0000920: Crash when catapult attack: Assertion `attackedStack || isCatapultAttack' failed.

View Issue Details [ Jump to Notes ]

[ Issue History ] [ Print ]

Project

Category

View Status

Date Submitted

Last Update

0000920

VCMI

Mechanics - Battles

public

2012-04-08 19:42

2014-05-30 17:41

Reporter

q4a

Assigned To

Tow

Priority

normal

Severity

crash

Reproducibility

sometimes

Status

closed

Resolution

fixed

Platform

amd64

Ubuntu

OS Version

12.04 beta2

Product Version

0.88

Target Version

Fixed in Version

0.89

Summary

0000920: Crash when catapult attack: Assertion `attackedStack || isCatapultAttack' failed.

Description

Tested on r2619 and some previous, with different versions of boost: 1.46.1 and 1.49. Reproducible sometimes/random. Cannot reproduce ob Windows.
I tested r2619: crash has happened booth with boost 1.46.1 and with boost 1.49.
Then i tested r2550 and 2570 with boost 1.46.1 to be sure, that it's bug of Ubuntu/gcc/boost or etc, but it works fine. I tested r2589-2607 (r2589,2595,2598,2599,2603,2607 - every time with "make clean"+"make distclean"), but I got build errors. On r2608,2611,2619 bug happened.

Client and server log attached. Console output:
CStupidAI [0x7fe7ec8fc8b0]: actionFinished called
CStupidAI [0x7fe7ec8fc8b0]: actionStarted called
lt-vcmiclient: ./BattleInterface/CBattleAnimations.cpp:146: CAttackAnimation::CAttackAnimation(CBattleInterface*, const CStack*, BattleHex, const CStack*): Assertion `attackedStack || isCatapultAttack' failed.
Emergency stop (core dumped)
1@qamd:~/game/vcmi$ lt-vcmiserver: /usr/include/boost/thread/pthread/condition_variable_fwd.hpp:46: boost::condition_variable::~condition_variable(): Assertion `!pthread_mutex_destroy(&internal_mutex)' failed.
Aborted (core dumped)

Steps To Reproduce

Start VCMI_Tests.h3m map from 0000003 (but bug happens on any map).
Attack Inferno town (Styx).
Sometimes(r2608,2611) I got crash at 1st catapult's turn.
Sometimes(r2619) I got crash at 2nd-3rd catapult's turn.
And rarely I did not get any crash.

Additional Information

Testing r2571-2588 useless, because there is no makefiles for Fuzzylite.
I'm going to test r2589 with patches from r2608, and if it will works fine, then this is really vcmi's bug.

Tags

No tags attached.

Attached Files

log.7z [^] (19,301 bytes) 2012-04-08 19:42

Relationships

Notes
(0002345) Tow (developer) 2012-04-08 20:51	I hate this assertion. Every time I fix it, I break sth else. :( r2620 should fix it, 0 is actually a valid wall part.

(0003376) joebraker (reporter) 2013-02-13 05:54	Probably, I have had the same case. Crashes, when i try to fire wall or towers by catapult. I use linux 64 deb package. If i miss catapult's turn - everything is ok.

Issue History
Date Modified	Username	Field	Change
2012-04-08 19:42	q4a	New Issue
2012-04-08 19:42	q4a	File Added: log.7z
2012-04-08 20:51	Tow	Note Added: 0002345
2012-04-08 20:51	Tow	Status	new => resolved
2012-04-08 20:51	Tow	Fixed in Version	=> 0.89
2012-04-08 20:51	Tow	Resolution	open => fixed
2012-04-08 20:51	Tow	Assigned To	=> Tow
2013-02-13 05:54	joebraker	Note Added: 0003376
2014-05-30 17:41	beegee	Status	resolved => closed

Site | Forums | Wiki | Slack | GitHub