Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757335AbYGASof (ORCPT ); Tue, 1 Jul 2008 14:44:35 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1757442AbYGASni (ORCPT ); Tue, 1 Jul 2008 14:43:38 -0400 Received: from smtp-outbound-1.vmware.com ([65.113.40.141]:49486 "EHLO smtp-outbound-1.vmware.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753424AbYGASnf (ORCPT ); Tue, 1 Jul 2008 14:43:35 -0400 Subject: [PATCH 04/05] x86: Merge tsc_init and clocksource code 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 11:43:34 -0700 Message-Id: <1214937814.20191.33.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: 19679 Lines: 720 [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 Unify the clocksource code. Unify the tsc_init code. Signed-off-by: Alok N Kataria Signed-off-by: Dan Hecht --- arch/x86/kernel/Makefile | 2 arch/x86/kernel/time_64.c | 32 ------- arch/x86/kernel/tsc.c | 212 ++++++++++++++++++++++++++++++++++++++++++++- arch/x86/kernel/tsc_32.c | 188 ---------------------------------------- arch/x86/kernel/tsc_64.c | 106 ----------------------- include/asm-x86/apic.h | 7 + include/asm-x86/delay.h | 4 + include/asm-x86/time.h | 2 include/asm-x86/tsc.h | 1 9 files changed, 224 insertions(+), 330 deletions(-) delete mode 100644 arch/x86/kernel/tsc_32.c delete mode 100644 arch/x86/kernel/tsc_64.c diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 631141b..8f60218 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 tsc.o +obj-y += tsc.o io_delay.o rtc.o obj-$(CONFIG_X86_TRAMPOLINE) += trampoline.o obj-y += process.o diff --git a/arch/x86/kernel/time_64.c b/arch/x86/kernel/time_64.c index c6ac4da..e3d49c5 100644 --- a/arch/x86/kernel/time_64.c +++ b/arch/x86/kernel/time_64.c @@ -56,7 +56,7 @@ static irqreturn_t timer_event_interrupt(int irq, void *dev_id) /* calibrate_cpu is used on systems with fixed rate TSCs to determine * processor frequency */ #define TICK_COUNT 100000000 -static unsigned long __init calibrate_cpu(void) +unsigned long __init calibrate_cpu(void) { int tsc_start, tsc_now; int i, no_ctr_free; @@ -114,41 +114,13 @@ void __init hpet_time_init(void) setup_irq(0, &irq0); } -extern void set_cyc2ns_scale(unsigned long cpu_khz, int cpu); - void __init time_init(void) { - int cpu; - - cpu_khz = calculate_cpu_khz(); - tsc_khz = cpu_khz; - - if (cpu_has(&boot_cpu_data, X86_FEATURE_CONSTANT_TSC) && - (boot_cpu_data.x86_vendor == X86_VENDOR_AMD)) - cpu_khz = calibrate_cpu(); - - lpj_fine = ((unsigned long)tsc_khz * 1000)/HZ; - - if (unsynchronized_tsc()) - mark_tsc_unstable("TSCs unsynchronized"); - + tsc_init(); if (cpu_has(&boot_cpu_data, X86_FEATURE_RDTSCP)) vgetcpu_mode = VGETCPU_RDTSCP; else vgetcpu_mode = VGETCPU_LSL; - printk(KERN_INFO "time.c: Detected %d.%03d MHz processor.\n", - cpu_khz / 1000, cpu_khz % 1000); - - /* - * Secondary CPUs do not run through tsc_init(), so set up - * all the scale factors for all CPUs, assuming the same - * speed as the bootup CPU. (cpufreq notifiers will fix this - * up if their speed diverges) - */ - for_each_possible_cpu(cpu) - set_cyc2ns_scale(cpu_khz, cpu); - - init_tsc_clocksource(); late_time_init = choose_time_init(); } diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index 595f78a..94c16bd 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -5,8 +5,16 @@ #include #include #include +#include +#include +#include +#include #include +#include +#include +#include +#include unsigned int cpu_khz; /* TSC clocks / usec, not used here */ EXPORT_SYMBOL(cpu_khz); @@ -16,12 +24,12 @@ EXPORT_SYMBOL(tsc_khz); /* * TSC can be unstable due to cpufreq or due to unsynced TSCs */ -int tsc_unstable; +static 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; +static int tsc_disabled = -1; /* * Scheduler clock - returns current time in nanosec units. @@ -241,7 +249,7 @@ EXPORT_SYMBOL(recalibrate_cpu_khz); DEFINE_PER_CPU(unsigned long, cyc2ns); -void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) +static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) { unsigned long long tsc_now, ns_now; unsigned long flags, *scale; @@ -329,3 +337,201 @@ static int __init cpufreq_tsc(void) core_initcall(cpufreq_tsc); #endif /* CONFIG_CPU_FREQ */ + +/* clocksource code */ + +static struct clocksource clocksource_tsc; + +/* + * We compare the TSC to the cycle_last value in the clocksource + * structure to avoid a nasty time-warp. This can be observed in a + * very small window right after one CPU updated cycle_last under + * xtime/vsyscall_gtod lock and the other CPU reads a TSC value which + * is smaller than the cycle_last reference value due to a TSC which + * is slighty behind. This delta is nowhere else observable, but in + * that case it results in a forward time jump in the range of hours + * due to the unsigned delta calculation of the time keeping core + * code, which is necessary to support wrapping clocksources like pm + * timer. + */ +static cycle_t read_tsc(void) +{ + cycle_t ret = (cycle_t)get_cycles(); + + return ret >= clocksource_tsc.cycle_last ? + ret : clocksource_tsc.cycle_last; +} + +static cycle_t __vsyscall_fn vread_tsc(void) +{ + cycle_t ret = (cycle_t)vget_cycles(); + + return ret >= __vsyscall_gtod_data.clock.cycle_last ? + ret : __vsyscall_gtod_data.clock.cycle_last; +} + +static struct clocksource clocksource_tsc = { + .name = "tsc", + .rating = 300, + .read = read_tsc, + .mask = CLOCKSOURCE_MASK(64), + .shift = 22, + .flags = CLOCK_SOURCE_IS_CONTINUOUS | + CLOCK_SOURCE_MUST_VERIFY, +#ifdef CONFIG_X86_64 + .vread = vread_tsc, +#endif +}; + +void mark_tsc_unstable(char *reason) +{ + if (!tsc_unstable) { + tsc_unstable = 1; + printk("Marking TSC unstable due to %s\n", reason); + /* Change only the rating, when not registered */ + if (clocksource_tsc.mult) + clocksource_change_rating(&clocksource_tsc, 0); + else + clocksource_tsc.rating = 0; + } +} + +EXPORT_SYMBOL_GPL(mark_tsc_unstable); + +static int __init dmi_mark_tsc_unstable(const struct dmi_system_id *d) +{ + printk(KERN_NOTICE "%s detected: marking TSC unstable.\n", + d->ident); + tsc_unstable = 1; + return 0; +} + +/* List of systems that have known TSC problems */ +static struct dmi_system_id __initdata bad_tsc_dmi_table[] = { + { + .callback = dmi_mark_tsc_unstable, + .ident = "IBM Thinkpad 380XD", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "IBM"), + DMI_MATCH(DMI_BOARD_NAME, "2635FA0"), + }, + }, + {} +}; + +/* + * Geode_LX - the OLPC CPU has a possibly a very reliable TSC + */ +#ifdef CONFIG_MGEODE_LX +/* RTSC counts during suspend */ +#define RTSC_SUSP 0x100 + +static void __init check_geode_tsc_reliable(void) +{ + unsigned long res_low, res_high; + + rdmsr_safe(MSR_GEODE_BUSCONT_CONF0, &res_low, &res_high); + if (res_low & RTSC_SUSP) + clocksource_tsc.flags &= ~CLOCK_SOURCE_MUST_VERIFY; +} +#else +static inline void check_geode_tsc_reliable(void) { } +#endif + +/* + * Make an educated guess if the TSC is trustworthy and synchronized + * over all CPUs. + */ +__cpuinit int unsynchronized_tsc(void) +{ + if (!cpu_has_tsc || tsc_unstable) + return 1; + +#ifdef CONFIG_SMP + if (apic_is_clustered_box()) + return 1; +#endif + + if (boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) + return 0; + /* + * Intel systems are normally all synchronized. + * Exceptions must mark TSC as unstable: + */ + if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) { + /* assume multi socket systems are not synchronized: */ + if (num_possible_cpus() > 1) + tsc_unstable = 1; + } + + return tsc_unstable; +} + +static void __init init_tsc_clocksource(void) +{ + clocksource_tsc.mult = clocksource_khz2mult(tsc_khz, + clocksource_tsc.shift); + /* lower the rating if we already know its unstable: */ + if (check_tsc_unstable()) { + clocksource_tsc.rating = 0; + clocksource_tsc.flags &= ~CLOCK_SOURCE_IS_CONTINUOUS; + } + clocksource_register(&clocksource_tsc); +} + +void __init tsc_init(void) +{ + u64 lpj; + int cpu; + + if (!cpu_has_tsc) + return; + + cpu_khz = calculate_cpu_khz(); + tsc_khz = cpu_khz; + + if (!cpu_khz) { + mark_tsc_unstable("could not calculate TSC khz"); + return; + } + +#ifdef CONFIG_X86_64 + if (cpu_has(&boot_cpu_data, X86_FEATURE_CONSTANT_TSC) && + (boot_cpu_data.x86_vendor == X86_VENDOR_AMD)) + cpu_khz = calibrate_cpu(); +#endif + + lpj = ((u64)tsc_khz * 1000); + do_div(lpj, HZ); + lpj_fine = lpj; + + printk("Detected %lu.%03lu MHz processor.\n", + (unsigned long)cpu_khz / 1000, + (unsigned long)cpu_khz % 1000); + + /* + * Secondary CPUs do not run through tsc_init(), so set up + * all the scale factors for all CPUs, assuming the same + * speed as the bootup CPU. (cpufreq notifiers will fix this + * up if their speed diverges) + */ + for_each_possible_cpu(cpu) + set_cyc2ns_scale(cpu_khz, cpu); + + if (tsc_disabled > 0) + return; + + /* now allow native_sched_clock() to use rdtsc */ + tsc_disabled = 0; + + use_tsc_delay(); + /* Check and install the TSC clocksource */ + dmi_check_system(bad_tsc_dmi_table); + + if (unsynchronized_tsc()) + mark_tsc_unstable("TSCs unsynchronized"); + + check_geode_tsc_reliable(); + init_tsc_clocksource(); +} + diff --git a/arch/x86/kernel/tsc_32.c b/arch/x86/kernel/tsc_32.c deleted file mode 100644 index bbc153d..0000000 --- a/arch/x86/kernel/tsc_32.c +++ /dev/null @@ -1,188 +0,0 @@ -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#include -#include -#include -#include - -#include "mach_timer.h" - -extern int tsc_unstable; -extern int tsc_disabled; - -/* clock source code */ - -static struct clocksource clocksource_tsc; - -/* - * We compare the TSC to the cycle_last value in the clocksource - * structure to avoid a nasty time-warp issue. This can be observed in - * a very small window right after one CPU updated cycle_last under - * xtime lock and the other CPU reads a TSC value which is smaller - * than the cycle_last reference value due to a TSC which is slighty - * behind. This delta is nowhere else observable, but in that case it - * results in a forward time jump in the range of hours due to the - * unsigned delta calculation of the time keeping core code, which is - * necessary to support wrapping clocksources like pm timer. - */ -static cycle_t read_tsc(void) -{ - cycle_t ret; - - rdtscll(ret); - - return ret >= clocksource_tsc.cycle_last ? - ret : clocksource_tsc.cycle_last; -} - -static struct clocksource clocksource_tsc = { - .name = "tsc", - .rating = 300, - .read = read_tsc, - .mask = CLOCKSOURCE_MASK(64), - .mult = 0, /* to be set */ - .shift = 22, - .flags = CLOCK_SOURCE_IS_CONTINUOUS | - CLOCK_SOURCE_MUST_VERIFY, -}; - -void mark_tsc_unstable(char *reason) -{ - if (!tsc_unstable) { - tsc_unstable = 1; - printk("Marking TSC unstable due to: %s.\n", reason); - /* Can be called before registration */ - if (clocksource_tsc.mult) - clocksource_change_rating(&clocksource_tsc, 0); - else - clocksource_tsc.rating = 0; - } -} -EXPORT_SYMBOL_GPL(mark_tsc_unstable); - -static int __init dmi_mark_tsc_unstable(const struct dmi_system_id *d) -{ - printk(KERN_NOTICE "%s detected: marking TSC unstable.\n", - d->ident); - tsc_unstable = 1; - return 0; -} - -/* List of systems that have known TSC problems */ -static struct dmi_system_id __initdata bad_tsc_dmi_table[] = { - { - .callback = dmi_mark_tsc_unstable, - .ident = "IBM Thinkpad 380XD", - .matches = { - DMI_MATCH(DMI_BOARD_VENDOR, "IBM"), - DMI_MATCH(DMI_BOARD_NAME, "2635FA0"), - }, - }, - {} -}; - -/* - * Make an educated guess if the TSC is trustworthy and synchronized - * over all CPUs. - */ -__cpuinit int unsynchronized_tsc(void) -{ - if (!cpu_has_tsc || tsc_unstable) - return 1; - - /* Anything with constant TSC should be synchronized */ - if (boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) - return 0; - - /* - * Intel systems are normally all synchronized. - * Exceptions must mark TSC as unstable: - */ - if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) { - /* assume multi socket systems are not synchronized: */ - if (num_possible_cpus() > 1) - tsc_unstable = 1; - } - return tsc_unstable; -} - -/* - * Geode_LX - the OLPC CPU has a possibly a very reliable TSC - */ -#ifdef CONFIG_MGEODE_LX -/* RTSC counts during suspend */ -#define RTSC_SUSP 0x100 - -static void __init check_geode_tsc_reliable(void) -{ - unsigned long res_low, res_high; - - rdmsr_safe(MSR_GEODE_BUSCONT_CONF0, &res_low, &res_high); - if (res_low & RTSC_SUSP) - clocksource_tsc.flags &= ~CLOCK_SOURCE_MUST_VERIFY; -} -#else -static inline void check_geode_tsc_reliable(void) { } -#endif - - -void __init tsc_init(void) -{ - int cpu; - u64 lpj; - - if (!cpu_has_tsc || tsc_disabled > 0) - return; - - cpu_khz = calculate_cpu_khz(); - tsc_khz = cpu_khz; - - if (!cpu_khz) { - mark_tsc_unstable("could not calculate TSC khz"); - return; - } - - lpj = ((u64)tsc_khz * 1000); - do_div(lpj, HZ); - lpj_fine = lpj; - - /* now allow native_sched_clock() to use rdtsc */ - tsc_disabled = 0; - - printk("Detected %lu.%03lu MHz processor.\n", - (unsigned long)cpu_khz / 1000, - (unsigned long)cpu_khz % 1000); - - /* - * Secondary CPUs do not run through tsc_init(), so set up - * all the scale factors for all CPUs, assuming the same - * speed as the bootup CPU. (cpufreq notifiers will fix this - * up if their speed diverges) - */ - for_each_possible_cpu(cpu) - set_cyc2ns_scale(cpu_khz, cpu); - - use_tsc_delay(); - - /* Check and install the TSC clocksource */ - dmi_check_system(bad_tsc_dmi_table); - - unsynchronized_tsc(); - check_geode_tsc_reliable(); - clocksource_tsc.mult = clocksource_khz2mult(tsc_khz, - clocksource_tsc.shift); - /* lower the rating if we already know its unstable: */ - if (check_tsc_unstable()) { - clocksource_tsc.rating = 0; - clocksource_tsc.flags &= ~CLOCK_SOURCE_IS_CONTINUOUS; - } - clocksource_register(&clocksource_tsc); -} diff --git a/arch/x86/kernel/tsc_64.c b/arch/x86/kernel/tsc_64.c deleted file mode 100644 index 80a274b..0000000 --- a/arch/x86/kernel/tsc_64.c +++ /dev/null @@ -1,106 +0,0 @@ -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#include -#include -#include -#include - -extern int tsc_unstable; -extern int tsc_disabled; - -/* - * Make an educated guess if the TSC is trustworthy and synchronized - * over all CPUs. - */ -__cpuinit int unsynchronized_tsc(void) -{ - if (tsc_unstable) - return 1; - -#ifdef CONFIG_SMP - if (apic_is_clustered_box()) - return 1; -#endif - - if (boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) - return 0; - - /* Assume multi socket systems are not synchronized */ - return num_present_cpus() > 1; -} - -static struct clocksource clocksource_tsc; - -/* - * We compare the TSC to the cycle_last value in the clocksource - * structure to avoid a nasty time-warp. This can be observed in a - * very small window right after one CPU updated cycle_last under - * xtime/vsyscall_gtod lock and the other CPU reads a TSC value which - * is smaller than the cycle_last reference value due to a TSC which - * is slighty behind. This delta is nowhere else observable, but in - * that case it results in a forward time jump in the range of hours - * due to the unsigned delta calculation of the time keeping core - * code, which is necessary to support wrapping clocksources like pm - * timer. - */ -static cycle_t read_tsc(void) -{ - cycle_t ret = (cycle_t)get_cycles(); - - return ret >= clocksource_tsc.cycle_last ? - ret : clocksource_tsc.cycle_last; -} - -static cycle_t __vsyscall_fn vread_tsc(void) -{ - cycle_t ret = (cycle_t)vget_cycles(); - - return ret >= __vsyscall_gtod_data.clock.cycle_last ? - ret : __vsyscall_gtod_data.clock.cycle_last; -} - -static struct clocksource clocksource_tsc = { - .name = "tsc", - .rating = 300, - .read = read_tsc, - .mask = CLOCKSOURCE_MASK(64), - .shift = 22, - .flags = CLOCK_SOURCE_IS_CONTINUOUS | - CLOCK_SOURCE_MUST_VERIFY, - .vread = vread_tsc, -}; - -void mark_tsc_unstable(char *reason) -{ - if (!tsc_unstable) { - tsc_unstable = 1; - printk("Marking TSC unstable due to %s\n", reason); - /* Change only the rating, when not registered */ - if (clocksource_tsc.mult) - clocksource_change_rating(&clocksource_tsc, 0); - else - clocksource_tsc.rating = 0; - } -} -EXPORT_SYMBOL_GPL(mark_tsc_unstable); - -void __init init_tsc_clocksource(void) -{ - if (tsc_disabled > 0) - return; - - clocksource_tsc.mult = clocksource_khz2mult(tsc_khz, - clocksource_tsc.shift); - if (check_tsc_unstable()) - clocksource_tsc.rating = 0; - - clocksource_register(&clocksource_tsc); -} diff --git a/include/asm-x86/apic.h b/include/asm-x86/apic.h index 9fe941c..99c3cfb 100644 --- a/include/asm-x86/apic.h +++ b/include/asm-x86/apic.h @@ -125,12 +125,17 @@ extern void enable_NMI_through_LVT0(void); */ #ifdef CONFIG_X86_64 extern void early_init_lapic_mapping(void); +extern int apic_is_clustered_box(void); +#else +static inline int apic_is_clustered_box(void) +{ + return 0; +} #endif extern u8 setup_APIC_eilvt_mce(u8 vector, u8 msg_type, u8 mask); extern u8 setup_APIC_eilvt_ibs(u8 vector, u8 msg_type, u8 mask); -extern int apic_is_clustered_box(void); #else /* !CONFIG_X86_LOCAL_APIC */ static inline void lapic_shutdown(void) { } diff --git a/include/asm-x86/delay.h b/include/asm-x86/delay.h index 409a649..bb80880 100644 --- a/include/asm-x86/delay.h +++ b/include/asm-x86/delay.h @@ -26,6 +26,10 @@ extern void __delay(unsigned long loops); ((n) > 20000 ? __bad_ndelay() : __const_udelay((n) * 5ul)) : \ __ndelay(n)) +#ifdef CONFIG_X86_32 void use_tsc_delay(void); +#else +#define use_tsc_delay() {} +#endif #endif /* _ASM_X86_DELAY_H */ diff --git a/include/asm-x86/time.h b/include/asm-x86/time.h index bce72d7..a17fa47 100644 --- a/include/asm-x86/time.h +++ b/include/asm-x86/time.h @@ -56,4 +56,6 @@ static inline int native_set_wallclock(unsigned long nowtime) #endif /* CONFIG_PARAVIRT */ +extern unsigned long __init calibrate_cpu(void); + #endif diff --git a/include/asm-x86/tsc.h b/include/asm-x86/tsc.h index 761054d..cb6f6ee 100644 --- a/include/asm-x86/tsc.h +++ b/include/asm-x86/tsc.h @@ -48,7 +48,6 @@ static __always_inline cycles_t vget_cycles(void) extern void tsc_init(void); extern void mark_tsc_unstable(char *reason); extern int unsynchronized_tsc(void); -extern void init_tsc_clocksource(void); int check_tsc_unstable(void); /* -- 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/