Date: Mon, 3 Mar 2014 11:56:01 GMT From: Petr Lampa <lampa@fit.vutbr.cz> To: freebsd-gnats-submit@FreeBSD.org Subject: bin/187221: fsck_ufs -p segmentation fault with SU+J Message-ID: <201403031156.s23Bu1Gr094701@cgiserv.freebsd.org> Resent-Message-ID: <201403031200.s23C00Pj026967@freefall.freebsd.org>
next in thread | raw e-mail | index | archive | help
>Number: 187221
>Category: bin
>Synopsis: fsck_ufs -p segmentation fault with SU+J
>Confidential: no
>Severity: non-critical
>Priority: low
>Responsible: freebsd-bugs
>State: open
>Quarter:
>Keywords:
>Date-Required:
>Class: sw-bug
>Submitter-Id: current-users
>Arrival-Date: Mon Mar 03 12:00:00 UTC 2014
>Closed-Date:
>Last-Modified:
>Originator: Petr Lampa
>Release: 9.2-STABLE
>Organization:
BUT FIT
>Environment:
FreeBSD xxx 9.2-STABLE FreeBSD 9.2-STABLE #23: Thu Jan 23 08:17:47 CET 2014 root@xxx:/usr/src/sys/amd64/compile/XXX amd64
>Description:
Some mismatch in SU+J journal was found during initial fsck -p
and it ends with segmentaion fault. Full fsck -y run without problems.
(gdb) where
#0 0x0000000000404fb0 in ckfini (markclean=0)
at /usr/src/sbin/fsck_ffs/fsutil.c:418
#1 0x000000000040548b in pfatal (fmt=<value optimized out>)
at /usr/src/sbin/fsck_ffs/fsutil.c:980
#2 0x00000000004047e1 in reply (question=0x417d07 "FALLBACK TO FULL FSCK")
at /usr/src/sbin/fsck_ffs/fsutil.c:106
#3 0x0000000000413462 in suj_check (filesys=0x41007928 "/dev/mirror/root")
at /usr/src/sbin/fsck_ffs/suj.c:2685
#4 0x0000000000408ca8 in main (argc=8193, argv=0x7fffffffdc50)
at /usr/src/sbin/fsck_ffs/main.c:403
(gdb) frame 0
#0 0x0000000000404fb0 in ckfini (markclean=0)
at /usr/src/sbin/fsck_ffs/fsutil.c:418
418 if (cgbufs[cnt].b_un.b_cg == NULL)
(gdb) l
413 free((char *)bp);
414 }
415 if (numbufs != cnt)
416 errx(EEXIT, "panic: lost %d buffers", numbufs - cnt);
417 for (cnt = 0; cnt < sblock.fs_ncg; cnt++) {
418 if (cgbufs[cnt].b_un.b_cg == NULL)
419 continue;
420 flush(fswritefd, &cgbufs[cnt]);
421 free(cgbufs[cnt].b_un.b_cg);
422 }
p cgbufs
$4 = (struct bufarea *) 0x0
(gdb)
It seems that cgbufs were not allocated, but they are flushed in ckfini() in
this case. This look similar problem like
http://docs.freebsd.org/cgi/getmsg.cgi?fetch=478954+0+current/svn-src-head
>How-To-Repeat:
>Fix:
>Release-Note:
>Audit-Trail:
>Unformatted:
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201403031156.s23Bu1Gr094701>