Bug 104363
Summary: | acroread abends on Find with no hit | ||
---|---|---|---|
Product: | [Retired] Red Hat Linux | Reporter: | Mike Vanecek <bug_rh> |
Component: | mozilla | Assignee: | Christopher Blizzard <blizzard> |
Status: | CLOSED NOTABUG | QA Contact: | |
Severity: | medium | Docs Contact: | |
Priority: | medium | ||
Version: | 9 | CC: | rmj |
Target Milestone: | --- | ||
Target Release: | --- | ||
Hardware: | i586 | ||
OS: | Linux | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | Bug Fix | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2003-09-19 19:09:53 UTC | Type: | --- |
Regression: | --- | Mount Type: | --- |
Documentation: | --- | CRM: | |
Verified Versions: | Category: | --- | |
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
Cloudforms Team: | --- | Target Upstream Version: | |
Embargoed: |
Description
Mike Vanecek
2003-09-13 14:30:31 UTC
We can't fix bugs in acrobat since it's not open source. I am not asking for you to fix acroread. It works on windows, it works on red hat 7.3, and may work on other distros. I am asking you to fix whatever is broken in red hat 9 so that an application that runs on serveral other systems also runs on red hat 9. Have you tried setting LANG=en_US before starting acroread? Yes, LANG is set correctly. [admin@www admin]$ echo $LANG en_US Except for the Find no-find problem, acroread works just fine (Searches that find something work, printing works, etc.). It is just the Find no-find abend that is a problem. This has been reported on other RH 8/9 systems on one of the mailing lists as well. Could this problem have anything to do with the migration from GCC 2.95 to 3.x? What is different on RH 8/9 from RH 7.3. BTW, this has also been reported to Adobe. No resolution from them either. FYI, for me, it crashes just the same on rh73 too. Just to add, I'm running around 40 desktops in various flavors of RedHat, and so can offer some potentially useful information. Adobe Acrobat Reader 5.0.x (tested x=6,7,8), on our systems running RedHat versions 8 and higher (tested 8,9,taroon), consistantly dumps core after the execution of a "Find" for text that does not exist in the document. The same acroread binaries (run from NFS) do not have this problem on any of our RedHat 7.1, 7.2 or 7.3 systems. Since it was a change in RedHat that caused the problem to appear, not an Adobe change), we were hoping that this would be an appropriate forum to discuss the issue, even though acrobat reader is not officially part of the distribution. I confirm this as well. Same acroread works fine in RH 7.1/7.3 but crashes in RH9 if the searched text is not found. Just attempt to reproduce it yourselves and see. strace ends with ioctl(3, FIONREAD, [0]) = 0 write(3, ";\3\7\0\32\1 \2\0\0\0\0:\1\371\0\t\0\332\0W\2\371\0\10"..., 144) = 144 gettimeofday({1073941705, 442533}, NULL) = 0 --- SIGSEGV (Segmentation fault) @ 0 (0) --- rt_sigprocmask(SIG_UNBLOCK, [SEGV], NULL, 8) = 0 mmap2(NULL, 2097152, PROT_NONE, MAP_PRIVATE|MAP_ANONYMOUS|MAP_NORESERVE, -1, 0) = 0x4115b000 munmap(0x4115b000, 675840) = 0 munmap(0x41300000, 372736) = 0 mprotect(0x41200000, 32768, PROT_READ|PROT_WRITE) = 0 rt_sigprocmask(SIG_UNBLOCK, [ABRT], NULL, 8) = 0 gettid() = 24890 tkill(24890, SIGABRT) = 0 --- SIGABRT (Aborted) @ 0 (0) --- +++ killed by SIGABRT +++ Someone in one of the Adobe forums suggested: export MALLOC_CHECK_=0 as a workaround for this problem. It seems to cure the problem for me on FC1, but I've no idea if there are any bad side effects. I tried export MALLOC_CHECK_=0 on my RH 9 system and the find still abends at the point where I click OK on the no find message. I wonder why it works with FC1 and not RH 9? Don't plan on leaving RH 9 for a couple more years or so and would like to see this work. I found the message to which you refer: Topic name: Acroread 5.0.6/7/8 on Linux abends with Find no hit Forum name: Acrobat UNIX - Message #14 Posted By: Ar Nehe - Mar 31, 2004 01:50 am Here is a workaround until Adobe has fixed the problem: Add following 2 lines to /usr/X11R6/bin/acroread. MALLOC_CHECK_=0 export MALLOC_CHECK_ (see also http://lists.suse.com/archive/suse-linux/2003-Nov/1455.html) I added the lines mentioned above to /usr/bin/acroread and indeed the error does stop. Not sure why it needs this way to work, but at least it works. |