November 2004 - Wine-Devel - list.winehq.org

Re: Unicodified fontdlg (next try)
by Robert Shearman Nov. 26, 2004

Nov. 26, 2004

Jacek Caban wrote: > Changelog: > Unicodified fontdlg > >------------------------------------------------------------------------ > >@@ -216,52 +213,60 @@ > */ > BOOL WINAPI ChooseFontA(LPCHOOSEFONTA lpChFont) > { >- LPCVOID template; >- HRSRC hResInfo; >- HINSTANCE hDlginst; >- HGLOBAL hDlgTmpl; >- >- if ( (lpChFont->Flags&CF_ENABLETEMPLATEHANDLE)!=0 ) >- { >- template=(LPCVOID)lpChFont->hInstance; >- } else >- { >- if ( (lpChFont->Flags&CF_ENABLETEMPLATE)!=0 ) >- { >- hDlginst=lpChFont->hInstance; >- if( !(hResInfo = FindResourceA(hDlginst, lpChFont->lpTemplateName, >- (LPSTR)RT_DIALOG))) >- { >- COMDLG32_SetCommDlgExtendedError(CDERR_FINDRESFAILURE); >- return FALSE; >- } >- } else >- { >- hDlginst=COMDLG32_hInstance; >- if (!(hResInfo = FindResourceA(hDlginst, "CHOOSE_FONT", (LPSTR)RT_DIALOG))) >- { >- COMDLG32_SetCommDlgExtendedError(CDERR_FINDRESFAILURE); >- return FALSE; >- } >- } >- if (!(hDlgTmpl = LoadResource(hDlginst, hResInfo )) || >- !(template = LockResource( hDlgTmpl ))) >- { >- COMDLG32_SetCommDlgExtendedError(CDERR_LOADRESFAILURE); >- return FALSE; >- } >+ CHOOSEFONTW chFontw; >+ LOGFONTW logFontw; >+ LPLOGFONTA lpLogFonta; >+ int len; >+ BOOL ret; >+ LPSTR lpszStyle; >+ LPCSTR lpTemplateName; >+ >+ memcpy(&chFontw, lpChFont, sizeof(CHOOSEFONTW)); >+ memcpy(&logFontw, lpChFont->lpLogFont, sizeof(LOGFONTA)); >+ chFontw.lpLogFont = &logFontw; >+ >+ MultiByteToWideChar(CP_ACP, 0, lpChFont->lpLogFont->lfFaceName, >+ LF_FACESIZE, chFontw.lpLogFont->lfFaceName, LF_FACESIZE); >+ >+ if(lpChFont->lpszStyle) { >+ len = MultiByteToWideChar(CP_ACP, 0, lpChFont->lpszStyle, -1, NULL, 0); >+ chFontw.lpszStyle = HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY, len*sizeof(WCHAR)); > > Initialising the whole string to zero just before it is about to be overwritten seems needlessly wasteful on CPU cycles and memory bandwidth. You seem to do this in a few places. Yes, MultiByteToWideChar can, in some cases, leave the string without a null terminator, but that only happens when there is insufficient buffer. In this case, the only way that can happen is if another thread modifies the buffer between the previous MultiByteToWideChar and the next one here. If you wanted to be really careful, you could always add "chFontw.lpszStyle[len-1] = '\0';" after to call to MultiByteToWideChar. >+ MultiByteToWideChar(CP_ACP, 0, lpChFont->lpszStyle, -1, chFontw.lpszStyle, len); >+ HeapFree(GetProcessHeap(), 0, lpChFont->lpszStyle); > > You seem to be freeing a buffer that the user passed in. How do you know it was allocated with HeapAlloc? >+ } >+ >+ if(lpChFont->lpTemplateName) { >+ len = MultiByteToWideChar(CP_ACP, 0, lpChFont->lpTemplateName, -1, NULL, 0); >+ chFontw.lpTemplateName = HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY, len*sizeof(WCHAR)); >+ MultiByteToWideChar(CP_ACP, 0, lpChFont->lpTemplateName, >+ -1, (LPWSTR)chFontw.lpTemplateName, len); >+ } >+ >+ ret = ChooseFontW(&chFontw); >+ >+ lpLogFonta = lpChFont->lpLogFont; >+ lpszStyle = lpChFont->lpszStyle; >+ lpTemplateName = lpChFont->lpTemplateName; >+ memcpy(lpChFont, &chFontw, sizeof(CHOOSEFONTA)); >+ lpChFont->lpLogFont = lpLogFonta; >+ lpChFont->lpszStyle = lpszStyle; >+ lpChFont->lpTemplateName = lpTemplateName; >+ memcpy(lpChFont->lpLogFont, &logFontw, sizeof(LOGFONTA)); >+ WideCharToMultiByte(CP_ACP, 0, logFontw.lfFaceName, >+ LF_FACESIZE, lpChFont->lpLogFont->lfFaceName, LF_FACESIZE, 0, 0); >+ >+ if(chFontw.lpszStyle) { >+ len = WideCharToMultiByte(CP_ACP, 0, chFontw.lpszStyle, -1, NULL, -1, 0, 0);; >+ WideCharToMultiByte(CP_ACP, 0, chFontw.lpszStyle, -1, lpChFont->lpszStyle, len, 0, 0); >+ HeapFree(GetProcessHeap(), 0, chFontw.lpszStyle); > } >- if (TRACE_ON(commdlg)) >- _dump_cf_flags(lpChFont->Flags); > >- if (lpChFont->Flags & (CF_SELECTSCRIPT | CF_NOVERTFONTS )) >- FIXME(": unimplemented flag (ignored)\n"); >+ if(chFontw.lpTemplateName) >+ HeapFree(GetProcessHeap(), 0, (LPBYTE)chFontw.lpTemplateName); > > You shouldn't need a cast here, unless lpTemplateName is a const. > >- return DialogBoxIndirectParamA(COMDLG32_hInstance, template, >- lpChFont->hwndOwner, FormatCharDlgProcA, (LPARAM)lpChFont ); >+ return ret; > } > >- > #define TEXT_EXTRAS 4 > #define TEXT_COLORS 16 > > > ... >@@ -1072,34 +1083,34 @@ > } > > /*********************************************************************** >- * FormatCharDlgProcA [internal] >+ * FormatCharDlgProc [internal] > */ >-INT_PTR CALLBACK FormatCharDlgProcA(HWND hDlg, UINT uMsg, WPARAM wParam, >+INT_PTR CALLBACK FormatCharDlgProc(HWND hDlg, UINT uMsg, WPARAM wParam, > LPARAM lParam) > { >- LPCHOOSEFONTA lpcf; >+ LPCHOOSEFONTW lpcf; > INT_PTR res = FALSE; > > if (uMsg!=WM_INITDIALOG) > { >- lpcf=(LPCHOOSEFONTA)GetPropA(hDlg, WINE_FONTDATA); >- if (!lpcf && uMsg != WM_MEASUREITEM) >+ lpcf=(LPCHOOSEFONTW)GetPropW(hDlg, strWineFontData); >+ if (!lpcf) > return FALSE; > if (CFn_HookCallChk32(lpcf)) >- res=CallWindowProcA((WNDPROC)lpcf->lpfnHook, hDlg, uMsg, wParam, lParam); >+ res=CallWindowProcW((WNDPROC)lpcf->lpfnHook, hDlg, uMsg, wParam, lParam); > > This isn't something wrong with your patch, but is in the original code too. We shouldn't need to cast here, because if lpfnHook isn't using the right calling conventions or doesn't have the same number of parameters then we shouldn't be using CallWindowProc to start with. > if (res) > return res; > } > else > { >- lpcf=(LPCHOOSEFONTA)lParam; >+ lpcf=(LPCHOOSEFONTW)lParam; > if (!CFn_WMInitDialog(hDlg, wParam, lParam, lpcf)) > { > TRACE("CFn_WMInitDialog returned FALSE\n"); > return FALSE; > } > if (CFn_HookCallChk32(lpcf)) >- return CallWindowProcA((WNDPROC)lpcf->lpfnHook,hDlg,WM_INITDIALOG,wParam,lParam); >+ return CallWindowProcW((WNDPROC)lpcf->lpfnHook,hDlg,WM_INITDIALOG,wParam,lParam); > } > switch (uMsg) > { > > Rob

1 0

Re: profiling GTA4. question about polling
by Brian Gunlogson Nov. 26, 2004

Nov. 26, 2004

The patch you gave me is too incomplete to run GTA :( I can start a game, but once the loading screen says it is done, the game crashes. Looked at the console, it says, "fixme:drawStridedFast TODO". The crash happens at the end of drawStridedFast(), inside glDrawElements(). Your patch will really speed things up when it is finished. In a fresh CVS tree, I disabled drawStridedSlow and applied 2 polling patches submitted to wine-patches. The game works almost as fast as in Windows!! Brian G. --- Raphael <fenix(a)club-internet.fr> wrote: > Here is it. > > But i have some problems with hardware vertex buffers (using vbo extension) > i have some 'transparent" objects (ie graphics bugs) > Anyway, If you encounter same problems you can uncomment this line: > //useSoft |= (0 != (Usage & D3DUSAGE_WRITEONLY)); > (dls/d3d8/vertexbuffer.c at line 71) __________________________________ Do you Yahoo!? Yahoo! Mail - You care about security. So do we. http://promotions.yahoo.com/new_mail

2 2

Internet Explorer 6 Installer script
by Hajime Segawa Nov. 26, 2004

Nov. 26, 2004

I wrote a tiny shell script that can install IE6sp1 and WMP7. I hope this script can help you. RealPlayer 10, MSN Messenger 6.4, Yahoo!Japan Messenger seem to work with this configuration. http://sidenet.ddo.jp/winetips/config.html I took some screenshots. http://sidenet.ddo.jp/winetips/files/

2 2

dlls/mscms/profile.c breakage
by Gerald Pfeifer Nov. 25, 2004

Nov. 25, 2004

The following change to dlls/mscms/profile.c revision 1.6 date: 2004/11/21 15:48:18; author: julliard; state: Exp; lines: +135 -3 Implement and test GetColorProfileElementTag, GetCountColorProfileElements and IsColorProfileTagPresent. Stub GetStandardColorSpaceProfile{A,W}. breaks FreeBSD 4.10 (with XFree86-4.3.0) as follows: gmake[2]: Entering directory `/.amd_mnt/nashira/files5/test/wine/dlls/mscms' /sw/gcc-3.3.4/bin/gcc -c -I. -I. -I../../include -I../../include -D__WINESRC__ -D_REENTRANT -fPIC -Wall -pipe -mpreferred-stack-boundary=2 -fno-strict-aliasing -gstabs+ -Wpointer-arith -g -O2 -o profile.o profile.c profile.c: In function `GetColorProfileElementTag': profile.c:142: error: `LCMSICCPROFILE' undeclared (first use in this function) profile.c:142: error: (Each undeclared identifier is reported only once profile.c:142: error: for each function it appears in.) profile.c:142: error: `cmsprofile' undeclared (first use in this function) profile.c:142: error: parse error before ')' token profile.c: In function `GetCountColorProfileElements': profile.c:174: error: `LCMSICCPROFILE' undeclared (first use in this function) profile.c:174: error: `cmsprofile' undeclared (first use in this function) profile.c:174: error: parse error before ')' token gmake[2]: *** [profile.o] Error 1 gmake[2]: Leaving directory `/.amd_mnt/nashira/files5/test/wine/dlls/mscms' Gerald -- Gerald Pfeifer (Jerry) gerald(a)pfeifer.com http://www.pfeifer.com/gerald/

4 13

Major Winelib Problems
by Sam Lauber Nov. 25, 2004

Nov. 25, 2004

I followed the tutorial for Winelib on my Linux system, building notepad. I examined the file "./notepad2" with readelf, and it said "Not an ELF file", and I do not have an a.out system. I opened it with my text editor (attached is the text I saw) and I was disappointed. Instead of a real "port", it was just a shell script that launches wine on a Windows program encapsulated in a ELF shared object file. I'm not happy. Samuel Lauber -- _____________________________________________________________ Web-based SMS services available at http://www.operamail.com. >From your mailbox to local or overseas cell phones. Powered by Outblaze

5 5

Wineconf committee
by Jeremy White Nov. 25, 2004

Nov. 25, 2004

Hey folks, Brian, Alexandre, and I have been talking with some kind folks that are volunteering to help host Wineconf in Stuttgart; we're thinking late spring for Stuttgart. We've also had a generous offer to help set up a meeting in conjunction with LinuxTag in Karlruhe in mid to late June. This is all great, but I realized suddenly that the three of us were starting to consider things and make decisions on behalf of the broader Wine community. While I enjoy exercising dictatorial impunity, I prefer to have a cover group to make my imperiousness less obvious <grin>. Seriously, I'd like to ask for volunteers to serve on a Wineconf planning committee. I think the requirements for the position are as follows: 1. Able and willing to attend the next Wineconf in Germany 2. Able and willing to engage in and follow the planning and decision making process 3. Have a credible background in the Wine community (or live in the right part of Germany <grin>). 4. Be willing to help in ways that may be required, such as showing up a day early to help prepare, or bringing A/V equipment, or whatever way makes sense. (This isn't a requirement, but I really would like folks that are at least willing to contribute a little extra, if they are able.) To volunteer, please email me privately, and I'll build a chain of emails. If we don't get any volunteers, then Brian, Alexandre, and I will consider ourselves empowered, and will dictate that Marcus does all the work <grin>. Cheers, Jeremy

2 2

Re: SetForegroundWindow() from WM_ACTIVATE broken
by Jon Griffiths Nov. 25, 2004

Nov. 25, 2004

Hi, More on this problem ... here what happens under native when the back window is clicked: Order of trace below is (msg, hwnd, wparam,lparam) The app is started, click on the client area of the background window WM_MOUSEACTIVATE back 001203aa 02010001 (MA_ACTIVATE) Front window is told its going inactive WM_NCACTIVATE front 00000000 back WM_ACTIVATE front WA_INACTIVE back SetWindowPos() is called to raise back window to the front of the zorder, nc client and background told to paint WM_WINDOWPOSCHANGING back 00000000 0012fea8 WM_NCPAINT back 00000001 00000000 WM_ERASEBKGND back 0101005a 00000000 WM_WINDOWPOSCHANGED back 00000000 0012fea8 Back window is told its going active WM_NCACTIVATE back 00000001 front WM_ACTIVATE back WA_CLICKACTIVE front Now the back window calls SetForegroundWindow(front), within its WM_ACTIVATE handler Back window is told its going inactive WM_NCACTIVATE back 00000000 front WM_ACTIVATE back 00000000 front Front window brought to the top of the z-order w/SetWindowPos WM_WINDOWPOSCHANGING front 00000000 0012fd0c WM_NCPAINT front 00000001 00000000 WM_ERASEBKGND front 0101005a 00000000 WM_WINDOWPOSCHANGED front 00000000 0012fd0c Front window is told its going active WM_NCACTIVATE front 00000001 back WM_ACTIVATE front 00000001 back Under wine the message stream doesn't have the SetWindowPos calls, and there are several hundred activate/ncactivate messages, ending with the back window in the front of the z-order, activated with the focus. Looking at the logic in dlls/user/focus.c, in set_active_window(), it seems that SetWindowPos() was not being called because the windows both have the WS_EX_MANAGED style set. Commenting out the line if (!(GetWindowLongW( hwnd, GWL_EXSTYLE ) & WS_EX_MANAGED)) Caused the front window to remain on top after clicking on the back window, as desired. However it did then not get the focus as it should, and titlebar clicks and double clicks in the client area still allow the back window to rise to the top of the z-order. I added a test to check if the window is the foreground window (i.e. that some other window hasn't been set in the interim) before setting the focus. This means my test app now appears to work just like native. The attached patch makes both of these changes. I'd be interested in any comments regarding whether the above is correct. Cheers, Jon ===== "Don't wait for the seas to part, or messiahs to come; Don't you sit around and waste this chance..." - Live jon_p_griffiths(a)yahoo.com __________________________________ Do you Yahoo!? The all-new My Yahoo! - Get yours free! http://my.yahoo.com

1 0

Breakage on Debian Sarge
by Michael Jung Nov. 25, 2004

Nov. 25, 2004

Hi, this patch (http://cvs.winehq.org/patch.py?id=14575) broke wine on a current Debian Sarge. Tested this on two machines. GUI applications do not respond any longer to mouse events. Tested regedit and winemine. Removing the patch solves the problem. Bye, Michael

3 2

Re: Adding debuginfo Support to Wine Makefiles
by Ivan Leo Puoti Nov. 25, 2004

Nov. 25, 2004

> It is already there. I added support for it a few months ago. I've did debug packages for mdk the day after AJ commited your patch, but is there any easy way I can test to see if the debug info works with winedbg? Ivan. ____________________________________________________________ Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazione. Abbonati subito su http://www.libero.it

1 0

Re: Winecfg -> browse for folder
by Rolf Kalbermatter Nov. 24, 2004

Nov. 24, 2004

Robert van Herk <robert(a)robertvanherk.nl> wrote: >Also, I am having some troubles into tricking the SHBrowseForFolder >thingy into using a different root folder, since I guess I need to use >SHParseDisplayName >(http://msdn.microsoft.com/library/default.asp?url=/library/en-us/shellcc/pl…) >for that, but that doesn't seem to be implemented in Wine. This shouldn't be to difficult I guess. There are helper functions in shell32.pidl.c which should be close to what this function does actually. Something like this in pidl.c should do the trick: HRESULT WINAPI SHParseDisplayNameW(LPCWSTR pszName, IBindCtx *pbc, LPITEMIDLIST *ppidl, SFGAOF sfgaoIn, SFGAOF *psfgaoOut) { LPSHELLFOLDER sf; HRESULT ret = E_FAIL; TRACE("%s %p 0x%08lx\n", debugstr_w(pszName), ppidl, attributes ? *attributes : 0); if (SUCCEEDED (SHGetDesktopFolder(&sf))) { ret = IShellFolder_ParseDisplayName(sf, 0, pbc, (LPOLESTR)pszName, NULL, ppidl, &sfgaoIn); if (succeeded && psfgaoOut) *psfgaoOut = sfgaoIn; IShellFolder_Release(sf); } return ret; } Or you could probably use SHILCreateFromPathW() too for your purposes, which is similar. >Finally, I don't know how to free PIDLS. In the MSDN documentation they >use C++ calls for that, but I don't know how to port that to C. ILFree() exported as ordinal 155 from shell32 should be the function to use. It was undocumented in Windows for a long time but exists in shell32 since at least Win95. Rolf Kalbermatter

1 0