break; } + + num = 1000000ULL * (400 * 3 + 100 * (ctrl_val1 & 0x3f)); + denom = 3 * pll_post_div; do_div(num, denom); + return (unsigned int)num; } -/* Calculate Frequency to the PIC from PLL. - * freq_out = ( ref_freq/2 * (6 + ctrl2[7:0]) + ctrl2[20:8]/2^13 ) / - * ((2^ctrl0[7:5]) * Table(ctrl0[26:24])) +static unsigned int nlm_xlp_get_core_frequency(int node, int core) +{ + unsigned int pll_divf, pll_divr, dfs_div, ext_div; + unsigned int rstval, dfsval, denom; + uint64_t num, sysbase; + + sysbase = nlm_get_node(node)->sysbase; + rstval = nlm_read_sys_reg(sysbase, SYS_POWER_ON_RESET_CFG); + dfsval = nlm_read_sys_reg(sysbase, SYS_CORE_DFS_DIV_VALUE); + pll_divf = ((rstval >> 10) & 0x7f) + 1; + pll_divr = ((rstval >> 8) & 0x3) + 1; + ext_div = ((rstval >> 30) & 0x3) + 1; + dfs_div = ((dfsval >> (core * 4)) & 0xf) + 1; + + num = 800000000ULL * pll_divf; + denom = 3 * pll_divr * ext_div * dfs_div; + do_div(num, denom); + + return (unsigned int)num; +} + +unsigned int nlm_get_core_frequency(int node, int core) +{ + if (cpu_is_xlpii()) + return nlm_xlp2_get_core_frequency(node, core); + else + return nlm_xlp_get_core_frequency(node, core); +} + +/* + * Calculate PIC frequency from PLL registers. + * freq_out = (ref_freq/2 * (6 + ctrl2[7:0]) + ctrl2[20:8]/2^13) / + * ((2^ctrl0[7:5]) * Table(ctrl0[26:24])) */ -static unsigned int nlm_2xx_get_pic_frequency(int node) +static unsigned int nlm_xlp2_get_pic_frequency(int node) { - u32 ctrl_val0, ctrl_val2, vco_post_div, pll_post_div; + u32 ctrl_val0, ctrl_val2, vco_post_div, pll_post_div, cpu_xlp9xx; u32 mdiv, fdiv, pll_out_freq_den, reg_select, ref_div, pic_div; - u64 ref_clk, sysbase, pll_out_freq_num, ref_clk_select; + u64 sysbase, pll_out_freq_num, ref_clk_select, clockbase, ref_clk; sysbase = nlm_get_node(node)->sysbase; + clockbase = nlm_get_clock_regbase(node); + cpu_xlp9xx = cpu_is_xlp9xx(); /* Find ref_clk_base */ - ref_clk_select = - (nlm_read_sys_reg(sysbase, SYS_POWER_ON_RESET_CFG) >> 18) & 0x3; + if (cpu_xlp9xx) + ref_clk_select = (nlm_read_sys_reg(sysbase, + SYS_9XX_POWER_ON_RESET_CFG) >> 18) & 0x3; + else + ref_clk_select = (nlm_read_sys_reg(sysbase, + SYS_POWER_ON_RESET_CFG) >> 18) & 0x3; switch (ref_clk_select) { case 0: ref_clk = 200000000ULL; @@ -239,30 +323,70 @@ static unsigned int nlm_2xx_get_pic_frequency(int node) } /* Find the clock source PLL device for PIC */ - reg_select = (nlm_read_sys_reg(sysbase, SYS_CLK_DEV_SEL) >> 22) & 0x3; - switch (reg_select) { - case 0: - ctrl_val0 = nlm_read_sys_reg(sysbase, SYS_PLL_CTRL0); - ctrl_val2 = nlm_read_sys_reg(sysbase, SYS_PLL_CTRL2); - break; - case 1: - ctrl_val0 = nlm_read_sys_reg(sysbase, SYS_PLL_CTRL0_DEVX(0)); - ctrl_val2 = nlm_read_sys_reg(sysbase, SYS_PLL_CTRL2_DEVX(0)); - break; - case 2: - ctrl_val0 = nlm_read_sys_reg(sysbase, SYS_PLL_CTRL0_DEVX(1)); - ctrl_val2 = nlm_read_sys_reg(sysbase, SYS_PLL_CTRL2_DEVX(1)); - break; - case 3: - ctrl_val0 = nlm_read_sys_reg(sysbase, SYS_PLL_CTRL0_DEVX(2)); - ctrl_val2 = nlm_read_sys_reg(sysbase, SYS_PLL_CTRL2_DEVX(2)); - break; + if (cpu_xlp9xx) { + reg_select = nlm_read_sys_reg(clockbase, + SYS_9XX_CLK_DEV_SEL) & 0x3; + switch (reg_select) { + case 0: + ctrl_val0 = nlm_read_sys_reg(clockbase, + SYS_9XX_PLL_CTRL0); + ctrl_val2 = nlm_read_sys_reg(clockbase, + SYS_9XX_PLL_CTRL2); + break; + case 1: + ctrl_val0 = nlm_read_sys_reg(clockbase, + SYS_9XX_PLL_CTRL0_DEVX(0)); + ctrl_val2 = nlm_read_sys_reg(clockbase, + SYS_9XX_PLL_CTRL2_DEVX(0)); + break; + case 2: + ctrl_val0 = nlm_read_sys_reg(clockbase, + SYS_9XX_PLL_CTRL0_DEVX(1)); + ctrl_val2 = nlm_read_sys_reg(clockbase, + SYS_9XX_PLL_CTRL2_DEVX(1)); + break; + case 3: + ctrl_val0 = nlm_read_sys_reg(clockbase, + SYS_9XX_PLL_CTRL0_DEVX(2)); + ctrl_val2 = nlm_read_sys_reg(clockbase, + SYS_9XX_PLL_CTRL2_DEVX(2)); + break; + } + } else { + reg_select = (nlm_read_sys_reg(sysbase, + SYS_CLK_DEV_SEL) >> 22) & 0x3; + switch (reg_select) { + case 0: + ctrl_val0 = nlm_read_sys_reg(sysbase, + SYS_PLL_CTRL0); + ctrl_val2 = nlm_read_sys_reg(sysbase, + SYS_PLL_CTRL2); + break; + case 1: + ctrl_val0 = nlm_read_sys_reg(sysbase, + SYS_PLL_CTRL0_DEVX(0)); + ctrl_val2 = nlm_read_sys_reg(sysbase, + SYS_PLL_CTRL2_DEVX(0)); + break; + case 2: + ctrl_val0 = nlm_read_sys_reg(sysbase, + SYS_PLL_CTRL0_DEVX(1)); + ctrl_val2 = nlm_read_sys_reg(sysbase, + SYS_PLL_CTRL2_DEVX(1)); + break; + case 3: + ctrl_val0 = nlm_read_sys_reg(sysbase, + SYS_PLL_CTRL0_DEVX(2)); + ctrl_val2 = nlm_read_sys_reg(sysbase, + SYS_PLL_CTRL2_DEVX(2)); + break; + } } vco_post_div = (ctrl_val0 >> 5) & 0x7; pll_post_div = (ctrl_val0 >> 24) & 0x7; mdiv = ctrl_val2 & 0xff; - fdiv = (ctrl_val2 >> 8) & 0xfff; + fdiv = (ctrl_val2 >> 8) & 0x1fff; /* Find PLL post divider value */ switch (pll_post_div) { @@ -292,7 +416,12 @@ static unsigned int nlm_2xx_get_pic_frequency(int node) do_div(pll_out_freq_num, pll_out_freq_den); /* PIC post divider, which happens after PLL */ - pic_div = (nlm_read_sys_reg(sysbase, SYS_CLK_DEV_DIV) >> 22) & 0x3; + if (cpu_xlp9xx) + pic_div = nlm_read_sys_reg(clockbase, + SYS_9XX_CLK_DEV_DIV) & 0x3; + else + pic_div = (nlm_read_sys_reg(sysbase, + SYS_CLK_DEV_DIV) >> 22) & 0x3; do_div(pll_out_freq_num, 1 << pic_div); return pll_out_freq_num; @@ -300,12 +429,8 @@ static unsigned int nlm_2xx_get_pic_frequency(int node) unsigned int nlm_get_pic_frequency(int node) { - /* TODO Has to calculate freq as like 2xx */ - if (cpu_is_xlp9xx()) - return 250000000; - if (cpu_is_xlpii()) - return nlm_2xx_get_pic_frequency(node); + return nlm_xlp2_get_pic_frequency(node); else return 133333333; } diff --git a/arch/mips/netlogic/xlp/setup.c b/arch/mips/netlogic/xlp/setup.c index 8c60a2dd9ef..4fdd9fd29d1 100644 --- a/arch/mips/netlogic/xlp/setup.c +++ b/arch/mips/netlogic/xlp/setup.c @@ -121,8 +121,9 @@ void __init plat_mem_setup(void) const char *get_system_type(void) { - switch (read_c0_prid() & 0xff00) { + switch (read_c0_prid() & PRID_IMP_MASK) { case PRID_IMP_NETLOGIC_XLP9XX: + case PRID_IMP_NETLOGIC_XLP5XX: case PRID_IMP_NETLOGIC_XLP2XX: return "Broadcom XLPII Series"; default: diff --git a/arch/mips/netlogic/xlp/wakeup.c b/arch/mips/netlogic/xlp/wakeup.c index 9a92617a2af..e5f44d2605a 100644 --- a/arch/mips/netlogic/xlp/wakeup.c +++ b/arch/mips/netlogic/xlp/wakeup.c @@ -135,11 +135,19 @@ static void xlp_enable_secondary_cores(const cpumask_t *wakeup_mask) if (cpu_is_xlp9xx()) { fusebase = nlm_get_fuse_regbase(n); fusemask = nlm_read_reg(fusebase, FUSE_9XX_DEVCFG6); - mask = 0xfffff; + switch (read_c0_prid() & PRID_IMP_MASK) { + case PRID_IMP_NETLOGIC_XLP5XX: + mask = 0xff; + break; + case PRID_IMP_NETLOGIC_XLP9XX: + default: + mask = 0xfffff; + break; + } } else { fusemask = nlm_read_sys_reg(nodep->sysbase, SYS_EFUSE_DEVICE_CFG_STATUS0); - switch (read_c0_prid() & 0xff00) { + switch (read_c0_prid() & PRID_IMP_MASK) { case PRID_IMP_NETLOGIC_XLP3XX: mask = 0xf; break; @@ -159,10 +167,6 @@ static void xlp_enable_secondary_cores(const cpumask_t *wakeup_mask) */ syscoremask = (1 << hweight32(~fusemask & mask)) - 1; - /* The boot cpu */ - if (n == 0) - nodep->coremask = 1; - pr_info("Node %d - SYS/FUSE coremask %x\n", n, syscoremask); for (core = 0; core < nlm_cores_per_node(); core++) { /* we will be on node 0 core 0 */ diff --git a/arch/mips/paravirt/Kconfig b/arch/mips/paravirt/Kconfig new file mode 100644 index 00000000000..ecae5861b60 --- /dev/null +++ b/arch/mips/paravirt/Kconfig @@ -0,0 +1,6 @@ +if MIPS_PARAVIRT + +config MIPS_PCI_VIRTIO + def_bool y + +endif # MIPS_PARAVIRT diff --git a/arch/mips/paravirt/Makefile b/arch/mips/paravirt/Makefile new file mode 100644 index 00000000000..5023af733a3 --- /dev/null +++ b/arch/mips/paravirt/Makefile @@ -0,0 +1,14 @@ +# +# Makefile for MIPS para-virtualized specific kernel interface routines +# under Linux. +# +# This file is subject to the terms and conditions of the GNU General Public +# License. See the file "COPYING" in the main directory of this archive +# for more details. +# +# Copyright (C) 2013 Cavium, Inc. +# + +obj-y := setup.o serial.o paravirt-irq.o + +obj-$(CONFIG_SMP) += paravirt-smp.o diff --git a/arch/mips/paravirt/Platform b/arch/mips/paravirt/Platform new file mode 100644 index 00000000000..7e76ef25ea1 --- /dev/null +++ b/arch/mips/paravirt/Platform @@ -0,0 +1,8 @@ +# +# Generic para-virtualized guest. +# +platform-$(CONFIG_MIPS_PARAVIRT) += paravirt/ +cflags-$(CONFIG_MIPS_PARAVIRT) += \ + -I$(srctree)/arch/mips/include/asm/mach-paravirt + +load-$(CONFIG_MIPS_PARAVIRT) = 0xffffffff80010000 diff --git a/arch/mips/paravirt/paravirt-irq.c b/arch/mips/paravirt/paravirt-irq.c new file mode 100644 index 00000000000..8987b06c9de --- /dev/null +++ b/arch/mips/paravirt/paravirt-irq.c @@ -0,0 +1,368 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2013 Cavium, Inc. + */ + +#include <linux/interrupt.h> +#include <linux/cpumask.h> +#include <linux/kernel.h> +#include <linux/mutex.h> + +#include <asm/io.h> + +#define MBOX_BITS_PER_CPU 2 + +static int cpunum_for_cpu(int cpu) +{ +#ifdef CONFIG_SMP + return cpu_logical_map(cpu); +#else + return get_ebase_cpunum(); +#endif +} + +struct core_chip_data { + struct mutex core_irq_mutex; + bool current_en; + bool desired_en; + u8 bit; +}; + +static struct core_chip_data irq_core_chip_data[8]; + +static void irq_core_ack(struct irq_data *data) +{ + struct core_chip_data *cd = irq_data_get_irq_chip_data(data); + unsigned int bit = cd->bit; + + /* + * We don't need to disable IRQs to make these atomic since + * they are already disabled earlier in the low level + * interrupt code. + */ + clear_c0_status(0x100 << bit); + /* The two user interrupts must be cleared manually. */ + if (bit < 2) + clear_c0_cause(0x100 << bit); +} + +static void irq_core_eoi(struct irq_data *data) +{ + struct core_chip_data *cd = irq_data_get_irq_chip_data(data); + + /* + * We don't need to disable IRQs to make these atomic since + * they are already disabled earlier in the low level + * interrupt code. + */ + set_c0_status(0x100 << cd->bit); +} + +static void irq_core_set_enable_local(void *arg) +{ + struct irq_data *data = arg; + struct core_chip_data *cd = irq_data_get_irq_chip_data(data); + unsigned int mask = 0x100 << cd->bit; + + /* + * Interrupts are already disabled, so these are atomic. + */ + if (cd->desired_en) + set_c0_status(mask); + else + clear_c0_status(mask); + +} + +static void irq_core_disable(struct irq_data *data) +{ + struct core_chip_data *cd = irq_data_get_irq_chip_data(data); + cd->desired_en = false; +} + +static void irq_core_enable(struct irq_data *data) +{ + struct core_chip_data *cd = irq_data_get_irq_chip_data(data); + cd->desired_en = true; +} + +static void irq_core_bus_lock(struct irq_data *data) +{ + struct core_chip_data *cd = irq_data_get_irq_chip_data(data); + + mutex_lock(&cd->core_irq_mutex); +} + +static void irq_core_bus_sync_unlock(struct irq_data *data) +{ + struct core_chip_data *cd = irq_data_get_irq_chip_data(data); + + if (cd->desired_en != cd->current_en) { + on_each_cpu(irq_core_set_enable_local, data, 1); + cd->current_en = cd->desired_en; + } + + mutex_unlock(&cd->core_irq_mutex); +} + +static struct irq_chip irq_chip_core = { + .name = "Core", + .irq_enable = irq_core_enable, + .irq_disable = irq_core_disable, + .irq_ack = irq_core_ack, + .irq_eoi = irq_core_eoi, + .irq_bus_lock = irq_core_bus_lock, + .irq_bus_sync_unlock = irq_core_bus_sync_unlock, + + .irq_cpu_online = irq_core_eoi, + .irq_cpu_offline = irq_core_ack, + .flags = IRQCHIP_ONOFFLINE_ENABLED, +}; + +static void __init irq_init_core(void) +{ + int i; + int irq; + struct core_chip_data *cd; + + /* Start with a clean slate */ + clear_c0_status(ST0_IM); + clear_c0_cause(CAUSEF_IP0 | CAUSEF_IP1); + + for (i = 0; i < ARRAY_SIZE(irq_core_chip_data); i++) { + cd = irq_core_chip_data + i; + cd->current_en = false; + cd->desired_en = false; + cd->bit = i; + mutex_init(&cd->core_irq_mutex); + + irq = MIPS_CPU_IRQ_BASE + i; + + switch (i) { + case 0: /* SW0 */ + case 1: /* SW1 */ + case 5: /* IP5 */ + case 6: /* IP6 */ + case 7: /* IP7 */ + irq_set_chip_data(irq, cd); + irq_set_chip_and_handler(irq, &irq_chip_core, + handle_percpu_irq); + break; + default: + break; + } + } +} + +static void __iomem *mips_irq_chip; +#define MIPS_IRQ_CHIP_NUM_BITS 0 +#define MIPS_IRQ_CHIP_REGS 8 + +static int mips_irq_cpu_stride; +static int mips_irq_chip_reg_raw; +static int mips_irq_chip_reg_src; +static int mips_irq_chip_reg_en; +static int mips_irq_chip_reg_raw_w1s; +static int mips_irq_chip_reg_raw_w1c; +static int mips_irq_chip_reg_en_w1s; +static int mips_irq_chip_reg_en_w1c; + +static void irq_pci_enable(struct irq_data *data) +{ + u32 mask = 1u << data->irq; + + __raw_writel(mask, mips_irq_chip + mips_irq_chip_reg_en_w1s); +} + +static void irq_pci_disable(struct irq_data *data) +{ + u32 mask = 1u << data->irq; + + __raw_writel(mask, mips_irq_chip + mips_irq_chip_reg_en_w1c); +} + +static void irq_pci_ack(struct irq_data *data) +{ +} + +static void irq_pci_mask(struct irq_data *data) +{ + u32 mask = 1u << data->irq; + + __raw_writel(mask, mips_irq_chip + mips_irq_chip_reg_en_w1c); +} + +static void irq_pci_unmask(struct irq_data *data) +{ + u32 mask = 1u << data->irq; + + __raw_writel(mask, mips_irq_chip + mips_irq_chip_reg_en_w1s); +} + +static struct irq_chip irq_chip_pci = { + .name = "PCI", + .irq_enable = irq_pci_enable, + .irq_disable = irq_pci_disable, + .irq_ack = irq_pci_ack, + .irq_mask = irq_pci_mask, + .irq_unmask = irq_pci_unmask, +}; + +static void irq_mbox_all(struct irq_data *data, void __iomem *base) +{ + int cpu; + unsigned int mbox = data->irq - MIPS_IRQ_MBOX0; + u32 mask; + + WARN_ON(mbox >= MBOX_BITS_PER_CPU); + + for_each_online_cpu(cpu) { + unsigned int cpuid = cpunum_for_cpu(cpu); + mask = 1 << (cpuid * MBOX_BITS_PER_CPU + mbox); + __raw_writel(mask, base + (cpuid * mips_irq_cpu_stride)); + } +} + +static void irq_mbox_enable(struct irq_data *data) +{ + irq_mbox_all(data, mips_irq_chip + mips_irq_chip_reg_en_w1s + sizeof(u32)); +} + +static void irq_mbox_disable(struct irq_data *data) +{ + irq_mbox_all(data, mips_irq_chip + mips_irq_chip_reg_en_w1c + sizeof(u32)); +} + +static void irq_mbox_ack(struct irq_data *data) +{ + u32 mask; + unsigned int mbox = data->irq - MIPS_IRQ_MBOX0; + + WARN_ON(mbox >= MBOX_BITS_PER_CPU); + + mask = 1 << (get_ebase_cpunum() * MBOX_BITS_PER_CPU + mbox); + __raw_writel(mask, mips_irq_chip + mips_irq_chip_reg_raw_w1c + sizeof(u32)); +} + +void irq_mbox_ipi(int cpu, unsigned int actions) +{ + unsigned int cpuid = cpunum_for_cpu(cpu); + u32 mask; + + WARN_ON(actions >= (1 << MBOX_BITS_PER_CPU)); + + mask = actions << (cpuid * MBOX_BITS_PER_CPU); + __raw_writel(mask, mips_irq_chip + mips_irq_chip_reg_raw_w1s + sizeof(u32)); +} + +static void irq_mbox_cpu_onoffline(struct irq_data *data, void __iomem *base) +{ + unsigned int mbox = data->irq - MIPS_IRQ_MBOX0; + unsigned int cpuid = get_ebase_cpunum(); + u32 mask; + + WARN_ON(mbox >= MBOX_BITS_PER_CPU); + + mask = 1 << (cpuid * MBOX_BITS_PER_CPU + mbox); + __raw_writel(mask, base + (cpuid * mips_irq_cpu_stride)); + +} + +static void irq_mbox_cpu_online(struct irq_data *data) +{ + irq_mbox_cpu_onoffline(data, mips_irq_chip + mips_irq_chip_reg_en_w1s + sizeof(u32)); +} + +static void irq_mbox_cpu_offline(struct irq_data *data) +{ + irq_mbox_cpu_onoffline(data, mips_irq_chip + mips_irq_chip_reg_en_w1c + sizeof(u32)); +} + +static struct irq_chip irq_chip_mbox = { + .name = "MBOX", + .irq_enable = irq_mbox_enable, + .irq_disable = irq_mbox_disable, + .irq_ack = irq_mbox_ack, + .irq_cpu_online = irq_mbox_cpu_online, + .irq_cpu_offline = irq_mbox_cpu_offline, + .flags = IRQCHIP_ONOFFLINE_ENABLED, +}; + +static void __init irq_pci_init(void) +{ + int i, stride; + u32 num_bits; + + mips_irq_chip = ioremap(0x1e010000, 4096); + + num_bits = __raw_readl(mips_irq_chip + MIPS_IRQ_CHIP_NUM_BITS); + stride = 8 * (1 + ((num_bits - 1) / 64)); + + + pr_notice("mips_irq_chip: %u bits, reg stride: %d\n", num_bits, stride); + mips_irq_chip_reg_raw = MIPS_IRQ_CHIP_REGS + 0 * stride; + mips_irq_chip_reg_raw_w1s = MIPS_IRQ_CHIP_REGS + 1 * stride; + mips_irq_chip_reg_raw_w1c = MIPS_IRQ_CHIP_REGS + 2 * stride; + mips_irq_chip_reg_src = MIPS_IRQ_CHIP_REGS + 3 * stride; + mips_irq_chip_reg_en = MIPS_IRQ_CHIP_REGS + 4 * stride; + mips_irq_chip_reg_en_w1s = MIPS_IRQ_CHIP_REGS + 5 * stride; + mips_irq_chip_reg_en_w1c = MIPS_IRQ_CHIP_REGS + 6 * stride; + mips_irq_cpu_stride = stride * 4; + + for (i = 0; i < 4; i++) + irq_set_chip_and_handler(i + MIPS_IRQ_PCIA, &irq_chip_pci, handle_level_irq); + + for (i = 0; i < 2; i++) + irq_set_chip_and_handler(i + MIPS_IRQ_MBOX0, &irq_chip_mbox, handle_percpu_irq); + + + set_c0_status(STATUSF_IP2); +} + +static void irq_pci_dispatch(void) +{ + unsigned int cpuid = get_ebase_cpunum(); + u32 en; + + en = __raw_readl(mips_irq_chip + mips_irq_chip_reg_src + + (cpuid * mips_irq_cpu_stride)); + + if (!en) { + en = __raw_readl(mips_irq_chip + mips_irq_chip_reg_src + (cpuid * mips_irq_cpu_stride) + sizeof(u32)); + en = (en >> (2 * cpuid)) & 3; + + if (!en) + spurious_interrupt(); + else + do_IRQ(__ffs(en) + MIPS_IRQ_MBOX0); /* MBOX type */ + } else { + do_IRQ(__ffs(en)); + } +} + + +void __init arch_init_irq(void) +{ + irq_init_core(); + irq_pci_init(); +} + +asmlinkage void plat_irq_dispatch(void) +{ + unsigned int pending = read_c0_cause() & read_c0_status() & ST0_IM; + int ip; + + if (unlikely(!pending)) { + spurious_interrupt(); + return; + } + + ip = ffs(pending) - 1 - STATUSB_IP0; + if (ip == 2) + irq_pci_dispatch(); + else + do_IRQ(MIPS_CPU_IRQ_BASE + ip); +} diff --git a/arch/mips/paravirt/paravirt-smp.c b/arch/mips/paravirt/paravirt-smp.c new file mode 100644 index 00000000000..0164b0c4835 --- /dev/null +++ b/arch/mips/paravirt/paravirt-smp.c @@ -0,0 +1,143 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2013 Cavium, Inc. + */ + +#include <linux/interrupt.h> +#include <linux/cpumask.h> +#include <linux/kernel.h> +#include <linux/sched.h> + +#include <asm/mipsregs.h> +#include <asm/setup.h> +#include <asm/time.h> +#include <asm/smp.h> + +/* + * Writing the sp releases the CPU, so writes must be ordered, gp + * first, then sp. + */ +unsigned long paravirt_smp_sp[NR_CPUS]; +unsigned long paravirt_smp_gp[NR_CPUS]; + +static int numcpus = 1; + +static int __init set_numcpus(char *str) +{ + int newval; + + if (get_option(&str, &newval)) { + if (newval < 1 || newval >= NR_CPUS) + goto bad; + numcpus = newval; + return 0; + } +bad: + return -EINVAL; +} +early_param("numcpus", set_numcpus); + + +static void paravirt_smp_setup(void) +{ + int id; + unsigned int cpunum = get_ebase_cpunum(); + + if (WARN_ON(cpunum >= NR_CPUS)) + return; + + /* The present CPUs are initially just the boot cpu (CPU 0). */ + for (id = 0; id < NR_CPUS; id++) { + set_cpu_possible(id, id == 0); + set_cpu_present(id, id == 0); + } + __cpu_number_map[cpunum] = 0; + __cpu_logical_map[0] = cpunum; + + for (id = 0; id < numcpus; id++) { + set_cpu_possible(id, true); + set_cpu_present(id, true); + __cpu_number_map[id] = id; + __cpu_logical_map[id] = id; + } +} + +void irq_mbox_ipi(int cpu, unsigned int actions); +static void paravirt_send_ipi_single(int cpu, unsigned int action) +{ + irq_mbox_ipi(cpu, action); +} + +static void paravirt_send_ipi_mask(const struct cpumask *mask, unsigned int action) +{ + unsigned int cpu; + + for_each_cpu_mask(cpu, *mask) + paravirt_send_ipi_single(cpu, action); +} + +static void paravirt_init_secondary(void) +{ + unsigned int sr; + + sr = set_c0_status(ST0_BEV); + write_c0_ebase((u32)ebase); + + sr |= STATUSF_IP2; /* Interrupt controller on IP2 */ + write_c0_status(sr); + + irq_cpu_online(); +} + +static void paravirt_smp_finish(void) +{ + /* to generate the first CPU timer interrupt */ + write_c0_compare(read_c0_count() + mips_hpt_frequency / HZ); + local_irq_enable(); +} + +static void paravirt_boot_secondary(int cpu, struct task_struct *idle) +{ + paravirt_smp_gp[cpu] = (unsigned long)task_thread_info(idle); + smp_wmb(); + paravirt_smp_sp[cpu] = __KSTK_TOS(idle); +} + +static irqreturn_t paravirt_reched_interrupt(int irq, void *dev_id) +{ + scheduler_ipi(); + return IRQ_HANDLED; +} + +static irqreturn_t paravirt_function_interrupt(int irq, void *dev_id) +{ + smp_call_function_interrupt(); + return IRQ_HANDLED; +} + +static void paravirt_prepare_cpus(unsigned int max_cpus) +{ + if (request_irq(MIPS_IRQ_MBOX0, paravirt_reched_interrupt, + IRQF_PERCPU | IRQF_NO_THREAD, "Scheduler", + paravirt_reched_interrupt)) { + panic("Cannot request_irq for SchedulerIPI"); + } + if (request_irq(MIPS_IRQ_MBOX1, paravirt_function_interrupt, + IRQF_PERCPU | IRQF_NO_THREAD, "SMP-Call", + paravirt_function_interrupt)) { + panic("Cannot request_irq for SMP-Call"); + } +} + +struct plat_smp_ops paravirt_smp_ops = { + .send_ipi_single = paravirt_send_ipi_single, + .send_ipi_mask = paravirt_send_ipi_mask, + .init_secondary = paravirt_init_secondary, + .smp_finish = paravirt_smp_finish, + .boot_secondary = paravirt_boot_secondary, + .smp_setup = paravirt_smp_setup, + .prepare_cpus = paravirt_prepare_cpus, +}; diff --git a/arch/mips/paravirt/serial.c b/arch/mips/paravirt/serial.c new file mode 100644 index 00000000000..02b665c0227 --- /dev/null +++ b/arch/mips/paravirt/serial.c @@ -0,0 +1,40 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2013 Cavium, Inc. + */ + +#include <linux/kernel.h> +#include <linux/virtio_console.h> +#include <linux/kvm_para.h> + +/* + * Emit one character to the boot console. + */ +int prom_putchar(char c) +{ + kvm_hypercall3(KVM_HC_MIPS_CONSOLE_OUTPUT, 0 /* port 0 */, + (unsigned long)&c, 1 /* len == 1 */); + + return 1; +} + +#ifdef CONFIG_VIRTIO_CONSOLE +static int paravirt_put_chars(u32 vtermno, const char *buf, int count) +{ + kvm_hypercall3(KVM_HC_MIPS_CONSOLE_OUTPUT, vtermno, + (unsigned long)buf, count); + + return count; +} + +static int __init paravirt_cons_init(void) +{ + virtio_cons_early_init(paravirt_put_chars); + return 0; +} +core_initcall(paravirt_cons_init); + +#endif diff --git a/arch/mips/paravirt/setup.c b/arch/mips/paravirt/setup.c new file mode 100644 index 00000000000..cb8448b373a --- /dev/null +++ b/arch/mips/paravirt/setup.c @@ -0,0 +1,67 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2013 Cavium, Inc. + */ + +#include <linux/kernel.h> +#include <linux/kvm_para.h> + +#include <asm/reboot.h> +#include <asm/bootinfo.h> +#include <asm/smp-ops.h> +#include <asm/time.h> + +extern struct plat_smp_ops paravirt_smp_ops; + +const char *get_system_type(void) +{ + return "MIPS Para-Virtualized Guest"; +} + +void __init plat_time_init(void) +{ + mips_hpt_frequency = kvm_hypercall0(KVM_HC_MIPS_GET_CLOCK_FREQ); + + preset_lpj = mips_hpt_frequency / (2 * HZ); +} + +static void pv_machine_halt(void) +{ + kvm_hypercall0(KVM_HC_MIPS_EXIT_VM); +} + +/* + * Early entry point for arch setup + */ +void __init prom_init(void) +{ + int i; + int argc = fw_arg0; + char **argv = (char **)fw_arg1; + +#ifdef CONFIG_32BIT + set_io_port_base(KSEG1ADDR(0x1e000000)); +#else /* CONFIG_64BIT */ + set_io_port_base(PHYS_TO_XKSEG_UNCACHED(0x1e000000)); +#endif + + for (i = 0; i < argc; i++) { + strlcat(arcs_cmdline, argv[i], COMMAND_LINE_SIZE); + if (i < argc - 1) + strlcat(arcs_cmdline, " ", COMMAND_LINE_SIZE); + } + _machine_halt = pv_machine_halt; + register_smp_ops(¶virt_smp_ops); +} + +void __init plat_mem_setup(void) +{ + /* Do nothing, the "mem=???" parser handles our memory. */ +} + +void __init prom_free_prom_memory(void) +{ +} diff --git a/arch/mips/pci/Makefile b/arch/mips/pci/Makefile index d61138a177c..ff8a5539b36 100644 --- a/arch/mips/pci/Makefile +++ b/arch/mips/pci/Makefile @@ -21,7 +21,7 @@ obj-$(CONFIG_BCM63XX) += pci-bcm63xx.o fixup-bcm63xx.o \ obj-$(CONFIG_MIPS_ALCHEMY) += pci-alchemy.o obj-$(CONFIG_SOC_AR71XX) += pci-ar71xx.o obj-$(CONFIG_PCI_AR724X) += pci-ar724x.o - +obj-$(CONFIG_MIPS_PCI_VIRTIO) += pci-virtio-guest.o # # These are still pretty much in the old state, watch, go blind. # diff --git a/arch/mips/pci/fixup-malta.c b/arch/mips/pci/fixup-malta.c index 2f9e52a1a75..40e920c653c 100644 --- a/arch/mips/pci/fixup-malta.c +++ b/arch/mips/pci/fixup-malta.c @@ -68,6 +68,7 @@ static void malta_piix_func0_fixup(struct pci_dev *pdev) { unsigned char reg_val; u32 reg_val32; + u16 reg_val16; /* PIIX PIRQC[A:D] irq mappings */ static int piixirqmap[PIIX4_FUNC0_PIRQRC_IRQ_ROUTING_MAX] = { 0, 0, 0, 3, @@ -107,6 +108,11 @@ static void malta_piix_func0_fixup(struct pci_dev *pdev) pci_read_config_byte(pdev, PIIX4_FUNC0_SERIRQC, ®_val); reg_val |= PIIX4_FUNC0_SERIRQC_EN | PIIX4_FUNC0_SERIRQC_CONT; pci_write_config_byte(pdev, PIIX4_FUNC0_SERIRQC, reg_val); + + /* Enable response to special cycles */ + pci_read_config_word(pdev, PCI_COMMAND, ®_val16); + pci_write_config_word(pdev, PCI_COMMAND, + reg_val16 | PCI_COMMAND_SPECIAL); } DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82371AB_0, diff --git a/arch/mips/pci/msi-octeon.c b/arch/mips/pci/msi-octeon.c index 2b91b0e6156..ab0c5d14c6f 100644 --- a/arch/mips/pci/msi-octeon.c +++ b/arch/mips/pci/msi-octeon.c @@ -15,6 +15,7 @@ #include <asm/octeon/cvmx-npi-defs.h> #include <asm/octeon/cvmx-pci-defs.h> #include <asm/octeon/cvmx-npei-defs.h> +#include <asm/octeon/cvmx-sli-defs.h> #include <asm/octeon/cvmx-pexp-defs.h> #include <asm/octeon/pci-octeon.h> @@ -162,6 +163,11 @@ msi_irq_allocated: msg.address_lo = (0 + CVMX_NPEI_PCIE_MSI_RCV) & 0xffffffff; msg.address_hi = (0 + CVMX_NPEI_PCIE_MSI_RCV) >> 32; break; + case OCTEON_DMA_BAR_TYPE_PCIE2: + /* When using PCIe2, Bar 0 is based at 0 */ + msg.address_lo = (0 + CVMX_SLI_PCIE_MSI_RCV) & 0xffffffff; + msg.address_hi = (0 + CVMX_SLI_PCIE_MSI_RCV) >> 32; + break; default: panic("arch_setup_msi_irq: Invalid octeon_dma_bar_type"); } diff --git a/arch/mips/pci/msi-xlp.c b/arch/mips/pci/msi-xlp.c index 3249685e03a..fa374fe3746 100644 --- a/arch/mips/pci/msi-xlp.c +++ b/arch/mips/pci/msi-xlp.c @@ -56,8 +56,8 @@ #include <asm/netlogic/xlp-hal/bridge.h> #define XLP_MSIVEC_PER_LINK 32 -#define XLP_MSIXVEC_TOTAL 32 -#define XLP_MSIXVEC_PER_LINK 8 +#define XLP_MSIXVEC_TOTAL (cpu_is_xlp9xx() ? 128 : 32) +#define XLP_MSIXVEC_PER_LINK (cpu_is_xlp9xx() ? 32 : 8) /* 128 MSI irqs per node, mapped starting at NLM_MSI_VEC_BASE */ static inline int nlm_link_msiirq(int link, int msivec) @@ -65,35 +65,44 @@ static inline int nlm_link_msiirq(int link, int msivec) return NLM_MSI_VEC_BASE + link * XLP_MSIVEC_PER_LINK + msivec; } +/* get the link MSI vector from irq number */ static inline int nlm_irq_msivec(int irq) { - return irq % XLP_MSIVEC_PER_LINK; + return (irq - NLM_MSI_VEC_BASE) % XLP_MSIVEC_PER_LINK; } +/* get the link from the irq number */ static inline int nlm_irq_msilink(int irq) { - return (irq % (XLP_MSIVEC_PER_LINK * PCIE_NLINKS)) / - XLP_MSIVEC_PER_LINK; + int total_msivec = XLP_MSIVEC_PER_LINK * PCIE_NLINKS; + + return ((irq - NLM_MSI_VEC_BASE) % total_msivec) / + XLP_MSIVEC_PER_LINK; } /* - * Only 32 MSI-X vectors are possible because there are only 32 PIC - * interrupts for MSI. We split them statically and use 8 MSI-X vectors - * per link - this keeps the allocation and lookup simple. + * For XLP 8xx/4xx/3xx/2xx, only 32 MSI-X vectors are possible because + * there are only 32 PIC interrupts for MSI. We split them statically + * and use 8 MSI-X vectors per link - this keeps the allocation and + * lookup simple. + * On XLP 9xx, there are 32 vectors per link, and the interrupts are + * not routed thru PIC, so we can use all 128 MSI-X vectors. */ static inline int nlm_link_msixirq(int link, int bit) { return NLM_MSIX_VEC_BASE + link * XLP_MSIXVEC_PER_LINK + bit; } +/* get the link MSI vector from irq number */ static inline int nlm_irq_msixvec(int irq) { - return irq % XLP_MSIXVEC_TOTAL; /* works when given xirq */ + return (irq - NLM_MSIX_VEC_BASE) % XLP_MSIXVEC_TOTAL; } -static inline int nlm_irq_msixlink(int irq) +/* get the link from MSIX vec */ +static inline int nlm_irq_msixlink(int msixvec) { - return nlm_irq_msixvec(irq) / XLP_MSIXVEC_PER_LINK; + return msixvec / XLP_MSIXVEC_PER_LINK; } /* @@ -129,7 +138,11 @@ static void xlp_msi_enable(struct irq_data *d) vec = nlm_irq_msivec(d->irq); spin_lock_irqsave(&md->msi_lock, flags); md->msi_enabled_mask |= 1u << vec; - nlm_write_reg(md->lnkbase, PCIE_MSI_EN, md->msi_enabled_mask); + if (cpu_is_xlp9xx()) + nlm_write_reg(md->lnkbase, PCIE_9XX_MSI_EN, + md->msi_enabled_mask); + else + nlm_write_reg(md->lnkbase, PCIE_MSI_EN, md->msi_enabled_mask); spin_unlock_irqrestore(&md->msi_lock, flags); } @@ -142,7 +155,11 @@ static void xlp_msi_disable(struct irq_data *d) vec = nlm_irq_msivec(d->irq); spin_lock_irqsave(&md->msi_lock, flags); md->msi_enabled_mask &= ~(1u << vec); - nlm_write_reg(md->lnkbase, PCIE_MSI_EN, md->msi_enabled_mask); + if (cpu_is_xlp9xx()) + nlm_write_reg(md->lnkbase, PCIE_9XX_MSI_EN, + md->msi_enabled_mask); + else + nlm_write_reg(md->lnkbase, PCIE_MSI_EN, md->msi_enabled_mask); spin_unlock_irqrestore(&md->msi_lock, flags); } @@ -156,11 +173,18 @@ static void xlp_msi_mask_ack(struct irq_data *d) xlp_msi_disable(d); /* Ack MSI on bridge */ - nlm_write_reg(md->lnkbase, PCIE_MSI_STATUS, 1u << vec); + if (cpu_is_xlp9xx()) + nlm_write_reg(md->lnkbase, PCIE_9XX_MSI_STATUS, 1u << vec); + else + nlm_write_reg(md->lnkbase, PCIE_MSI_STATUS, 1u << vec); /* Ack at eirr and PIC */ ack_c0_eirr(PIC_PCIE_LINK_MSI_IRQ(link)); - nlm_pic_ack(md->node->picbase, PIC_IRT_PCIE_LINK_INDEX(link)); + if (cpu_is_xlp9xx()) + nlm_pic_ack(md->node->picbase, + PIC_9XX_IRT_PCIE_LINK_INDEX(link)); + else + nlm_pic_ack(md->node->picbase, PIC_IRT_PCIE_LINK_INDEX(link)); } static struct irq_chip xlp_msi_chip = { @@ -172,30 +196,45 @@ static struct irq_chip xlp_msi_chip = { }; /* - * The MSI-X interrupt handling is different from MSI, there are 32 - * MSI-X interrupts generated by the PIC and each of these correspond - * to a MSI-X vector (0-31) that can be assigned. + * XLP8XX/4XX/3XX/2XX: + * The MSI-X interrupt handling is different from MSI, there are 32 MSI-X + * interrupts generated by the PIC and each of these correspond to a MSI-X + * vector (0-31) that can be assigned. * - * We divide the MSI-X vectors to 8 per link and do a per-link - * allocation + * We divide the MSI-X vectors to 8 per link and do a per-link allocation + * + * XLP9XX: + * 32 MSI-X vectors are available per link, and the interrupts are not routed + * thru the PIC. PIC ack not needed. * * Enable and disable done using standard MSI functions. */ static void xlp_msix_mask_ack(struct irq_data *d) { - struct xlp_msi_data *md = irq_data_get_irq_handler_data(d); + struct xlp_msi_data *md; int link, msixvec; + uint32_t status_reg, bit; msixvec = nlm_irq_msixvec(d->irq); - link = nlm_irq_msixlink(d->irq); + link = nlm_irq_msixlink(msixvec); mask_msi_irq(d); + md = irq_data_get_irq_handler_data(d); /* Ack MSI on bridge */ - nlm_write_reg(md->lnkbase, PCIE_MSIX_STATUS, 1u << msixvec); + if (cpu_is_xlp9xx()) { + status_reg = PCIE_9XX_MSIX_STATUSX(link); + bit = msixvec % XLP_MSIXVEC_PER_LINK; + } else { + status_reg = PCIE_MSIX_STATUS; + bit = msixvec; + } + nlm_write_reg(md->lnkbase, status_reg, 1u << bit); /* Ack at eirr and PIC */ ack_c0_eirr(PIC_PCIE_MSIX_IRQ(link)); - nlm_pic_ack(md->node->picbase, PIC_IRT_PCIE_MSIX_INDEX(msixvec)); + if (!cpu_is_xlp9xx()) + nlm_pic_ack(md->node->picbase, + PIC_IRT_PCIE_MSIX_INDEX(msixvec)); } static struct irq_chip xlp_msix_chip = { @@ -219,10 +258,18 @@ static void xlp_config_link_msi(uint64_t lnkbase, int lirq, uint64_t msiaddr) { u32 val; - val = nlm_read_reg(lnkbase, PCIE_INT_EN0); - if ((val & 0x200) == 0) { - val |= 0x200; /* MSI Interrupt enable */ - nlm_write_reg(lnkbase, PCIE_INT_EN0, val); + if (cpu_is_xlp9xx()) { + val = nlm_read_reg(lnkbase, PCIE_9XX_INT_EN0); + if ((val & 0x200) == 0) { + val |= 0x200; /* MSI Interrupt enable */ + nlm_write_reg(lnkbase, PCIE_9XX_INT_EN0, val); + } + } else { + val = nlm_read_reg(lnkbase, PCIE_INT_EN0); + if ((val & 0x200) == 0) { + val |= 0x200; + nlm_write_reg(lnkbase, PCIE_INT_EN0, val); + } } val = nlm_read_reg(lnkbase, 0x1); /* CMD */ @@ -269,9 +316,12 @@ static int xlp_setup_msi(uint64_t lnkbase, int node, int link, spin_lock_irqsave(&md->msi_lock, flags); if (md->msi_alloc_mask == 0) { - /* switch the link IRQ to MSI range */ xlp_config_link_msi(lnkbase, lirq, msiaddr); - irt = PIC_IRT_PCIE_LINK_INDEX(link); + /* switch the link IRQ to MSI range */ + if (cpu_is_xlp9xx()) + irt = PIC_9XX_IRT_PCIE_LINK_INDEX(link); + else + irt = PIC_IRT_PCIE_LINK_INDEX(link); nlm_setup_pic_irq(node, lirq, lirq, irt); nlm_pic_init_irt(nlm_get_node(node)->picbase, irt, lirq, node * nlm_threads_per_node(), 1 /*en */); @@ -311,10 +361,19 @@ static void xlp_config_link_msix(uint64_t lnkbase, int lirq, uint64_t msixaddr) val |= 0x80000000U; nlm_write_reg(lnkbase, 0x2C, val); } - val = nlm_read_reg(lnkbase, PCIE_INT_EN0); - if ((val & 0x200) == 0) { - val |= 0x200; /* MSI Interrupt enable */ - nlm_write_reg(lnkbase, PCIE_INT_EN0, val); + + if (cpu_is_xlp9xx()) { + val = nlm_read_reg(lnkbase, PCIE_9XX_INT_EN0); + if ((val & 0x200) == 0) { + val |= 0x200; /* MSI Interrupt enable */ + nlm_write_reg(lnkbase, PCIE_9XX_INT_EN0, val); + } + } else { + val = nlm_read_reg(lnkbase, PCIE_INT_EN0); + if ((val & 0x200) == 0) { + val |= 0x200; /* MSI Interrupt enable */ + nlm_write_reg(lnkbase, PCIE_INT_EN0, val); + } } val = nlm_read_reg(lnkbase, 0x1); /* CMD */ @@ -329,10 +388,19 @@ static void xlp_config_link_msix(uint64_t lnkbase, int lirq, uint64_t msixaddr) val |= (1 << 8) | lirq; nlm_write_pci_reg(lnkbase, 0xf, val); - /* MSI-X addresses */ - nlm_write_reg(lnkbase, PCIE_BRIDGE_MSIX_ADDR_BASE, msixaddr >> 8); - nlm_write_reg(lnkbase, PCIE_BRIDGE_MSIX_ADDR_LIMIT, - (msixaddr + MSI_ADDR_SZ) >> 8); + if (cpu_is_xlp9xx()) { + /* MSI-X addresses */ + nlm_write_reg(lnkbase, PCIE_9XX_BRIDGE_MSIX_ADDR_BASE, + msixaddr >> 8); + nlm_write_reg(lnkbase, PCIE_9XX_BRIDGE_MSIX_ADDR_LIMIT, + (msixaddr + MSI_ADDR_SZ) >> 8); + } else { + /* MSI-X addresses */ + nlm_write_reg(lnkbase, PCIE_BRIDGE_MSIX_ADDR_BASE, + msixaddr >> 8); + nlm_write_reg(lnkbase, PCIE_BRIDGE_MSIX_ADDR_LIMIT, + (msixaddr + MSI_ADDR_SZ) >> 8); + } } /* @@ -369,6 +437,7 @@ static int xlp_setup_msix(uint64_t lnkbase, int node, int link, xirq += t; msixvec = nlm_irq_msixvec(xirq); + msg.address_hi = msixaddr >> 32; msg.address_lo = msixaddr & 0xffffffff; msg.data = 0xc00 | msixvec; @@ -409,7 +478,7 @@ void __init xlp_init_node_msi_irqs(int node, int link) { struct nlm_soc_info *nodep; struct xlp_msi_data *md; - int irq, i, irt, msixvec; + int irq, i, irt, msixvec, val; pr_info("[%d %d] Init node PCI IRT\n", node, link); nodep = nlm_get_node(node); @@ -430,19 +499,28 @@ void __init xlp_init_node_msi_irqs(int node, int link) irq_set_handler_data(i, md); } - for (i = 0; i < XLP_MSIXVEC_PER_LINK; i++) { - /* Initialize MSI-X irts to generate one interrupt per link */ - msixvec = link * XLP_MSIXVEC_PER_LINK + i; - irt = PIC_IRT_PCIE_MSIX_INDEX(msixvec); - nlm_pic_init_irt(nodep->picbase, irt, PIC_PCIE_MSIX_IRQ(link), - node * nlm_threads_per_node(), 1 /* enable */); + for (i = 0; i < XLP_MSIXVEC_PER_LINK ; i++) { + if (cpu_is_xlp9xx()) { + val = ((node * nlm_threads_per_node()) << 7 | + PIC_PCIE_MSIX_IRQ(link) << 1 | 0 << 0); + nlm_write_pcie_reg(md->lnkbase, PCIE_9XX_MSIX_VECX(i + + (link * XLP_MSIXVEC_PER_LINK)), val); + } else { + /* Initialize MSI-X irts to generate one interrupt + * per link + */ + msixvec = link * XLP_MSIXVEC_PER_LINK + i; + irt = PIC_IRT_PCIE_MSIX_INDEX(msixvec); + nlm_pic_init_irt(nodep->picbase, irt, + PIC_PCIE_MSIX_IRQ(link), + node * nlm_threads_per_node(), 1); + } /* Initialize MSI-X extended irq space for the link */ irq = nlm_irq_to_xirq(node, nlm_link_msixirq(link, i)); irq_set_chip_and_handler(irq, &xlp_msix_chip, handle_level_irq); irq_set_handler_data(irq, md); } - } void nlm_dispatch_msi(int node, int lirq) @@ -454,7 +532,11 @@ void nlm_dispatch_msi(int node, int lirq) link = lirq - PIC_PCIE_LINK_MSI_IRQ_BASE; irqbase = nlm_irq_to_xirq(node, nlm_link_msiirq(link, 0)); md = irq_get_handler_data(irqbase); - status = nlm_read_reg(md->lnkbase, PCIE_MSI_STATUS) & + if (cpu_is_xlp9xx()) + status = nlm_read_reg(md->lnkbase, PCIE_9XX_MSI_STATUS) & + md->msi_enabled_mask; + else + status = nlm_read_reg(md->lnkbase, PCIE_MSI_STATUS) & md->msi_enabled_mask; while (status) { i = __ffs(status); @@ -472,10 +554,14 @@ void nlm_dispatch_msix(int node, int lirq) link = lirq - PIC_PCIE_MSIX_IRQ_BASE; irqbase = nlm_irq_to_xirq(node, nlm_link_msixirq(link, 0)); md = irq_get_handler_data(irqbase); - status = nlm_read_reg(md->lnkbase, PCIE_MSIX_STATUS); + if (cpu_is_xlp9xx()) + status = nlm_read_reg(md->lnkbase, PCIE_9XX_MSIX_STATUSX(link)); + else + status = nlm_read_reg(md->lnkbase, PCIE_MSIX_STATUS); /* narrow it down to the MSI-x vectors for our link */ - status = (status >> (link * XLP_MSIXVEC_PER_LINK)) & + if (!cpu_is_xlp9xx()) + status = (status >> (link * XLP_MSIXVEC_PER_LINK)) & ((1 << XLP_MSIXVEC_PER_LINK) - 1); while (status) { diff --git a/arch/mips/pci/ops-pmcmsp.c b/arch/mips/pci/ops-pmcmsp.c index 3d27800edba..50034f985be 100644 --- a/arch/mips/pci/ops-pmcmsp.c +++ b/arch/mips/pci/ops-pmcmsp.c @@ -7,7 +7,7 @@ * Author: Jun Sun, jsun@mvista.com or jsun@junsun.net * * Much of the code is derived from the original DDB5074 port by - * Geert Uytterhoeven <geert@sonycom.com> + * Geert Uytterhoeven <geert@linux-m68k.org> * * This program is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License as published by the diff --git a/arch/mips/pci/ops-tx3927.c b/arch/mips/pci/ops-tx3927.c index 02d64f77e96..d35dc9c9ab9 100644 --- a/arch/mips/pci/ops-tx3927.c +++ b/arch/mips/pci/ops-tx3927.c @@ -11,7 +11,7 @@ * Define the pci_ops for TX3927. * * Much of the code is derived from the original DDB5074 port by - * Geert Uytterhoeven <geert@sonycom.com> + * Geert Uytterhoeven <geert@linux-m68k.org> * * This program is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License as published by the diff --git a/arch/mips/pci/ops-tx4927.c b/arch/mips/pci/ops-tx4927.c index 3d5df514d02..0e046d82e4e 100644 --- a/arch/mips/pci/ops-tx4927.c +++ b/arch/mips/pci/ops-tx4927.c @@ -202,17 +202,20 @@ char *tx4927_pcibios_setup(char *str) unsigned long val; if (!strncmp(str, "trdyto=", 7)) { - if (strict_strtoul(str + 7, 0, &val) == 0) + u8 val = 0; + if (kstrtou8(str + 7, 0, &val) == 0) tx4927_pci_opts.trdyto = val; return NULL; } if (!strncmp(str, "retryto=", 8)) { - if (strict_strtoul(str + 8, 0, &val) == 0) + u8 val = 0; + if (kstrtou8(str + 8, 0, &val) == 0) tx4927_pci_opts.retryto = val; return NULL; } if (!strncmp(str, "gbwc=", 5)) { - if (strict_strtoul(str + 5, 0, &val) == 0) + u16 val; + if (kstrtou16(str + 5, 0, &val) == 0) tx4927_pci_opts.gbwc = val; return NULL; } diff --git a/arch/mips/pci/pci-virtio-guest.c b/arch/mips/pci/pci-virtio-guest.c new file mode 100644 index 00000000000..40a078bc461 --- /dev/null +++ b/arch/mips/pci/pci-virtio-guest.c @@ -0,0 +1,131 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2013 Cavium, Inc. + */ + +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/interrupt.h> +#include <linux/pci.h> + +#include <uapi/asm/bitfield.h> +#include <asm/byteorder.h> +#include <asm/io.h> + +#define PCI_CONFIG_ADDRESS 0xcf8 +#define PCI_CONFIG_DATA 0xcfc + +union pci_config_address { + struct { + __BITFIELD_FIELD(unsigned enable_bit : 1, /* 31 */ + __BITFIELD_FIELD(unsigned reserved : 7, /* 30 .. 24 */ + __BITFIELD_FIELD(unsigned bus_number : 8, /* 23 .. 16 */ + __BITFIELD_FIELD(unsigned devfn_number : 8, /* 15 .. 8 */ + __BITFIELD_FIELD(unsigned register_number : 8, /* 7 .. 0 */ + ))))); + }; + u32 w; +}; + +int pcibios_plat_dev_init(struct pci_dev *dev) +{ + return 0; +} + +int pcibios_map_irq(const struct pci_dev *dev, u8 slot, u8 pin) +{ + return ((pin + slot) % 4)+ MIPS_IRQ_PCIA; +} + +static void pci_virtio_guest_write_config_addr(struct pci_bus *bus, + unsigned int devfn, int reg) +{ + union pci_config_address pca = { .w = 0 }; + + pca.register_number = reg; + pca.devfn_number = devfn; + pca.bus_number = bus->number; + pca.enable_bit = 1; + + outl(pca.w, PCI_CONFIG_ADDRESS); +} + +static int pci_virtio_guest_write_config(struct pci_bus *bus, + unsigned int devfn, int reg, int size, u32 val) +{ + pci_virtio_guest_write_config_addr(bus, devfn, reg); + + switch (size) { + case 1: + outb(val, PCI_CONFIG_DATA + (reg & 3)); + break; + case 2: + outw(val, PCI_CONFIG_DATA + (reg & 2)); + break; + case 4: + outl(val, PCI_CONFIG_DATA); + break; + } + + return PCIBIOS_SUCCESSFUL; +} + +static int pci_virtio_guest_read_config(struct pci_bus *bus, unsigned int devfn, + int reg, int size, u32 *val) +{ + pci_virtio_guest_write_config_addr(bus, devfn, reg); + + switch (size) { + case 1: + *val = inb(PCI_CONFIG_DATA + (reg & 3)); + break; + case 2: + *val = inw(PCI_CONFIG_DATA + (reg & 2)); + break; + case 4: + *val = inl(PCI_CONFIG_DATA); + break; + } + return PCIBIOS_SUCCESSFUL; +} + +static struct pci_ops pci_virtio_guest_ops = { + .read = pci_virtio_guest_read_config, + .write = pci_virtio_guest_write_config, +}; + +static struct resource pci_virtio_guest_mem_resource = { + .name = "Virtio MEM", + .flags = IORESOURCE_MEM, + .start = 0x10000000, + .end = 0x1dffffff +}; + +static struct resource pci_virtio_guest_io_resource = { + .name = "Virtio IO", + .flags = IORESOURCE_IO, + .start = 0, + .end = 0xffff +}; + +static struct pci_controller pci_virtio_guest_controller = { + .pci_ops = &pci_virtio_guest_ops, + .mem_resource = &pci_virtio_guest_mem_resource, + .io_resource = &pci_virtio_guest_io_resource, +}; + +static int __init pci_virtio_guest_setup(void) +{ + pr_err("pci_virtio_guest_setup\n"); + + /* Virtio comes pre-assigned */ + pci_set_flags(PCI_PROBE_ONLY); + + pci_virtio_guest_controller.io_map_base = mips_io_port_base; + register_pci_controller(&pci_virtio_guest_controller); + return 0; +} +arch_initcall(pci_virtio_guest_setup); diff --git a/arch/mips/pmcs-msp71xx/Makefile b/arch/mips/pmcs-msp71xx/Makefile index 9201c8b3858..d4f7220f248 100644 --- a/arch/mips/pmcs-msp71xx/Makefile +++ b/arch/mips/pmcs-msp71xx/Makefile @@ -10,4 +10,3 @@ obj-$(CONFIG_PCI) += msp_pci.o obj-$(CONFIG_MSP_HAS_MAC) += msp_eth.o obj-$(CONFIG_MSP_HAS_USB) += msp_usb.o obj-$(CONFIG_MIPS_MT_SMP) += msp_smp.o -obj-$(CONFIG_MIPS_MT_SMTC) += msp_smtc.o diff --git a/arch/mips/pmcs-msp71xx/msp_eth.c b/arch/mips/pmcs-msp71xx/msp_eth.c index c584df393de..15679b427f4 100644 --- a/arch/mips/pmcs-msp71xx/msp_eth.c +++ b/arch/mips/pmcs-msp71xx/msp_eth.c @@ -38,73 +38,6 @@ #define MSP_ETHERNET_GPIO1 15 #define MSP_ETHERNET_GPIO2 16 -#ifdef CONFIG_MSP_HAS_TSMAC -#define MSP_TSMAC_SIZE 0x10020 -#define MSP_TSMAC_ID "pmc_tsmac" - -static struct resource msp_tsmac0_resources[] = { - [0] = { - .start = MSP_MAC0_BASE, - .end = MSP_MAC0_BASE + MSP_TSMAC_SIZE - 1, - .flags = IORESOURCE_MEM, - }, - [1] = { - .start = MSP_INT_MAC0, - .end = MSP_INT_MAC0, - .flags = IORESOURCE_IRQ, - }, -}; - -static struct resource msp_tsmac1_resources[] = { - [0] = { - .start = MSP_MAC1_BASE, - .end = MSP_MAC1_BASE + MSP_TSMAC_SIZE - 1, - .flags = IORESOURCE_MEM, - }, - [1] = { - .start = MSP_INT_MAC1, - .end = MSP_INT_MAC1, - .flags = IORESOURCE_IRQ, - }, -}; -static struct resource msp_tsmac2_resources[] = { - [0] = { - .start = MSP_MAC2_BASE, - .end = MSP_MAC2_BASE + MSP_TSMAC_SIZE - 1, - .flags = IORESOURCE_MEM, - }, - [1] = { - .start = MSP_INT_SAR, - .end = MSP_INT_SAR, - .flags = IORESOURCE_IRQ, - }, -}; - - -static struct platform_device tsmac_device[] = { - [0] = { - .name = MSP_TSMAC_ID, - .id = 0, - .num_resources = ARRAY_SIZE(msp_tsmac0_resources), - .resource = msp_tsmac0_resources, - }, - [1] = { - .name = MSP_TSMAC_ID, - .id = 1, - .num_resources = ARRAY_SIZE(msp_tsmac1_resources), - .resource = msp_tsmac1_resources, - }, - [2] = { - .name = MSP_TSMAC_ID, - .id = 2, - .num_resources = ARRAY_SIZE(msp_tsmac2_resources), - .resource = msp_tsmac2_resources, - }, -}; -#define msp_eth_devs tsmac_device - -#else -/* If it is not TSMAC assume MSP_ETH (100Mbps) */ #define MSP_ETH_ID "pmc_mspeth" #define MSP_ETH_SIZE 0xE0 static struct resource msp_eth0_resources[] = { @@ -152,7 +85,6 @@ static struct platform_device mspeth_device[] = { }; #define msp_eth_devs mspeth_device -#endif int __init msp_eth_setup(void) { int i, ret = 0; @@ -161,14 +93,6 @@ int __init msp_eth_setup(void) msp_gpio_pin_mode(MSP_GPIO_OUTPUT, MSP_ETHERNET_GPIO0); msp_gpio_pin_hi(MSP_ETHERNET_GPIO0); -#ifdef CONFIG_MSP_HAS_TSMAC - /* 3 phys on boards with TSMAC */ - msp_gpio_pin_mode(MSP_GPIO_OUTPUT, MSP_ETHERNET_GPIO1); - msp_gpio_pin_hi(MSP_ETHERNET_GPIO1); - - msp_gpio_pin_mode(MSP_GPIO_OUTPUT, MSP_ETHERNET_GPIO2); - msp_gpio_pin_hi(MSP_ETHERNET_GPIO2); -#endif for (i = 0; i < ARRAY_SIZE(msp_eth_devs); i++) { ret = platform_device_register(&msp_eth_devs[i]); printk(KERN_INFO "device: %d, return value = %d\n", i, ret); diff --git a/arch/mips/pmcs-msp71xx/msp_irq.c b/arch/mips/pmcs-msp71xx/msp_irq.c index 9da5619c00a..941744aabb5 100644 --- a/arch/mips/pmcs-msp71xx/msp_irq.c +++ b/arch/mips/pmcs-msp71xx/msp_irq.c @@ -32,7 +32,7 @@ extern void msp_vsmp_int_init(void); /* vectored interrupt implementation */ -/* SW0/1 interrupts are used for SMP/SMTC */ +/* SW0/1 interrupts are used for SMP */ static inline void mac0_int_dispatch(void) { do_IRQ(MSP_INT_MAC0); } static inline void mac1_int_dispatch(void) { do_IRQ(MSP_INT_MAC1); } static inline void mac2_int_dispatch(void) { do_IRQ(MSP_INT_SAR); } @@ -138,14 +138,6 @@ void __init arch_init_irq(void) set_vi_handler(MSP_INT_SEC, sec_int_dispatch); #ifdef CONFIG_MIPS_MT_SMP msp_vsmp_int_init(); -#elif defined CONFIG_MIPS_MT_SMTC - /*Set hwmask for all platform devices */ - irq_hwmask[MSP_INT_MAC0] = C_IRQ0; - irq_hwmask[MSP_INT_MAC1] = C_IRQ1; - irq_hwmask[MSP_INT_USB] = C_IRQ2; - irq_hwmask[MSP_INT_SAR] = C_IRQ3; - irq_hwmask[MSP_INT_SEC] = C_IRQ5; - #endif /* CONFIG_MIPS_MT_SMP */ #endif /* CONFIG_MIPS_MT */ /* setup the cascaded interrupts */ @@ -153,8 +145,10 @@ void __init arch_init_irq(void) setup_irq(MSP_INT_PER, &per_cascade_msp); #else - /* setup the 2nd-level SLP register based interrupt controller */ - /* VSMP /SMTC support support is not enabled for SLP */ + /* + * Setup the 2nd-level SLP register based interrupt controller. + * VSMP support support is not enabled for SLP. + */ msp_slp_irq_init(); /* setup the cascaded SLP/PER interrupts */ diff --git a/arch/mips/pmcs-msp71xx/msp_irq_cic.c b/arch/mips/pmcs-msp71xx/msp_irq_cic.c index e49b499f66d..b8df2f7b332 100644 --- a/arch/mips/pmcs-msp71xx/msp_irq_cic.c +++ b/arch/mips/pmcs-msp71xx/msp_irq_cic.c @@ -120,10 +120,9 @@ static void msp_cic_irq_ack(struct irq_data *d) * hurt for the others */ *CIC_STS_REG = (1 << (d->irq - MSP_CIC_INTBASE)); - smtc_im_ack_irq(d->irq); } -/*Note: Limiting to VSMP . Not tested in SMTC */ +/* Note: Limiting to VSMP. */ #ifdef CONFIG_MIPS_MT_SMP static int msp_cic_irq_set_affinity(struct irq_data *d, @@ -183,10 +182,6 @@ void __init msp_cic_irq_init(void) for (i = MSP_CIC_INTBASE ; i < MSP_CIC_INTBASE + 32 ; i++) { irq_set_chip_and_handler(i, &msp_cic_irq_controller, handle_level_irq); -#ifdef CONFIG_MIPS_MT_SMTC - /* Mask of CIC interrupt */ - irq_hwmask[i] = C_IRQ4; -#endif } /* Initialize the PER interrupt sub-system */ diff --git a/arch/mips/pmcs-msp71xx/msp_irq_per.c b/arch/mips/pmcs-msp71xx/msp_irq_per.c index d1fd530479d..a111836bcec 100644 --- a/arch/mips/pmcs-msp71xx/msp_irq_per.c +++ b/arch/mips/pmcs-msp71xx/msp_irq_per.c @@ -113,9 +113,6 @@ void __init msp_per_irq_init(void) /* initialize all the IRQ descriptors */ for (i = MSP_PER_INTBASE; i < MSP_PER_INTBASE + 32; i++) { irq_set_chip(i, &msp_per_irq_controller); -#ifdef CONFIG_MIPS_MT_SMTC - irq_hwmask[i] = C_IRQ4; -#endif } } diff --git a/arch/mips/pmcs-msp71xx/msp_setup.c b/arch/mips/pmcs-msp71xx/msp_setup.c index 7e980767679..4f925e06c41 100644 --- a/arch/mips/pmcs-msp71xx/msp_setup.c +++ b/arch/mips/pmcs-msp71xx/msp_setup.c @@ -27,7 +27,6 @@ #endif extern void msp_serial_setup(void); -extern void pmctwiled_setup(void); #if defined(CONFIG_PMC_MSP7120_EVAL) || \ defined(CONFIG_PMC_MSP7120_GW) || \ @@ -148,8 +147,6 @@ void __init plat_mem_setup(void) pm_power_off = msp_power_off; } -extern struct plat_smp_ops msp_smtc_smp_ops; - void __init prom_init(void) { unsigned long family; @@ -230,17 +227,5 @@ void __init prom_init(void) */ msp_serial_setup(); - if (register_vsmp_smp_ops()) { -#ifdef CONFIG_MIPS_MT_SMTC - register_smp_ops(&msp_smtc_smp_ops); -#endif - } - -#ifdef CONFIG_PMCTWILED - /* - * Setup LED states before the subsys_initcall loads other - * dependent drivers/modules. - */ - pmctwiled_setup(); -#endif + register_vsmp_smp_ops(); } diff --git a/arch/mips/pmcs-msp71xx/msp_smtc.c b/arch/mips/pmcs-msp71xx/msp_smtc.c deleted file mode 100644 index 6b5607fce27..00000000000 --- a/arch/mips/pmcs-msp71xx/msp_smtc.c +++ /dev/null @@ -1,104 +0,0 @@ -/* - * MSP71xx Platform-specific hooks for SMP operation - */ -#include <linux/irq.h> -#include <linux/init.h> - -#include <asm/mipsmtregs.h> -#include <asm/mipsregs.h> -#include <asm/smtc.h> -#include <asm/smtc_ipi.h> - -/* VPE/SMP Prototype implements platform interfaces directly */ - -/* - * Cause the specified action to be performed on a targeted "CPU" - */ - -static void msp_smtc_send_ipi_single(int cpu, unsigned int action) -{ - /* "CPU" may be TC of same VPE, VPE of same CPU, or different CPU */ - smtc_send_ipi(cpu, LINUX_SMP_IPI, action); -} - -static void msp_smtc_send_ipi_mask(const struct cpumask *mask, - unsigned int action) -{ - unsigned int i; - - for_each_cpu(i, mask) - msp_smtc_send_ipi_single(i, action); -} - -/* - * Post-config but pre-boot cleanup entry point - */ -static void msp_smtc_init_secondary(void) -{ - int myvpe; - - /* Don't enable Malta I/O interrupts (IP2) for secondary VPEs */ - myvpe = read_c0_tcbind() & TCBIND_CURVPE; - if (myvpe > 0) - change_c0_status(ST0_IM, STATUSF_IP0 | STATUSF_IP1 | - STATUSF_IP6 | STATUSF_IP7); - smtc_init_secondary(); -} - -/* - * Platform "CPU" startup hook - */ -static void msp_smtc_boot_secondary(int cpu, struct task_struct *idle) -{ - smtc_boot_secondary(cpu, idle); -} - -/* - * SMP initialization finalization entry point - */ -static void msp_smtc_smp_finish(void) -{ - smtc_smp_finish(); -} - -/* - * Hook for after all CPUs are online - */ - -static void msp_smtc_cpus_done(void) -{ -} - -/* - * Platform SMP pre-initialization - * - * As noted above, we can assume a single CPU for now - * but it may be multithreaded. - */ - -static void __init msp_smtc_smp_setup(void) -{ - /* - * we won't get the definitive value until - * we've run smtc_prepare_cpus later, but - */ - - if (read_c0_config3() & (1 << 2)) - smp_num_siblings = smtc_build_cpu_map(0); -} - -static void __init msp_smtc_prepare_cpus(unsigned int max_cpus) -{ - smtc_prepare_cpus(max_cpus); -} - -struct plat_smp_ops msp_smtc_smp_ops = { - .send_ipi_single = msp_smtc_send_ipi_single, - .send_ipi_mask = msp_smtc_send_ipi_mask, - .init_secondary = msp_smtc_init_secondary, - .smp_finish = msp_smtc_smp_finish, - .cpus_done = msp_smtc_cpus_done, - .boot_secondary = msp_smtc_boot_secondary, - .smp_setup = msp_smtc_smp_setup, - .prepare_cpus = msp_smtc_prepare_cpus, -}; diff --git a/arch/mips/pmcs-msp71xx/msp_usb.c b/arch/mips/pmcs-msp71xx/msp_usb.c index 4dab915696e..c87c5f810cd 100644 --- a/arch/mips/pmcs-msp71xx/msp_usb.c +++ b/arch/mips/pmcs-msp71xx/msp_usb.c @@ -75,47 +75,6 @@ static struct mspusb_device msp_usbhost0_device = { .resource = msp_usbhost0_resources, }, }; - -/* MSP7140/MSP82XX has two USB2 hosts. */ -#ifdef CONFIG_MSP_HAS_DUAL_USB -static u64 msp_usbhost1_dma_mask = 0xffffffffUL; - -static struct resource msp_usbhost1_resources[] = { - [0] = { /* EHCI-HS operational and capabilities registers */ - .start = MSP_USB1_HS_START, - .end = MSP_USB1_HS_END, - .flags = IORESOURCE_MEM, - }, - [1] = { - .start = MSP_INT_USB, - .end = MSP_INT_USB, - .flags = IORESOURCE_IRQ, - }, - [2] = { /* MSBus-to-AMBA bridge register space */ - .start = MSP_USB1_MAB_START, - .end = MSP_USB1_MAB_END, - .flags = IORESOURCE_MEM, - }, - [3] = { /* Identification and general hardware parameters */ - .start = MSP_USB1_ID_START, - .end = MSP_USB1_ID_END, - .flags = IORESOURCE_MEM, - }, -}; - -static struct mspusb_device msp_usbhost1_device = { - .dev = { - .name = "pmcmsp-ehci", - .id = 1, - .dev = { - .dma_mask = &msp_usbhost1_dma_mask, - .coherent_dma_mask = 0xffffffffUL, - }, - .num_resources = ARRAY_SIZE(msp_usbhost1_resources), - .resource = msp_usbhost1_resources, - }, -}; -#endif /* CONFIG_MSP_HAS_DUAL_USB */ #endif /* CONFIG_USB_EHCI_HCD */ #if defined(CONFIG_USB_GADGET) @@ -157,46 +116,6 @@ static struct mspusb_device msp_usbdev0_device = { .resource = msp_usbdev0_resources, }, }; - -#ifdef CONFIG_MSP_HAS_DUAL_USB -static struct resource msp_usbdev1_resources[] = { - [0] = { /* EHCI-HS operational and capabilities registers */ - .start = MSP_USB1_HS_START, - .end = MSP_USB1_HS_END, - .flags = IORESOURCE_MEM, - }, - [1] = { - .start = MSP_INT_USB, - .end = MSP_INT_USB, - .flags = IORESOURCE_IRQ, - }, - [2] = { /* MSBus-to-AMBA bridge register space */ - .start = MSP_USB1_MAB_START, - .end = MSP_USB1_MAB_END, - .flags = IORESOURCE_MEM, - }, - [3] = { /* Identification and general hardware parameters */ - .start = MSP_USB1_ID_START, - .end = MSP_USB1_ID_END, - .flags = IORESOURCE_MEM, - }, -}; - -/* This may need to be converted to a mspusb_device, too. */ -static struct mspusb_device msp_usbdev1_device = { - .dev = { - .name = "msp71xx_udc", - .id = 0, - .dev = { - .dma_mask = &msp_usbdev_dma_mask, - .coherent_dma_mask = 0xffffffffUL, - }, - .num_resources = ARRAY_SIZE(msp_usbdev1_resources), - .resource = msp_usbdev1_resources, - }, -}; - -#endif /* CONFIG_MSP_HAS_DUAL_USB */ #endif /* CONFIG_USB_GADGET */ static int __init msp_usb_setup(void) @@ -231,10 +150,6 @@ static int __init msp_usb_setup(void) #if defined(CONFIG_USB_EHCI_HCD) msp_devs[0] = &msp_usbhost0_device.dev; ppfinit("platform add USB HOST done %s.\n", msp_devs[0]->name); -#ifdef CONFIG_MSP_HAS_DUAL_USB - msp_devs[1] = &msp_usbhost1_device.dev; - ppfinit("platform add USB HOST done %s.\n", msp_devs[1]->name); -#endif #else ppfinit("%s: echi_hcd not supported\n", __FILE__); #endif /* CONFIG_USB_EHCI_HCD */ @@ -244,11 +159,6 @@ static int __init msp_usb_setup(void) msp_devs[0] = &msp_usbdev0_device.dev; ppfinit("platform add USB DEVICE done %s.\n" , msp_devs[0]->name); -#ifdef CONFIG_MSP_HAS_DUAL_USB - msp_devs[1] = &msp_usbdev1_device.dev; - ppfinit("platform add USB DEVICE done %s.\n" - , msp_devs[1]->name); -#endif #else ppfinit("%s: usb_gadget not supported\n", __FILE__); #endif /* CONFIG_USB_GADGET */ diff --git a/arch/mips/pnx833x/common/platform.c b/arch/mips/pnx833x/common/platform.c index 2b7e837dc2e..b4b774bc317 100644 --- a/arch/mips/pnx833x/common/platform.c +++ b/arch/mips/pnx833x/common/platform.c @@ -33,11 +33,6 @@ #include <linux/mtd/nand.h> #include <linux/mtd/partitions.h> -#ifdef CONFIG_I2C_PNX0105 -/* Until i2c driver available in kernel.*/ -#include <linux/i2c-pnx0105.h> -#endif - #include <irq.h> #include <irq-mapping.h> #include <pnx833x.h> @@ -134,70 +129,6 @@ static struct platform_device pnx833x_usb_ehci_device = { .resource = pnx833x_usb_ehci_resources, }; -#ifdef CONFIG_I2C_PNX0105 -static struct resource pnx833x_i2c0_resources[] = { - { - .start = PNX833X_I2C0_PORTS_START, - .end = PNX833X_I2C0_PORTS_END, - .flags = IORESOURCE_MEM, - }, - { - .start = PNX833X_PIC_I2C0_INT, - .end = PNX833X_PIC_I2C0_INT, - .flags = IORESOURCE_IRQ, - }, -}; - -static struct resource pnx833x_i2c1_resources[] = { - { - .start = PNX833X_I2C1_PORTS_START, - .end = PNX833X_I2C1_PORTS_END, - .flags = IORESOURCE_MEM, - }, - { - .start = PNX833X_PIC_I2C1_INT, - .end = PNX833X_PIC_I2C1_INT, - .flags = IORESOURCE_IRQ, - }, -}; - -static struct i2c_pnx0105_dev pnx833x_i2c_dev[] = { - { - .base = PNX833X_I2C0_PORTS_START, - .irq = -1, /* should be PNX833X_PIC_I2C0_INT but polling is faster */ - .clock = 6, /* 0 == 400 kHz, 4 == 100 kHz(Maximum HDMI), 6 = 50kHz(Preferred HDCP) */ - .bus_addr = 0, /* no slave support */ - }, - { - .base = PNX833X_I2C1_PORTS_START, - .irq = -1, /* on high freq, polling is faster */ - /*.irq = PNX833X_PIC_I2C1_INT,*/ - .clock = 4, /* 0 == 400 kHz, 4 == 100 kHz. 100 kHz seems a safe default for now */ - .bus_addr = 0, /* no slave support */ - }, -}; - -static struct platform_device pnx833x_i2c0_device = { - .name = "i2c-pnx0105", - .id = 0, - .dev = { - .platform_data = &pnx833x_i2c_dev[0], - }, - .num_resources = ARRAY_SIZE(pnx833x_i2c0_resources), - .resource = pnx833x_i2c0_resources, -}; - -static struct platform_device pnx833x_i2c1_device = { - .name = "i2c-pnx0105", - .id = 1, - .dev = { - .platform_data = &pnx833x_i2c_dev[1], - }, - .num_resources = ARRAY_SIZE(pnx833x_i2c1_resources), - .resource = pnx833x_i2c1_resources, -}; -#endif - static u64 ethernet_dmamask = DMA_BIT_MASK(32); static struct resource pnx833x_ethernet_resources[] = { @@ -294,10 +225,6 @@ static struct platform_device pnx833x_flash_nand = { static struct platform_device *pnx833x_platform_devices[] __initdata = { &pnx833x_uart_device, &pnx833x_usb_ehci_device, -#ifdef CONFIG_I2C_PNX0105 - &pnx833x_i2c0_device, - &pnx833x_i2c1_device, -#endif &pnx833x_ethernet_device, &pnx833x_sata_device, &pnx833x_flash_nand, diff --git a/arch/mips/sgi-ip22/ip22-gio.c b/arch/mips/sgi-ip22/ip22-gio.c index ab0e379dc7e..8e52446286c 100644 --- a/arch/mips/sgi-ip22/ip22-gio.c +++ b/arch/mips/sgi-ip22/ip22-gio.c @@ -19,6 +19,9 @@ static struct { } gio_name_table[] = { { .name = "SGI Impact", .id = 0x10 }, { .name = "Phobos G160", .id = 0x35 }, + { .name = "Phobos G130", .id = 0x36 }, + { .name = "Phobos G100", .id = 0x37 }, + { .name = "Set Engineering GFE", .id = 0x38 }, /* fake IDs */ { .name = "SGI Newport", .id = 0x7e }, { .name = "SGI GR2/GR3", .id = 0x7f }, @@ -293,7 +296,16 @@ static int ip22_gio_id(unsigned long addr, u32 *res) * data matches */ ptr8 = (void *)CKSEG1ADDR(addr + 3); - get_dbe(tmp8, ptr8); + if (get_dbe(tmp8, ptr8)) { + /* + * 32bit access worked, but 8bit doesn't + * so we don't see phantom reads on + * a pipelined bus, but a real card which + * doesn't support 8 bit reads + */ + *res = tmp32; + return 1; + } ptr16 = (void *)CKSEG1ADDR(addr + 2); get_dbe(tmp16, ptr16); if (tmp8 == (tmp16 & 0xff) && @@ -324,7 +336,7 @@ static int ip22_is_gr2(unsigned long addr) } -static void ip22_check_gio(int slotno, unsigned long addr) +static void ip22_check_gio(int slotno, unsigned long addr, int irq) { const char *name = "Unknown"; struct gio_device *gio_dev; @@ -338,9 +350,9 @@ static void ip22_check_gio(int slotno, unsigned long addr) else { if (!ip22_gio_id(addr, &tmp)) { /* - * no GIO signature at start address of slot, but - * Newport doesn't have one, so let's check usea - * status register + * no GIO signature at start address of slot + * since Newport doesn't have one, we check if + * user status register is readable */ if (ip22_gio_id(addr + NEWPORT_USTATUS_OFFS, &tmp)) tmp = 0x7e; @@ -369,6 +381,7 @@ static void ip22_check_gio(int slotno, unsigned long addr) gio_dev->resource.start = addr; gio_dev->resource.end = addr + 0x3fffff; gio_dev->resource.flags = IORESOURCE_MEM; + gio_dev->irq = irq; dev_set_name(&gio_dev->dev, "%d", slotno); gio_device_register(gio_dev); } else @@ -408,16 +421,17 @@ int __init ip22_gio_init(void) request_resource(&iomem_resource, &gio_bus_resource); printk(KERN_INFO "GIO: Probing bus...\n"); - if (ip22_is_fullhouse() || - !get_dbe(pbdma, (unsigned int *)&hpc3c1->pbdma[1])) { - /* Indigo2 and ChallengeS */ - ip22_check_gio(0, GIO_SLOT_GFX_BASE); - ip22_check_gio(1, GIO_SLOT_EXP0_BASE); + if (ip22_is_fullhouse()) { + /* Indigo2 */ + ip22_check_gio(0, GIO_SLOT_GFX_BASE, SGI_GIO_1_IRQ); + ip22_check_gio(1, GIO_SLOT_EXP0_BASE, SGI_GIO_1_IRQ); } else { - /* Indy */ - ip22_check_gio(0, GIO_SLOT_GFX_BASE); - ip22_check_gio(1, GIO_SLOT_EXP0_BASE); - ip22_check_gio(2, GIO_SLOT_EXP1_BASE); + /* Indy/Challenge S */ + if (get_dbe(pbdma, (unsigned int *)&hpc3c1->pbdma[1])) + ip22_check_gio(0, GIO_SLOT_GFX_BASE, + SGI_GIO_0_IRQ); + ip22_check_gio(1, GIO_SLOT_EXP0_BASE, SGI_GIOEXP0_IRQ); + ip22_check_gio(2, GIO_SLOT_EXP1_BASE, SGI_GIOEXP1_IRQ); } } else device_unregister(&gio_bus); diff --git a/arch/mips/sgi-ip22/ip22-int.c b/arch/mips/sgi-ip22/ip22-int.c index 58b40ae5933..c66889fc491 100644 --- a/arch/mips/sgi-ip22/ip22-int.c +++ b/arch/mips/sgi-ip22/ip22-int.c @@ -119,9 +119,14 @@ static void indy_local0_irqdispatch(void) } else irq = lc0msk_to_irqnr[mask]; - /* if irq == 0, then the interrupt has already been cleared */ + /* + * workaround for INT2 bug; if irq == 0, INT2 has seen a fifo full + * irq, but failed to latch it into status register + */ if (irq) do_IRQ(irq); + else + do_IRQ(SGINT_LOCAL0 + 0); } static void indy_local1_irqdispatch(void) diff --git a/arch/mips/sgi-ip27/ip27-smp.c b/arch/mips/sgi-ip27/ip27-smp.c index f4ea8aa79ba..f9ae6a8fa7c 100644 --- a/arch/mips/sgi-ip27/ip27-smp.c +++ b/arch/mips/sgi-ip27/ip27-smp.c @@ -186,10 +186,6 @@ static void ip27_smp_finish(void) local_irq_enable(); } -static void __init ip27_cpus_done(void) -{ -} - /* * Launch a slave into smp_bootstrap(). It doesn't take an argument, and we * set sp to the kernel stack of the newly created idle process, gp to the proc @@ -236,7 +232,6 @@ struct plat_smp_ops ip27_smp_ops = { .send_ipi_mask = ip27_send_ipi_mask, .init_secondary = ip27_init_secondary, .smp_finish = ip27_smp_finish, - .cpus_done = ip27_cpus_done, .boot_secondary = ip27_boot_secondary, .smp_setup = ip27_smp_setup, .prepare_cpus = ip27_prepare_cpus, diff --git a/arch/mips/sibyte/bcm1480/irq.c b/arch/mips/sibyte/bcm1480/irq.c index 59cfe265977..373fbbc8425 100644 --- a/arch/mips/sibyte/bcm1480/irq.c +++ b/arch/mips/sibyte/bcm1480/irq.c @@ -347,19 +347,8 @@ asmlinkage void plat_irq_dispatch(void) unsigned int cpu = smp_processor_id(); unsigned int pending; -#ifdef CONFIG_SIBYTE_BCM1480_PROF - /* Set compare to count to silence count/compare timer interrupts */ - write_c0_compare(read_c0_count()); -#endif - pending = read_c0_cause() & read_c0_status(); -#ifdef CONFIG_SIBYTE_BCM1480_PROF - if (pending & CAUSEF_IP7) /* Cpu performance counter interrupt */ - sbprof_cpu_intr(); - else -#endif - if (pending & CAUSEF_IP4) do_IRQ(K_BCM1480_INT_TIMER_0 + cpu); #ifdef CONFIG_SMP diff --git a/arch/mips/sibyte/bcm1480/smp.c b/arch/mips/sibyte/bcm1480/smp.c index 70d9182b26f..af7d44edd9a 100644 --- a/arch/mips/sibyte/bcm1480/smp.c +++ b/arch/mips/sibyte/bcm1480/smp.c @@ -115,13 +115,6 @@ static void bcm1480_smp_finish(void) } /* - * Final cleanup after all secondaries booted - */ -static void bcm1480_cpus_done(void) -{ -} - -/* * Setup the PC, SP, and GP of a secondary processor and start it * running! */ @@ -170,7 +163,6 @@ struct plat_smp_ops bcm1480_smp_ops = { .send_ipi_mask = bcm1480_send_ipi_mask, .init_secondary = bcm1480_init_secondary, .smp_finish = bcm1480_smp_finish, - .cpus_done = bcm1480_cpus_done, .boot_secondary = bcm1480_boot_secondary, .smp_setup = bcm1480_smp_setup, .prepare_cpus = bcm1480_prepare_cpus, diff --git a/arch/mips/sibyte/sb1250/smp.c b/arch/mips/sibyte/sb1250/smp.c index db976117dd4..c0c4b3f88a0 100644 --- a/arch/mips/sibyte/sb1250/smp.c +++ b/arch/mips/sibyte/sb1250/smp.c @@ -103,13 +103,6 @@ static void sb1250_smp_finish(void) } /* - * Final cleanup after all secondaries booted - */ -static void sb1250_cpus_done(void) -{ -} - -/* * Setup the PC, SP, and GP of a secondary processor and start it * running! */ @@ -158,7 +151,6 @@ struct plat_smp_ops sb_smp_ops = { .send_ipi_mask = sb1250_send_ipi_mask, .init_secondary = sb1250_init_secondary, .smp_finish = sb1250_smp_finish, - .cpus_done = sb1250_cpus_done, .boot_secondary = sb1250_boot_secondary, .smp_setup = sb1250_smp_setup, .prepare_cpus = sb1250_prepare_cpus, diff --git a/arch/mips/txx9/generic/setup.c b/arch/mips/txx9/generic/setup.c index 2b0b83c171e..dd2cf25b5ae 100644 --- a/arch/mips/txx9/generic/setup.c +++ b/arch/mips/txx9/generic/setup.c @@ -309,8 +309,8 @@ static void __init preprocess_cmdline(void) txx9_board_vec = find_board_byname(str + 6); continue; } else if (strncmp(str, "masterclk=", 10) == 0) { - unsigned long val; - if (strict_strtoul(str + 10, 10, &val) == 0) + unsigned int val; + if (kstrtouint(str + 10, 10, &val) == 0) txx9_master_clock = val; continue; } else if (strcmp(str, "icdisable") == 0) { diff --git a/drivers/cpuidle/Kconfig b/drivers/cpuidle/Kconfig index f04e25f6c98..1b96fb91d32 100644 --- a/drivers/cpuidle/Kconfig +++ b/drivers/cpuidle/Kconfig @@ -35,6 +35,11 @@ depends on ARM source "drivers/cpuidle/Kconfig.arm" endmenu +menu "MIPS CPU Idle Drivers" +depends on MIPS +source "drivers/cpuidle/Kconfig.mips" +endmenu + menu "POWERPC CPU Idle Drivers" depends on PPC source "drivers/cpuidle/Kconfig.powerpc" diff --git a/drivers/cpuidle/Kconfig.mips b/drivers/cpuidle/Kconfig.mips new file mode 100644 index 00000000000..0e70ee28a5c --- /dev/null +++ b/drivers/cpuidle/Kconfig.mips @@ -0,0 +1,17 @@ +# +# MIPS CPU Idle Drivers +# +config MIPS_CPS_CPUIDLE + bool "CPU Idle driver for MIPS CPS platforms" + depends on CPU_IDLE + depends on SYS_SUPPORTS_MIPS_CPS + select ARCH_NEEDS_CPU_IDLE_COUPLED if MIPS_MT + select GENERIC_CLOCKEVENTS_BROADCAST if SMP + select MIPS_CPS_PM + default y + help + Select this option to enable processor idle state management + through cpuidle for systems built around the MIPS Coherent + Processing System (CPS) architecture. In order to make use of + the deepest idle states you will need to ensure that you are + also using the CONFIG_MIPS_CPS SMP implementation. diff --git a/drivers/cpuidle/Makefile b/drivers/cpuidle/Makefile index 9b5b2b560d7..d8bb1ff7256 100644 --- a/drivers/cpuidle/Makefile +++ b/drivers/cpuidle/Makefile @@ -18,6 +18,10 @@ obj-$(CONFIG_ARM_AT91_CPUIDLE) += cpuidle-at91.o obj-$(CONFIG_ARM_EXYNOS_CPUIDLE) += cpuidle-exynos.o ############################################################################### +# MIPS drivers +obj-$(CONFIG_MIPS_CPS_CPUIDLE) += cpuidle-cps.o + +############################################################################### # POWERPC drivers obj-$(CONFIG_PSERIES_CPUIDLE) += cpuidle-pseries.o obj-$(CONFIG_POWERNV_CPUIDLE) += cpuidle-powernv.o diff --git a/drivers/cpuidle/cpuidle-cps.c b/drivers/cpuidle/cpuidle-cps.c new file mode 100644 index 00000000000..fc7b62720de --- /dev/null +++ b/drivers/cpuidle/cpuidle-cps.c @@ -0,0 +1,186 @@ +/* + * Copyright (C) 2014 Imagination Technologies + * Author: Paul Burton <paul.burton@imgtec.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + */ + +#include <linux/cpu_pm.h> +#include <linux/cpuidle.h> +#include <linux/init.h> + +#include <asm/idle.h> +#include <asm/pm-cps.h> + +/* Enumeration of the various idle states this driver may enter */ +enum cps_idle_state { + STATE_WAIT = 0, /* MIPS wait instruction, coherent */ + STATE_NC_WAIT, /* MIPS wait instruction, non-coherent */ + STATE_CLOCK_GATED, /* Core clock gated */ + STATE_POWER_GATED, /* Core power gated */ + STATE_COUNT +}; + +static int cps_nc_enter(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) +{ + enum cps_pm_state pm_state; + int err; + + /* + * At least one core must remain powered up & clocked in order for the + * system to have any hope of functioning. + * + * TODO: don't treat core 0 specially, just prevent the final core + * TODO: remap interrupt affinity temporarily + */ + if (!cpu_data[dev->cpu].core && (index > STATE_NC_WAIT)) + index = STATE_NC_WAIT; + + /* Select the appropriate cps_pm_state */ + switch (index) { + case STATE_NC_WAIT: + pm_state = CPS_PM_NC_WAIT; + break; + case STATE_CLOCK_GATED: + pm_state = CPS_PM_CLOCK_GATED; + break; + case STATE_POWER_GATED: + pm_state = CPS_PM_POWER_GATED; + break; + default: + BUG(); + return -EINVAL; + } + + /* Notify listeners the CPU is about to power down */ + if ((pm_state == CPS_PM_POWER_GATED) && cpu_pm_enter()) + return -EINTR; + + /* Enter that state */ + err = cps_pm_enter_state(pm_state); + + /* Notify listeners the CPU is back up */ + if (pm_state == CPS_PM_POWER_GATED) + cpu_pm_exit(); + + return err ?: index; +} + +static struct cpuidle_driver cps_driver = { + .name = "cpc_cpuidle", + .owner = THIS_MODULE, + .states = { + [STATE_WAIT] = MIPS_CPUIDLE_WAIT_STATE, + [STATE_NC_WAIT] = { + .enter = cps_nc_enter, + .exit_latency = 200, + .target_residency = 450, + .flags = CPUIDLE_FLAG_TIME_VALID, + .name = "nc-wait", + .desc = "non-coherent MIPS wait", + }, + [STATE_CLOCK_GATED] = { + .enter = cps_nc_enter, + .exit_latency = 300, + .target_residency = 700, + .flags = CPUIDLE_FLAG_TIME_VALID | + CPUIDLE_FLAG_TIMER_STOP, + .name = "clock-gated", + .desc = "core clock gated", + }, + [STATE_POWER_GATED] = { + .enter = cps_nc_enter, + .exit_latency = 600, + .target_residency = 1000, + .flags = CPUIDLE_FLAG_TIME_VALID | + CPUIDLE_FLAG_TIMER_STOP, + .name = "power-gated", + .desc = "core power gated", + }, + }, + .state_count = STATE_COUNT, + .safe_state_index = 0, +}; + +static void __init cps_cpuidle_unregister(void) +{ + int cpu; + struct cpuidle_device *device; + + for_each_possible_cpu(cpu) { + device = &per_cpu(cpuidle_dev, cpu); + cpuidle_unregister_device(device); + } + + cpuidle_unregister_driver(&cps_driver); +} + +static int __init cps_cpuidle_init(void) +{ + int err, cpu, core, i; + struct cpuidle_device *device; + + /* Detect supported states */ + if (!cps_pm_support_state(CPS_PM_POWER_GATED)) + cps_driver.state_count = STATE_CLOCK_GATED + 1; + if (!cps_pm_support_state(CPS_PM_CLOCK_GATED)) + cps_driver.state_count = STATE_NC_WAIT + 1; + if (!cps_pm_support_state(CPS_PM_NC_WAIT)) + cps_driver.state_count = STATE_WAIT + 1; + + /* Inform the user if some states are unavailable */ + if (cps_driver.state_count < STATE_COUNT) { + pr_info("cpuidle-cps: limited to "); + switch (cps_driver.state_count - 1) { + case STATE_WAIT: + pr_cont("coherent wait\n"); + break; + case STATE_NC_WAIT: + pr_cont("non-coherent wait\n"); + break; + case STATE_CLOCK_GATED: + pr_cont("clock gating\n"); + break; + } + } + + /* + * Set the coupled flag on the appropriate states if this system + * requires it. + */ + if (coupled_coherence) + for (i = STATE_NC_WAIT; i < cps_driver.state_count; i++) + cps_driver.states[i].flags |= CPUIDLE_FLAG_COUPLED; + + err = cpuidle_register_driver(&cps_driver); + if (err) { + pr_err("Failed to register CPS cpuidle driver\n"); + return err; + } + + for_each_possible_cpu(cpu) { + core = cpu_data[cpu].core; + device = &per_cpu(cpuidle_dev, cpu); + device->cpu = cpu; +#ifdef CONFIG_MIPS_MT + cpumask_copy(&device->coupled_cpus, &cpu_sibling_map[cpu]); +#endif + + err = cpuidle_register_device(device); + if (err) { + pr_err("Failed to register CPU%d cpuidle device\n", + cpu); + goto err_out; + } + } + + return 0; +err_out: + cps_cpuidle_unregister(); + return err; +} +device_initcall(cps_cpuidle_init); diff --git a/drivers/usb/host/ehci-pmcmsp.c b/drivers/usb/host/ehci-pmcmsp.c index af3974a5e7c..7d75465d97c 100644 --- a/drivers/usb/host/ehci-pmcmsp.c +++ b/drivers/usb/host/ehci-pmcmsp.c @@ -68,9 +68,6 @@ static void usb_hcd_tdi_set_mode(struct ehci_hcd *ehci) /* set TWI GPIO USB_HOST_DEV pin high */ gpio_direction_output(MSP_PIN_USB0_HOST_DEV, 1); -#ifdef CONFIG_MSP_HAS_DUAL_USB - gpio_direction_output(MSP_PIN_USB1_HOST_DEV, 1); -#endif } /* called during probe() after chip reset completes */ @@ -248,33 +245,6 @@ void usb_hcd_msp_remove(struct usb_hcd *hcd, struct platform_device *dev) usb_put_hcd(hcd); } -#ifdef CONFIG_MSP_HAS_DUAL_USB -/* - * Wrapper around the main ehci_irq. Since both USB host controllers are - * sharing the same IRQ, need to first determine whether we're the intended - * recipient of this interrupt. - */ -static irqreturn_t ehci_msp_irq(struct usb_hcd *hcd) -{ - u32 int_src; - struct device *dev = hcd->self.controller; - struct platform_device *pdev; - struct mspusb_device *mdev; - struct ehci_hcd *ehci = hcd_to_ehci(hcd); - /* need to reverse-map a couple of containers to get our device */ - pdev = to_platform_device(dev); - mdev = to_mspusb_device(pdev); - - /* Check to see if this interrupt is for this host controller */ - int_src = ehci_readl(ehci, &mdev->mab_regs->int_stat); - if (int_src & (1 << pdev->id)) - return ehci_irq(hcd); - - /* Not for this device */ - return IRQ_NONE; -} -#endif /* DUAL_USB */ - static const struct hc_driver ehci_msp_hc_driver = { .description = hcd_name, .product_desc = "PMC MSP EHCI", @@ -283,11 +253,7 @@ static const struct hc_driver ehci_msp_hc_driver = { /* * generic hardware linkage */ -#ifdef CONFIG_MSP_HAS_DUAL_USB - .irq = ehci_msp_irq, -#else .irq = ehci_irq, -#endif .flags = HCD_MEMORY | HCD_USB2 | HCD_BH, /* @@ -334,9 +300,6 @@ static int ehci_hcd_msp_drv_probe(struct platform_device *pdev) return -ENODEV; gpio_request(MSP_PIN_USB0_HOST_DEV, "USB0_HOST_DEV_GPIO"); -#ifdef CONFIG_MSP_HAS_DUAL_USB - gpio_request(MSP_PIN_USB1_HOST_DEV, "USB1_HOST_DEV_GPIO"); -#endif ret = usb_hcd_msp_probe(&ehci_msp_hc_driver, pdev); @@ -351,9 +314,6 @@ static int ehci_hcd_msp_drv_remove(struct platform_device *pdev) /* free TWI GPIO USB_HOST_DEV pin */ gpio_free(MSP_PIN_USB0_HOST_DEV); -#ifdef CONFIG_MSP_HAS_DUAL_USB - gpio_free(MSP_PIN_USB1_HOST_DEV); -#endif return 0; } diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h index c51a436135c..25e0df6155a 100644 --- a/include/linux/cpuidle.h +++ b/include/linux/cpuidle.h @@ -84,6 +84,7 @@ struct cpuidle_device { }; DECLARE_PER_CPU(struct cpuidle_device *, cpuidle_devices); +DECLARE_PER_CPU(struct cpuidle_device, cpuidle_dev); /** * cpuidle_get_last_residency - retrieves the last state's residency time diff --git a/include/uapi/linux/kvm_para.h b/include/uapi/linux/kvm_para.h index 2841f86eae0..bf6cd7d5cac 100644 --- a/include/uapi/linux/kvm_para.h +++ b/include/uapi/linux/kvm_para.h @@ -20,6 +20,9 @@ #define KVM_HC_FEATURES 3 #define KVM_HC_PPC_MAP_MAGIC_PAGE 4 #define KVM_HC_KICK_CPU 5 +#define KVM_HC_MIPS_GET_CLOCK_FREQ 6 +#define KVM_HC_MIPS_EXIT_VM 7 +#define KVM_HC_MIPS_CONSOLE_OUTPUT 8 /* * hypercalls use architecture specific |