Skip to contentSkip to conversation

gcc coredump with 2.6.12+ kernels

DISCUSSION Johnny Stenback 2 participants 5 messages Sep 3, 2005 – Oct 30, 2005

lore
Share & tools
lore
#0 lore raw

Hey all,

I just attempted to upgrade my kernel to 2.6.13. The kernel appears to boot and run just fine, but when I try to build any larger projects like Mozilla or the Linux kernel I constantly get segfaults from gcc. All other apps *seem* to work fine. I remember seeing this with 2.6.12 too when I tried to upgrade to it too but I didn't have the time to investigate at all then, but now I see the same problem with 2.6.13. The last version I've used that didn't show this problem is 2.6.11.3, and that's running with no problems here.

When gcc segfaults I get the following messages in the messages log:

cc1[16775]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffaaf0a0 error 4 cc1[17086]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffc4dfc0 error 4 cc1[17788]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffd777e0 error 4 cc1[17823]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffc4d630 error 4 cc1[17895]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007ffffffd2330 error 4

I'm on a dual AMD Opteron system, running x86_64 code. Using Fedora Core 2 (yeah, old, I know...) and gcc 3.3.3 20040412.

I've ran memtest etc on this system, and with 2.6.11.3 and older kernels I've ran this system under moderate load for months w/o rebooting and I've never had any problems like this, so I'm fairly certain that I don't have bad hardware.

Anyone have any clues as to what could be causing this problem?

Signature · 1 line
jst
#1 lore raw
Quoted 24 lines · On Sat, Sep 03, 2005 at 10:25:37AM -0700 Johnny Stenback wrote:

Hey all,

I just attempted to upgrade my kernel to 2.6.13. The kernel appears to boot and run just fine, but when I try to build any larger projects like Mozilla or the Linux kernel I constantly get segfaults from gcc. All other apps *seem* to work fine. I remember seeing this with 2.6.12 too when I tried to upgrade to it too but I didn't have the time to investigate at all then, but now I see the same problem with 2.6.13. The last version I've used that didn't show this problem is 2.6.11.3, and that's running with no problems here.

When gcc segfaults I get the following messages in the messages log:

cc1[16775]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffaaf0a0 error 4
cc1[17086]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffc4dfc0 error 4
cc1[17788]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffd777e0 error 4
cc1[17823]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffc4d630 error 4
cc1[17895]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007ffffffd2330 error 4

Quoted 2 lines

I'm on a dual AMD Opteron system, running x86_64 code. Using Fedora Core 2 (yeah, old, I know...) and gcc 3.3.3 20040412.

Does it still happen if you run:

echo 0 > /proc/sys/kernel/randomize_va_space

#2 lore raw
Quoted 14 lines · Alexander Nyberg
Quoted 12 lines · On Sat, Sep 03, 2005 at 10:25:37AM -0700 Johnny Stenback wrote:

Hey all,

I just attempted to upgrade my kernel to 2.6.13. The kernel appears to boot and run just fine, but when I try to build any larger projects like Mozilla or the Linux kernel I constantly get segfaults from gcc. All other apps *seem* to work fine. I remember seeing this with 2.6.12 too when I tried to upgrade to it too but I didn't have the time to investigate at all then, but now I see the same problem with 2.6.13. The last version I've used that didn't show this problem is 2.6.11.3, and that's running with no problems here.

When gcc segfaults I get the following messages in the messages log:

cc1[16775]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffaaf0a0 error 4 cc1[17086]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffc4dfc0 error 4 cc1[17788]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffd777e0 error 4 cc1[17823]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffc4d630 error 4 cc1[17895]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007ffffffd2330 error 4

Quoted 5 lines
Quoted 1 line

I'm on a dual AMD Opteron system, running x86_64 code. Using Fedora Core 2 (yeah, old, I know...) and gcc 3.3.3 20040412.

Does it still happen if you run:

echo 0 > /proc/sys/kernel/randomize_va_space

Just tried that, and I still get the same error, and the same error in the log too (just a different address):

cc1plus[2961]: segfault at 0000000000000000 rip
00000036f2b0119e rsp 00007fffffffdbb0 error 4

Anything else I can try?

Quoted 6 lines

-
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/

Signature · 1 line
jst
#3 lore raw

Hello again,

Got some more info that may be of value here. After I saw gcc coredump on me with 2.6.13 I tried halting my system and I saw a kernel panic while shutting down. The stack trace it showed me was:

flush_tlb_mm+201
exit_mmap+249
mmput+49
do_exit+434
sys_reboot+264
__group_send_sig_info+184
group_send_info+118
thread_return
thread_return+82
filp_close+110
system_call+126

All typed down by hand since it never made it into a log, so there can be omissions and typos in there (and I didn't feel like typing down all those 64-bit hex numbers, sorry).

Quoted 35 lines · Johnny Stenback
Quoted 14 lines · Alexander Nyberg
Quoted 12 lines · On Sat, Sep 03, 2005 at 10:25:37AM -0700 Johnny Stenback wrote:

Hey all,

I just attempted to upgrade my kernel to 2.6.13. The kernel appears to boot and run just fine, but when I try to build any larger projects like Mozilla or the Linux kernel I constantly get segfaults from gcc. All other apps *seem* to work fine. I remember seeing this with 2.6.12 too when I tried to upgrade to it too but I didn't have the time to investigate at all then, but now I see the same problem with 2.6.13. The last version I've used that didn't show this problem is 2.6.11.3, and that's running with no problems here.

When gcc segfaults I get the following messages in the messages log:

cc1[16775]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffaaf0a0 error 4 cc1[17086]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffc4dfc0 error 4 cc1[17788]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffd777e0 error 4 cc1[17823]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffc4d630 error 4 cc1[17895]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007ffffffd2330 error 4

Quoted 5 lines
Quoted 1 line

I'm on a dual AMD Opteron system, running x86_64 code. Using Fedora Core 2 (yeah, old, I know...) and gcc 3.3.3 20040412.

Does it still happen if you run:

echo 0 > /proc/sys/kernel/randomize_va_space

Just tried that, and I still get the same error, and the same error in the log too (just a different address):

cc1plus[2961]: segfault at 0000000000000000 rip
00000036f2b0119e rsp 00007fffffffdbb0 error 4

Anything else I can try?

Quoted 6 lines

-
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/

Signature · 1 line
jst
#4 lore raw
In reply to Alexander Nyberg

FWIW, this issue was resolved in 2.6.14. I don't know off hand which patch did it, but I don't see this problem now after upgrading to 2.6.14.

Thanks to whoever figured this out :)

Quoted 25 lines · Alexander Nyberg
Quoted 6 lines · On Sat, Sep 03, 2005 at 10:25:37AM -0700 Johnny Stenback wrote:

Hey all,

I just attempted to upgrade my kernel to 2.6.13. The kernel appears to boot and run just fine, but when I try to build any larger projects like Mozilla or the Linux kernel I constantly get segfaults from gcc. All other apps *seem* to work fine. I remember seeing this with 2.6.12 too when I tried to upgrade to it too but I didn't have the time to investigate at all then, but now I see the same problem with 2.6.13. The last version I've used that didn't show this problem is 2.6.11.3, and that's running with no problems here.

When gcc segfaults I get the following messages in the messages log:

Quoted 7 lines

cc1[16775]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffaaf0a0 error 4 cc1[17086]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffc4dfc0 error 4 cc1[17788]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffd777e0 error 4 cc1[17823]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007fffffc4d630 error 4 cc1[17895]: segfault at 0000000000000000 rip 00000036f2b0119e rsp 00007ffffffd2330 error 4

I'm on a dual AMD Opteron system, running x86_64 code. Using Fedora Core 2 (yeah, old, I know...) and gcc 3.3.3 20040412.

Does it still happen if you run:

echo 0 > /proc/sys/kernel/randomize_va_space
-
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/

Signature · 1 line
jst