public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH] Timespec normalize off by one errors
@ 2005-11-11 21:16 George Anzinger
       [not found] ` <1131744325.9957.20.camel@localhost>
  2005-11-11 22:25 ` George Anzinger
  0 siblings, 2 replies; 3+ messages in thread
From: George Anzinger @ 2005-11-11 21:16 UTC (permalink / raw)
  To: Andrew Morton; +Cc: lkml

[-- Attachment #1: Type: text/plain, Size: 165 bytes --]

Found three places were we use ">" instead of ">=".
-- 
George Anzinger   george@mvista.com
HRT (High-res-timers):  http://sourceforge.net/projects/high-res-timers/

[-- Attachment #2: time-spec2.patch --]
[-- Type: text/plain, Size: 1633 bytes --]

Source: MontaVista Software, Inc. 
Type: Defect Fix 

It would appear that the timespec normalize code has an off by one error.
Found in three places.

Signed-off-by: George Anzinger<george@mvista.com>

 include/linux/time.h  |    2 +-
 kernel/posix-timers.c |   10 +++-------
 2 files changed, 4 insertions(+), 8 deletions(-)

Index: linux-2.6.15-rc/include/linux/time.h
===================================================================
--- linux-2.6.15-rc.orig/include/linux/time.h
+++ linux-2.6.15-rc/include/linux/time.h
@@ -101,7 +101,7 @@ extern struct timespec timespec_trunc(st
 static inline void
 set_normalized_timespec (struct timespec *ts, time_t sec, long nsec)
 {
-	while (nsec > NSEC_PER_SEC) {
+	while (nsec >= NSEC_PER_SEC) {
 		nsec -= NSEC_PER_SEC;
 		++sec;
 	}
Index: linux-2.6.15-rc/kernel/posix-timers.c
===================================================================
--- linux-2.6.15-rc.orig/kernel/posix-timers.c
+++ linux-2.6.15-rc/kernel/posix-timers.c
@@ -270,7 +270,7 @@ static void tstojiffie(struct timespec *
 	long sec = tp->tv_sec;
 	long nsec = tp->tv_nsec + res - 1;
 
-	if (nsec > NSEC_PER_SEC) {
+	if (nsec >= NSEC_PER_SEC) {
 		sec++;
 		nsec -= NSEC_PER_SEC;
 	}
@@ -1209,13 +1209,9 @@ static int do_posix_clock_monotonic_get(
 
 	do_posix_clock_monotonic_gettime_parts(tp, &wall_to_mono);
 
-	tp->tv_sec += wall_to_mono.tv_sec;
-	tp->tv_nsec += wall_to_mono.tv_nsec;
+	set_normalized_timespec(tp, tp->tv_sec += wall_to_mono.tv_sec,
+				tv_nsec += wall_to_mono.tv_nsec);
 
-	if ((tp->tv_nsec - NSEC_PER_SEC) > 0) {
-		tp->tv_nsec -= NSEC_PER_SEC;
-		tp->tv_sec++;
-	}
 	return 0;
 }
 

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] Timespec normalize off by one errors
       [not found] ` <1131744325.9957.20.camel@localhost>
@ 2005-11-11 22:15   ` George Anzinger
  0 siblings, 0 replies; 3+ messages in thread
From: George Anzinger @ 2005-11-11 22:15 UTC (permalink / raw)
  To: Joe Perches, Andrew Morton, lkml

Joe Perches wrote:
> On Fri, 2005-11-11 at 13:16 -0800, George Anzinger wrote:
> 
>>@@ -1209,13 +1209,9 @@ static int do_posix_clock_monotonic_get(
>> 
>> 	do_posix_clock_monotonic_gettime_parts(tp, &wall_to_mono);
>> 
>>-	tp->tv_sec += wall_to_mono.tv_sec;
>>-	tp->tv_nsec += wall_to_mono.tv_nsec;
>>+	set_normalized_timespec(tp, tp->tv_sec += wall_to_mono.tv_sec,
>>+				tv_nsec += wall_to_mono.tv_nsec);
>> 
>>-	if ((tp->tv_nsec - NSEC_PER_SEC) > 0) {
>>-		tp->tv_nsec -= NSEC_PER_SEC;
>>-		tp->tv_sec++;
>>-	}
>> 	return 0;
>> }
> 
> 
> This is extremely ugly.
> 
> tp->tv_sec += wall_to_mono.tv_sec;
> tp->tv_nsec += wall_to_mono.tv_nsec;
> set_normalized_timespec(tp, tv->sec, tv_nsec);
> 
> is much more intelligible.
> 
> Besides, I think you forgot the "tp->" indirection in the
> last argument of the set_normalized_timespec.

Oh SHIT.  That is wrong.  Think I will compile the next version...  Stand by.
> 
> cheers,  Joe
> 

-- 
George Anzinger   george@mvista.com
HRT (High-res-timers):  http://sourceforge.net/projects/high-res-timers/

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] Timespec normalize off by one errors
  2005-11-11 21:16 [PATCH] Timespec normalize off by one errors George Anzinger
       [not found] ` <1131744325.9957.20.camel@localhost>
@ 2005-11-11 22:25 ` George Anzinger
  1 sibling, 0 replies; 3+ messages in thread
From: George Anzinger @ 2005-11-11 22:25 UTC (permalink / raw)
  Cc: Andrew Morton, lkml, benh, joe

[-- Attachment #1: Type: text/plain, Size: 219 bytes --]

George Anzinger wrote:
> Found three places were we use ">" instead of ">=".
> 
Lets just try this again.
-- 
George Anzinger   george@mvista.com
HRT (High-res-timers):  http://sourceforge.net/projects/high-res-timers/

[-- Attachment #2: time-spec2.patch --]
[-- Type: text/plain, Size: 1635 bytes --]

Source: MontaVista Software, Inc. 
Type: Defect Fix 

It would appear that the timespec normalize code has an off by one error.
Found in three places.

Signed-off-by: George Anzinger<george@mvista.com>

 include/linux/time.h  |    2 +-
 kernel/posix-timers.c |   10 +++-------
 2 files changed, 4 insertions(+), 8 deletions(-)

Index: linux-2.6.15-rc/include/linux/time.h
===================================================================
--- linux-2.6.15-rc.orig/include/linux/time.h
+++ linux-2.6.15-rc/include/linux/time.h
@@ -101,7 +101,7 @@ extern struct timespec timespec_trunc(st
 static inline void
 set_normalized_timespec (struct timespec *ts, time_t sec, long nsec)
 {
-	while (nsec > NSEC_PER_SEC) {
+	while (nsec >= NSEC_PER_SEC) {
 		nsec -= NSEC_PER_SEC;
 		++sec;
 	}
Index: linux-2.6.15-rc/kernel/posix-timers.c
===================================================================
--- linux-2.6.15-rc.orig/kernel/posix-timers.c
+++ linux-2.6.15-rc/kernel/posix-timers.c
@@ -270,7 +270,7 @@ static void tstojiffie(struct timespec *
 	long sec = tp->tv_sec;
 	long nsec = tp->tv_nsec + res - 1;
 
-	if (nsec > NSEC_PER_SEC) {
+	if (nsec >= NSEC_PER_SEC) {
 		sec++;
 		nsec -= NSEC_PER_SEC;
 	}
@@ -1209,13 +1209,9 @@ static int do_posix_clock_monotonic_get(
 
 	do_posix_clock_monotonic_gettime_parts(tp, &wall_to_mono);
 
-	tp->tv_sec += wall_to_mono.tv_sec;
-	tp->tv_nsec += wall_to_mono.tv_nsec;
+	set_normalized_timespec(tp, tp->tv_sec + wall_to_mono.tv_sec,
+				tp->tv_nsec + wall_to_mono.tv_nsec);
 
-	if ((tp->tv_nsec - NSEC_PER_SEC) > 0) {
-		tp->tv_nsec -= NSEC_PER_SEC;
-		tp->tv_sec++;
-	}
 	return 0;
 }
 

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2005-11-11 22:25 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2005-11-11 21:16 [PATCH] Timespec normalize off by one errors George Anzinger
     [not found] ` <1131744325.9957.20.camel@localhost>
2005-11-11 22:15   ` George Anzinger
2005-11-11 22:25 ` George Anzinger

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox