Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757457AbYGASni (ORCPT ); Tue, 1 Jul 2008 14:43:38 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754469AbYGASnU (ORCPT ); Tue, 1 Jul 2008 14:43:20 -0400 Received: from smtp-outbound-1.vmware.com ([65.113.40.141]:49346 "EHLO smtp-outbound-1.vmware.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754097AbYGASnS (ORCPT ); Tue, 1 Jul 2008 14:43:18 -0400 Subject: [PATCH 01/05] x86: merge sched_clock handling From: Alok Kataria Reply-To: akataria@vmware.com To: Ingo Molnar , Thomas Gleixner , the arch/x86 maintainers Cc: Dan Hecht , LKML Content-Type: text/plain Organization: VMware INC. Date: Tue, 01 Jul 2008 11:43:18 -0700 Message-Id: <1214937798.20191.30.camel@promb-2n-dhcp368.eng.vmware.com> Mime-Version: 1.0 X-Mailer: Evolution 2.8.0 (2.8.0-40.el5) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 9930 Lines: 366 [VMware, Inc. is providing the following patch to you under the terms of the GPL version 2 and no later version. This patch is provided as is, with no warranties or support. VMware disclaims all liability in connection with the use/inability to use this patch. Any use of the attached is considered acceptance of the above.] From: Alok N Kataria Move the basic global variable definitions and sched_clock handling in the common "tsc.c" file. - Unify notsc kernel command line handling for 32 bit and 64bit. - Functional changes for 64bit. - "tsc_disabled" is updated if "notsc" is passed at boottime. - Fallback to jiffies for sched_clock, incase notsc is passed on commandline. Signed-off-by: Alok N Kataria Signed-off-by: Dan Hecht --- arch/x86/kernel/Makefile | 2 + arch/x86/kernel/time_32.c | 3 -- arch/x86/kernel/tsc.c | 86 +++++++++++++++++++++++++++++++++++++++++++++ arch/x86/kernel/tsc_32.c | 86 +-------------------------------------------- arch/x86/kernel/tsc_64.c | 67 ++++++----------------------------- 5 files changed, 100 insertions(+), 144 deletions(-) create mode 100644 arch/x86/kernel/tsc.c diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index a33513d..631141b 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -34,7 +34,7 @@ obj-$(CONFIG_X86_64) += syscall_64.o vsyscall_64.o obj-y += bootflag.o e820.o obj-y += pci-dma.o quirks.o i8237.o topology.o kdebugfs.o obj-y += alternative.o i8253.o pci-nommu.o -obj-y += tsc_$(BITS).o io_delay.o rtc.o +obj-y += tsc_$(BITS).o io_delay.o rtc.o tsc.o obj-$(CONFIG_X86_TRAMPOLINE) += trampoline.o obj-y += process.o diff --git a/arch/x86/kernel/time_32.c b/arch/x86/kernel/time_32.c index 5f29f12..059ca6e 100644 --- a/arch/x86/kernel/time_32.c +++ b/arch/x86/kernel/time_32.c @@ -39,9 +39,6 @@ #include "do_timer.h" -unsigned int cpu_khz; /* Detected as we calibrate the TSC */ -EXPORT_SYMBOL(cpu_khz); - int timer_ack; unsigned long profile_pc(struct pt_regs *regs) diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c new file mode 100644 index 0000000..5d0be77 --- /dev/null +++ b/arch/x86/kernel/tsc.c @@ -0,0 +1,86 @@ +#include +#include +#include +#include + +unsigned int cpu_khz; /* TSC clocks / usec, not used here */ +EXPORT_SYMBOL(cpu_khz); +unsigned int tsc_khz; +EXPORT_SYMBOL(tsc_khz); + +/* + * TSC can be unstable due to cpufreq or due to unsynced TSCs + */ +int tsc_unstable; + +/* native_sched_clock() is called before tsc_init(), so + we must start with the TSC soft disabled to prevent + erroneous rdtsc usage on !cpu_has_tsc processors */ +int tsc_disabled = -1; + +/* + * Scheduler clock - returns current time in nanosec units. + */ +u64 native_sched_clock(void) +{ + u64 this_offset; + + /* + * Fall back to jiffies if there's no TSC available: + * ( But note that we still use it if the TSC is marked + * unstable. We do this because unlike Time Of Day, + * the scheduler clock tolerates small errors and it's + * very important for it to be as fast as the platform + * can achive it. ) + */ + if (unlikely(tsc_disabled)) { + /* No locking but a rare wrong value is not a big deal: */ + return (jiffies_64 - INITIAL_JIFFIES) * (1000000000 / HZ); + } + + /* read the Time Stamp Counter: */ + rdtscll(this_offset); + + /* return the value in ns */ + return cycles_2_ns(this_offset); +} + +/* We need to define a real function for sched_clock, to override the + weak default version */ +#ifdef CONFIG_PARAVIRT +unsigned long long sched_clock(void) +{ + return paravirt_sched_clock(); +} +#else +unsigned long long +sched_clock(void) __attribute__((alias("native_sched_clock"))); +#endif + +int check_tsc_unstable(void) +{ + return tsc_unstable; +} +EXPORT_SYMBOL_GPL(check_tsc_unstable); + +#ifdef CONFIG_X86_TSC +int __init notsc_setup(char *str) +{ + printk(KERN_WARNING "notsc: Kernel compiled with CONFIG_X86_TSC, " + "cannot disable TSC completely.\n"); + tsc_disabled = 1; + return 1; +} +#else +/* + * disable flag for tsc. Takes effect by clearing the TSC cpu flag + * in cpu/common.c + */ +int __init notsc_setup(char *str) +{ + setup_clear_cpu_cap(X86_FEATURE_TSC); + return 1; +} +#endif + +__setup("notsc", notsc_setup); diff --git a/arch/x86/kernel/tsc_32.c b/arch/x86/kernel/tsc_32.c index 6240922..dc89900 100644 --- a/arch/x86/kernel/tsc_32.c +++ b/arch/x86/kernel/tsc_32.c @@ -15,52 +15,8 @@ #include "mach_timer.h" -/* native_sched_clock() is called before tsc_init(), so - we must start with the TSC soft disabled to prevent - erroneous rdtsc usage on !cpu_has_tsc processors */ -static int tsc_disabled = -1; - -/* - * On some systems the TSC frequency does not - * change with the cpu frequency. So we need - * an extra value to store the TSC freq - */ -unsigned int tsc_khz; -EXPORT_SYMBOL_GPL(tsc_khz); - -#ifdef CONFIG_X86_TSC -static int __init tsc_setup(char *str) -{ - printk(KERN_WARNING "notsc: Kernel compiled with CONFIG_X86_TSC, " - "cannot disable TSC completely.\n"); - tsc_disabled = 1; - return 1; -} -#else -/* - * disable flag for tsc. Takes effect by clearing the TSC cpu flag - * in cpu/common.c - */ -static int __init tsc_setup(char *str) -{ - setup_clear_cpu_cap(X86_FEATURE_TSC); - return 1; -} -#endif - -__setup("notsc", tsc_setup); - -/* - * code to mark and check if the TSC is unstable - * due to cpufreq or due to unsynced TSCs - */ -static int tsc_unstable; - -int check_tsc_unstable(void) -{ - return tsc_unstable; -} -EXPORT_SYMBOL_GPL(check_tsc_unstable); +extern int tsc_unstable; +extern int tsc_disabled; /* Accelerators for sched_clock() * convert from cycles(64bits) => nanoseconds (64bits) @@ -109,44 +65,6 @@ static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) local_irq_restore(flags); } -/* - * Scheduler clock - returns current time in nanosec units. - */ -unsigned long long native_sched_clock(void) -{ - unsigned long long this_offset; - - /* - * Fall back to jiffies if there's no TSC available: - * ( But note that we still use it if the TSC is marked - * unstable. We do this because unlike Time Of Day, - * the scheduler clock tolerates small errors and it's - * very important for it to be as fast as the platform - * can achive it. ) - */ - if (unlikely(tsc_disabled)) - /* No locking but a rare wrong value is not a big deal: */ - return (jiffies_64 - INITIAL_JIFFIES) * (1000000000 / HZ); - - /* read the Time Stamp Counter: */ - rdtscll(this_offset); - - /* return the value in ns */ - return cycles_2_ns(this_offset); -} - -/* We need to define a real function for sched_clock, to override the - weak default version */ -#ifdef CONFIG_PARAVIRT -unsigned long long sched_clock(void) -{ - return paravirt_sched_clock(); -} -#else -unsigned long long sched_clock(void) - __attribute__((alias("native_sched_clock"))); -#endif - unsigned long native_calculate_cpu_khz(void) { unsigned long long start, end; diff --git a/arch/x86/kernel/tsc_64.c b/arch/x86/kernel/tsc_64.c index 9898fb0..69cbe4c 100644 --- a/arch/x86/kernel/tsc_64.c +++ b/arch/x86/kernel/tsc_64.c @@ -13,12 +13,8 @@ #include #include -static int notsc __initdata = 0; - -unsigned int cpu_khz; /* TSC clocks / usec, not used here */ -EXPORT_SYMBOL(cpu_khz); -unsigned int tsc_khz; -EXPORT_SYMBOL(tsc_khz); +extern int tsc_unstable; +extern int tsc_disabled; /* Accelerators for sched_clock() * convert from cycles(64bits) => nanoseconds (64bits) @@ -41,6 +37,7 @@ EXPORT_SYMBOL(tsc_khz); * * -johnstul@us.ibm.com "math is hard, lets go shopping!" */ + DEFINE_PER_CPU(unsigned long, cyc2ns); static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) @@ -63,41 +60,6 @@ static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) local_irq_restore(flags); } -unsigned long long native_sched_clock(void) -{ - unsigned long a = 0; - - /* Could do CPU core sync here. Opteron can execute rdtsc speculatively, - * which means it is not completely exact and may not be monotonous - * between CPUs. But the errors should be too small to matter for - * scheduling purposes. - */ - - rdtscll(a); - return cycles_2_ns(a); -} - -/* We need to define a real function for sched_clock, to override the - weak default version */ -#ifdef CONFIG_PARAVIRT -unsigned long long sched_clock(void) -{ - return paravirt_sched_clock(); -} -#else -unsigned long long -sched_clock(void) __attribute__((alias("native_sched_clock"))); -#endif - - -static int tsc_unstable; - -int check_tsc_unstable(void) -{ - return tsc_unstable; -} -EXPORT_SYMBOL_GPL(check_tsc_unstable); - #ifdef CONFIG_CPU_FREQ /* Frequency scaling support. Adjust the TSC based timer when the cpu frequency @@ -281,14 +243,6 @@ __cpuinit int unsynchronized_tsc(void) return num_present_cpus() > 1; } -int __init notsc_setup(char *s) -{ - notsc = 1; - return 1; -} - -__setup("notsc", notsc_setup); - static struct clocksource clocksource_tsc; /* @@ -346,12 +300,13 @@ EXPORT_SYMBOL_GPL(mark_tsc_unstable); void __init init_tsc_clocksource(void) { - if (!notsc) { - clocksource_tsc.mult = clocksource_khz2mult(tsc_khz, - clocksource_tsc.shift); - if (check_tsc_unstable()) - clocksource_tsc.rating = 0; + if (tsc_disabled > 0) + return; - clocksource_register(&clocksource_tsc); - } + clocksource_tsc.mult = clocksource_khz2mult(tsc_khz, + clocksource_tsc.shift); + if (check_tsc_unstable()) + clocksource_tsc.rating = 0; + + clocksource_register(&clocksource_tsc); } -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/