From 9a0ebfbe3f1007008d198ccc6b86783cdb312fec Mon Sep 17 00:00:00 2001 From: Daniel J Blueman Date: Mon, 5 Dec 2011 16:20:36 +0800 Subject: x86: Make flat_init_apic_ldr() available Allow flat_init_apic_ldr() to be used outside the compilation unit for similar APIC implementations. Signed-off-by: Daniel J Blueman Cc: Steffen Persvold Cc: Jesse Barnes Link: http://lkml.kernel.org/r/1323073238-32686-1-git-send-email-daniel@numascale-asia.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/apic/apic_flat_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel/apic') diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c index f7a41e4cae4..57c1f4135fa 100644 --- a/arch/x86/kernel/apic/apic_flat_64.c +++ b/arch/x86/kernel/apic/apic_flat_64.c @@ -62,7 +62,7 @@ static void flat_vector_allocation_domain(int cpu, struct cpumask *retmask) * an APIC. See e.g. "AP-388 82489DX User's Manual" (Intel * document number 292116). So here it goes... */ -static void flat_init_apic_ldr(void) +void flat_init_apic_ldr(void) { unsigned long val; unsigned long num, id; -- cgit v1.2.3-70-g09d2 From 44b111b519160e33fdc41eadb39af86a24707edf Mon Sep 17 00:00:00 2001 From: Steffen Persvold Date: Tue, 6 Dec 2011 00:07:26 +0800 Subject: x86: Add NumaChip support Adds support for Numascale NumaChip large-SMP systems. It is needed to enable the booting of more than ~168 cores. v2: - [Steffen] enumerate only accessible northbridges - [Daniel] rediffed and validated against 3.1-rc10 v3: - [Daniel] use x86_init core numbering override - [Daniel] cleanups as per feedback v4: - [Daniel] use updated x86_cpuinit override v5: - drop disabling interrupts locally, as ISR write is atomic; drop delay - added read-mostly annotations where appropriate - require CONFIG_SMP, so drop conditional path Workload tested on 96 cores/16 sockets. Signed-off-by: Steffen Persvold Signed-off-by: Daniel J Blueman Cc: Jesse Barnes Link: http://lkml.kernel.org/r/1323101246-2400-1-git-send-email-daniel@numascale-asia.com Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 13 ++ arch/x86/include/asm/numachip/numachip_csr.h | 167 +++++++++++++++ arch/x86/kernel/apic/Makefile | 1 + arch/x86/kernel/apic/apic_numachip.c | 294 +++++++++++++++++++++++++++ 4 files changed, 475 insertions(+) create mode 100644 arch/x86/include/asm/numachip/numachip_csr.h create mode 100644 arch/x86/kernel/apic/apic_numachip.c (limited to 'arch/x86/kernel/apic') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index cb9a1044a77..7b9eaa1ae10 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -343,6 +343,7 @@ config X86_EXTENDED_PLATFORM If you enable this option then you'll be able to select support for the following (non-PC) 64 bit x86 platforms: + Numascale NumaChip ScaleMP vSMP SGI Ultraviolet @@ -351,6 +352,18 @@ config X86_EXTENDED_PLATFORM endif # This is an alphabetically sorted list of 64 bit extended platforms # Please maintain the alphabetic order if and when there are additions +config X86_NUMACHIP + bool "Numascale NumaChip" + depends on X86_64 + depends on X86_EXTENDED_PLATFORM + depends on NUMA + depends on SMP + depends on X86_X2APIC + depends on !EDAC_AMD64 + ---help--- + Adds support for Numascale NumaChip large-SMP systems. Needed to + enable more than ~168 cores. + If you don't have one of these, you should say N here. config X86_VSMP bool "ScaleMP vSMP" diff --git a/arch/x86/include/asm/numachip/numachip_csr.h b/arch/x86/include/asm/numachip/numachip_csr.h new file mode 100644 index 00000000000..660f843df92 --- /dev/null +++ b/arch/x86/include/asm/numachip/numachip_csr.h @@ -0,0 +1,167 @@ +/* + * 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. + * + * Numascale NumaConnect-Specific Header file + * + * Copyright (C) 2011 Numascale AS. All rights reserved. + * + * Send feedback to + * + */ + +#ifndef _ASM_X86_NUMACHIP_NUMACHIP_CSR_H +#define _ASM_X86_NUMACHIP_NUMACHIP_CSR_H + +#include +#include +#include +#include +#include +#include + +#define CSR_NODE_SHIFT 16 +#define CSR_NODE_BITS(p) (((unsigned long)(p)) << CSR_NODE_SHIFT) +#define CSR_NODE_MASK 0x0fff /* 4K nodes */ + +/* 32K CSR space, b15 indicates geo/non-geo */ +#define CSR_OFFSET_MASK 0x7fffUL + +/* Global CSR space covers all 4K possible nodes with 64K CSR space per node */ +#define NUMACHIP_GCSR_BASE 0x3fff00000000ULL +#define NUMACHIP_GCSR_LIM 0x3fff0fffffffULL +#define NUMACHIP_GCSR_SIZE (NUMACHIP_GCSR_LIM - NUMACHIP_GCSR_BASE + 1) + +/* + * Local CSR space starts in global CSR space with "nodeid" = 0xfff0, however + * when using the direct mapping on x86_64, both start and size needs to be + * aligned with PMD_SIZE which is 2M + */ +#define NUMACHIP_LCSR_BASE 0x3ffffe000000ULL +#define NUMACHIP_LCSR_LIM 0x3fffffffffffULL +#define NUMACHIP_LCSR_SIZE (NUMACHIP_LCSR_LIM - NUMACHIP_LCSR_BASE + 1) + +static inline void *gcsr_address(int node, unsigned long offset) +{ + return __va(NUMACHIP_GCSR_BASE | (1UL << 15) | + CSR_NODE_BITS(node & CSR_NODE_MASK) | (offset & CSR_OFFSET_MASK)); +} + +static inline void *lcsr_address(unsigned long offset) +{ + return __va(NUMACHIP_LCSR_BASE | (1UL << 15) | + CSR_NODE_BITS(0xfff0) | (offset & CSR_OFFSET_MASK)); +} + +static inline unsigned int read_gcsr(int node, unsigned long offset) +{ + return swab32(readl(gcsr_address(node, offset))); +} + +static inline void write_gcsr(int node, unsigned long offset, unsigned int val) +{ + writel(swab32(val), gcsr_address(node, offset)); +} + +static inline unsigned int read_lcsr(unsigned long offset) +{ + return swab32(readl(lcsr_address(offset))); +} + +static inline void write_lcsr(unsigned long offset, unsigned int val) +{ + writel(swab32(val), lcsr_address(offset)); +} + +/* ========================================================================= */ +/* CSR_G0_STATE_CLEAR */ +/* ========================================================================= */ + +#define CSR_G0_STATE_CLEAR (0x000 + (0 << 12)) +union numachip_csr_g0_state_clear { + unsigned int v; + struct numachip_csr_g0_state_clear_s { + unsigned int _state:2; + unsigned int _rsvd_2_6:5; + unsigned int _lost:1; + unsigned int _rsvd_8_31:24; + } s; +}; + +/* ========================================================================= */ +/* CSR_G0_NODE_IDS */ +/* ========================================================================= */ + +#define CSR_G0_NODE_IDS (0x008 + (0 << 12)) +union numachip_csr_g0_node_ids { + unsigned int v; + struct numachip_csr_g0_node_ids_s { + unsigned int _initialid:16; + unsigned int _nodeid:12; + unsigned int _rsvd_28_31:4; + } s; +}; + +/* ========================================================================= */ +/* CSR_G3_EXT_IRQ_GEN */ +/* ========================================================================= */ + +#define CSR_G3_EXT_IRQ_GEN (0x030 + (3 << 12)) +union numachip_csr_g3_ext_irq_gen { + unsigned int v; + struct numachip_csr_g3_ext_irq_gen_s { + unsigned int _vector:8; + unsigned int _msgtype:3; + unsigned int _index:5; + unsigned int _destination_apic_id:16; + } s; +}; + +/* ========================================================================= */ +/* CSR_G3_EXT_IRQ_STATUS */ +/* ========================================================================= */ + +#define CSR_G3_EXT_IRQ_STATUS (0x034 + (3 << 12)) +union numachip_csr_g3_ext_irq_status { + unsigned int v; + struct numachip_csr_g3_ext_irq_status_s { + unsigned int _result:32; + } s; +}; + +/* ========================================================================= */ +/* CSR_G3_EXT_IRQ_DEST */ +/* ========================================================================= */ + +#define CSR_G3_EXT_IRQ_DEST (0x038 + (3 << 12)) +union numachip_csr_g3_ext_irq_dest { + unsigned int v; + struct numachip_csr_g3_ext_irq_dest_s { + unsigned int _irq:8; + unsigned int _rsvd_8_31:24; + } s; +}; + +/* ========================================================================= */ +/* CSR_G3_NC_ATT_MAP_SELECT */ +/* ========================================================================= */ + +#define CSR_G3_NC_ATT_MAP_SELECT (0x7fc + (3 << 12)) +union numachip_csr_g3_nc_att_map_select { + unsigned int v; + struct numachip_csr_g3_nc_att_map_select_s { + unsigned int _upper_address_bits:4; + unsigned int _select_ram:4; + unsigned int _rsvd_8_31:24; + } s; +}; + +/* ========================================================================= */ +/* CSR_G3_NC_ATT_MAP_SELECT_0-255 */ +/* ========================================================================= */ + +#define CSR_G3_NC_ATT_MAP_SELECT_0 (0x800 + (3 << 12)) + +#endif /* _ASM_X86_NUMACHIP_NUMACHIP_CSR_H */ + diff --git a/arch/x86/kernel/apic/Makefile b/arch/x86/kernel/apic/Makefile index 767fd04f284..0ae0323b1f9 100644 --- a/arch/x86/kernel/apic/Makefile +++ b/arch/x86/kernel/apic/Makefile @@ -10,6 +10,7 @@ obj-$(CONFIG_SMP) += ipi.o ifeq ($(CONFIG_X86_64),y) # APIC probe will depend on the listing order here +obj-$(CONFIG_X86_NUMACHIP) += apic_numachip.o obj-$(CONFIG_X86_UV) += x2apic_uv_x.o obj-$(CONFIG_X86_X2APIC) += x2apic_phys.o obj-$(CONFIG_X86_X2APIC) += x2apic_cluster.o diff --git a/arch/x86/kernel/apic/apic_numachip.c b/arch/x86/kernel/apic/apic_numachip.c new file mode 100644 index 00000000000..09d3d8c1cd9 --- /dev/null +++ b/arch/x86/kernel/apic/apic_numachip.c @@ -0,0 +1,294 @@ +/* + * 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. + * + * Numascale NumaConnect-Specific APIC Code + * + * Copyright (C) 2011 Numascale AS. All rights reserved. + * + * Send feedback to + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +static int numachip_system __read_mostly; + +static struct apic apic_numachip __read_mostly; + +static unsigned int get_apic_id(unsigned long x) +{ + unsigned long value; + unsigned int id; + + rdmsrl(MSR_FAM10H_NODE_ID, value); + id = ((x >> 24) & 0xffU) | ((value << 2) & 0x3f00U); + + return id; +} + +static unsigned long set_apic_id(unsigned int id) +{ + unsigned long x; + + x = ((id & 0xffU) << 24); + return x; +} + +static unsigned int read_xapic_id(void) +{ + return get_apic_id(apic_read(APIC_ID)); +} + +static int numachip_apic_id_registered(void) +{ + return physid_isset(read_xapic_id(), phys_cpu_present_map); +} + +static int numachip_phys_pkg_id(int initial_apic_id, int index_msb) +{ + return initial_apic_id >> index_msb; +} + +static const struct cpumask *numachip_target_cpus(void) +{ + return cpu_online_mask; +} + +static void numachip_vector_allocation_domain(int cpu, struct cpumask *retmask) +{ + cpumask_clear(retmask); + cpumask_set_cpu(cpu, retmask); +} + +static int __cpuinit numachip_wakeup_secondary(int phys_apicid, unsigned long start_rip) +{ + union numachip_csr_g3_ext_irq_gen int_gen; + + int_gen.s._destination_apic_id = phys_apicid; + int_gen.s._vector = 0; + int_gen.s._msgtype = APIC_DM_INIT >> 8; + int_gen.s._index = 0; + + write_lcsr(CSR_G3_EXT_IRQ_GEN, int_gen.v); + + int_gen.s._msgtype = APIC_DM_STARTUP >> 8; + int_gen.s._vector = start_rip >> 12; + + write_lcsr(CSR_G3_EXT_IRQ_GEN, int_gen.v); + + atomic_set(&init_deasserted, 1); + return 0; +} + +static void numachip_send_IPI_one(int cpu, int vector) +{ + union numachip_csr_g3_ext_irq_gen int_gen; + int apicid = per_cpu(x86_cpu_to_apicid, cpu); + + int_gen.s._destination_apic_id = apicid; + int_gen.s._vector = vector; + int_gen.s._msgtype = (vector == NMI_VECTOR ? APIC_DM_NMI : APIC_DM_FIXED) >> 8; + int_gen.s._index = 0; + + write_lcsr(CSR_G3_EXT_IRQ_GEN, int_gen.v); +} + +static void numachip_send_IPI_mask(const struct cpumask *mask, int vector) +{ + unsigned int cpu; + + for_each_cpu(cpu, mask) + numachip_send_IPI_one(cpu, vector); +} + +static void numachip_send_IPI_mask_allbutself(const struct cpumask *mask, + int vector) +{ + unsigned int this_cpu = smp_processor_id(); + unsigned int cpu; + + for_each_cpu(cpu, mask) { + if (cpu != this_cpu) + numachip_send_IPI_one(cpu, vector); + } +} + +static void numachip_send_IPI_allbutself(int vector) +{ + unsigned int this_cpu = smp_processor_id(); + unsigned int cpu; + + for_each_online_cpu(cpu) { + if (cpu != this_cpu) + numachip_send_IPI_one(cpu, vector); + } +} + +static void numachip_send_IPI_all(int vector) +{ + numachip_send_IPI_mask(cpu_online_mask, vector); +} + +static void numachip_send_IPI_self(int vector) +{ + __default_send_IPI_shortcut(APIC_DEST_SELF, vector, APIC_DEST_PHYSICAL); +} + +static unsigned int numachip_cpu_mask_to_apicid(const struct cpumask *cpumask) +{ + int cpu; + + /* + * We're using fixed IRQ delivery, can only return one phys APIC ID. + * May as well be the first. + */ + cpu = cpumask_first(cpumask); + if (likely((unsigned)cpu < nr_cpu_ids)) + return per_cpu(x86_cpu_to_apicid, cpu); + + return BAD_APICID; +} + +static unsigned int +numachip_cpu_mask_to_apicid_and(const struct cpumask *cpumask, + const struct cpumask *andmask) +{ + int cpu; + + /* + * We're using fixed IRQ delivery, can only return one phys APIC ID. + * May as well be the first. + */ + for_each_cpu_and(cpu, cpumask, andmask) { + if (cpumask_test_cpu(cpu, cpu_online_mask)) + break; + } + return per_cpu(x86_cpu_to_apicid, cpu); +} + +static int __init numachip_probe(void) +{ + return apic == &apic_numachip; +} + +static void __init map_csrs(void) +{ + printk(KERN_INFO "NumaChip: Mapping local CSR space (%016llx - %016llx)\n", + NUMACHIP_LCSR_BASE, NUMACHIP_LCSR_BASE + NUMACHIP_LCSR_SIZE - 1); + init_extra_mapping_uc(NUMACHIP_LCSR_BASE, NUMACHIP_LCSR_SIZE); + + printk(KERN_INFO "NumaChip: Mapping global CSR space (%016llx - %016llx)\n", + NUMACHIP_GCSR_BASE, NUMACHIP_GCSR_BASE + NUMACHIP_GCSR_SIZE - 1); + init_extra_mapping_uc(NUMACHIP_GCSR_BASE, NUMACHIP_GCSR_SIZE); +} + +static void fixup_cpu_id(struct cpuinfo_x86 *c, int node) +{ + c->phys_proc_id = node; + per_cpu(cpu_llc_id, smp_processor_id()) = node; +} + +static int __init numachip_system_init(void) +{ + unsigned int val; + + if (!numachip_system) + return 0; + + x86_cpuinit.fixup_cpu_id = fixup_cpu_id; + + map_csrs(); + + val = read_lcsr(CSR_G0_NODE_IDS); + printk(KERN_INFO "NumaChip: Local NodeID = %08x\n", val); + + return 0; +} +early_initcall(numachip_system_init); + +static int numachip_acpi_madt_oem_check(char *oem_id, char *oem_table_id) +{ + if (!strncmp(oem_id, "NUMASC", 6)) { + numachip_system = 1; + return 1; + } + + return 0; +} + +static struct apic apic_numachip __refconst = { + + .name = "NumaConnect system", + .probe = numachip_probe, + .acpi_madt_oem_check = numachip_acpi_madt_oem_check, + .apic_id_registered = numachip_apic_id_registered, + + .irq_delivery_mode = dest_Fixed, + .irq_dest_mode = 0, /* physical */ + + .target_cpus = numachip_target_cpus, + .disable_esr = 0, + .dest_logical = 0, + .check_apicid_used = NULL, + .check_apicid_present = NULL, + + .vector_allocation_domain = numachip_vector_allocation_domain, + .init_apic_ldr = flat_init_apic_ldr, + + .ioapic_phys_id_map = NULL, + .setup_apic_routing = NULL, + .multi_timer_check = NULL, + .cpu_present_to_apicid = default_cpu_present_to_apicid, + .apicid_to_cpu_present = NULL, + .setup_portio_remap = NULL, + .check_phys_apicid_present = default_check_phys_apicid_present, + .enable_apic_mode = NULL, + .phys_pkg_id = numachip_phys_pkg_id, + .mps_oem_check = NULL, + + .get_apic_id = get_apic_id, + .set_apic_id = set_apic_id, + .apic_id_mask = 0xffU << 24, + + .cpu_mask_to_apicid = numachip_cpu_mask_to_apicid, + .cpu_mask_to_apicid_and = numachip_cpu_mask_to_apicid_and, + + .send_IPI_mask = numachip_send_IPI_mask, + .send_IPI_mask_allbutself = numachip_send_IPI_mask_allbutself, + .send_IPI_allbutself = numachip_send_IPI_allbutself, + .send_IPI_all = numachip_send_IPI_all, + .send_IPI_self = numachip_send_IPI_self, + + .wakeup_secondary_cpu = numachip_wakeup_secondary, + .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, + .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, + .wait_for_init_deassert = NULL, + .smp_callin_clear_local_apic = NULL, + .inquire_remote_apic = NULL, /* REMRD not supported */ + + .read = native_apic_mem_read, + .write = native_apic_mem_write, + .icr_read = native_apic_icr_read, + .icr_write = native_apic_icr_write, + .wait_icr_idle = native_apic_wait_icr_idle, + .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, +}; +apic_driver(apic_numachip); + -- cgit v1.2.3-70-g09d2 From 346b46be5f10e4d247160ea94ac34450be60ce1e Mon Sep 17 00:00:00 2001 From: Fernando Luis Vázquez Cao Date: Tue, 13 Dec 2011 11:51:53 +0900 Subject: x86: Add per-cpu stat counter for APIC ICR read tries MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In the IPI delivery slow path (NMI delivery) we retry the ICR read to check for delivery completion a limited number of times. [ The reason for the limited retries is that some of the places where it is used (cpu boot, kdump, etc) IPI delivery might not succeed (due to a firmware bug or system crash, for example) and in such a case it is better to give up and resume execution of other code. ] This patch adds a new entry to /proc/interrupts, RTR, which tells user space the number of times we retried the ICR read in the IPI delivery slow path. This should give some insight into how well the APIC message delivery hardware is working - if the counts are way too large then we are hitting a (very-) slow path way too often. Signed-off-by: Fernando Luis Vazquez Cao Cc: Jörn Engel Cc: Suresh Siddha Link: http://lkml.kernel.org/n/tip-vzsp20lo2xdzh5f70g0eis2s@git.kernel.org [ extended the changelog ] Signed-off-by: Ingo Molnar --- arch/x86/include/asm/apic.h | 3 +++ arch/x86/kernel/apic/apic.c | 6 ++++++ arch/x86/kernel/irq.c | 5 +++++ 3 files changed, 14 insertions(+) (limited to 'arch/x86/kernel/apic') diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h index 1a6c09af048..5fe0bd57475 100644 --- a/arch/x86/include/asm/apic.h +++ b/arch/x86/include/asm/apic.h @@ -410,6 +410,9 @@ extern int wakeup_secondary_cpu_via_nmi(int apicid, unsigned long start_eip); #endif #ifdef CONFIG_X86_LOCAL_APIC + +DECLARE_PER_CPU(unsigned, icr_read_retry_count); + static inline u32 apic_read(u32 reg) { return apic->read(reg); diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index f98d84caf94..2942794a9a5 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -79,6 +79,11 @@ DEFINE_EARLY_PER_CPU(u16, x86_bios_cpu_apicid, BAD_APICID); EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_apicid); EXPORT_EARLY_PER_CPU_SYMBOL(x86_bios_cpu_apicid); +/* + * ICR read retry counter + */ +DEFINE_PER_CPU(unsigned, icr_read_retry_count); + #ifdef CONFIG_X86_32 /* @@ -250,6 +255,7 @@ u32 native_safe_apic_wait_icr_idle(void) send_status = apic_read(APIC_ICR) & APIC_ICR_BUSY; if (!send_status) break; + percpu_inc(icr_read_retry_count); udelay(100); } while (timeout++ < 1000); diff --git a/arch/x86/kernel/irq.c b/arch/x86/kernel/irq.c index 429e0c92924..4bbf1627905 100644 --- a/arch/x86/kernel/irq.c +++ b/arch/x86/kernel/irq.c @@ -74,6 +74,10 @@ int arch_show_interrupts(struct seq_file *p, int prec) for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->apic_irq_work_irqs); seq_printf(p, " IRQ work interrupts\n"); + seq_printf(p, "%*s: ", prec, "RTR"); + for_each_online_cpu(j) + seq_printf(p, "%10u ", per_cpu(icr_read_retry_count, j)); + seq_printf(p, " APIC ICR read retries\n"); #endif if (x86_platform_ipi_callback) { seq_printf(p, "%*s: ", prec, "PLT"); @@ -136,6 +140,7 @@ u64 arch_irq_stat_cpu(unsigned int cpu) sum += irq_stats(cpu)->irq_spurious_count; sum += irq_stats(cpu)->apic_perf_irqs; sum += irq_stats(cpu)->apic_irq_work_irqs; + sum += per_cpu(icr_read_retry_count, cpu); #endif if (x86_platform_ipi_callback) sum += irq_stats(cpu)->x86_platform_ipis; -- cgit v1.2.3-70-g09d2 From b49d7d877ff96428c8cd2076b33ba72bf85ceaba Mon Sep 17 00:00:00 2001 From: Fernando Luis Vazquez Cao Date: Thu, 15 Dec 2011 11:32:24 +0900 Subject: x86: Convert per-cpu counter icr_read_retry_count into a member of irq_stat MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit LAPIC related statistics are grouped inside the per-cpu structure irq_stat, so there is no need for icr_read_retry_count to be a standalone per-cpu variable. This patch moves icr_read_retry_count to where it belongs. Suggested-y: Thomas Gleixner Signed-off-by: Fernando Luis Vazquez Cao Cc: Jörn Engel Signed-off-by: Ingo Molnar --- arch/x86/include/asm/apic.h | 2 -- arch/x86/include/asm/hardirq.h | 1 + arch/x86/kernel/apic/apic.c | 7 +------ arch/x86/kernel/irq.c | 4 ++-- 4 files changed, 4 insertions(+), 10 deletions(-) (limited to 'arch/x86/kernel/apic') diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h index 5fe0bd57475..a0f541a3094 100644 --- a/arch/x86/include/asm/apic.h +++ b/arch/x86/include/asm/apic.h @@ -411,8 +411,6 @@ extern int wakeup_secondary_cpu_via_nmi(int apicid, unsigned long start_eip); #ifdef CONFIG_X86_LOCAL_APIC -DECLARE_PER_CPU(unsigned, icr_read_retry_count); - static inline u32 apic_read(u32 reg) { return apic->read(reg); diff --git a/arch/x86/include/asm/hardirq.h b/arch/x86/include/asm/hardirq.h index 55e4de613f0..da0b3ca815b 100644 --- a/arch/x86/include/asm/hardirq.h +++ b/arch/x86/include/asm/hardirq.h @@ -11,6 +11,7 @@ typedef struct { #ifdef CONFIG_X86_LOCAL_APIC unsigned int apic_timer_irqs; /* arch dependent */ unsigned int irq_spurious_count; + unsigned int icr_read_retry_count; #endif unsigned int x86_platform_ipis; /* arch dependent */ unsigned int apic_perf_irqs; diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 2942794a9a5..07832363b72 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -79,11 +79,6 @@ DEFINE_EARLY_PER_CPU(u16, x86_bios_cpu_apicid, BAD_APICID); EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_apicid); EXPORT_EARLY_PER_CPU_SYMBOL(x86_bios_cpu_apicid); -/* - * ICR read retry counter - */ -DEFINE_PER_CPU(unsigned, icr_read_retry_count); - #ifdef CONFIG_X86_32 /* @@ -255,7 +250,7 @@ u32 native_safe_apic_wait_icr_idle(void) send_status = apic_read(APIC_ICR) & APIC_ICR_BUSY; if (!send_status) break; - percpu_inc(icr_read_retry_count); + inc_irq_stat(icr_read_retry_count); udelay(100); } while (timeout++ < 1000); diff --git a/arch/x86/kernel/irq.c b/arch/x86/kernel/irq.c index 4bbf1627905..ef54ed4e307 100644 --- a/arch/x86/kernel/irq.c +++ b/arch/x86/kernel/irq.c @@ -76,7 +76,7 @@ int arch_show_interrupts(struct seq_file *p, int prec) seq_printf(p, " IRQ work interrupts\n"); seq_printf(p, "%*s: ", prec, "RTR"); for_each_online_cpu(j) - seq_printf(p, "%10u ", per_cpu(icr_read_retry_count, j)); + seq_printf(p, "%10u ", irq_stats(j)->icr_read_retry_count); seq_printf(p, " APIC ICR read retries\n"); #endif if (x86_platform_ipi_callback) { @@ -140,7 +140,7 @@ u64 arch_irq_stat_cpu(unsigned int cpu) sum += irq_stats(cpu)->irq_spurious_count; sum += irq_stats(cpu)->apic_perf_irqs; sum += irq_stats(cpu)->apic_irq_work_irqs; - sum += per_cpu(icr_read_retry_count, cpu); + sum += irq_stats(cpu)->icr_read_retry_count; #endif if (x86_platform_ipi_callback) sum += irq_stats(cpu)->x86_platform_ipis; -- cgit v1.2.3-70-g09d2 From e8524b2f43ab6747518aef81c709d104c478b1cd Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 21 Dec 2011 17:45:15 -0800 Subject: x86, apic: Add probe() for apic_flat Currently we start with the default apic_flat mode and switch to some other apic model depending on the apic drivers acpi_madt_oem_check() routines and later followed by the apic drivers probe() routines. Once we selected non flat mode there was no case where we fall back to flat mode again. Upcoming changes allow bios-enabled x2apic mode to be disabled by the OS if interrupt-remapping etc is not setup properly by the bios. We now has a case for the apic to fall back to legacy flat mode during apic driver probe() seqeuence. Add a simple flat_probe() which allows the apic_flat mode to be the last fallback option. Signed-off-by: Yinghai Lu Link: http://lkml.kernel.org/r/20111222014632.484984298@sbsiddha-desk.sc.intel.com Signed-off-by: Suresh Siddha Signed-off-by: H. Peter Anvin --- arch/x86/kernel/apic/apic_flat_64.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'arch/x86/kernel/apic') diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c index 57c1f4135fa..8c3cdded6f2 100644 --- a/arch/x86/kernel/apic/apic_flat_64.c +++ b/arch/x86/kernel/apic/apic_flat_64.c @@ -171,9 +171,14 @@ static int flat_phys_pkg_id(int initial_apic_id, int index_msb) return initial_apic_id >> index_msb; } +static int flat_probe(void) +{ + return 1; +} + static struct apic apic_flat = { .name = "flat", - .probe = NULL, + .probe = flat_probe, .acpi_madt_oem_check = flat_acpi_madt_oem_check, .apic_id_registered = flat_apic_id_registered, -- cgit v1.2.3-70-g09d2 From fb209bd891645bb87b9618b724f0b4928e0df3de Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 21 Dec 2011 17:45:17 -0800 Subject: x86, x2apic: Fallback to xapic when BIOS doesn't setup interrupt-remapping On some of the recent Intel SNB platforms, by default bios is pre-enabling x2apic mode in the cpu with out setting up interrupt-remapping. This case was resulting in the kernel to panic as the cpu is already in x2apic mode but the OS was not able to enable interrupt-remapping (which is a pre-req for using x2apic capability). On these platforms all the apic-ids are < 255 and the kernel can fallback to xapic mode if the bios has not enabled interrupt-remapping (which is mostly the case if the bios has not exported interrupt-remapping tables to the OS). Reported-by: Berck E. Nash Signed-off-by: Yinghai Lu Link: http://lkml.kernel.org/r/20111222014632.600418637@sbsiddha-desk.sc.intel.com Signed-off-by: Suresh Siddha Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/apic.h | 4 +++ arch/x86/include/asm/apicdef.h | 1 + arch/x86/kernel/apic/apic.c | 73 +++++++++++++++++++++++++++++++----------- arch/x86/kernel/apic/io_apic.c | 4 +++ 4 files changed, 64 insertions(+), 18 deletions(-) (limited to 'arch/x86/kernel/apic') diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h index a0f541a3094..a12d57193fe 100644 --- a/arch/x86/include/asm/apic.h +++ b/arch/x86/include/asm/apic.h @@ -176,6 +176,7 @@ static inline u64 native_x2apic_icr_read(void) } extern int x2apic_phys; +extern int x2apic_preenabled; extern void check_x2apic(void); extern void enable_x2apic(void); extern void x2apic_icr_write(u32 low, u32 id); @@ -198,6 +199,9 @@ static inline void x2apic_force_phys(void) x2apic_phys = 1; } #else +static inline void disable_x2apic(void) +{ +} static inline void check_x2apic(void) { } diff --git a/arch/x86/include/asm/apicdef.h b/arch/x86/include/asm/apicdef.h index 3925d800786..134bba00df0 100644 --- a/arch/x86/include/asm/apicdef.h +++ b/arch/x86/include/asm/apicdef.h @@ -144,6 +144,7 @@ #define APIC_BASE (fix_to_virt(FIX_APIC_BASE)) #define APIC_BASE_MSR 0x800 +#define XAPIC_ENABLE (1UL << 11) #define X2APIC_ENABLE (1UL << 10) #ifdef CONFIG_X86_32 diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 07832363b72..2c07aebbb6f 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -146,7 +146,8 @@ __setup("apicpmtimer", setup_apicpmtimer); int x2apic_mode; #ifdef CONFIG_X86_X2APIC /* x2apic enabled before OS handover */ -static int x2apic_preenabled; +int x2apic_preenabled; +static int x2apic_disabled; static __init int setup_nox2apic(char *str) { if (x2apic_enabled()) { @@ -1432,6 +1433,40 @@ void __init bsp_end_local_APIC_setup(void) } #ifdef CONFIG_X86_X2APIC +/* + * Need to disable xapic and x2apic at the same time and then enable xapic mode + */ +static inline void __disable_x2apic(u64 msr) +{ + wrmsrl(MSR_IA32_APICBASE, + msr & ~(X2APIC_ENABLE | XAPIC_ENABLE)); + wrmsrl(MSR_IA32_APICBASE, msr & ~X2APIC_ENABLE); +} + +static void disable_x2apic(void) +{ + u64 msr; + + if (!cpu_has_x2apic) + return; + + rdmsrl(MSR_IA32_APICBASE, msr); + if (msr & X2APIC_ENABLE) { + u32 x2apic_id = read_apic_id(); + + if (x2apic_id >= 255) + panic("Cannot disable x2apic, id: %08x\n", x2apic_id); + + pr_info("Disabling x2apic\n"); + __disable_x2apic(msr); + + x2apic_disabled = 1; + x2apic_mode = 0; + + register_lapic_address(mp_lapic_addr); + } +} + void check_x2apic(void) { if (x2apic_enabled()) { @@ -1442,15 +1477,20 @@ void check_x2apic(void) void enable_x2apic(void) { - int msr, msr2; + u64 msr; + + rdmsrl(MSR_IA32_APICBASE, msr); + if (x2apic_disabled) { + __disable_x2apic(msr); + return; + } if (!x2apic_mode) return; - rdmsr(MSR_IA32_APICBASE, msr, msr2); if (!(msr & X2APIC_ENABLE)) { printk_once(KERN_INFO "Enabling x2apic\n"); - wrmsr(MSR_IA32_APICBASE, msr | X2APIC_ENABLE, msr2); + wrmsrl(MSR_IA32_APICBASE, msr | X2APIC_ENABLE); } } #endif /* CONFIG_X86_X2APIC */ @@ -1487,7 +1527,7 @@ void __init enable_IR_x2apic(void) ret = save_ioapic_entries(); if (ret) { pr_info("Saving IO-APIC state failed: %d\n", ret); - goto out; + return; } local_irq_save(flags); @@ -1499,13 +1539,19 @@ void __init enable_IR_x2apic(void) else ret = enable_IR(); + if (!x2apic_supported()) + goto nox2apic; + if (ret < 0) { /* IR is required if there is APIC ID > 255 even when running * under KVM */ if (max_physical_apicid > 255 || - !hypervisor_x2apic_available()) + !hypervisor_x2apic_available()) { + if (x2apic_preenabled) + disable_x2apic(); goto nox2apic; + } /* * without IR all CPUs can be addressed by IOAPIC/MSI * only in physical mode @@ -1513,8 +1559,10 @@ void __init enable_IR_x2apic(void) x2apic_force_phys(); } - if (ret == IRQ_REMAP_XAPIC_MODE) + if (ret == IRQ_REMAP_XAPIC_MODE) { + pr_info("x2apic not enabled, IRQ remapping is in xapic mode\n"); goto nox2apic; + } x2apic_enabled = 1; @@ -1529,17 +1577,6 @@ nox2apic: restore_ioapic_entries(); legacy_pic->restore_mask(); local_irq_restore(flags); - -out: - if (x2apic_enabled || !x2apic_supported()) - return; - - if (x2apic_preenabled) - panic("x2apic: enabled by BIOS but kernel init failed."); - else if (ret == IRQ_REMAP_XAPIC_MODE) - pr_info("x2apic not enabled, IRQ remapping is in xapic mode\n"); - else if (ret < 0) - pr_info("x2apic not enabled, IRQ remapping init failed\n"); } #ifdef CONFIG_X86_64 diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 6d939d7847e..45b461fdb34 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -2948,6 +2948,10 @@ static inline void __init check_timer(void) } local_irq_disable(); apic_printk(APIC_QUIET, KERN_INFO "..... failed :(.\n"); + if (x2apic_preenabled) + apic_printk(APIC_QUIET, KERN_INFO + "Perhaps problem with the pre-enabled x2apic mode\n" + "Try booting with x2apic and interrupt-remapping disabled in the bios.\n"); panic("IO-APIC + timer doesn't work! Boot with apic=debug and send a " "report. Then try booting with the 'noapic' option.\n"); out: -- cgit v1.2.3-70-g09d2 From a31bc32760992a2c68f3d6bf7da9f760c0fd7c41 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Fri, 23 Dec 2011 11:01:43 -0800 Subject: x86, x2apic: Allow "nox2apic" to disable x2apic mode setup by BIOS Currently "nox2apic" boot parameter was not enabling x2apic mode if the cpu, kernel are all capable of enabling x2apic mode and the OS handover happened in xapic mode. However If the bios enabled x2apic prior to OS handover, using "nox2apic" boot parameter had no effect. If the boot cpu's apicid is < 255, enable "nox2apic" boot parameter to disable the x2apic mode setup by the bios. This will enable the kernel to fallback to xapic mode and bringup only the cpu's which has apic-id < 255. -v2: fix patch error and two compiling warning make disable_x2apic to be __init Signed-off-by: Yinghai Lu Signed-off-by: Suresh Siddha Link: http://lkml.kernel.org/r/CAE9FiQUeB-3uxJAMiHsz=uPWoFv5Hg1pVepz7aU6YtqOxMC-=Q@mail.gmail.com Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/apic.h | 1 + arch/x86/kernel/apic/apic.c | 37 +++++++++++++++++++++++++++---------- 2 files changed, 28 insertions(+), 10 deletions(-) (limited to 'arch/x86/kernel/apic') diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h index a12d57193fe..3ab9bdd87e7 100644 --- a/arch/x86/include/asm/apic.h +++ b/arch/x86/include/asm/apic.h @@ -216,6 +216,7 @@ static inline void x2apic_force_phys(void) { } +#define nox2apic 0 #define x2apic_preenabled 0 #define x2apic_supported() 0 #endif diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 2c07aebbb6f..ff69d5d79ca 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -148,15 +148,24 @@ int x2apic_mode; /* x2apic enabled before OS handover */ int x2apic_preenabled; static int x2apic_disabled; +static int nox2apic; static __init int setup_nox2apic(char *str) { if (x2apic_enabled()) { - pr_warning("Bios already enabled x2apic, " - "can't enforce nox2apic"); - return 0; - } + int apicid = native_apic_msr_read(APIC_ID); + + if (apicid >= 255) { + pr_warning("Apicid: %08x, cannot enforce nox2apic\n", + apicid); + return 0; + } + + pr_warning("x2apic already enabled. will disable it\n"); + } else + setup_clear_cpu_cap(X86_FEATURE_X2APIC); + + nox2apic = 1; - setup_clear_cpu_cap(X86_FEATURE_X2APIC); return 0; } early_param("nox2apic", setup_nox2apic); @@ -1443,7 +1452,7 @@ static inline void __disable_x2apic(u64 msr) wrmsrl(MSR_IA32_APICBASE, msr & ~X2APIC_ENABLE); } -static void disable_x2apic(void) +static __init void disable_x2apic(void) { u64 msr; @@ -1460,6 +1469,11 @@ static void disable_x2apic(void) pr_info("Disabling x2apic\n"); __disable_x2apic(msr); + if (nox2apic) { + clear_cpu_cap(&cpu_data(0), X86_FEATURE_X2APIC); + setup_clear_cpu_cap(X86_FEATURE_X2APIC); + } + x2apic_disabled = 1; x2apic_mode = 0; @@ -1534,13 +1548,16 @@ void __init enable_IR_x2apic(void) legacy_pic->mask_all(); mask_ioapic_entries(); + if (x2apic_preenabled && nox2apic) + disable_x2apic(); + if (dmar_table_init_ret) ret = -1; else ret = enable_IR(); if (!x2apic_supported()) - goto nox2apic; + goto skip_x2apic; if (ret < 0) { /* IR is required if there is APIC ID > 255 even when running @@ -1550,7 +1567,7 @@ void __init enable_IR_x2apic(void) !hypervisor_x2apic_available()) { if (x2apic_preenabled) disable_x2apic(); - goto nox2apic; + goto skip_x2apic; } /* * without IR all CPUs can be addressed by IOAPIC/MSI @@ -1561,7 +1578,7 @@ void __init enable_IR_x2apic(void) if (ret == IRQ_REMAP_XAPIC_MODE) { pr_info("x2apic not enabled, IRQ remapping is in xapic mode\n"); - goto nox2apic; + goto skip_x2apic; } x2apic_enabled = 1; @@ -1572,7 +1589,7 @@ void __init enable_IR_x2apic(void) pr_info("Enabled x2apic\n"); } -nox2apic: +skip_x2apic: if (ret < 0) /* IR enabling failed */ restore_ioapic_entries(); legacy_pic->restore_mask(); -- cgit v1.2.3-70-g09d2