Bug 1031333

Summary: [abrt] alpine-2.10-4.fc19: fs_give: Process /usr/bin/alpine was killed by signal 6 (SIGABRT)
Product: [Fedora] Fedora Reporter: Paul DeStefano <paul.destefano-redhat2>
Component: alpineAssignee: Joshua Daniel Franklin <joshuadfranklin>
Status: CLOSED ERRATA QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 19CC: chappa, jima, joshuadfranklin, paul.destefano-redhat2, pwouters, rdieter
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Unspecified   
URL: https://retrace.fedoraproject.org/faf/reports/bthash/88fffcfff012a992de09bd568644753f81542e36
Whiteboard: abrt_hash:127ccc86ce4bb61fa1c89396a5cc4188c4b87073
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2013-12-07 14:30:08 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Attachments:
Description Flags
File: backtrace
none
File: cgroup
none
File: core_backtrace
none
File: dso_list
none
File: environ
none
File: limits
none
File: maps
none
File: open_fds
none
File: proc_pid_status
none
File: var_log_messages
none
.pinerc none

Description Paul DeStefano 2013-11-16 22:59:59 UTC 
Description of problem:
I think I just hit 'r' to reply to a message.  But, was stands out is that a few commands eariler, I tried to open a URL that was huge and my web-browser never loaded the page.  Normally, that works fine.  But, I've noticed this before with very long URLs.

Version-Release number of selected component:
alpine-2.10-4.fc19

Additional info:
reporter:       libreport-2.1.9
backtrace_rating: 4
crash_function: fs_give
executable:     /usr/bin/alpine
kernel:         3.11.4-201.fc19.x86_64
runlevel:       N 5
type:           CCpp
uid:            13013

Truncated backtrace:
Thread no. 1 (10 frames)
 #5 fs_give at fs_unix.c:68
 #6 free_color_table at color.c:1072
 #7 pico_toggle_color at color.c:1231
 #8 pico_endcolor at color.c:1272
 #9 end_screen at termout.unx.c:333
 #10 ttyfix at signal.c:737
 #11 pipe_callback at signal.c:563
 #12 open_system_pipe at pipe.c:336
 #13 dfilter at dispfilt.c:102
 #14 detach at detach.c:315
Comment 1 Paul DeStefano 2013-11-16 23:00:09 UTC 
Created attachment 824997 [details]
File: backtrace
Comment 2 Paul DeStefano 2013-11-16 23:00:14 UTC 
Created attachment 824998 [details]
File: cgroup
Comment 3 Paul DeStefano 2013-11-16 23:00:18 UTC 
Created attachment 824999 [details]
File: core_backtrace
Comment 4 Paul DeStefano 2013-11-16 23:00:21 UTC 
Created attachment 825000 [details]
File: dso_list
Comment 5 Paul DeStefano 2013-11-16 23:00:25 UTC 
Created attachment 825001 [details]
File: environ
Comment 6 Paul DeStefano 2013-11-16 23:00:28 UTC 
Created attachment 825002 [details]
File: limits
Comment 7 Paul DeStefano 2013-11-16 23:00:31 UTC 
Created attachment 825003 [details]
File: maps
Comment 8 Paul DeStefano 2013-11-16 23:00:34 UTC 
Created attachment 825004 [details]
File: open_fds
Comment 9 Paul DeStefano 2013-11-16 23:00:38 UTC 
Created attachment 825005 [details]
File: proc_pid_status
Comment 10 Paul DeStefano 2013-11-16 23:00:41 UTC 
Created attachment 825006 [details]
File: var_log_messages
Comment 11 Joshua Daniel Franklin 2013-11-17 20:04:59 UTC 
Hi,

I couldn't reproduce the crash with a the following huge URL:

http://www.thelongestlistofthelongeststuffatthelongestdomainnameatlonglast.com/wearejustdoingthistobestupidnowsincethiscangoonforeverandeverandeverbutitstilllookskindaneatinthebrowsereventhoughitsabigwasteoftimeandenergyandhasnorealpointbutwehadtodoitanyways.html

Does that crash for you? Can you attach your .pinerc (feel free to remove usernames etc)
Comment 12 Paul DeStefano 2013-11-19 21:51:52 UTC 
Nope, that's way to short.  I consistently have problems from this one site that sends me appointment reminders and the URL is takes up six lines.  Unfortunately, I don't have any e-mails from them right now.  I'll probably receive a new one soon-ish, though.

Perhaps that was a red herring, anyway.  The crash didn't happen until a few commands after I tried to open that particular URL.  Was the crash not helpful?
Comment 13 Paul DeStefano 2013-11-19 21:53:03 UTC 
Created attachment 826330 [details]
.pinerc
Comment 14 Joshua Daniel Franklin 2013-12-07 14:30:08 UTC 

*** This bug has been marked as a duplicate of bug 914999 ***
Comment 15 Eduardo Chappa 2013-12-11 06:53:25 UTC 
Joshua,

  This bug is not a duplicate of bug 914999. I still do not understand it completely but Bug 914999 is a crash during string conversion between charsets. The data is not complete. Paul, when reporting bugs obfuscating information is not a good idea, because the obfuscated information might be the culprit of the crash. Please do not do so in the future.

  This bug has to do with the fact that color support is not thread safe, so there are two ways to make this problem go away that do not involve patching Alpine. One is to disable thread support at the configure stage, and the other turn colors off. The harder thing to do is to modify Alpine to make color support thread safe. This is not trivial. Color support is so pervasive in Alpine that modifying this takes several days. I made some progress on this during last summer, and might work on it again later, bit do not hold your breath for it.

  Just to be clear, this bug has nothing to do with URL length. Paul, my hunch is that if you disable color support that you will not be able to repeat this crash.
Comment 16 Paul DeStefano 2013-12-18 10:19:56 UTC 
Hi Eduadro,

Interesting.  Thanks so much for digging so deep.  I understand what you are saying.  I think I'll just live with it; it's infrequent enough that I'd rather have the colors.  I'll not report this in the future if the backtrace includes color stuff.

Josh,

Maybe you should leave this as ASSIGNED or CLOSED ERRATA, just for posterity, since it's not really a dup.
Comment 17 Joshua Daniel Franklin 2013-12-18 11:57:29 UTC 
Sure, done