July 2011 - Wine-devel - list.winehq.org

Winhelp crash in assertion(0) if .hlp file is not found
by Bruno Jesus 31 Jul '11

31 Jul '11

Hello, I've been in the list for a long time as a reader and after years I finally managed to get a development computer and restarted playing with current wine git. There are about 15 bugs on bugzilla but none of them are related to this. I'm trying the list first before opening a bug. It's quite simple to reproduce, although I'm not sure if I can attach the hlp file here since it's from windows. The crash happens because the help file needs another help file. There is a button on the screen called Glossary that needs a file called glossary.hlp. If this file does not exist wine asks me to point it, if I answer No or yes and then cancel file selection wine crashes on an assert. The following lines are displayed on the console: fixme:winhelp:WINHELP_GetWindowInfo Couldn't find window info for glossary winhelp.c:270: WINHELP_GetWindowInfo: Assertion `0' failed. wine: Assertion failed at address 0xb78f0424 (thread 003b), starting debugger... The following lines are from file winhelp.c, around line 270. ... if (strcmp(name, "main") != 0) { WINE_FIXME("Couldn't find window info for %s\n", name); assert(0); return NULL; } ... The function WINHELP_OpenHelpWindow does not resist WINHELP_GetWindowInfo returning NULL. That was why the assert(0) was added inside it to abort the program execution (as far as I see). There are 3 calls like this (one on winhelp.c and the others on macro.c): ... WINHELP_OpenHelpWindow(HLPFILE_PageByHash, hlpfile, lHash, WINHELP_GetWindowInfo(hlpfile, wndname), show); ... My approach to fix this is quite simple: instead of splitting the above lines on different places and add a check for each of them the easier way is to simply make OpenHelpWindow resist null pointers and do not open anything because the user was already warned and asked for the nonexistent file. The proposed solution (if I may call adding one "if" a solution) is attached. If anyone has the problematic hlp file that was the cause to add the assert(0) please send me. Looks like the assert is there since wine 1.0 according to http://source.winehq.org/source/programs/winhlp32/winhelp.c?v=wine-1.0 Best wishes, Bruno -- universe* god::bigbang (void); //and then it all began...

2 4

Re: [PATCH 1/5] mshtml: Added semi-stub implementation of IHTMLWindow2::status property
by Marvin 31 Jul '11

31 Jul '11

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=13084 Your paranoid android. === WVISTAADM (32 bit dom) === Failure running script in VM: The specified guest user must be logged in interactively to perform this operation

1 0

Re: mshtml: Added semi-stub implementation of IHTMLWindow2::status property
by Jacek Caban 31 Jul '11

31 Jul '11

Please ignore this one. I will send proper version with other mshtml patches. Thanks, Jacek

1 0

Re: [1/2] cmd/tests: Add FOR /D tests (try 2)
by Octavian Voicu 30 Jul '11

30 Jul '11

2011/7/29 Frédéric Delanoy <frederic.delanoy(a)gmail.com>: > +rem for /d %%i in (ba*) do echo %%i>> tmp > +rem sort < tmp > +rem del tmp I believe you can also use this syntax [put for instruction in parentheses, add | sort at the end]: (for ... ...) | sort Btw, it's a horrible hack, but until you'll have a wine sort you can test with (for ... ...) | Z:\usr\bin\sort :) Octavian

2 2

Re: [PATCH] shell32: Disable WOW64 redirection when converting pidls to UNIX paths. (try 2, resend)
by Paul Vriens 29 Jul '11

29 Jul '11

On 07/29/2011 01:29 AM, Octavian Voicu wrote: > + else > + win_skip("Not on WOW64, skipping test.\n"); Why a win_skip()? Isn't a normal skip() more appropriate? -- Cheers, Paul.

2 1

Re: [PATCH 8/9] wscript: implemented Host_get_Application (resend)
by Marvin 29 Jul '11

29 Jul '11

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=13007 Your paranoid android. === W7PROX64 (32 bit) === run.c:180: Test failed: test.JS: L"WScript.FullName = " run.c:180: Test failed: test.JS: L"WScript.Path = "

1 0

Re: [PATCH 7/9] wscript: implemented Host_put_Interactive (resend)
by Marvin 29 Jul '11

29 Jul '11

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=13006 Your paranoid android. === W7PROX64 (32 bit) === run.c:180: Test failed: test.JS: L"WScript.FullName = " run.c:180: Test failed: test.JS: L"WScript.Path = "

1 0

Re: [PATCH 6/9] wscript: implemented Host_get_Interactive (resend)
by Marvin 29 Jul '11

29 Jul '11

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=13005 Your paranoid android. === W7PROX64 (32 bit) === run.c:180: Test failed: test.JS: L"WScript.FullName = " run.c:180: Test failed: test.JS: L"WScript.Path = "

1 0

Re: [PATCH 5/9] wscript: implemented Arguments2_get_length (resend)
by Marvin 29 Jul '11

29 Jul '11

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=13004 Your paranoid android. === W7PROX64 (32 bit) === run.c:180: Test failed: test.JS: L"WScript.FullName = " run.c:180: Test failed: test.JS: L"WScript.Path = "

1 0

Re: [PATCH 4/9] wscript: implemented Arguments2_Count (resend)
by Marvin 29 Jul '11

29 Jul '11

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=13003 Your paranoid android. === W7PROX64 (32 bit) === run.c:180: Test failed: test.JS: L"WScript.FullName = " run.c:180: Test failed: test.JS: L"WScript.Path = "

1 0