Bug 1205588 - [abrt] stockfish: Position::gives_check(): stockfish killed by SIGABRT
Summary: [abrt] stockfish: Position::gives_check(): stockfish killed by SIGABRT
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Fedora
Classification: Fedora
Component: stockfish
Version: 21
Hardware: x86_64
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Raphael Groner
QA Contact: Fedora Extras Quality Assurance
URL: https://retrace.fedoraproject.org/faf...
Whiteboard: abrt_hash:f28f36573e9f9bf590f5d940fa0...
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2015-03-25 09:19 UTC by Joachim Katzer
Modified: 2015-05-27 16:12 UTC (History)
1 user (show)

Fixed In Version: stockfish-7-0.3.20150506git2e86d1f.fc21
Clone Of:
Environment:
Last Closed: 2015-05-27 16:12:37 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
File: backtrace (17.59 KB, text/plain)
2015-03-25 09:19 UTC, Joachim Katzer
no flags Details
File: cgroup (190 bytes, text/plain)
2015-03-25 09:19 UTC, Joachim Katzer
no flags Details
File: core_backtrace (7.86 KB, text/plain)
2015-03-25 09:19 UTC, Joachim Katzer
no flags Details
File: dso_list (588 bytes, text/plain)
2015-03-25 09:19 UTC, Joachim Katzer
no flags Details
File: environ (907 bytes, text/plain)
2015-03-25 09:19 UTC, Joachim Katzer
no flags Details
File: limits (1.29 KB, text/plain)
2015-03-25 09:19 UTC, Joachim Katzer
no flags Details
File: maps (3.69 KB, text/plain)
2015-03-25 09:19 UTC, Joachim Katzer
no flags Details
File: open_fds (207 bytes, text/plain)
2015-03-25 09:19 UTC, Joachim Katzer
no flags Details
File: proc_pid_status (948 bytes, text/plain)
2015-03-25 09:19 UTC, Joachim Katzer
no flags Details

Description Joachim Katzer 2015-03-25 09:19:35 UTC
Description of problem:
Using Stockfish as analysis engine for SCID.
Happened when analyzing a variation of a game.

Not reproducible

Version-Release number of selected component:
stockfish-6-4.20150228git1e6d21d.fc21

Additional info:
reporter:       libreport-2.3.0
backtrace_rating: 4
cmdline:        /usr/bin/stockfish
crash_function: Position::gives_check
executable:     /usr/bin/stockfish
kernel:         3.19.1-201.fc21.x86_64
runlevel:       N 5
type:           CCpp
uid:            1001

Truncated backtrace:
Thread no. 1 (6 frames)
 #4 Position::gives_check at /usr/src/debug/Stockfish-1e6d21dbb6918a2d5f2f09730b0c30e3a4895d5c/src/position.cpp:631
 #5 Position::do_move at /usr/src/debug/Stockfish-1e6d21dbb6918a2d5f2f09730b0c30e3a4895d5c/src/position.cpp:693
 #6 Search::RootMove::extract_ponder_from_tt at /usr/src/debug/Stockfish-1e6d21dbb6918a2d5f2f09730b0c30e3a4895d5c/src/search.cpp:1509
 #7 Search::think at /usr/src/debug/Stockfish-1e6d21dbb6918a2d5f2f09730b0c30e3a4895d5c/src/search.cpp:279
 #8 MainThread::idle_loop at /usr/src/debug/Stockfish-1e6d21dbb6918a2d5f2f09730b0c30e3a4895d5c/src/thread.cpp:272
 #9 (anonymous namespace)::start_routine at /usr/src/debug/Stockfish-1e6d21dbb6918a2d5f2f09730b0c30e3a4895d5c/src/thread.cpp:39

Comment 1 Joachim Katzer 2015-03-25 09:19:39 UTC
Created attachment 1006215 [details]
File: backtrace

Comment 2 Joachim Katzer 2015-03-25 09:19:40 UTC
Created attachment 1006216 [details]
File: cgroup

Comment 3 Joachim Katzer 2015-03-25 09:19:42 UTC
Created attachment 1006217 [details]
File: core_backtrace

Comment 4 Joachim Katzer 2015-03-25 09:19:44 UTC
Created attachment 1006218 [details]
File: dso_list

Comment 5 Joachim Katzer 2015-03-25 09:19:46 UTC
Created attachment 1006219 [details]
File: environ

Comment 6 Joachim Katzer 2015-03-25 09:19:48 UTC
Created attachment 1006220 [details]
File: limits

Comment 7 Joachim Katzer 2015-03-25 09:19:50 UTC
Created attachment 1006221 [details]
File: maps

Comment 8 Joachim Katzer 2015-03-25 09:19:51 UTC
Created attachment 1006222 [details]
File: open_fds

Comment 9 Joachim Katzer 2015-03-25 09:19:53 UTC
Created attachment 1006223 [details]
File: proc_pid_status

Comment 10 Raphael Groner 2015-03-25 13:07:11 UTC
Hi Joachim,

thanks for your report. Is it possible that stockfish tried to execute an illegal move? The call of gives_check in the backtrace indicates that a possible check move was done but not allowed and fails therefore in the assert condition. So stockfish seems not to be able to detect a deadlock situation in a smoother way than crashing. Could you try to provide the original game situation?

Comment 11 Joachim Katzer 2015-03-28 00:09:51 UTC
I am using stockfish as chess analysis engine for scid-4.5.2-6 and for chessx-1.2.2. I have reported this bug because of the report/abrt interface.

The crash may occur when I change the position by clicking on a different move in the SCID game notation window. From your explanation I suspect that SCID may send sporadically a wrong FEN position to the engine and then an illegal move. 
I'll enable debug logging of the stockfish engine next time.

In combination with chessx 1.2.2 (svn commit 1940) I have not yet encountered this problem.

Comment 12 Fedora Update System 2015-05-08 22:06:58 UTC
stockfish-7-0.3.20150506git2e86d1f.fc21 has been submitted as an update for Fedora 21.
https://admin.fedoraproject.org/updates/stockfish-7-0.3.20150506git2e86d1f.fc21

Comment 13 Raphael Groner 2015-05-08 22:09:18 UTC
Please try again with the new build. Can you reproduce any regression?

Comment 14 Fedora Update System 2015-05-10 23:37:51 UTC
Package stockfish-7-0.3.20150506git2e86d1f.fc21:
* should fix your issue,
* was pushed to the Fedora 21 testing repository,
* should be available at your local mirror within two days.
Update it with:
# su -c 'yum update --enablerepo=updates-testing stockfish-7-0.3.20150506git2e86d1f.fc21'
as soon as you are able to.
Please go to the following url:
https://admin.fedoraproject.org/updates/FEDORA-2015-7695/stockfish-7-0.3.20150506git2e86d1f.fc21
then log in and leave karma (feedback).

Comment 15 Joachim Katzer 2015-05-15 20:19:55 UTC
Thanks for providing the update. The new version from git2e86d1f apparently has resolved my issue. I didn't encounter any crashes after installation of stockfish-7-0.3.20150506git2e86d1f.fc21.

But why gets the new package the version number 7?
Upstream has not yet published a version 7 and I couldn't found any announcement in upstream's blog, github or Google discussion group https://groups.google.com/forum/?fromgroups=#!forum/fishcooking

Comment 16 Raphael Groner 2015-05-25 07:31:21 UTC
Thanks for testing.

(In reply to Joachim Katzer from comment #15)
…
> But why gets the new package the version number 7?
> Upstream has not yet published a version 7 and I couldn't found any
> announcement in upstream's blog, github or Google discussion group
> https://groups.google.com/forum/?fromgroups=#!forum/fishcooking

This is a prerelease of latest development snapshot from GitHub (as of 2015-05-06, commit 2e86d1febc163e076320a6a21060fc5a1c7c0dff). We assume that the futuristic major release should be with version 7, this build logic is already in Fedora22.

Comment 17 Fedora Update System 2015-05-27 16:12:37 UTC
stockfish-7-0.3.20150506git2e86d1f.fc21 has been pushed to the Fedora 21 stable repository.  If problems still persist, please make note of it in this bug report.


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