Skip to contentSkip to conversation

time/ntp fix

GIT PULL Ingo Molnar 5 participants 7 messages Feb 20, 2015 – Feb 22, 2015 Timekeeping, clocksource core, NTP, alarmtimer

lore
Share & tools
lore
#0 lore raw

Linus,

Please pull the latest timers-urgent-for-linus git tree from:

git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git timers-urgent-for-linus

# HEAD: 29183a70b0b828500816bd794b3fe192fce89f73 ntp: Fixup adjtimex freq validation on 32-bit systems

An adjtimex interface regression fix for 32-bit systems.

Thanks,

Ingo

------------------>
John Stultz (1):
ntp: Fixup adjtimex freq validation on 32-bit systems

 kernel/time/ntp.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)
kernel/time/ntp.c +7 -3
index 4b585e0fdd22..0f60b08a4f07 100644
@@ -633,10 +633,14 @@ int ntp_validate_timex(struct timex *txc)
 	if ((txc->modes & ADJ_SETOFFSET) && (!capable(CAP_SYS_TIME)))
 		return -EPERM;
 
-	if (txc->modes & ADJ_FREQUENCY) {
-		if (LONG_MIN / PPM_SCALE > txc->freq)
+	/*
+	 * Check for potential multiplication overflows that can
+	 * only happen on 64-bit systems:
+	 */
+	if ((txc->modes & ADJ_FREQUENCY) && (BITS_PER_LONG == 64)) {
+		if (LLONG_MIN / PPM_SCALE > txc->freq)
 			return -EINVAL;
-		if (LONG_MAX / PPM_SCALE < txc->freq)
+		if (LLONG_MAX / PPM_SCALE < txc->freq)
 			return -EINVAL;
 	}
 
#1 lore raw
Quoted 27 lines · Ingo Molnar

Linus,

Please pull the latest timers-urgent-for-linus git tree from:

git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git timers-urgent-for-linus

# HEAD: 29183a70b0b828500816bd794b3fe192fce89f73 ntp: Fixup adjtimex freq validation on 32-bit systems

An adjtimex interface regression fix for 32-bit systems.

Thanks,

Ingo

------------------>
John Stultz (1):
ntp: Fixup adjtimex freq validation on 32-bit systems

 kernel/time/ntp.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)
kernel/time/ntp.c +0 -0
index 4b585e0fdd22..0f60b08a4f07 100644

kernel/time/ntp.c int ntp_validate_timex(struct timex *txc) · line 638

Quoted 8 lines
kernel/time/ntp.c +3 -2
@@ -633,10 +633,14 @@ int ntp_validate_timex(struct timex *txc)
        if ((txc->modes & ADJ_SETOFFSET) && (!capable(CAP_SYS_TIME)))
                return -EPERM;
 
-       if (txc->modes & ADJ_FREQUENCY) {
-               if (LONG_MIN / PPM_SCALE > txc->freq)
+       /*
+        * Check for potential multiplication overflows that can
+        * only happen on 64-bit systems:

Typo ?

Quoted 14 lines
kernel/time/ntp.c +4 -2
+        */
+       if ((txc->modes & ADJ_FREQUENCY) && (BITS_PER_LONG == 64)) {
+               if (LLONG_MIN / PPM_SCALE > txc->freq)
                        return -EINVAL;
-               if (LONG_MAX / PPM_SCALE < txc->freq)
+               if (LLONG_MAX / PPM_SCALE < txc->freq)
                        return -EINVAL;
        }
 
--

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

#2 lore raw
In reply to Ingo Molnar
Quoted 3 lines · Ingo Molnar

John Stultz (1):
ntp: Fixup adjtimex freq validation on 32-bit systems

This is confusing. 32-bit?

kernel/time/ntp.c int ntp_validate_timex(struct timex *txc) · line 638

Quoted 3 lines
kernel/time/ntp.c +3 -0
+       /*
+        * Check for potential multiplication overflows that can
+        * only happen on 64-bit systems:

64-bit?

Quoted 1 line

+ if ((txc->modes & ADJ_FREQUENCY) && (BITS_PER_LONG == 64)) {

Hmm. The code clearly says "&& (BITS_PER_LONG == 64)"

But:

kernel/time/ntp.c int ntp_validate_timex(struct timex *txc) · line 644

Quoted 5 lines
kernel/time/ntp.c +2 -1
+               if (LLONG_MIN / PPM_SCALE > txc->freq)
                        return -EINVAL;
-               if (LONG_MAX / PPM_SCALE < txc->freq)
+               if (LLONG_MAX / PPM_SCALE < txc->freq)
                        return -EINVAL;

The difference between LONG_MAX and LLONG_MAX only matters if BITS_PER_LONG would be 32.

So the changes are confusing to begin with and the commit log description doesn't match them anyway.

I'm not pulling this without clarifications.

Linus

#3 lore raw
Quoted 6 lines · Linus Torvalds
Quoted 3 lines · Ingo Molnar

John Stultz (1):
ntp: Fixup adjtimex freq validation on 32-bit systems

This is confusing. 32-bit?

Right, so the check that was added in a previous commit is really only a concern for 64bit systems, but was applied to both 32 and 64bit systems, which results in breaking 32bit systems.

Thus the "fix" here is to make the check only apply to 64bit systems.

Quoted 12 lines

kernel/time/ntp.c int ntp_validate_timex(struct timex *txc) · line 638

Quoted 3 lines
kernel/time/ntp.c +3 -0
+       /*
+        * Check for potential multiplication overflows that can
+        * only happen on 64-bit systems:

64-bit?

Quoted 1 line

+ if ((txc->modes & ADJ_FREQUENCY) && (BITS_PER_LONG == 64)) {

Hmm. The code clearly says "&& (BITS_PER_LONG == 64)"

But:

Quoted 8 lines

kernel/time/ntp.c int ntp_validate_timex(struct timex *txc) · line 644

Quoted 5 lines
kernel/time/ntp.c +2 -1
+               if (LLONG_MIN / PPM_SCALE > txc->freq)
                        return -EINVAL;
-               if (LONG_MAX / PPM_SCALE < txc->freq)
+               if (LLONG_MAX / PPM_SCALE < txc->freq)
                        return -EINVAL;

The difference between LONG_MAX and LLONG_MAX only matters if BITS_PER_LONG would be 32.

Right. So v1 of this fix just changed the LONG_MIX/MAX to be LLONG_MIN/MAX. This resolves the issue, but with some compiler checks we get warnings since LLONG_MIN/MAX is always smaller/larger then a 32bit value. Thus the extra BITS_PER_LONG check (which I think is a bit ugly, but I didn't get any better suggestions) was added to avoid the build warnings as well.

So the fix is somewhat redundant, but I do think the LLONG specifier is more exact.

Quoted 4 lines

So the changes are confusing to begin with and the commit log description doesn't match them anyway.

I'm not pulling this without clarifications.

Is the above clarification sufficient? Or would you like me to reword the commit message as well?

thanks
-john

#4 lore raw
Quoted 3 lines · John Stultz

Is the above clarification sufficient? Or would you like me to reword the commit message as well?

It's ok. And I'll add it to the merge message so that it's mentioned somewhere at least.

Thanks,
Linus

#5 lore raw
In reply to John Stultz
Quoted 9 lines · John Stultz
Quoted 6 lines · Linus Torvalds
Quoted 3 lines · Ingo Molnar

John Stultz (1):
ntp: Fixup adjtimex freq validation on 32-bit systems

This is confusing. 32-bit?

Quoted 7 lines

Right, so the check that was added in a previous commit is really only a concern for 64bit systems, but was applied to both 32 and 64bit systems, which results in breaking 32bit systems.

Thus the "fix" here is to make the check only apply to 64bit systems.

Yeah, perhaps a better commit title would have been to write:

time/ntp: Fix adjtimex freq validation code build warning on 32-bit systems

To make it clear that the problem fixed is a 32-bit warning, and that the fix for that is to only check on 64-bit systems.

I agreed with your BITS_PER_LONG check when I reviewed your patch, people usually do an ugly #ifdef, I think this in line check form is nicer.

Thanks,

Ingo

#6 lore raw

Hi Ingo,

Quoted 24 lines · Ingo Molnar
Quoted 9 lines · John Stultz
Quoted 6 lines · Linus Torvalds
Quoted 3 lines · Ingo Molnar

John Stultz (1):
ntp: Fixup adjtimex freq validation on 32-bit systems

This is confusing. 32-bit?

Quoted 7 lines

Right, so the check that was added in a previous commit is really only a concern for 64bit systems, but was applied to both 32 and 64bit systems, which results in breaking 32bit systems.

Thus the "fix" here is to make the check only apply to 64bit systems.

Yeah, perhaps a better commit title would have been to write:

time/ntp: Fix adjtimex freq validation code build warning on 32-bit systems

Quoted 7 lines

To make it clear that the problem fixed is a 32-bit warning, and that the fix for that is to only check on 64-bit systems.

I agreed with your BITS_PER_LONG check when I reviewed your patch, people usually do an ugly #ifdef, I think this in line check form is nicer.

Unfortunately it doesn't help with all compiler versions. With gcc 4.1.2 for m68k:

kernel/time/ntp.c: In function ‘ntp_validate_timex’:
kernel/time/ntp.c:641: warning: comparison is always false due to limited range of data type
kernel/time/ntp.c:643: warning: comparison is always false due to limited range of data type

Gcc 4.6.3 and 4.9.0 are OK (for m68k).

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds