March 2005 - Wine-devel - list.winehq.org

winecfg
by Chris Morgan 06 Mar '05

06 Mar '05

I noticed that winecfg is still reporting that it "does not modify your configuration", and in winecfg.h WINE_KEY_ROOT isn't set to the configuration key. Is there some reason why we haven't enabled writing to the actual configuration yet? Are there some bugs or lack of features that need to be resolved before it goes active? Without a ~/.wine/config file it seems like winecfg is the only way to let users easily adjust their configuration. Chris

2 1

Regression in GetModuleHandle16
by Vitaliy Margolen 06 Mar '05

06 Mar '05

This patch http://cvs.winehq.org/patch.py?id=16245 broke my wine (changes to "winbase.h"). It fails to start. Could someone take a look at it? I'm sorry, my brain failing me this weekend too, and I couldn't find the cause myself. I tried running without ~/.wine - didn't help. I'm running wine from the tree without installing it. Could this be a reason why no one else seeing the problem? 0009:trace:module:load_dll Found loaded module L"c:\\windows\\system\\kernel32.dll" for L"kernel32.dll" at 0x404c0000, count=13 0009:trace:heap:RtlFreeHeap (0x40380000,00000002,40380f78): returning TRUE 0009:trace:module:GetModuleHandle16 (krnl386.exe) 0009:trace:module:MODULE_LoadModule16 Trying built-in 'krnl386.exe' 0009:trace:global:GLOBAL_Alloc 56 flags=0000 0009:trace:heap:RtlAllocateHeap (0x40380000,00000002,00000040): returning 40380f78 0009:trace:module:NE_InitResourceHandler InitResourceHandler[1026] 0009:trace:module:GetModuleHandle16 (KERNEL) 0009:trace:module:GetProcAddress16 0000 'DefResourceHandler' 0009:trace:loaddll:MODULE_LoadModule16 Loaded module "krnl386.exe" : builtin 0009:trace:module:GetModuleHandle16 (krnl386.exe) 0009:err:module:MODULE_LoadModule16 Serious trouble. Just loaded module 'krnl386.exe' (hinst=0x1026), but can't get module handle. Filename too long ? -- Best regards, Vitaliy

1 0

dbghelp problem
by Robert Reif 06 Mar '05

06 Mar '05

I can't get a stack trace with symbol information from an exe file and tracked the problem down to this. dbghelp tries to load the symbols for the executable file by looking up the file name without the .exe extension and fails because the file is not found. This patch adds the .exe back on to the file name and tries to look it up again before failing. I doubt this is the right solution but it works for me. --- pe_module.c 2005-03-01 06:20:19.000000000 -0500 +++ /mnt/floppy/pe_module.c 2005-03-04 11:06:30.000000000 -0500 @@ -332,6 +332,7 @@ HANDLE hMap; void* mapping; char loaded_name[MAX_PATH]; + TRACE("(%p %s %p %08lx %08lx)\n", pcs, name, hFile, base, size); loaded_name[0] = '\0'; if (!hFile) @@ -341,13 +342,22 @@ /* FIXME SetLastError */ return NULL; } - if ((hFile = FindExecutableImage(name, NULL, loaded_name)) == NULL) - return NULL; + if ((hFile = FindExecutableImage(name, NULL, loaded_name)) == NULL) { + char full_name[MAX_PATH]; + TRACE("file: %s not found\n", name); + strcpy(full_name, name); + strcat(full_name, ".exe"); + if ((hFile = FindExecutableImage(full_name, NULL, loaded_name)) == NULL) { + TRACE("file: %s not found\n", full_name); + return NULL; + } + } opened = TRUE; } else if (name) strcpy(loaded_name, name); else if (dbghelp_options & SYMOPT_DEFERRED_LOADS) FIXME("Trouble ahead (no module name passed in deferred mode)\n"); + TRACE("loaded_name = %s\n", loaded_name); if (!(module = module_find_by_name(pcs, loaded_name, DMT_PE)) && (hMap = CreateFileMappingA(hFile, NULL, PAGE_READONLY, 0, 0, NULL)) != NULL) {

2 1

Picasa2- slideshow mode fixes [patch #1]
by Matthew Mastracci 06 Mar '05

06 Mar '05

I've been trying to get the slideshow mode in Picasa2 to work, but it seems like it's relying on IDirectDrawSurface->Lock() to return DDERR_SURFACEBUSY if the surface is already locked. If this call doesn't return the error, it seems to keep hammering the function until it does. It then proceeds to hit a stack overflow condition somehow and dies. :) I think it's a bug in Picasa, but my ddraw-surface-lock.diff patch in wine-patches fixes this behaviour (but not the feature, unfortunately). Note that now the slideshow mode now just terminates without error. I'll keep on looking to see what's happening. Matt.

1 0

Re: More destinations for wineoss's mixer
by wino＠piments.com 05 Mar '05

05 Mar '05

On 05 Mar 2005 12:34:26 -0500, Vincent Béron <vberon(a)mecano.gme.usherb.ca> wrote: > According to MSDN, DST_HEADPHONES can be the same as DST_SPEAKERS, > DST_LINE and DST_VOICEIN as DST_WAVEIN. > > Peter, does this one helps Dragon NS 7? > > Changelog: > Make wineoss's mixer know about DST_HEADPHONES, DST_LINE and > DST_VOICEIN. > Use proper constants instead of 0 and 1. > > Vincent one more thing to add: it seems your patch has done some good since I now notice that entering the audio setup now sets the mike to zero just before it gets stuck. I went back to try NS4 and I did now manage to go through the mike setup routine and sig/noise check. Although the mike setting was not waiting for me to speak before returning. Another brick is in place. 8) -- Opera 7 mail on Linux

1 0

Re: More destinations for wineoss's mixer
by peter＠piments.com 05 Mar '05

05 Mar '05

On 05 Mar 2005 12:34:26 -0500, Vincent Béron <vberon(a)mecano.gme.usherb.ca> wrote: > According to MSDN, DST_HEADPHONES can be the same as DST_SPEAKERS, > DST_LINE and DST_VOICEIN as DST_WAVEIN. > > Peter, does this one helps Dragon NS 7? > > Changelog: > Make wineoss's mixer know about DST_HEADPHONES, DST_LINE and > DST_VOICEIN. > Use proper constants instead of 0 and 1. > > Vincent Thanks , that got rid of the errors but did not make it work, sadly. I can get it to install reliably using Installshield. What I thought was blocking it was that there is a OK/Cancel dlg coming up _behind_ the splash screen. This is obviously a bug in that the dlg is modal but not visible when it comes up. The trick is to move the splash to one side before clicking Next. The prog now starts and works as much as it can without sound and a speach profile , ie not much but it can close cleanly. Attempting to set the mike volume is still the killer but there are now no messages, the set up dlg hangs and I have to close it using the X button on the frame. I can then exit NS. Progress none the less Thanks again. -- Using Opera e-mail on Gentoo Linux

1 0

Re: MultiByteToWideCharacter problem with CP 950 (Traditional Chinese)
by Andreas Mohr 05 Mar '05

05 Mar '05

Hi, On Sun, Mar 06, 2005 at 04:49:13AM +0900, Mike McCormack wrote: > 1st problem: (pre-existing) No comments. > 2nd problem: (regression) > > There's another serious problem, which is a regression. Starting Office > 2000 (any language)with the Big5 code page as below, causes serious > trouble with my X server (X.org 6.7.0). The X cpu usage goes to 95%, > and the install hangs, more or less. > > LC_ALL=zh_TW.Big5 wine m:\setup > > The problem happens whether or not the Chinese truetype font is installed. This probably is the same issue that has prevented success for any attempt to use some relatively recent Wine versions on the notebook of my girlfriend (P3/450, Debian, Chinese simplified, non-Unicode setup: GB2312, XF86 4.3.x): whenever I started Wine, everything slowed to a crawl and I had veritable trouble trying to kill it (definitely points towards XF86 overload). I could offer to do some tests if need be. Haven't tried the new "non-official" Debian Wine packages, BTW. Andreas Mohr

1 0

Re: USER32: Error checking for sysparams tests
by Paul Vriens 05 Mar '05

05 Mar '05

On Sat, 2005-03-05 at 06:42, jchevrier(a)nexicom.net wrote: > As discussed on wine-devel there are a handful of calls in the > SystemParametersInfo tests that don't check for returned errors and continue > on with testing a given parameter. This patch ensures that all initial calls > to SystemParametersInfoA in each parameter test have their returned errors > checked. To ease adding more error cases and reduce code duplication I have > consolidated error message handling into a single function. > > Changelog: > > * Consolidate error message handling > * Add error checking to all initial SystemParametersInfoA calls in each > parameter test to make sure that if a given parameter isn't supported or has > failed in some other way that we return and don't continue on with the test > * Remove a couple of TABS and align registry keys/values at the top of the > file Hi Justin, what about the attached one? Added/changed the following: - Always set LastError before calling test_error_msg - SPI_{GET,SET}ICONTITLEWRAP is supported on Win98, but was clouded by an error that was set after a previous call failed. Changelog: Justin Chevrier <jchevrier(a)nexicom.net> * Consolidate error message handling * Add error checking to all initial SystemParametersInfoA calls in each parameter test to make sure that if a given parameter isn't supported or has failed in some other way that we return and don't continue on with the test * SetLastError before calling test_error_msg * Remove a couple of TABS and align registry keys/values at the top of the file * SPI_{GET,SET}ICONTITLEWRAP is supported on Win98 Cheers, Paul.

2 1

Any resources about threading problems?
by Bill Medland 05 Mar '05

05 Mar '05

Does anyone know any site or book or something with a decent discussion about the problems of the different threading models and their implementations? In the Developer's Guide in section 8.3 there are references to "the threading chapter". Is this a section that has not yet been written? As I understand it we (in the pthreads version) "override the native pthreads library". What does this mean? I am trying to research the issues associated with threading so that I understand the complications. The fact that Linux has two models and that Wine, FreeDCE and Samba (that I know of) seem to have found it necessary to bypass/hack/modify/override them indicates to me that there is something to understand. -- Bill Medland mailto:billmedland(a)mercuryspeed.com http://webhome.idirect.com/~kbmed

2 3

installshield and Wine
by wino＠piments.com 05 Mar '05

05 Mar '05

What is the current state of play? It seems that InstallShield is pretty much and essential element in a working Wine env. But I tried to install Dragon NS 7 yesterday and it hung completely with a system wide modal in the middle of the screen. I had to kill the parent. There has been some regressions in Wine recently it seems , what would be the best wine version to install to get around this issue.? Thanks. -- Opera 7 mail on Linux

5 6