June 2005 - Wine-Devel - list.winehq.org

Jedi Academy regression?
by Fernando Vilas June 26, 2005

June 26, 2005

Hi, In wine-20050524, Jedi Academy works well. After an upgrade to wine-cvs, now it doesn't go fullscreen and doesn't get keyboard focus. wine-20050524: (works) $ wine jasp fixme:opengl:query_function_pbuffer gl_version is: "1.5.3 NVIDIA 71.67" fixme:opengl:query_function_pbuffer glx_exts is: "GLX_EXT_visual_info GLX_EXT_visual_rating GLX_SGIX_fbconfig GLX_SGIX_pbuffer GLX_SGI_video_sync GLX_SGI_swap_control GLX_ARB_multisample GLX_NV_float_buffer GLX_ARB_get_proc_address " fixme:opengl:query_function_pbuffer gl_version is: "1.5.3 NVIDIA 71.67" fixme:opengl:query_function_pbuffer glx_exts is: "GLX_EXT_visual_info GLX_EXT_visual_rating GLX_SGIX_fbconfig GLX_SGIX_pbuffer GLX_SGI_video_sync GLX_SGI_swap_control GLX_ARB_multisample GLX_NV_float_buffer GLX_ARB_get_proc_address " fixme:wave:DSD_CreateSecondaryBuffer (0x7fdc03b8,0x7fc5eb10,10024,0,0x7fdc331c,0x7fdc342c,0x7fdc32f8): stub wine-cvs: (doesn't) $ wine jasp err:x11settings:X11DRV_ChangeDisplaySettingsExW No matching mode found! (NoRes) fixme:wave:DSD_CreateSecondaryBuffer (0x7fdc0fc0,0x7fc5eb10,10024,0,0x7fdc19fc,0x7fdc1b0c,0x7fdc19d8): stub This started about June 13th. (give or take) after spending some time to see if there was some configuration variable that changed, I installed the older version of wine for comparison. I believe this is a regression, but if there's some thing I should have done on my box to fix it, please let me know. Relevant info: Slackware-current Linux 2.6.12 nVidia GeForce 5900 (7664 drivers) using xinerama, not twinview AMD TBird 1.4GHz I can get you config files if those would help. -- Thanks, Fernando Vilas fvilasatinamedotcom

2 1

Wine tests
by Glen Kaukola June 26, 2005

June 26, 2005

Hi there, So I'm not seeing a test for MoveFileEx at all. I wouldn't mind taking a stab at creating one, and for that matter for anything else I happen to find not being tested. Sound like a plan? Glen

2 2

lotus notes regression
by Paul R Streitman June 25, 2005

June 25, 2005

Patch 14725, applied 2004/12/07 11:31:53, breaks lotus notes. The symptom is a hang with the CPU at 100%, and it happens immediately if you try to click on a notes 'bookmark', or shortly after you try to do much of anything else if you try to avoid the 'bookmarks'. Sorry that it took me so long to report this, when I decided that it wasn't just a temporary instability I eventually had to iteratively compile (many) CVS snapshots to narrow down when the problem was introduced because I found it impossible to untangle the various patches that went in around that time frame. I also ran afoul of the jack problems near the end. I do not know how to narrow the problem down further. I tried running winedbg (at CVS level a couple of days ago), and that seems to be broken too -- it wanted to run things from the wine build tree? There are no unusual console messages or anything when the problem happens. Thanks, Paul z/OS core components development Internet: prs(a)us.ibm.com

7 8

Font matching headaches
by Andreas Mohr June 25, 2005

June 25, 2005

Hi, another visitor on LinuxTag was having huge issues with a heavy-duty app doing batches of calculations, and every time after the end of those calculations there was an X server spike going up to 80% CPU. Turns out that spike was caused by the WARN("Huge font size %d pixels has been reduced to %d\n", h, MAX_FONT_SIZ E); message in dlls/x11drv/xfont.c/LFD_ComposeLFD() By binary-hacking x11drv.so to reduce it to 20 instead, I was able to totally eliminate the CPU wait period, since it then didn't render 1000 pixel fonts any more. Needless to say this is not exactly satisfying, since it's only yet another bandaid for this problem (while it helps in this situation, it would be useful to get this issues fully fixed in Wine). X11 version is XFree86 4.3.0-dfsg-1 (i.e., Debian). Is there anything that can be done to prevent this situation? (I know that it's an inconvenient situation in a workaround handler for even more inconvenient behaviour, but still...). I can ask for further machine specific information if it turns out that it would be useful... Another thing that I found there is the following loop: do { LFD_ComposeLFD( pfo, fm.height, lpLFD, uRelaxLevel++ ); if( (pfo->fs = safe_XLoadQueryFont( gdi_display, lpLFD )) ) brea k; } while( uRelaxLevel ); which *might* perhaps run amok since it continually grows without any effective check (but the break; should hopefully be sufficient to stop this reliably). Thanks, Andreas Mohr

1 0

Re: shell32: fix infinite loop in SHGetPathFromIDListW
by Vitaliy Margolen June 25, 2005

June 25, 2005

Thank you. This patch does fix the infinite loop bug. Vitaliy Margolen Saturday, June 25, 2005, 5:14:27, Michael Jung wrote: > Hi, > With a seemingly unrelated change in his local tree Vigaliy Margolen triggered > an infinite loop in SHGetPathFromIDList (which was introduced by one of my > patches): > SHGetPathFromIDList (pidl.c:1256) calls SHBindToParent, which calls Desktop->>BindToObject (pidl.c:1315), which calls SHELL32_BindToChild > (shfldr_desktop.c:348), which binds to MyComputer (shlfolder.c:289) and then calls MyComputer->>BindToObject (shlfolder.c:314), which calls > SHELL32_BindToChild (shfldr_mycomp.c:346), which calls > SHELL32_GetCustomFolderAttribute (shlfolder.c:299), which calls > SHGetPathFromIDList (shlfolder.c:91). Phew. > Since the path is already known in SHELL32_BindToChild, we can break the loop > easily. > I don't understand, why we don't see a stack overflow with wine cvs, but the > size of a +shell log shrinks 90% with this patch applied compared to cvs. So > we do have the loop, but it is terminated somehow before the stack overflows. > Changelog: > Fix a lurking infinite loop in SHGetPathFromIDList. > Bye,

1 0

loop in SHGetPathFromIDList
by Michael Jung June 24, 2005

June 24, 2005

Hi Vitaliy, your patch to shlfolder.c is what produces the loop. It is the line, which I commented out below: if (dwAttributes & FILE_ATTRIBUTE_DIRECTORY) ;//*pdwAttributes |= (SFGAO_FOLDER | SFGAO_HASSUBFOLDER | SFGAO_FILESYSANCESTOR); else *pdwAttributes &= ~(SFGAO_FOLDER | SFGAO_HASSUBFOLDER | SFGAO_FILESYSANCESTOR); It works fine if the line is commented out. I don't understand at the moment why this is the fact, but the loop seems in fact to be a problem with SHGetPathFromIDList. I didn't figure it out tonight, but I will investigate this further over the weekend. Bye, -- Michael Jung mjung(a)iss.tu-darmstadt.de

1 0

X11 -- how do you prevent window from closing if user closes it
by Peter Quiring June 24, 2005

June 24, 2005

I trying to use X11 and let's say I create a text editor. Let's say the user is using the text editor, makes some changes and then clicks on the [X] to close the window before saving changes. I want to popup a "are you sure" before the window is closed but X11 doesn't allow this. How does Wine go about this problem? In Windows this is easy, in the WindowProc() just ignore WM_CLOSE. FYI :Porting some code over to Linux and need a little help. Thanks.

3 2

Re: cabinet.dll status ?
by Gerold J. Wucherpfennig June 24, 2005

June 24, 2005

On Friday 24 June 2005 14:32, Tom Wickline wrote: > Hello Gerold, > > I currently list cabinet.dll at 80% completion can you let me know > what you feel > the current status should be? > > Would you like for me to add you as the current worker? I just looked > through CVS and seem to be the main developer now for cabinet.dll > > http://www.winehq.org/site/status_dlls > > Thanks, > Tom Hi Tom, I thank you very much for asking me about cabinet.dll. :) I would really like to be added as the current worker. :) My personal opinion about the completition status, not specifically about cabinet.dll, is that the status should be roughly equal to the amount of code that is implemented compared to the 100% of code . Right now the status based on the percieved amount of usage of the respective API functions. That of course is a bit optimistic, because API functions that are used most will be implemented first and the amount of work that is required to complete the remaining API functions may be underestimated. Regarding cabinet.dll: circa half of the required code and half of the required functions is in CVS right now. Nevertheless I think that you should keep the current completition status as it is right now, because first it would look strange when the status would decrease and second I hope that my current patch will be committed soon which will put the completition status to what it is listed right now. Regards, Gerold

1 0

very old com tech spec doc
by Vijay Kiran Kamuju June 24, 2005

June 24, 2005

very old com tech spec doc. have a look http://www.geocities.com/infyquest/comspec10.zip

1 0

Re: shell32: Implement SHGetPathFromIDList based on GetDisplayNameOf
by Vitaliy Margolen June 24, 2005

June 24, 2005

Alexandre, could you back this patch out? This patch broke shell ParseDisplayName (created call loop) from IShellFolder_ParseDisplayName to SHELL32_ParseNextElement back to IShellFolder_ParseDisplayName Vitaliy Margolen Tuesday, June 14, 2005, 14:08:31, Vitaliy Margolen wrote: > Changelog: > Implement SHGetPathFromIDList based on GetDisplayNameOf > Bye,

2 3