From d22616942804798105e61428afa41a9132421bb9 Mon Sep 17 00:00:00 2001 From: Huang Weiyi Date: Mon, 6 Apr 2009 17:16:46 +0800 Subject: ACPI: cpufreq: remove dupilcated #include Remove dupilicated #include in arch/x86/kernel/cpu/cpufreq/longhaul.c. Signed-off-by: Huang Weiyi Signed-off-by: Len Brown --- arch/x86/kernel/cpu/cpufreq/longhaul.c | 1 - 1 file changed, 1 deletion(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/cpufreq/longhaul.c b/arch/x86/kernel/cpu/cpufreq/longhaul.c index 0bd48e65a0c..ce2ed3e4aad 100644 --- a/arch/x86/kernel/cpu/cpufreq/longhaul.c +++ b/arch/x86/kernel/cpu/cpufreq/longhaul.c @@ -33,7 +33,6 @@ #include #include #include -#include #include #include -- cgit v1.2.3-70-g09d2 From e4f6937222dbb61b8b8e62caca3d32e648b3b14b Mon Sep 17 00:00:00 2001 From: Venkatesh Pallipadi Date: Mon, 6 Apr 2009 11:26:07 -0700 Subject: ACPI x86: Cleanup acpi_cpufreq structures related to aperf/mperf Change structure name to make the code cleaner and simpler. No functionality change in this patch. Signed-off-by: Venkatesh Pallipadi Signed-off-by: Len Brown --- arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c | 42 ++++++++++++++---------------- 1 file changed, 20 insertions(+), 22 deletions(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index 19f6b9d27e8..340bdbebba0 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -241,23 +241,23 @@ static u32 get_cur_val(const struct cpumask *mask) return cmd.val; } -struct perf_cur { +struct perf_pair { union { struct { u32 lo; u32 hi; } split; u64 whole; - } aperf_cur, mperf_cur; + } aperf, mperf; }; static long read_measured_perf_ctrs(void *_cur) { - struct perf_cur *cur = _cur; + struct perf_pair *cur = _cur; - rdmsr(MSR_IA32_APERF, cur->aperf_cur.split.lo, cur->aperf_cur.split.hi); - rdmsr(MSR_IA32_MPERF, cur->mperf_cur.split.lo, cur->mperf_cur.split.hi); + rdmsr(MSR_IA32_APERF, cur->aperf.split.lo, cur->aperf.split.hi); + rdmsr(MSR_IA32_MPERF, cur->mperf.split.lo, cur->mperf.split.hi); wrmsr(MSR_IA32_APERF, 0, 0); wrmsr(MSR_IA32_MPERF, 0, 0); @@ -281,7 +281,7 @@ static long read_measured_perf_ctrs(void *_cur) static unsigned int get_measured_perf(struct cpufreq_policy *policy, unsigned int cpu) { - struct perf_cur cur; + struct perf_pair cur; unsigned int perf_percent; unsigned int retval; @@ -294,39 +294,37 @@ static unsigned int get_measured_perf(struct cpufreq_policy *policy, * Get an approximate value. Return failure in case we cannot get * an approximate value. */ - if (unlikely(cur.aperf_cur.split.hi || cur.mperf_cur.split.hi)) { + if (unlikely(cur.aperf.split.hi || cur.mperf.split.hi)) { int shift_count; u32 h; - h = max_t(u32, cur.aperf_cur.split.hi, cur.mperf_cur.split.hi); + h = max_t(u32, cur.aperf.split.hi, cur.mperf.split.hi); shift_count = fls(h); - cur.aperf_cur.whole >>= shift_count; - cur.mperf_cur.whole >>= shift_count; + cur.aperf.whole >>= shift_count; + cur.mperf.whole >>= shift_count; } - if (((unsigned long)(-1) / 100) < cur.aperf_cur.split.lo) { + if (((unsigned long)(-1) / 100) < cur.aperf.split.lo) { int shift_count = 7; - cur.aperf_cur.split.lo >>= shift_count; - cur.mperf_cur.split.lo >>= shift_count; + cur.aperf.split.lo >>= shift_count; + cur.mperf.split.lo >>= shift_count; } - if (cur.aperf_cur.split.lo && cur.mperf_cur.split.lo) - perf_percent = (cur.aperf_cur.split.lo * 100) / - cur.mperf_cur.split.lo; + if (cur.aperf.split.lo && cur.mperf.split.lo) + perf_percent = (cur.aperf.split.lo * 100) / cur.mperf.split.lo; else perf_percent = 0; #else - if (unlikely(((unsigned long)(-1) / 100) < cur.aperf_cur.whole)) { + if (unlikely(((unsigned long)(-1) / 100) < cur.aperf.whole)) { int shift_count = 7; - cur.aperf_cur.whole >>= shift_count; - cur.mperf_cur.whole >>= shift_count; + cur.aperf.whole >>= shift_count; + cur.mperf.whole >>= shift_count; } - if (cur.aperf_cur.whole && cur.mperf_cur.whole) - perf_percent = (cur.aperf_cur.whole * 100) / - cur.mperf_cur.whole; + if (cur.aperf.whole && cur.mperf.whole) + perf_percent = (cur.aperf.whole * 100) / cur.mperf.whole; else perf_percent = 0; -- cgit v1.2.3-70-g09d2 From 18b2646fe3babeb40b34a0c1751e0bf5adfdc64c Mon Sep 17 00:00:00 2001 From: Venkatesh Pallipadi Date: Mon, 6 Apr 2009 11:26:08 -0700 Subject: ACPI x86: Make aperf/mperf MSR access in acpi_cpufreq read_only Do not write zeroes to APERF and MPERF by ondemand governor. With this change, other users can share these MSRs for reads. Signed-off-by: Venkatesh Pallipadi Signed-off-by: Len Brown --- arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index 340bdbebba0..9d3af380c6b 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -68,6 +68,7 @@ struct acpi_cpufreq_data { unsigned int max_freq; unsigned int resume; unsigned int cpu_feature; + u64 saved_aperf, saved_mperf; }; static DEFINE_PER_CPU(struct acpi_cpufreq_data *, drv_data); @@ -259,9 +260,6 @@ static long read_measured_perf_ctrs(void *_cur) rdmsr(MSR_IA32_APERF, cur->aperf.split.lo, cur->aperf.split.hi); rdmsr(MSR_IA32_MPERF, cur->mperf.split.lo, cur->mperf.split.hi); - wrmsr(MSR_IA32_APERF, 0, 0); - wrmsr(MSR_IA32_MPERF, 0, 0); - return 0; } @@ -281,13 +279,20 @@ static long read_measured_perf_ctrs(void *_cur) static unsigned int get_measured_perf(struct cpufreq_policy *policy, unsigned int cpu) { - struct perf_pair cur; + struct perf_pair readin, cur; unsigned int perf_percent; unsigned int retval; - if (!work_on_cpu(cpu, read_measured_perf_ctrs, &cur)) + if (!work_on_cpu(cpu, read_measured_perf_ctrs, &readin)) return 0; + cur.aperf.whole = readin.aperf.whole - + per_cpu(drv_data, cpu)->saved_aperf; + cur.mperf.whole = readin.mperf.whole - + per_cpu(drv_data, cpu)->saved_mperf; + per_cpu(drv_data, cpu)->saved_aperf = readin.aperf.whole; + per_cpu(drv_data, cpu)->saved_mperf = readin.mperf.whole; + #ifdef __i386__ /* * We dont want to do 64 bit divide with 32 bit kernel -- cgit v1.2.3-70-g09d2 From db954b5898dd3ef3ef93f4144158ea8f97deb058 Mon Sep 17 00:00:00 2001 From: Venkatesh Pallipadi Date: Mon, 6 Apr 2009 18:51:29 -0700 Subject: x86 ACPI: Add support for Always Running APIC timer Add support for Always Running APIC timer, CPUID_0x6_EAX_Bit2. This bit means the APIC timer continues to run even when CPU is in deep C-states. The advantage is that we can use LAPIC timer on these CPUs always, and there is no need for "slow to read and program" external timers (HPET/PIT) and the timer broadcast logic and related code in C-state entry and exit. Signed-off-by: Venkatesh Pallipadi Acked-by: H. Peter Anvin Signed-off-by: Len Brown --- arch/x86/include/asm/cpufeature.h | 1 + arch/x86/kernel/apic/apic.c | 6 ++++++ arch/x86/kernel/cpu/addon_cpuid_features.c | 1 + drivers/acpi/processor_idle.c | 3 +++ 4 files changed, 11 insertions(+) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h index 0beba0d1468..bb83b1c397a 100644 --- a/arch/x86/include/asm/cpufeature.h +++ b/arch/x86/include/asm/cpufeature.h @@ -154,6 +154,7 @@ * CPUID levels like 0x6, 0xA etc */ #define X86_FEATURE_IDA (7*32+ 0) /* Intel Dynamic Acceleration */ +#define X86_FEATURE_ARAT (7*32+ 1) /* Always Running APIC Timer */ /* Virtualization flags: Linux defined */ #define X86_FEATURE_TPR_SHADOW (8*32+ 0) /* Intel TPR Shadow */ diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 098ec84b8c0..f2870920f24 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -431,6 +431,12 @@ static void __cpuinit setup_APIC_timer(void) { struct clock_event_device *levt = &__get_cpu_var(lapic_events); + if (cpu_has(¤t_cpu_data, X86_FEATURE_ARAT)) { + lapic_clockevent.features &= ~CLOCK_EVT_FEAT_C3STOP; + /* Make LAPIC timer preferrable over percpu HPET */ + lapic_clockevent.rating = 150; + } + memcpy(levt, &lapic_clockevent, sizeof(*levt)); levt->cpumask = cpumask_of(smp_processor_id()); diff --git a/arch/x86/kernel/cpu/addon_cpuid_features.c b/arch/x86/kernel/cpu/addon_cpuid_features.c index 8220ae69849..c965e521271 100644 --- a/arch/x86/kernel/cpu/addon_cpuid_features.c +++ b/arch/x86/kernel/cpu/addon_cpuid_features.c @@ -31,6 +31,7 @@ void __cpuinit init_scattered_cpuid_features(struct cpuinfo_x86 *c) static const struct cpuid_bit __cpuinitconst cpuid_bits[] = { { X86_FEATURE_IDA, CR_EAX, 1, 0x00000006 }, + { X86_FEATURE_ARAT, CR_EAX, 2, 0x00000006 }, { 0, 0, 0, 0 } }; diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 4e6e758bd39..6fe121434ff 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -145,6 +145,9 @@ static void acpi_timer_check_state(int state, struct acpi_processor *pr, struct acpi_processor_power *pwr = &pr->power; u8 type = local_apic_timer_c2_ok ? ACPI_STATE_C3 : ACPI_STATE_C2; + if (cpu_has(&cpu_data(pr->id), X86_FEATURE_ARAT)) + return; + /* * Check, if one of the previous states already marked the lapic * unstable -- cgit v1.2.3-70-g09d2 From f20ab9c38fb85b4dde8b4139788ab9e735a35279 Mon Sep 17 00:00:00 2001 From: Jaswinder Singh Rajput Date: Wed, 8 Apr 2009 22:49:46 +0530 Subject: x86: cpu_debug remove execute permission It seems by mistake these files got execute permissions so removing it. Signed-off-by: Jaswinder Singh Rajput LKML-Reference: <1239211186.9037.2.camel@ht.satnam> Signed-off-by: Ingo Molnar --- arch/x86/include/asm/cpu_debug.h | 0 arch/x86/kernel/cpu/cpu_debug.c | 0 2 files changed, 0 insertions(+), 0 deletions(-) mode change 100755 => 100644 arch/x86/include/asm/cpu_debug.h mode change 100755 => 100644 arch/x86/kernel/cpu/cpu_debug.c (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/include/asm/cpu_debug.h b/arch/x86/include/asm/cpu_debug.h old mode 100755 new mode 100644 diff --git a/arch/x86/kernel/cpu/cpu_debug.c b/arch/x86/kernel/cpu/cpu_debug.c old mode 100755 new mode 100644 -- cgit v1.2.3-70-g09d2 From 01599fca6758d2cd133e78f87426fc851c9ea725 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Mon, 13 Apr 2009 10:27:49 -0700 Subject: cpufreq: use smp_call_function_[single|many]() in acpi-cpufreq.c Atttempting to rid us of the problematic work_on_cpu(). Just use smp_call_fuction_single() here. This repairs a 10% sysbench(oltp)+mysql regression which Mike reported, due to commit 6b44003e5ca66a3fffeb5bc90f40ada2c4340896 Author: Andrew Morton Date: Thu Apr 9 09:50:37 2009 -0600 work_on_cpu(): rewrite it to create a kernel thread on demand It seems that the kernel calls these acpi-cpufreq functions at a quite high frequency. Valdis Kletnieks also reports that this causes 70-90 forks per second on his hardware. Cc: Valdis.Kletnieks@vt.edu Cc: Rusty Russell Cc: Venkatesh Pallipadi Cc: Len Brown Cc: Zhao Yakui Acked-by: Dave Jones Cc: Thomas Gleixner Tested-by: Mike Galbraith Cc: "Zhang, Yanmin" Signed-off-by: Andrew Morton Acked-by: Ingo Molnar [ Made it use smp_call_function_many() instead of looping over cpu's with smp_call_function_single() - Linus ] Signed-off-by: Linus Torvalds --- arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c | 24 +++++++++--------------- 1 file changed, 9 insertions(+), 15 deletions(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index 9d3af380c6b..3e3cd3db7a0 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -153,7 +153,8 @@ struct drv_cmd { u32 val; }; -static long do_drv_read(void *_cmd) +/* Called via smp_call_function_single(), on the target CPU */ +static void do_drv_read(void *_cmd) { struct drv_cmd *cmd = _cmd; u32 h; @@ -170,10 +171,10 @@ static long do_drv_read(void *_cmd) default: break; } - return 0; } -static long do_drv_write(void *_cmd) +/* Called via smp_call_function_many(), on the target CPUs */ +static void do_drv_write(void *_cmd) { struct drv_cmd *cmd = _cmd; u32 lo, hi; @@ -192,23 +193,18 @@ static long do_drv_write(void *_cmd) default: break; } - return 0; } static void drv_read(struct drv_cmd *cmd) { cmd->val = 0; - work_on_cpu(cpumask_any(cmd->mask), do_drv_read, cmd); + smp_call_function_single(cpumask_any(cmd->mask), do_drv_read, cmd, 1); } static void drv_write(struct drv_cmd *cmd) { - unsigned int i; - - for_each_cpu(i, cmd->mask) { - work_on_cpu(i, do_drv_write, cmd); - } + smp_call_function_many(cmd->mask, do_drv_write, cmd, 1); } static u32 get_cur_val(const struct cpumask *mask) @@ -252,15 +248,13 @@ struct perf_pair { } aperf, mperf; }; - -static long read_measured_perf_ctrs(void *_cur) +/* Called via smp_call_function_single(), on the target CPU */ +static void read_measured_perf_ctrs(void *_cur) { struct perf_pair *cur = _cur; rdmsr(MSR_IA32_APERF, cur->aperf.split.lo, cur->aperf.split.hi); rdmsr(MSR_IA32_MPERF, cur->mperf.split.lo, cur->mperf.split.hi); - - return 0; } /* @@ -283,7 +277,7 @@ static unsigned int get_measured_perf(struct cpufreq_policy *policy, unsigned int perf_percent; unsigned int retval; - if (!work_on_cpu(cpu, read_measured_perf_ctrs, &readin)) + if (smp_call_function_single(cpu, read_measured_perf_ctrs, &cur, 1)) return 0; cur.aperf.whole = readin.aperf.whole - -- cgit v1.2.3-70-g09d2 From 1c98aa7424ff163637d8321674ec58dee28152d4 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 13 Apr 2009 18:09:20 -0700 Subject: Fix quilt merge error in acpi-cpufreq.c We ended up incorrectly using '&cur' instead of '&readin' in the work_on_cpu() -> smp_call_function_single() transformation in commit 01599fca6758d2cd133e78f87426fc851c9ea725 ("cpufreq: use smp_call_function_[single|many]() in acpi-cpufreq.c"). Andrew explains: "OK, the acpi tree went and had conflicting changes merged into it after I'd written the patch and it appears that I incorrectly reverted part of 18b2646fe3babeb40b34a0c1751e0bf5adfdc64c while fixing the resulting rejects. Switching it to `readin' looks correct." Acked-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index 3e3cd3db7a0..837c2c4cc20 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -277,7 +277,7 @@ static unsigned int get_measured_perf(struct cpufreq_policy *policy, unsigned int perf_percent; unsigned int retval; - if (smp_call_function_single(cpu, read_measured_perf_ctrs, &cur, 1)) + if (smp_call_function_single(cpu, read_measured_perf_ctrs, &readin, 1)) return 0; cur.aperf.whole = readin.aperf.whole - -- cgit v1.2.3-70-g09d2 From ea34f43a074af85823e49b9bf62f47d8d3f0e81a Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Wed, 15 Apr 2009 08:05:13 -0700 Subject: acpi-cpufreq: fix 'smp_call_function_many()' confusion It turns out that 'smp_call_function_many()' doesn't work at all like 'smp_call_function_single()', and my change to Andrew's patch to use it rather than a loop over all CPU's acpi-cpufreq doesn't work. My bad. 'smp_call_function_many()' has two "features" (aka "documented bugs"): (a) it needs to be called with preemption disabled, because it uses smp_processor_id() without guarding the CPU lookup with 'get_cpu()' and 'put_cpu()' like the 'single' variant does. (b) even if the current CPU is part of the CPU mask, it won't do the call on that CPU. Still, we're better off trying to use 'smp_call_function_many()' than looping over CPU's, since it at least in theory allows us to use a broadcast IPI and do it all in parallel. So let's just work around the silly semantic bugs in that function. Reported-and-tested-by: Ali Gholami Rudi Cc: Ingo Molnar Cc: Andrew Morton , Cc: Rusty Russell Cc: Dave Jones Signed-off-by: Linus Torvalds --- arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index 837c2c4cc20..ecdb682ab51 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -204,7 +204,13 @@ static void drv_read(struct drv_cmd *cmd) static void drv_write(struct drv_cmd *cmd) { + int this_cpu; + + this_cpu = get_cpu(); + if (cpumask_test_cpu(this_cpu, cmd->mask)) + do_drv_write(cmd); smp_call_function_many(cmd->mask, do_drv_write, cmd, 1); + put_cpu(); } static u32 get_cur_val(const struct cpumask *mask) -- cgit v1.2.3-70-g09d2 From 093f13e23137b9e5f7629dd5932ceea1419e2b61 Mon Sep 17 00:00:00 2001 From: "Pallipadi, Venkatesh" Date: Wed, 15 Apr 2009 10:37:33 -0700 Subject: x86, acpi_cpufreq: Fix the NULL pointer dereference in get_measured_perf Fix for a regression that was introduced by earlier commit 18b2646fe3babeb40b34a0c1751e0bf5adfdc64c on Mon Apr 6 11:26:08 2009 Regression resulted in the below error happened on systems with software coordination where per_cpu acpi data will not be initiated for secondary CPUs in a P-state domain. On Tue, 2009-04-14 at 23:01 -0700, Zhang, Yanmin wrote: My machine hanged with kernel 2.6.30-rc2 when script read > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor. > > opps happens in get_measured_perf: > > cur.aperf.whole = readin.aperf.whole - > per_cpu(drv_data, cpu)->saved_aperf; > > Because per_cpu(drv_data, cpu)=NULL. > > So function get_measured_perf should check if (per_cpu(drv_data, > cpu)==NULL) > and return 0 if it's NULL. --------------sys log------------------ BUG: unable to handle kernel NULL pointer dereference at 0000000000000020 IP: [] get_measured_perf+0x4a/0xf9 PGD a7dd88067 PUD a7ccf5067 PMD 0 Oops: 0000 [#1] SMP last sysfs file: /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor CPU 0 Modules linked in: video output Pid: 2091, comm: kondemand/0 Not tainted 2.6.30-rc2 #1 MP Server RIP: 0010:[] [] get_measured_perf+0x4a/0xf9 RSP: 0018:ffff880a7d56de20 EFLAGS: 00010246 RAX: 0000000000000000 RBX: 00000046241a42b6 RCX: ffff88004d219000 RDX: 000000000000b660 RSI: 0000000000000020 RDI: 0000000000000001 RBP: ffff880a7f052000 R08: 00000046241a42b6 R09: ffffffff807639f0 R10: 00000000ffffffea R11: ffffffff802207f4 R12: ffff880a7f052000 R13: ffff88004d20e460 R14: 0000000000ddd5a6 R15: 0000000000000001 FS: 0000000000000000(0000) GS:ffff88004d200000(0000) knlGS:0000000000000000 CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b CR2: 0000000000000020 CR3: 0000000a7f1bf000 CR4: 00000000000006e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process kondemand/0 (pid: 2091, threadinfo ffff880a7d56c000, task ffff880a7d4d18c0) Stack: ffff880a7f052078 ffffffff803efd54 00000046241a42b6 000000462ffa9e95 0000000000000001 0000000000000001 00000000ffffffea ffffffff8064f41a 0000000000000012 0000000000000012 ffff880a7f052000 ffffffff80650547 Call Trace: [] ? kobject_get+0x12/0x17 [] ? __cpufreq_driver_getavg+0x42/0x57 [] ? do_dbs_timer+0x147/0x272 [] ? do_dbs_timer+0x0/0x272 [] ? worker_thread+0x15b/0x1f5 [] ? autoremove_wake_function+0x0/0x2e [] ? worker_thread+0x0/0x1f5 [] ? kthread+0x54/0x83 [] ? child_rip+0xa/0x20 [] ? kthread+0x0/0x83 [] ? child_rip+0x0/0x20 Code: 99 a6 03 00 31 c9 85 c0 0f 85 c3 00 00 00 89 df 4c 8b 44 24 10 48 c7 c2 60 b6 00 00 48 8b 0c fd e0 30 a5 80 4c 89 c3 48 8b 04 0a <48> 2b 58 20 48 8b 44 24 18 48 89 1c 24 48 8b 34 0a 48 2b 46 28 RIP [] get_measured_perf+0x4a/0xf9 RSP CR2: 0000000000000020 ---[ end trace 2b8fac9a49e19ad4 ]--- Tested-by: "Zhang, Yanmin" Signed-off-by: Venkatesh Pallipadi Signed-off-by: Len Brown --- arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index ecdb682ab51..dd0bd76d14c 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -68,11 +68,16 @@ struct acpi_cpufreq_data { unsigned int max_freq; unsigned int resume; unsigned int cpu_feature; - u64 saved_aperf, saved_mperf; }; static DEFINE_PER_CPU(struct acpi_cpufreq_data *, drv_data); +struct acpi_msr_data { + u64 saved_aperf, saved_mperf; +}; + +static DEFINE_PER_CPU(struct acpi_msr_data, msr_data); + DEFINE_TRACE(power_mark); /* acpi_perf_data is a pointer to percpu data. */ @@ -287,11 +292,11 @@ static unsigned int get_measured_perf(struct cpufreq_policy *policy, return 0; cur.aperf.whole = readin.aperf.whole - - per_cpu(drv_data, cpu)->saved_aperf; + per_cpu(msr_data, cpu).saved_aperf; cur.mperf.whole = readin.mperf.whole - - per_cpu(drv_data, cpu)->saved_mperf; - per_cpu(drv_data, cpu)->saved_aperf = readin.aperf.whole; - per_cpu(drv_data, cpu)->saved_mperf = readin.mperf.whole; + per_cpu(msr_data, cpu).saved_mperf; + per_cpu(msr_data, cpu).saved_aperf = readin.aperf.whole; + per_cpu(msr_data, cpu).saved_mperf = readin.mperf.whole; #ifdef __i386__ /* -- cgit v1.2.3-70-g09d2 From e0e8c4e512e92bc25c19bd8d4926de17d2f8fbf2 Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Fri, 17 Apr 2009 16:22:06 +0200 Subject: acpi-cpufreq: Cleanup: Use printk_once Signed-off-by: Thomas Renninger Signed-off-by: Len Brown --- arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index dd0bd76d14c..4eab747a896 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -693,13 +693,9 @@ static int acpi_cpufreq_cpu_init(struct cpufreq_policy *policy) /* Check for high latency (>20uS) from buggy BIOSes, like on T42 */ if (perf->control_register.space_id == ACPI_ADR_SPACE_FIXED_HARDWARE && policy->cpuinfo.transition_latency > 20 * 1000) { - static int print_once; policy->cpuinfo.transition_latency = 20 * 1000; - if (!print_once) { - print_once = 1; - printk(KERN_INFO "Capping off P-state tranision latency" - " at 20 uS\n"); - } + printk_once(KERN_INFO "Capping off P-state tranision" + " latency at 20 uS\n"); } data->max_freq = perf->states[0].core_frequency * 1000; -- cgit v1.2.3-70-g09d2 From d91758f5ddb80e91176fa2cf80c88c1633950b3d Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Fri, 17 Apr 2009 16:22:07 +0200 Subject: acpi-cpufreq: style-only: add parens to math expression Signed-off-by: Thomas Renninger Signed-off-by: Len Brown --- arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index 4eab747a896..aec3161abed 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -340,7 +340,7 @@ static unsigned int get_measured_perf(struct cpufreq_policy *policy, #endif - retval = per_cpu(drv_data, policy->cpu)->max_freq * perf_percent / 100; + retval = (per_cpu(drv_data, policy->cpu)->max_freq * perf_percent) / 100; return retval; } -- cgit v1.2.3-70-g09d2 From d876dfbbf5c8728102fb4f683450fa9ae3259cda Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Fri, 17 Apr 2009 16:22:08 +0200 Subject: acpi-cpufreq: Do not let get_measured perf depend on internal variable Take already available policy->cpuinfo.max_freq and get rid of acpi-cpufreq specific max_freq variable. This implies that P0 is always the highest frequency which should always be true as ACPI spec says: As a result, the zeroth entry describes the highest performance state Signed-off-by: Thomas Renninger Acked-by: Venkatesh Pallipadi Signed-off-by: Len Brown --- arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index aec3161abed..208ecf6643d 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -65,7 +65,6 @@ enum { struct acpi_cpufreq_data { struct acpi_processor_performance *acpi_data; struct cpufreq_frequency_table *freq_table; - unsigned int max_freq; unsigned int resume; unsigned int cpu_feature; }; @@ -340,7 +339,7 @@ static unsigned int get_measured_perf(struct cpufreq_policy *policy, #endif - retval = (per_cpu(drv_data, policy->cpu)->max_freq * perf_percent) / 100; + retval = (policy->cpuinfo.max_freq * perf_percent) / 100; return retval; } @@ -698,7 +697,6 @@ static int acpi_cpufreq_cpu_init(struct cpufreq_policy *policy) " latency at 20 uS\n"); } - data->max_freq = perf->states[0].core_frequency * 1000; /* table init */ for (i = 0; i < perf->state_count; i++) { if (i > 0 && perf->states[i].core_frequency >= @@ -717,6 +715,9 @@ static int acpi_cpufreq_cpu_init(struct cpufreq_policy *policy) if (result) goto err_freqfree; + if (perf->states[0].core_frequency * 1000 != policy->cpuinfo.max_freq) + printk(KERN_WARNING FW_WARN "P-state 0 is not max freq\n"); + switch (perf->control_register.space_id) { case ACPI_ADR_SPACE_SYSTEM_IO: /* Current speed is unknown and not detectable by IO port */ -- cgit v1.2.3-70-g09d2 From 6298c512bc1007c3ff5c9ce20e6996781651cc45 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Thu, 9 Apr 2009 12:28:22 +0200 Subject: x86, mce: make polling timer interval per CPU The polling timer while running per CPU still uses a global next_interval variable, which lead to some CPUs either polling too fast or too slow. This was not a serious problem because all errors get picked up eventually, but it's still better to avoid it. Turn next_interval into a per cpu variable. v2: Fix check_interval == 0 case (Hidetoshi Seto) [ Impact: minor bug fix ] Signed-off-by: Andi Kleen Reviewed-by: Hidetoshi Seto Signed-off-by: H. Peter Anvin --- arch/x86/kernel/cpu/mcheck/mce_64.c | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/mcheck/mce_64.c b/arch/x86/kernel/cpu/mcheck/mce_64.c index 863f89568b1..82614f1b923 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_64.c @@ -452,13 +452,14 @@ void mce_log_therm_throt_event(__u64 status) */ static int check_interval = 5 * 60; /* 5 minutes */ -static int next_interval; /* in jiffies */ +static DEFINE_PER_CPU(int, next_interval); /* in jiffies */ static void mcheck_timer(unsigned long); static DEFINE_PER_CPU(struct timer_list, mce_timer); static void mcheck_timer(unsigned long data) { struct timer_list *t = &per_cpu(mce_timer, data); + int *n; WARN_ON(smp_processor_id() != data); @@ -470,14 +471,14 @@ static void mcheck_timer(unsigned long data) * Alert userspace if needed. If we logged an MCE, reduce the * polling interval, otherwise increase the polling interval. */ + n = &__get_cpu_var(next_interval); if (mce_notify_user()) { - next_interval = max(next_interval/2, HZ/100); + *n = max(*n/2, HZ/100); } else { - next_interval = min(next_interval * 2, - (int)round_jiffies_relative(check_interval*HZ)); + *n = min(*n*2, (int)round_jiffies_relative(check_interval*HZ)); } - t->expires = jiffies + next_interval; + t->expires = jiffies + *n; add_timer(t); } @@ -632,14 +633,13 @@ static void mce_cpu_features(struct cpuinfo_x86 *c) static void mce_init_timer(void) { struct timer_list *t = &__get_cpu_var(mce_timer); + int *n = &__get_cpu_var(next_interval); - /* data race harmless because everyone sets to the same value */ - if (!next_interval) - next_interval = check_interval * HZ; - if (!next_interval) + *n = check_interval * HZ; + if (!*n) return; setup_timer(t, mcheck_timer, smp_processor_id()); - t->expires = round_jiffies(jiffies + next_interval); + t->expires = round_jiffies(jiffies + *n); add_timer(t); } @@ -907,7 +907,6 @@ static void mce_cpu_restart(void *data) /* Reinit MCEs after user configuration changes */ static void mce_restart(void) { - next_interval = check_interval * HZ; on_each_cpu(mce_cpu_restart, NULL, 1); } @@ -1110,7 +1109,8 @@ static int __cpuinit mce_cpu_callback(struct notifier_block *nfb, break; case CPU_DOWN_FAILED: case CPU_DOWN_FAILED_FROZEN: - t->expires = round_jiffies(jiffies + next_interval); + t->expires = round_jiffies(jiffies + + __get_cpu_var(next_interval)); add_timer_on(t, cpu); smp_call_function_single(cpu, mce_reenable_cpu, &action, 1); break; -- cgit v1.2.3-70-g09d2 From 5679af4c1625a1534a4321e1ecc3c48a1cf65eb8 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Tue, 7 Apr 2009 17:06:55 +0200 Subject: x86, mce: fix boot logging logic The earlier patch to change the poller to a separate function subtly broke the boot logging logic. This could lead to machine checks getting logged at boot even when disabled or defaulting to off on some systems. Fix that. [ Impact: bug fix - avoid spurious MCE in log ] Signed-off-by: Andi Kleen Reviewed-by: Hidetoshi Seto Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/mce.h | 1 + arch/x86/kernel/cpu/mcheck/mce_64.c | 9 +++++---- 2 files changed, 6 insertions(+), 4 deletions(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/include/asm/mce.h b/arch/x86/include/asm/mce.h index 563933e06a3..4f8c199584e 100644 --- a/arch/x86/include/asm/mce.h +++ b/arch/x86/include/asm/mce.h @@ -137,6 +137,7 @@ DECLARE_PER_CPU(mce_banks_t, mce_poll_banks); enum mcp_flags { MCP_TIMESTAMP = (1 << 0), /* log time stamp */ MCP_UC = (1 << 1), /* log uncorrected errors */ + MCP_DONTLOG = (1 << 2), /* only clear, don't log */ }; extern void machine_check_poll(enum mcp_flags flags, mce_banks_t *b); diff --git a/arch/x86/kernel/cpu/mcheck/mce_64.c b/arch/x86/kernel/cpu/mcheck/mce_64.c index 82614f1b923..6fb0b359d2a 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_64.c @@ -239,9 +239,10 @@ void machine_check_poll(enum mcp_flags flags, mce_banks_t *b) * Don't get the IP here because it's unlikely to * have anything to do with the actual error location. */ - - mce_log(&m); - add_taint(TAINT_MACHINE_CHECK); + if (!(flags & MCP_DONTLOG)) { + mce_log(&m); + add_taint(TAINT_MACHINE_CHECK); + } /* * Clear state for this bank. @@ -585,7 +586,7 @@ static void mce_init(void *dummy) * Log the machine checks left over from the previous reset. */ bitmap_fill(all_banks, MAX_NR_BANKS); - machine_check_poll(MCP_UC, &all_banks); + machine_check_poll(MCP_UC|(!mce_bootlog ? MCP_DONTLOG : 0), &all_banks); set_in_cr4(X86_CR4_MCE); -- cgit v1.2.3-70-g09d2 From f9a196b8dceba3c1e5fe885b81e45043ad7c60fc Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 1 May 2009 20:59:25 +0200 Subject: x86: initialize io_bitmap_base on 32bit commit db949bba3c7cf2e664ac12e237c6d4c914f0c69d (x86-32: use non-lazy io bitmap context switching) broke ioperm for 32bit because it removed the lazy initialization of io_bitmap_base and did not set it to the real bitmap offset. [ Impact: fix non-working sys_ioperm() on 32-bit kernels ] Signed-off-by: Thomas Gleixner --- arch/x86/kernel/cpu/common.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index c4f667896c2..c1caefc82e6 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1203,6 +1203,8 @@ void __cpuinit cpu_init(void) load_TR_desc(); load_LDT(&init_mm.context); + t->x86_tss.io_bitmap_base = offsetof(struct tss_struct, io_bitmap); + #ifdef CONFIG_DOUBLEFAULT /* Set up doublefault TSS pointer in the GDT */ __set_tss_desc(cpu, GDT_ENTRY_DOUBLEFAULT_TSS, &doublefault_tss); -- cgit v1.2.3-70-g09d2 From 35d11680a9d82c93eb92f08f9702b72877427b4a Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Mon, 4 May 2009 20:28:59 +0200 Subject: x86: show number of core_siblings instead of thread_siblings in /proc/cpuinfo Commit 7ad728f98162cb1af06a85b2a5fc422dddd4fb78 (cpumask: x86: convert cpu_sibling_map/cpu_core_map to cpumask_var_t) changed the output of /proc/cpuinfo for siblings: Example on an AMD Phenom: physical id : 0 siblings : 1 core id : 3 cpu cores : 4 Before that commit it was: physical id : 0 siblings : 4 core id : 3 cpu cores : 4 Instead of cpu_core_mask it now uses cpu_sibling_mask to count siblings. This is due to the following hunk of above commit: | --- a/arch/x86/kernel/cpu/proc.c | +++ b/arch/x86/kernel/cpu/proc.c | @@ -14,7 +14,7 @@ static void show_cpuinfo_core(struct seq_file *m, struct cpuinf | if (c->x86_max_cores * smp_num_siblings > 1) { | seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); | seq_printf(m, "siblings\t: %d\n", | - cpus_weight(per_cpu(cpu_core_map, cpu))); | + cpumask_weight(cpu_sibling_mask(cpu))); | seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); | seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); | seq_printf(m, "apicid\t\t: %d\n", c->apicid); This was a mistake, because the impact line shows that this side-effect was not anticipated: Impact: reduce per-cpu size for CONFIG_CPUMASK_OFFSTACK=y So revert the respective hunk to restore the old behavior. [ Impact: fix sibling-info regression in /proc/cpuinfo ] Signed-off-by: Andreas Herrmann Cc: Rusty Russell LKML-Reference: <20090504182859.GA29045@alberich.amd.com> Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/proc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/proc.c b/arch/x86/kernel/cpu/proc.c index f93047fed79..d5e30397246 100644 --- a/arch/x86/kernel/cpu/proc.c +++ b/arch/x86/kernel/cpu/proc.c @@ -14,7 +14,7 @@ static void show_cpuinfo_core(struct seq_file *m, struct cpuinfo_x86 *c, if (c->x86_max_cores * smp_num_siblings > 1) { seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); seq_printf(m, "siblings\t: %d\n", - cpumask_weight(cpu_sibling_mask(cpu))); + cpumask_weight(cpu_core_mask(cpu))); seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); seq_printf(m, "apicid\t\t: %d\n", c->apicid); -- cgit v1.2.3-70-g09d2 From e5299926d7459d9fa7c7f856983147817aedb10e Mon Sep 17 00:00:00 2001 From: Hidetoshi Seto Date: Fri, 8 May 2009 17:28:40 +0900 Subject: x86: MCE: make cmci_discover_lock irq-safe Lockdep reports the warning below when Li tries to offline one cpu: [ 110.835487] ================================= [ 110.835616] [ INFO: inconsistent lock state ] [ 110.835688] 2.6.30-rc4-00336-g8c9ed89 #52 [ 110.835757] --------------------------------- [ 110.835828] inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage. [ 110.835908] swapper/0 [HC1[1]:SC0[0]:HE0:SE1] takes: [ 110.835982] (cmci_discover_lock){?.+...}, at: [] cmci_clear+0x30/0x9b cmci_clear() can be called via smp_call_function_single(). It is better to disable interrupt while holding cmci_discover_lock, to turn it into an irq-safe lock - we can deadlock otherwise. [ Impact: fix possible deadlock in the MCE code ] Reported-by: Shaohua Li Signed-off-by: Hidetoshi Seto Cc: Andi Kleen Cc: Andrew Morton LKML-Reference: <4A03ED38.8000700@jp.fujitsu.com> Signed-off-by: Ingo Molnar Reported-by: Shaohua Li --- arch/x86/kernel/cpu/mcheck/mce_intel_64.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) (limited to 'arch/x86/kernel/cpu') diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c index d6b72df89d6..cef3ee30744 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c @@ -151,10 +151,11 @@ static void print_update(char *type, int *hdr, int num) static void cmci_discover(int banks, int boot) { unsigned long *owned = (void *)&__get_cpu_var(mce_banks_owned); + unsigned long flags; int hdr = 0; int i; - spin_lock(&cmci_discover_lock); + spin_lock_irqsave(&cmci_discover_lock, flags); for (i = 0; i < banks; i++) { u64 val; @@ -184,7 +185,7 @@ static void cmci_discover(int banks, int boot) WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks))); } } - spin_unlock(&cmci_discover_lock); + spin_unlock_irqrestore(&cmci_discover_lock, flags); if (hdr) printk(KERN_CONT "\n"); } @@ -211,13 +212,14 @@ void cmci_recheck(void) */ void cmci_clear(void) { + unsigned long flags; int i; int banks; u64 val; if (!cmci_supported(&banks)) return; - spin_lock(&cmci_discover_lock); + spin_lock_irqsave(&cmci_discover_lock, flags); for (i = 0; i < banks; i++) { if (!test_bit(i, __get_cpu_var(mce_banks_owned))) continue; @@ -227,7 +229,7 @@ void cmci_clear(void) wrmsrl(MSR_IA32_MC0_CTL2 + i, val); __clear_bit(i, __get_cpu_var(mce_banks_owned)); } - spin_unlock(&cmci_discover_lock); + spin_unlock_irqrestore(&cmci_discover_lock, flags); } /* -- cgit v1.2.3-70-g09d2