Bug 1125326 - ksh segfault
Summary: ksh segfault
Keywords:
Status: CLOSED DUPLICATE of bug 1112306
Alias: None
Product: Red Hat Enterprise Linux 6
Classification: Red Hat
Component: ksh
Version: 6.4
Hardware: All
OS: Linux
high
high
Target Milestone: rc
: ---
Assignee: Michal Hlavinka
QA Contact: BaseOS QE - Apps
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2014-07-31 15:03 UTC by jstephen
Modified: 2014-08-04 14:45 UTC (History)
0 users

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2014-08-04 10:44:09 UTC


Attachments (Terms of Use)
ksh core (1.56 MB, application/octet-stream)
2014-07-31 15:05 UTC, jstephen
no flags Details

Description jstephen 2014-07-31 15:03:52 UTC
Description of problem:
The crash is in ksh's version of free under job_chksave - the "segment" asked for in regionof() is null.

Version-Release number of selected component (if applicable):
ksh-20120801-10.el6_5.5.x86_64

How reproducible:
It is happening intermittently, not every time a job is run.

Steps to Reproduce:
1. No identifiable reproducer at this time
2.
3.

Actual results:
ksh crashes
[400075.213273] ksh[10862]: segfault at 0 ip 00000000004dbbb4 sp 00007fff29e5bb50 error 4 in ksh93[400000+15b000]

Expected results:
No crash

Additional info:
Core dump attached

The crash is in ksh's version of free under job_chksave - the "segment" asked for in regionof() is null.

#0  0x00000000004dbbb4 in regionof (addr=0x7f08c1921860) at /usr/src/debug/ksh-20120801/src/lib/libast/vmalloc/malloc.c:347
#1  0x00000000004dca78 in _ast_free (data=0x7f08c1921860) at /usr/src/debug/ksh-20120801/src/lib/libast/vmalloc/malloc.c:594
#2  0x0000000000428154 in job_chksave (pid=<value optimized out>) at /usr/src/debug/ksh-20120801/src/cmd/ksh93/sh/jobs.c:1980
#3  0x000000000042ac30 in job_post (shp=0x76cba0, pid=11334, join=<value optimized out>) at /usr/src/debug/ksh-20120801/src/cmd/ksh93/sh/jobs.c:1396


#0  0x00000000004dbbb4 in regionof (addr=0x7f08c1921860) at /usr/src/debug/ksh-20120801/src/lib/libast/vmalloc/malloc.c:347
347                     Vmdata_t *vd = SEG(BLOCK(addr))->vmdt;

#define SEG(b)          ((b)->head.head.seg.seg)
#define BLOCK(d)        ((Block_t*)((char*)(d) - sizeof(Head_t)) )

(gdb) ptype Block_t
type = struct _block_s {
    Head_t head;
    Body_t body;
}

(gdb) p ((Block_t *)((char *)addr - sizeof(Head_t)))
$28 = (Block_t *) 0x7f08c1921850

(gdb) p ((Block_t *)((char *)addr - sizeof(Head_t)))->head.head.seg.seg
$30 = (Seg_t *) 0x0

I suspect that this is a kind of double-free that is happening in job_post/job_save.

I found the following BZ's that have very similar backtraces, it could be another manifestation of these problems in the job list code:
https://bugzilla.redhat.com/show_bug.cgi?id=1110063
https://bugzilla.redhat.com/show_bug.cgi?id=825520

Comment 1 jstephen 2014-07-31 15:05:20 UTC
Created attachment 922958 [details]
ksh core

Comment 2 Michal Hlavinka 2014-08-04 10:44:09 UTC
This is most likely duplicate of bug #1112306 but as there are no reproducers for either of these bugs, it's hard to say.

Anyway, closing as a duplicate at least for now - until you can reproduce this crash with fixed ksh package referenced in the duplicate bug. If you can do that, feel free to reopen this bug.

*** This bug has been marked as a duplicate of bug 1112306 ***


Note You need to log in before you can comment on or make changes to this bug.