Re: [PATCH v3 6/8] x86-64: Move vread_tsc and vread_hpet into thevDSO

From: H. Peter Anvin
Date: Wed Jul 13 2011 - 23:40:14 EST

Next message: tip-bot for Tejun Heo: "[tip:x86/memblock] memblock: Replace memblock_find_base() with memblock_find_in_range()"
Previous message: tip-bot for Tejun Heo: "[tip:x86/memblock] memblock: Kill MEMBLOCK_ERROR"
In reply to: Andy Lutomirski: "[PATCH v3 6/8] x86-64: Move vread_tsc and vread_hpet into the vDSO"
Next in thread: Andy Lutomirski: "Re: [PATCH v3] x86-64: Move vread_tsc and vread_hpet into the vDSO"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On 07/13/2011 06:24 AM, Andy Lutomirski wrote:
> The vsyscall page now consists entirely of trap instructions.
>
> Cc: John Stultz <johnstul@xxxxxxxxxx>
> Signed-off-by: Andy Lutomirski <luto@xxxxxxx>

This patch causes a build failure on x86-64 allnoconfig:

/home/hpa/kernel/linux-2.6-tip.vdso/arch/x86/vdso/vclock_gettime.c: In
function ‘vread_hpet’:
/home/hpa/kernel/linux-2.6-tip.vdso/arch/x86/vdso/vclock_gettime.c:62:
error: implicit declaration of function ‘fix_to_virt’
/home/hpa/kernel/linux-2.6-tip.vdso/arch/x86/vdso/vclock_gettime.c:62:
error: ‘VSYSCALL_HPET’ undeclared (first use in this function)
/home/hpa/kernel/linux-2.6-tip.vdso/arch/x86/vdso/vclock_gettime.c:62:
error: (Each undeclared identifier is reported only once
/home/hpa/kernel/linux-2.6-tip.vdso/arch/x86/vdso/vclock_gettime.c:62:
error: for each function it appears in.)
make[2]: *** [arch/x86/vdso/vclock_gettime.o] Error 1
make[1]: *** [arch/x86/vdso/vclock_gettime.o] Error 2
make: *** [sub-make] Error 2

> ---
> arch/x86/include/asm/clocksource.h | 6 +++-
> arch/x86/include/asm/tsc.h | 4 ---
> arch/x86/include/asm/vgtod.h | 2 +-
> arch/x86/include/asm/vsyscall.h | 4 ---
> arch/x86/kernel/Makefile | 7 +----
> arch/x86/kernel/alternative.c | 8 -----
> arch/x86/kernel/hpet.c | 9 +-----
> arch/x86/kernel/tsc.c | 2 +-
> arch/x86/kernel/vmlinux.lds.S | 3 --
> arch/x86/kernel/vread_tsc_64.c | 36 -------------------------
> arch/x86/kernel/vsyscall_64.c | 2 +-
> arch/x86/vdso/vclock_gettime.c | 52 +++++++++++++++++++++++++++++++----
> 12 files changed, 56 insertions(+), 79 deletions(-)
> delete mode 100644 arch/x86/kernel/vread_tsc_64.c
>
> diff --git a/arch/x86/include/asm/clocksource.h b/arch/x86/include/asm/clocksource.h
> index a5df33f..3882c65 100644
> --- a/arch/x86/include/asm/clocksource.h
> +++ b/arch/x86/include/asm/clocksource.h
> @@ -7,8 +7,12 @@
>
> #define __ARCH_HAS_CLOCKSOURCE_DATA
>
> +#define VCLOCK_NONE 0 /* No vDSO clock available. */
> +#define VCLOCK_TSC 1 /* vDSO should use vread_tsc. */
> +#define VCLOCK_HPET 2 /* vDSO should use vread_hpet. */
> +
> struct arch_clocksource_data {
> - cycle_t (*vread)(void);
> + int vclock_mode;
> };
>
> #endif /* CONFIG_X86_64 */
> diff --git a/arch/x86/include/asm/tsc.h b/arch/x86/include/asm/tsc.h
> index 9db5583..83e2efd 100644
> --- a/arch/x86/include/asm/tsc.h
> +++ b/arch/x86/include/asm/tsc.h
> @@ -51,10 +51,6 @@ extern int unsynchronized_tsc(void);
> extern int check_tsc_unstable(void);
> extern unsigned long native_calibrate_tsc(void);
>
> -#ifdef CONFIG_X86_64
> -extern cycles_t vread_tsc(void);
> -#endif
> -
> /*
> * Boot-time check whether the TSCs are synchronized across
> * all CPUs/cores:
> diff --git a/arch/x86/include/asm/vgtod.h b/arch/x86/include/asm/vgtod.h
> index aa5add8..815285b 100644
> --- a/arch/x86/include/asm/vgtod.h
> +++ b/arch/x86/include/asm/vgtod.h
> @@ -13,7 +13,7 @@ struct vsyscall_gtod_data {
>
> struct timezone sys_tz;
> struct { /* extract of a clocksource struct */
> - cycle_t (*vread)(void);
> + int vclock_mode;
> cycle_t cycle_last;
> cycle_t mask;
> u32 mult;
> diff --git a/arch/x86/include/asm/vsyscall.h b/arch/x86/include/asm/vsyscall.h
> index d555973..6010707 100644
> --- a/arch/x86/include/asm/vsyscall.h
> +++ b/arch/x86/include/asm/vsyscall.h
> @@ -16,10 +16,6 @@ enum vsyscall_num {
> #ifdef __KERNEL__
> #include <linux/seqlock.h>
>
> -/* Definitions for CONFIG_GENERIC_TIME definitions */
> -#define __vsyscall_fn \
> - __attribute__ ((unused, __section__(".vsyscall_fn"))) notrace
> -
> #define VGETCPU_RDTSCP 1
> #define VGETCPU_LSL 2
>
> diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile
> index cc0469a..2deef3d 100644
> --- a/arch/x86/kernel/Makefile
> +++ b/arch/x86/kernel/Makefile
> @@ -24,17 +24,12 @@ endif
> nostackp := $(call cc-option, -fno-stack-protector)
> CFLAGS_vsyscall_64.o := $(PROFILING) -g0 $(nostackp)
> CFLAGS_hpet.o := $(nostackp)
> -CFLAGS_vread_tsc_64.o := $(nostackp)
> CFLAGS_paravirt.o := $(nostackp)
> GCOV_PROFILE_vsyscall_64.o := n
> GCOV_PROFILE_hpet.o := n
> GCOV_PROFILE_tsc.o := n
> -GCOV_PROFILE_vread_tsc_64.o := n
> GCOV_PROFILE_paravirt.o := n
>
> -# vread_tsc_64 is hot and should be fully optimized:
> -CFLAGS_REMOVE_vread_tsc_64.o = -pg -fno-optimize-sibling-calls
> -
> obj-y := process_$(BITS).o signal.o entry_$(BITS).o
> obj-y += traps.o irq.o irq_$(BITS).o dumpstack_$(BITS).o
> obj-y += time.o ioport.o ldt.o dumpstack.o
> @@ -43,7 +38,7 @@ obj-$(CONFIG_IRQ_WORK) += irq_work.o
> obj-y += probe_roms.o
> obj-$(CONFIG_X86_32) += sys_i386_32.o i386_ksyms_32.o
> obj-$(CONFIG_X86_64) += sys_x86_64.o x8664_ksyms_64.o
> -obj-$(CONFIG_X86_64) += syscall_64.o vsyscall_64.o vread_tsc_64.o
> +obj-$(CONFIG_X86_64) += syscall_64.o vsyscall_64.o
> obj-$(CONFIG_X86_64) += vsyscall_emu_64.o
> obj-y += bootflag.o e820.o
> obj-y += pci-dma.o quirks.o topology.o kdebugfs.o
> diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c
> index ddb207b..c638228 100644
> --- a/arch/x86/kernel/alternative.c
> +++ b/arch/x86/kernel/alternative.c
> @@ -14,7 +14,6 @@
> #include <asm/pgtable.h>
> #include <asm/mce.h>
> #include <asm/nmi.h>
> -#include <asm/vsyscall.h>
> #include <asm/cacheflush.h>
> #include <asm/tlbflush.h>
> #include <asm/io.h>
> @@ -250,7 +249,6 @@ static void __init_or_module add_nops(void *insns, unsigned int len)
>
> extern struct alt_instr __alt_instructions[], __alt_instructions_end[];
> extern s32 __smp_locks[], __smp_locks_end[];
> -extern char __vsyscall_0;
> void *text_poke_early(void *addr, const void *opcode, size_t len);
>
> /* Replace instructions with better alternatives for this CPU type.
> @@ -294,12 +292,6 @@ void __init_or_module apply_alternatives(struct alt_instr *start,
> add_nops(insnbuf + a->replacementlen,
> a->instrlen - a->replacementlen);
>
> -#ifdef CONFIG_X86_64
> - /* vsyscall code is not mapped yet. resolve it manually. */
> - if (instr >= (u8 *)VSYSCALL_START && instr < (u8*)VSYSCALL_END) {
> - instr = __va(instr - (u8*)VSYSCALL_START + (u8*)__pa_symbol(&__vsyscall_0));
> - }
> -#endif
> text_poke_early(instr, insnbuf, a->instrlen);
> }
> }
> diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c
> index 0e07257..d10cc00 100644
> --- a/arch/x86/kernel/hpet.c
> +++ b/arch/x86/kernel/hpet.c
> @@ -738,13 +738,6 @@ static cycle_t read_hpet(struct clocksource *cs)
> return (cycle_t)hpet_readl(HPET_COUNTER);
> }
>
> -#ifdef CONFIG_X86_64
> -static cycle_t __vsyscall_fn vread_hpet(void)
> -{
> - return readl((const void __iomem *)fix_to_virt(VSYSCALL_HPET) + 0xf0);
> -}
> -#endif
> -
> static struct clocksource clocksource_hpet = {
> .name = "hpet",
> .rating = 250,
> @@ -753,7 +746,7 @@ static struct clocksource clocksource_hpet = {
> .flags = CLOCK_SOURCE_IS_CONTINUOUS,
> .resume = hpet_resume_counter,
> #ifdef CONFIG_X86_64
> - .archdata = { .vread = vread_hpet },
> + .archdata = { .vclock_mode = VCLOCK_HPET },
> #endif
> };
>
> diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
> index e7a74b8..56c633a 100644
> --- a/arch/x86/kernel/tsc.c
> +++ b/arch/x86/kernel/tsc.c
> @@ -777,7 +777,7 @@ static struct clocksource clocksource_tsc = {
> .flags = CLOCK_SOURCE_IS_CONTINUOUS |
> CLOCK_SOURCE_MUST_VERIFY,
> #ifdef CONFIG_X86_64
> - .archdata = { .vread = vread_tsc },
> + .archdata = { .vclock_mode = VCLOCK_TSC },
> #endif
> };
>
> diff --git a/arch/x86/kernel/vmlinux.lds.S b/arch/x86/kernel/vmlinux.lds.S
> index 8017471..4aa9c54 100644
> --- a/arch/x86/kernel/vmlinux.lds.S
> +++ b/arch/x86/kernel/vmlinux.lds.S
> @@ -169,9 +169,6 @@ SECTIONS
> .vsyscall : AT(VLOAD(.vsyscall)) {
> *(.vsyscall_0)
>
> - . = ALIGN(L1_CACHE_BYTES);
> - *(.vsyscall_fn)
> -
> . = 1024;
> *(.vsyscall_1)
>
> diff --git a/arch/x86/kernel/vread_tsc_64.c b/arch/x86/kernel/vread_tsc_64.c
> deleted file mode 100644
> index a81aa9e..0000000
> --- a/arch/x86/kernel/vread_tsc_64.c
> +++ /dev/null
> @@ -1,36 +0,0 @@
> -/* This code runs in userspace. */
> -
> -#define DISABLE_BRANCH_PROFILING
> -#include <asm/vgtod.h>
> -
> -notrace cycle_t __vsyscall_fn vread_tsc(void)
> -{
> - cycle_t ret;
> - u64 last;
> -
> - /*
> - * Empirically, a fence (of type that depends on the CPU)
> - * before rdtsc is enough to ensure that rdtsc is ordered
> - * with respect to loads. The various CPU manuals are unclear
> - * as to whether rdtsc can be reordered with later loads,
> - * but no one has ever seen it happen.
> - */
> - rdtsc_barrier();
> - ret = (cycle_t)vget_cycles();
> -
> - last = VVAR(vsyscall_gtod_data).clock.cycle_last;
> -
> - if (likely(ret >= last))
> - return ret;
> -
> - /*
> - * GCC likes to generate cmov here, but this branch is extremely
> - * predictable (it's just a funciton of time and the likely is
> - * very likely) and there's a data dependence, so force GCC
> - * to generate a branch instead. I don't barrier() because
> - * we don't actually need a barrier, and if this function
> - * ever gets inlined it will generate worse code.
> - */
> - asm volatile ("");
> - return last;
> -}
> diff --git a/arch/x86/kernel/vsyscall_64.c b/arch/x86/kernel/vsyscall_64.c
> index 12d488f..dda7dff 100644
> --- a/arch/x86/kernel/vsyscall_64.c
> +++ b/arch/x86/kernel/vsyscall_64.c
> @@ -74,7 +74,7 @@ void update_vsyscall(struct timespec *wall_time, struct timespec *wtm,
> write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
>
> /* copy vsyscall data */
> - vsyscall_gtod_data.clock.vread = clock->archdata.vread;
> + vsyscall_gtod_data.clock.vclock_mode = clock->archdata.vclock_mode;
> vsyscall_gtod_data.clock.cycle_last = clock->cycle_last;
> vsyscall_gtod_data.clock.mask = clock->mask;
> vsyscall_gtod_data.clock.mult = mult;
> diff --git a/arch/x86/vdso/vclock_gettime.c b/arch/x86/vdso/vclock_gettime.c
> index cf54813..9869bac 100644
> --- a/arch/x86/vdso/vclock_gettime.c
> +++ b/arch/x86/vdso/vclock_gettime.c
> @@ -25,6 +25,43 @@
>
> #define gtod (&VVAR(vsyscall_gtod_data))
>
> +notrace static cycle_t vread_tsc(void)
> +{
> + cycle_t ret;
> + u64 last;
> +
> + /*
> + * Empirically, a fence (of type that depends on the CPU)
> + * before rdtsc is enough to ensure that rdtsc is ordered
> + * with respect to loads. The various CPU manuals are unclear
> + * as to whether rdtsc can be reordered with later loads,
> + * but no one has ever seen it happen.
> + */
> + rdtsc_barrier();
> + ret = (cycle_t)vget_cycles();
> +
> + last = VVAR(vsyscall_gtod_data).clock.cycle_last;
> +
> + if (likely(ret >= last))
> + return ret;
> +
> + /*
> + * GCC likes to generate cmov here, but this branch is extremely
> + * predictable (it's just a funciton of time and the likely is
> + * very likely) and there's a data dependence, so force GCC
> + * to generate a branch instead. I don't barrier() because
> + * we don't actually need a barrier, and if this function
> + * ever gets inlined it will generate worse code.
> + */
> + asm volatile ("");
> + return last;
> +}
> +
> +static notrace cycle_t vread_hpet(void)
> +{
> + return readl((const void __iomem *)fix_to_virt(VSYSCALL_HPET) + 0xf0);
> +}
> +
> notrace static long vdso_fallback_gettime(long clock, struct timespec *ts)
> {
> long ret;
> @@ -36,9 +73,12 @@ notrace static long vdso_fallback_gettime(long clock, struct timespec *ts)
> notrace static inline long vgetns(void)
> {
> long v;
> - cycles_t (*vread)(void);
> - vread = gtod->clock.vread;
> - v = (vread() - gtod->clock.cycle_last) & gtod->clock.mask;
> + cycles_t cycles;
> + if (gtod->clock.vclock_mode == VCLOCK_TSC)
> + cycles = vread_tsc();
> + else
> + cycles = vread_hpet();
> + v = (cycles - gtod->clock.cycle_last) & gtod->clock.mask;
> return (v * gtod->clock.mult) >> gtod->clock.shift;
> }
>
> @@ -118,11 +158,11 @@ notrace int __vdso_clock_gettime(clockid_t clock, struct timespec *ts)
> {
> switch (clock) {
> case CLOCK_REALTIME:
> - if (likely(gtod->clock.vread))
> + if (likely(gtod->clock.vclock_mode != VCLOCK_NONE))
> return do_realtime(ts);
> break;
> case CLOCK_MONOTONIC:
> - if (likely(gtod->clock.vread))
> + if (likely(gtod->clock.vclock_mode != VCLOCK_NONE))
> return do_monotonic(ts);
> break;
> case CLOCK_REALTIME_COARSE:
> @@ -139,7 +179,7 @@ int clock_gettime(clockid_t, struct timespec *)
> notrace int __vdso_gettimeofday(struct timeval *tv, struct timezone *tz)
> {
> long ret;
> - if (likely(gtod->clock.vread)) {
> + if (likely(gtod->clock.vclock_mode != VCLOCK_NONE)) {
> if (likely(tv != NULL)) {
> BUILD_BUG_ON(offsetof(struct timeval, tv_usec) !=
> offsetof(struct timespec, tv_nsec) ||

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Next message: tip-bot for Tejun Heo: "[tip:x86/memblock] memblock: Replace memblock_find_base() with memblock_find_in_range()"
Previous message: tip-bot for Tejun Heo: "[tip:x86/memblock] memblock: Kill MEMBLOCK_ERROR"
In reply to: Andy Lutomirski: "[PATCH v3 6/8] x86-64: Move vread_tsc and vread_hpet into the vDSO"
Next in thread: Andy Lutomirski: "Re: [PATCH v3] x86-64: Move vread_tsc and vread_hpet into the vDSO"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]