- Wine-Devel - list.winehq.org

regression of patch 8366
by Oleg Prokhorov June 30, 2003

June 30, 2003

Hello , This patch http://cvs.winehq.com/patch.py?root=/home/winehq/opt/cvs-commit&id=8366 Caused regression with chess.net application. (http://www.chess.net/cnfw/download/cnfw254.exe) If you click on a help button it will drop a messagebox which tells that 'No browser found', it is normal. The anomality is that it is not possible to close messagebox by clicking "ok" button with a mouse. But both pressing "enter" and clicking close window icon work. The line which produce the problem is (dlls/ole2/compobj.c) apt->win = CreateWindowA(aptWinClass, NULL, 0, 0, 0, 0, 0, 0, 0, OLE32_hInstance, NULL); -- Best regards, Oleg mailto:xolegpro(a)rbcmail.ru

1 0

Re: Known Wine ToDo's
by Shachar Shemesh June 30, 2003

June 30, 2003

Tom wrote: > Hello Shachar, > > I am going to update the ToDo page : > http://www.winehq.com/?page=status_todo > > And im wondering if you can send me a list of ToDo's that your > aware of ? > > Thanks > > Tom > The tasks I'm talking about here are long term. Not all of them have been discussed on wine-devel yet. I'm CCing wine-devel in case anyone wants to object. National Language Support - Remove "Hebrew and Arabic" - it's just part of BiDi if technology is what wer'e talking about. If interface localization is what wer'e talking about, then just add "add localization to more languages". As a subitem of keyboard support, add "allow using a keyboard with keys outside the current locale codepage". Tools - Add "perform Windows' reboot operations automatically when required". Under "Low priority" add "Windows ACL support". -- Shachar Shemesh Open Source integration consultant Home page & resume - http://www.shemesh.biz/

1 0

Re: [MSRLE32] Some bugs and improvements.
by Tom June 29, 2003

June 29, 2003

Michael Günnewig wrote: >Changelog: > - Fixed interframe compression - now works for RLE4 and RLE8. > - Fixed alignment of RLE4 in absolute mode. > - Improved compression (smaller output). > > Michael Günnewig > > > >------------------------------------------------------------------------ > >Index: dlls/msvideo/msrle32/msrle32.c >=================================================================== >RCS file: /home/wine/wine/dlls/msvideo/msrle32/msrle32.c,v >retrieving revision 1.4 >diff -d -u -r1.4 msrle32.c >--- dlls/msvideo/msrle32/msrle32.c 13 Jun 2003 23:17:01 -0000 1.4 >+++ dlls/msvideo/msrle32/msrle32.c 29 Jun 2003 17:32:06 -0000 >@@ -1,5 +1,5 @@ > /* >- * Copyright 2002 Michael Günnewig >+ * Copyright 2002-2003 Michael Günnewig > * > * This library is free software; you can redistribute it and/or > * modify it under the terms of the GNU Lesser General Public >@@ -17,8 +17,7 @@ > */ > > /* TODO: >- * - compression of RLE4 is buggy -- see FIXME's >- * - many improvements possible >+ * - some improvements possible > * - implement DecompressSetPalette? -- does we need it for anything? > */ > >@@ -297,10 +296,13 @@ > if (ColorCmp(clr1, clr3) <= lDist && > ColorCmp(clr2, clr4) <= lDist) { > /* diff at end? -- look-ahead for atleast ?? more encodable pixel */ >- >- /* FIXME */ >- >- return count; >+ if (pos + 2 < width && ColorCmp(clr1,lpB[pos+1]) <= lDist && >+ ColorCmp(clr2,lpB[pos+2]) <= lDist) { >+ if (pos + 4 < width && ColorCmp(lpB[pos+1],lpB[pos+3]) <= lDist && >+ ColorCmp(lpB[pos+2],lpB[pos+4]) <= lDist) >+ return count - 3; /* followed by atleast 4 encodable pixels */ >+ return count - 2; >+ } > } else if (lpP != NULL && ColorCmp(lpP[pos], lpB[pos]) <= lDist) { > /* 'compare' with previous frame for end of diff */ > INT count2 = 0; >@@ -331,9 +333,11 @@ > > for (count = 0; pos < width; pos++, count++) { > if (ColorCmp(lpA[pos], lpB[pos]) <= lDist) { >- /* diff at end? -- look-ahead for atleast one more encodable pixel */ >- if (pos + 1 < width && ColorCmp(lpA[pos], lpB[pos+1]) <= lDist) >- return count; >+ /* diff at end? -- look-ahead for some more encodable pixel */ >+ if (pos + 1 < width && ColorCmp(lpB[pos], lpB[pos+1]) <= lDist) >+ return count - 1; >+ if (pos + 2 < width && ColorCmp(lpB[pos+1], lpB[pos+2]) <= lDist) >+ return count - 1; > } else if (lpP != NULL && ColorCmp(lpP[pos], lpB[pos]) <= lDist) { > /* 'compare' with previous frame for end of diff */ > INT count2 = 0; >@@ -382,6 +386,8 @@ > /* add some pixel for absoluting if possible */ > count += countDiffRLE4(lpP, lpC - 1, lpC, pos-1, lDist, lpbi->biWidth); > >+ assert(count > 0); >+ > /* check for near end of line */ > if (x + count > lpbi->biWidth) > count = lpbi->biWidth - x; >@@ -390,9 +396,11 @@ > while (count > 2) { > INT i; > INT size = min(count, 254); >- BOOL extra_byte = (size / 2) % 2 || (size % 2); >+ int bytes = ((size + 1) & (~1)) / 2; >+ BOOL extra_byte = bytes & 0x01; > >- *lpSizeImage += 2 + size/2 + (size % 2) + extra_byte; >+ *lpSizeImage += 2 + bytes + extra_byte; >+ assert(((*lpSizeImage) % 2) == 0); > count -= size; > *lpOut++ = 0; > *lpOut++ = size; >@@ -468,6 +476,8 @@ > /* add some more pixels for absoluting if possible */ > count += countDiffRLE8(lpP, lpC - 1, lpC, pos-1, lDist, lpbi->biWidth); > >+ assert(count > 0); >+ > /* check for over end of line */ > if (x + count > lpbi->biWidth) > count = lpbi->biWidth - x; >@@ -663,6 +673,10 @@ > assert(lpOut == lpOutStart + lpbiOut->biSizeImage); > } > } >+ >+ /* add EOL -- will be changed to EOI */ >+ lpbiOut->biSizeImage += 2; >+ *((LPWORD)lpOut)++ = 0; > } > > /* change EOL to EOI -- end of image */ >@@ -789,11 +803,14 @@ > if (jumpy == 0) { > /* add EOL -- end of line */ > lpbiOut->biSizeImage += 2; >- *lpOut++ = 0; >- *lpOut++ = 0; >+ *((LPDWORD)lpOut)++ = 0; > assert(lpOut == (lpOutStart + lpbiOut->biSizeImage)); > } > } >+ >+ /* add EOL -- will be changed to EOI */ >+ lpbiOut->biSizeImage += 2; >+ *((LPDWORD)lpOut)++ = 0; > } > > /* change EOL to EOI -- end of image */ >@@ -847,7 +864,7 @@ > } > break; > default: /* absolute mode */ >- extra_byte = (code1 / 2) & 0x01 || (code1 % 2); >+ extra_byte = (((code1 + 1) & (~1)) / 2) & 0x01; > > if (pixel_ptr/bytes_per_pixel + code1 > lpbi->biWidth) > return ICERR_ERROR; >@@ -1468,9 +1485,7 @@ > > if (lpic->lpbiOutput->biBitCount == 4) > MSRLE32_CompressRLE4(pi, lpic->lpbiInput, (LPBYTE)lpic->lpInput, >- lpic->lpbiOutput, (LPBYTE)lpic->lpOutput, TRUE); >- /*MSRLE32_CompressRLE4(pi, lpic->lpbiInput, (LPBYTE)lpic->lpInput, >- lpic->lpbiOutput, (LPBYTE)lpic->lpOutput, (lpic->dwFlags & ICCOMPRESS_KEYFRAME) != 0);*/ >+ lpic->lpbiOutput, (LPBYTE)lpic->lpOutput, (lpic->dwFlags & ICCOMPRESS_KEYFRAME) != 0); > else > MSRLE32_CompressRLE8(pi, lpic->lpbiInput, (LPBYTE)lpic->lpInput, > lpic->lpbiOutput, (LPBYTE)lpic->lpOutput, (lpic->dwFlags & ICCOMPRESS_KEYFRAME) != 0); > >

1 0

Re: Uninitialised variable in TASK_Create
by Saulius Krasuckas June 29, 2003

June 29, 2003

> Hi, > > There is an unintialised string (the name variable) in TASK_Create in > task.c (visible if you run almost any program with --debugmsg +task). > I would try and fix the problem but I have no idea what the correct > fix is. Is this just a case of adding *name = 0? > > Rob hello, (heh, i've lost the race.. :P) cents: i have found patch, which made my valgrind to scream bout the same issue: http://cvs.winehq.com/patch.py?root=/home/winehq/opt/cvs-commit&id=8245 Index: wine/loader/task.c diff -u wine/loader/task.c:1.132 wine/loader/task.c:1.133 --- wine/loader/task.c:1.132 Sun Jun 29 20:22:43 2003 +++ wine/loader/task.c Sun Jun 29 20:22:43 2003 <skipped> @@ -299,13 +300,16 @@ /* Copy the module name */ - GetModuleName16( pModule->self, name, sizeof(name) ); - strncpy( pTask->module_name, name, sizeof(pTask->module_name) ); + if (hModule) + { + GetModuleName16( hModule, name, sizeof(name) ); + strncpy( pTask->module_name, name, sizeof(pTask->module_name) ); + } in addition to null-initializing the "name", i'd add this at several page-downs: - pTask->compat_flags = GetProfileIntA( "Compatibility", name, 0 ); + if(name) + { + pTask->compat_flags = GetProfileIntA( "Compatibility", name, 0 ); + } IMHO we'd "save" two api-calls (GetProfileIntA() ant GetPrivateProfileIntA()) using such check.

1 0

Uninitialised variable in TASK_Create
by Robert Shearman June 29, 2003

June 29, 2003

Hi, There is an unintialised string (the name variable) in TASK_Create in task.c (visible if you run almost any program with --debugmsg +task). I would try and fix the problem but I have no idea what the correct fix is. Is this just a case of adding *name = 0? Rob

1 0

CVS Install error: version mismatch 113/109
by Moreno June 29, 2003

June 29, 2003

Hi, i download last wine source. i compiling and when i use wineinstall i have the follow error if i execute wine i have the same error wine client error:9: version mismatch 113/109. Your wine binary was not upgraded correctly, or you have an older one somewhere in your PATH. Or maybe the wrong wineserver is still running? Registry install failed. what can i do? -- Moreno <morenoca(a)freemail.it>

2 4

Re: QueryPerformanceFrequency bugfix
by Andreas Mohr June 29, 2003

June 29, 2003

Hi, On Sun, Jun 29, 2003 at 04:05:04PM +0100, Mike Hearn wrote: > This patch stops the demo of Settlers 4 crashing at startup. > > ChangeLog: > Add some traces > A Mhz is 1000Hz, not 100000 Whaaaaa!? First, it's 1000000, which is one zero more than what you wrote (and what the code used: +1000000.0). Second, one MHz *IS* 1000000Hz. So there are two possibilities: a) the code should calculate the result in units of kHz (then your observation would be correct) b) you are (completely? ;) wrong Andreas Mohr

2 1

Re: QueryPerformanceFrequency bugfix
by Tom June 29, 2003

June 29, 2003

Mike Hearn wrote: >Yes, OK, so as just pointed out to me, neither me nor my flatmate with >whom I double checked this can count :D > >On Sun, 2003-06-29 at 16:05, Mike Hearn wrote: > > >>This patch stops the demo of Settlers 4 crashing at startup. >> >>ChangeLog: >>Add some traces >>A Mhz is 1000Hz, not 100000 >> >> :) Here ya go : http://searchnetworking.techtarget.com/sDefinition/0,,sid7_gci212562,00.html Tom

1 0

Re: PATCH: more capabilities for linux/cpu flags
by Philipp Wollermann June 29, 2003

June 29, 2003

Hi, according to MSDN (http://msdn.microsoft.com/library/default.asp?url=/library/en-us/sysinfo/ba se/isprocessorfeaturepresent.asp): PF_XMMI_INSTRUCTIONS_AVAILABLE The SSE instruction set is available. PF_XMMI64_INSTRUCTIONS_AVAILABLE The SSE2 instruction set is available. Philipp ----- Original Message ----- From: "Marcus Meissner" <marcus(a)jet.franken.de> To: <wine-patches(a)winehq.com> Sent: Sunday, June 29, 2003 12:35 PM Subject: PATCH: more capabilities for linux/cpu flags > Hi, > > This adds 2 more checks for PF_xxx flag CPU detection for Linux. > > Ciao, Marcus > > Changelog: > Add tests for PF_AMD3D and PF_XMMI instructions. (Not sure if > XMMI instructions are SSE or SSE2). > > Index: misc/cpu.c > =================================================================== > RCS file: /home/wine/wine/misc/cpu.c,v > retrieving revision 1.32 > diff -u -r1.32 cpu.c > --- misc/cpu.c 21 May 2003 18:24:34 -0000 1.32 > +++ misc/cpu.c 29 Jun 2003 10:32:11 -0000 > @@ -315,6 +315,13 @@ > PF[PF_MMX_INSTRUCTIONS_AVAILABLE] = TRUE; > if (strstr(value,"tsc")) > PF[PF_RDTSC_INSTRUCTION_AVAILABLE] = TRUE; > + if (strstr(value,"3dnow")) > + PF[PF_AMD3D_INSTRUCTIONS_AVAILABLE] = TRUE; > + /* Not sure if XMMI is sse2 or sse, stay on > + * the safe side. > + */ > + if (strstr(value,"sse2")) > + PF[PF_XMMI_INSTRUCTIONS_AVAILABLE] = TRUE; > > } > } >

1 0

link updates #4
by Tom Wickline June 29, 2003

June 29, 2003

Hello, With these three fixes all failed urls are fixed and there is only one issue of WWN in the past 13 weeks that has a failed url. The fixes to cvs.shtml, opengl-problems.shtml, wine-debugger.shtml are in #2 and #3. I was going to try and fix most/all the WWN links but after looking into this I see it could be a time consuming task. And i'm not sure how substantial the end benefit would really be.. Anyone have any comments about this ? If you guy's think its a good thing ill fix as many as I can.. If you think its a no win situation I will just look at the past 13 issues in the future.. Tom

1 0