[PATCH] [44/48] x86_64: vsyscall_gtod_data diet and vgettimeofday() fix
From: Andi Kleen
Date: Sun Apr 29 2007 - 06:59:40 EST
From: Eric Dumazet <dada1@xxxxxxxxxxxxx>
Current vsyscall_gtod_data is large (3 or 4 cache lines dirtied at timer
interrupt). We can shrink it to exactly 64 bytes (1 cache line on AMD64)
Instead of copying a whole struct clocksource, we copy only needed fields.
I deleted an unused field : offset_base
This patch fixes one oddity in vgettimeofday(): It can returns a timeval with
tv_usec = 1000000. Maybe not a bug, but why not doing the right thing ?
Signed-off-by: Eric Dumazet <dada1@xxxxxxxxxxxxx>
Signed-off-by: Andi Kleen <ak@xxxxxxx>
---
arch/x86_64/kernel/vsyscall.c | 53 ++++++++++++++++++++++++++++--------------
1 file changed, 36 insertions(+), 17 deletions(-)
Index: linux/arch/x86_64/kernel/vsyscall.c
===================================================================
--- linux.orig/arch/x86_64/kernel/vsyscall.c
+++ linux/arch/x86_64/kernel/vsyscall.c
@@ -51,13 +51,28 @@
asm("" : "=r" (v) : "0" (x)); \
((v - VSYSCALL_FIRST_PAGE) + __pa_symbol(&__vsyscall_0)); })
+/*
+ * vsyscall_gtod_data contains data that is :
+ * - readonly from vsyscalls
+ * - writen by timer interrupt or systcl (/proc/sys/kernel/vsyscall64)
+ * Try to keep this structure as small as possible to avoid cache line ping pongs
+ */
struct vsyscall_gtod_data_t {
- seqlock_t lock;
- int sysctl_enabled;
- struct timeval wall_time_tv;
+ seqlock_t lock;
+
+ /* open coded 'struct timespec' */
+ time_t wall_time_sec;
+ u32 wall_time_nsec;
+
+ int sysctl_enabled;
struct timezone sys_tz;
- cycle_t offset_base;
- struct clocksource clock;
+ struct { /* extract of a clocksource struct */
+ cycle_t (*vread)(void);
+ cycle_t cycle_last;
+ cycle_t mask;
+ u32 mult;
+ u32 shift;
+ } clock;
};
int __vgetcpu_mode __section_vgetcpu_mode;
@@ -73,9 +88,13 @@ void update_vsyscall(struct timespec *wa
write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
/* copy vsyscall data */
- vsyscall_gtod_data.clock = *clock;
- vsyscall_gtod_data.wall_time_tv.tv_sec = wall_time->tv_sec;
- vsyscall_gtod_data.wall_time_tv.tv_usec = wall_time->tv_nsec/1000;
+ vsyscall_gtod_data.clock.vread = clock->vread;
+ vsyscall_gtod_data.clock.cycle_last = clock->cycle_last;
+ vsyscall_gtod_data.clock.mask = clock->mask;
+ vsyscall_gtod_data.clock.mult = clock->mult;
+ vsyscall_gtod_data.clock.shift = clock->shift;
+ vsyscall_gtod_data.wall_time_sec = wall_time->tv_sec;
+ vsyscall_gtod_data.wall_time_nsec = wall_time->tv_nsec;
vsyscall_gtod_data.sys_tz = sys_tz;
write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
}
@@ -110,7 +129,8 @@ static __always_inline long time_syscall
static __always_inline void do_vgettimeofday(struct timeval * tv)
{
cycle_t now, base, mask, cycle_delta;
- unsigned long seq, mult, shift, nsec_delta;
+ unsigned seq;
+ unsigned long mult, shift, nsec;
cycle_t (*vread)(void);
do {
seq = read_seqbegin(&__vsyscall_gtod_data.lock);
@@ -126,21 +146,20 @@ static __always_inline void do_vgettimeo
mult = __vsyscall_gtod_data.clock.mult;
shift = __vsyscall_gtod_data.clock.shift;
- *tv = __vsyscall_gtod_data.wall_time_tv;
-
+ tv->tv_sec = __vsyscall_gtod_data.wall_time_sec;
+ nsec = __vsyscall_gtod_data.wall_time_nsec;
} while (read_seqretry(&__vsyscall_gtod_data.lock, seq));
/* calculate interval: */
cycle_delta = (now - base) & mask;
/* convert to nsecs: */
- nsec_delta = (cycle_delta * mult) >> shift;
+ nsec += (cycle_delta * mult) >> shift;
- /* convert to usecs and add to timespec: */
- tv->tv_usec += nsec_delta / NSEC_PER_USEC;
- while (tv->tv_usec > USEC_PER_SEC) {
+ while (nsec >= NSEC_PER_SEC) {
tv->tv_sec += 1;
- tv->tv_usec -= USEC_PER_SEC;
+ nsec -= NSEC_PER_SEC;
}
+ tv->tv_usec = nsec / NSEC_PER_USEC;
}
int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * tz)
@@ -159,7 +178,7 @@ time_t __vsyscall(1) vtime(time_t *t)
time_t result;
if (unlikely(!__vsyscall_gtod_data.sysctl_enabled))
return time_syscall(t);
- result = __vsyscall_gtod_data.wall_time_tv.tv_sec;
+ result = __vsyscall_gtod_data.wall_time_sec;
if (t)
*t = result;
return result;
-
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/