Bug 92687 – Segfoult after changeing shmmax

• Status: RESOLVED TEST-REQUEST
• Product: Gentoo Linux

[Description Sebastian Fohler 2005-05-15 04:08:11 UTC]

I changed my shmmax in /proc/sys/kernel to the value 1342117728
after that change I get segfaults on using some different programms, one is especially the use of emerge sync where I get this: 

Call Trace:
 [<c016bcec>] vfs_lstat+0x4c/0x60
 [<c016c3eb>] sys_lstat64+0x1b/0x40
 [<c0115740>] do_page_fault+0x0/0x649
 [<c01032af>] syscall_call+0x7/0xb
Code: 45 3c 5b 5e 5f 5d c3 8d 74 26 00 83 ec 18 8b 44 24 20 89 74 24 10 8b 74 24 24 89 5c 24 0c 89 7c 24 14 8b 58 10 8b 93 a4 00 00 00 <8b> 4a 3c 85 c9 74 29 89 44 24 04 8b 44 24 1c 89 74 24 08 89 04
 <1>Unable to handle kernel paging request at virtual address 5046297c
 printing eip:
c016bbb0
*pde = 00000000
Oops: 0000 [#22]
PREEMPT SMP
Modules linked in: e1000
CPU:    0
EIP:    0060:[<c016bbb0>]    Not tainted VLI
EFLAGS: 00010286   (2.6.11-gentoo-r6)
EIP is at vfs_getattr+0x20/0xb0
eax: f3b65830   ebx: f3b68dc8   ecx: 00000000   edx: 50462940
esi: df289f6c   edi: 000081a4   ebp: df289000   esp: df289ef0
ds: 007b   es: 007b   ss: 0068
Process rsync (pid: 9773, threadinfo=df289000 task=f7dba550)
Stack: c592e680 dd631000 df289f6c 00000000 df289f18 000081a4 c016bcec c5940180
       f3b65830 df289f6c f3b65830 c5940180 00000000 00020000 00000008 00000000
       00000001 00000000 4255a09b 00000000 4267de32 00000000 00016ca6 00000000
Call Trace:
 [<c016bcec>] vfs_lstat+0x4c/0x60
 [<c016c3eb>] sys_lstat64+0x1b/0x40
 [<c01181a4>] sched_fork+0x24/0xc0
 [<c0115740>] do_page_fault+0x0/0x649
 [<c01032af>] syscall_call+0x7/0xb
Code: 45 3c 5b 5e 5f 5d c3 8d 74 26 00 83 ec 18 8b 44 24 20 89 74 24 10 8b 74 24 24 89 5c 24 0c 89 7c 24 14 8b 58 10 8b 93 a4 00 00 00 <8b> 4a 3c 85 c9 74 29 89 44 24 04 8b 44 24 1c 89 74 24 08 89 04
 <1>Unable to handle kernel paging request at virtual address 20666f20
 printing eip:
20666f20
*pde = 00000000
Oops: 0000 [#23]
PREEMPT SMP
Modules linked in: e1000
CPU:    0
EIP:    0060:[<20666f20>]    Not tainted VLI
EFLAGS: 00010202   (2.6.11-gentoo-r6)
EIP is at 0x20666f20
eax: c5940180   ebx: e3b68d48   ecx: 20666f20   edx: f0b69940
esi: dc86cf6c   edi: bffff2ac   ebp: dc86c000   esp: dc86ceec
ds: 007b   es: 007b   ss: 0068
Process updatedb (pid: 22672, threadinfo=dc86c000 task=f7205a80)
Stack: c016bbc9 c5940180 e3b67b90 dc86cf6c 00000000 dc86cf18 bffff2ac c016bcec
       c5940180 e3b67b90 dc86cf6c e3b67b90 c5940180 00000000 00020000 00000058
       00000000 00000001 00000000 42680b59 00000000 42680b59 00000000 0004731d
Call Trace:
 [<c016bbc9>] vfs_getattr+0x39/0xb0
 [<c016bcec>] vfs_lstat+0x4c/0x60
 [<c016c3eb>] sys_lstat64+0x1b/0x40
 [<c01781a4>] __posix_lock_file+0x2f4/0x640
 [<c01032af>] syscall_call+0x7/0xb
Code:  Bad EIP value.

if I call dmesg after emerge sync.
I use Gentoo Kernel 2.6.11-r6

[Comment 1 Daniel Drake (RETIRED) 2005-05-18 15:04:43 UTC]

Please try and reproduce on vanilla-sources-2.6.12_rc4

[Comment 2 Sebastian Fohler 2005-05-18 17:55:06 UTC]

That could be a problem because its a productive system, but I will try

[Comment 3 Daniel Drake (RETIRED) 2005-05-31 16:09:54 UTC]

OK - please reopen when you are able to test the latest development kernel
(currently 2.6.12-rc5). If the problem exists there then we can form a bug
report to send upstream.