June 2003 - Wine-devel - list.winehq.org

Re: XEMBED System tray support
by Mike Hearn 28 Jun '03

28 Jun '03

This one has been dropped a couple of times now. Anything I need to do to it (other than the obvious cosmetic improvements/work around gnome bugs etc) On Mon, 2003-06-23 at 20:53, Mike Hearn wrote: > bump > > ChangeLog: > Implement XEMBED based system tray support for KDE 3.1+ and GNOME 2.2+

1 0

imagehlp.h question - building with Mingw or MS_VC headers
by Steven Edwards 28 Jun '03

28 Jun '03

around line 560 in imagehlp.h we have this /*********************************************************************** * Wine specific */ extern HANDLE IMAGEHLP_hHeap; Can this be moved to a internal header or done in some other way so we can build with other headers? Thanks Steven

2 1

Re: PATCH: dosmem access via ReadProcessMemory
by Alexandre Julliard 28 Jun '03

28 Jun '03

Marcus Meissner <meissner(a)suse.de> writes: > An EEPROM programmer from BP Microsystems uses ReadProcessMemory to access > the BIOS data segment (to find out parallel ports). What happens if you run it with Windows version set to NT? -- Alexandre Julliard julliard(a)winehq.com

1 0

Re: RFC: dsound patch
by Robert Reif 27 Jun '03

27 Jun '03

I just sent a rather large patch to wine-patches that addresses a number of 3D sound problems and also fixes a problem I was having when switching between sound cards quickly. I added more tests for 3D, COM and reference counts and have run the tests on win2k with dx8.1 with the same results. I have also tested with the sample programs from the DX8 SDK. Please test it with real apps and send me feedback. Thanks, Bob.

1 0

Wine Status Changelog
by Tom 27 Jun '03

27 Jun '03

Hello, Here is what I have for the Changelog thus far. There is 49 entries so far and I could have over looked something. If anything is out of place, incorrect just let me know and ill fix it. Tom

4 7

[Patch] dlls/oleaut32/typelib.c - Via a CVS search, reverse regression to get Kazza working
by Stefan Jones 27 Jun '03

27 Jun '03

On upgrading to Wine 20030618 I found that kazaa lite will no longer work and gave me the following error: No debug information in 32bit DLL 'C:\WINDOWS\SYSTEM\SHDOCVW.DLL' (0x71000000) Unhandled exception: page fault on read access to 0x00000004 in 32-bit code (0x411ed9c9). In 32-bit mode. 0x411ed9c9 (ITypeInfo_fnGetContainingTypeLib+0x49 [typelib.c:4801] in oleaut32.dll.so): call *0x4(%edx) 4802 TRACE("returning ppTLib=%p", *ppTLib); Wine-dbg>WineDbg terminated on pid a using in wine.config: [AppDefaults\\KazaaLite.kpp\\DllOverrides] "*" = "builtin, native, so" ;"shdoclc" = "native" "shdocvw" = "native" "shlwapi" = "native" "commctrl" = "native" ;"comdlg32" = "native" I looked though the CVS logs and mail archives and I found the following: http://www.winehq.org/hypermail/wine-devel/2003/06/0414.html and the patch below which was applied to dlls/oleaut32/typelib.c This caused the crash, reversing it fixes the error. please fix / apply, or I will cry!!! anyway, how do you use winedbg, I am not thick but it refuses to do normal gdb stuff like the run command or a backtrace! Is there a URL I missed? Stefan revision 1.95 date: 2003/05/13 22:15:26; author: julliard; state: Exp; lines: +0 -2 Dmitry Timoshkov <dmitry(a)codeweavers.com> Remove redundant ITypeLib2_AddRef's. Makes LoadTypeLib/ITypeLib2_Release work. Index: typelib.c =================================================================== RCS file: /home/wine/wine/dlls/oleaut32/typelib.c,v retrieving revision 1.94 retrieving revision 1.95 diff -u -r1.94 -r1.95 --- typelib.c 13 May 2003 00:41:58 -0000 1.94 +++ typelib.c 13 May 2003 22:15:26 -0000 1.95 @@ -1844,7 +1844,6 @@ pcx->pTblDir->pTypeInfoTab.offset+count*sizeof(tiBase)); /* this is where we are coming from */ ptiRet->pTypeLib = pLibInfo; - ITypeLib2_AddRef((ITypeLib2 *)pLibInfo); ptiRet->index=count; /* fill in the typeattr fields */ FIXME("Assign constructor/destructor memid\n"); @@ -3055,7 +3054,6 @@ } *ppTypeInfoImpl = (ITypeInfoImpl*)ITypeInfo_Constructor(); (*ppTypeInfoImpl)->pTypeLib = pTypeLibImpl; - ITypeLib2_AddRef((ITypeLib2 *)pTypeLibImpl); (*ppTypeInfoImpl)->index = i; (*ppTypeInfoImpl)->Name = TLB_MultiByteToBSTR( pOtherTypeInfoBlks[i].name_offs +

1 0

Re: winedos / Preparing for DPMI IRQs
by Sylvain Petreolle 27 Jun '03

27 Jun '03

Hi, Could you give examples of what has been enabled here ? I will be happy to do tests if some free programs uses this. ===== Sylvain Petreolle (spetreolle at users dot sourceforge dot net) ICQ #170597259 "What if tomorrow the War could be over ?" Morpheus, in "Reloaded". For the Law of Oil and Fire, Im an European that lives in France. For all my Brothers and friends, Im a human living on Earth. ___________________________________________________________ Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français ! Yahoo! Mail : http://fr.mail.yahoo.com

3 7

Wintab: Update on debugging & demo patch.
by Robert North 26 Jun '03

26 Jun '03

Ok, Firstly, I'm including a patch that demonstrates fixes to the crashing in Painter, and the recursion in Photoshop. The patch assumes that you have already applied Aric's patch from April 23rd. I'd like feedback on wether the patch is OK, as it was generated on from my local subversion repository, not CVS. Please bear in mind that the code is designed as quick demonstration fixes, not for quality code. Having said that comments are welcome. ******** Secondly, I've done some more investigation into why Photoshop was not seeing tablet pressure. It turns out that this is due to Photoshop requesting tablet events from the desktop window. In the x11drv/event.c file, X11 events are interrogated, and the X11 window converted to a win32 window. This is done using the X11 feature of contexts: Application specific info mapping from an X11 window to user defined value. It appears that the X11 window has no X11 context, and therefore the XInput event handler is passed a hWnd of NULL. This causes the FindOpenContext not match any context. (Hence the recursion) Without a context, no event is posted (Hence the lack of pressure info) Here's a typical debug trace showing the problem: trace:event:EVENT_ProcessEvent called. trace:event:EVENT_ProcessEvent Got extention event for hwnd/window (nil)/37, GetFocus()=0x40042 trace:event:X11DRV_ProcessTabletEvent Recieved tablet motion event ((nil)) trace:wintab32:X11DRV_ProcessTabletEvent Recieved tablet motion event ((nil)) trace:wintab32:WTPacketsPeek (0xc00, 64, 0x43c828c4) trace:wintab32:WTPacketsGet (0xc00, 64, 0x1098388) trace:wintab32:TABLET_WindowProc Incomming Message 0x7ff0 (0x00000000, 0x00000000) trace:wintab32:AddPacketToContextQueue Trying Queue c00 (0 10020) trace:wintab32:AddPacketToContextQueue Done ((nil)) ******** Thirdly, I've found out why Painter wasn't scaling correctly. There appear to be 2 issues. a. Tablet coordinates should be relative to the window the wintab context is attached to. Currently they are absoloute. b. The default system context as returned by WTInfo contains zero in the lcSysExtX,lcSysExtY fields These fields should contain the I know how to fix the above problems, and will do so. P.S. Aric, I can now see the race condtion problem you were talking about in Painter. Cheers -Rob. Index: dlls/wintab32/context.c =================================================================== --- dlls/wintab32/context.c (.../file:///usr/local/svnroot/wine/trunk) (revision 73) +++ dlls/wintab32/context.c (revision 73) @@ -3,6 +3,7 @@ * * Copyright 2002 Patrik Stridvall * Copyright 2003 Codeweavers, Aric Stewart + * Copyright 2003 Robert North * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public @@ -168,11 +169,16 @@ ptr = gOpenContexts; while (ptr) { - TRACE("Trying Context %lx (%x %x)\n",(LONG)ptr->handle,(INT)hwnd, + TRACE("Trying Context %lx hwnd(%x %x)\n",(LONG)ptr->handle,(INT)hwnd, (INT)ptr->hwndOwner); - if (ptr->hwndOwner == hwnd) + if (ptr->hwndOwner == hwnd) return ptr; + /* Rob 19-6-2003 Added recursion down pointer chein. + to avoid infinite loop. + Use TABLET_FindOpenContext as template. + */ + ptr = ptr->next; } return NULL; } @@ -251,7 +257,13 @@ int static inline CopyTabletData(LPVOID target, LPVOID src, INT size) { - memcpy(target,src,size); +/* + Rob 19-6-2003 + Return size only, if target is NULL: + Was causing horrible crashes in MetaCreation's Painter 5. +*/ + if(target != NULL) + memcpy(target,src,size); return(size); } @@ -318,39 +330,42 @@ static VOID TABLET_BlankPacketData(LPOPENCONTEXT context, LPVOID lpPkt, INT n) { - int rc = 0; + if(lpPkt != NULL) + { + int rc = 0; - if (context->context.lcPktData & PK_CONTEXT) - rc +=sizeof(HCTX); - if (context->context.lcPktData & PK_STATUS) - rc +=sizeof(UINT); - if (context->context.lcPktData & PK_TIME) - rc += sizeof(LONG); - if (context->context.lcPktData & PK_CHANGED) - rc += sizeof(WTPKT); - if (context->context.lcPktData & PK_SERIAL_NUMBER) - rc += sizeof(UINT); - if (context->context.lcPktData & PK_CURSOR) - rc += sizeof(UINT); - if (context->context.lcPktData & PK_BUTTONS) - rc += sizeof(DWORD); - if (context->context.lcPktData & PK_X) - rc += sizeof(DWORD); - if (context->context.lcPktData & PK_Y) - rc += sizeof(DWORD); - if (context->context.lcPktData & PK_Z) - rc += sizeof(DWORD); - if (context->context.lcPktData & PK_NORMAL_PRESSURE) - rc += sizeof(UINT); - if (context->context.lcPktData & PK_TANGENT_PRESSURE) - rc += sizeof(UINT); - if (context->context.lcPktData & PK_ORIENTATION) - rc += sizeof(ORIENTATION); - if (context->context.lcPktData & PK_ROTATION) - rc += sizeof(ROTATION); + if (context->context.lcPktData & PK_CONTEXT) + rc +=sizeof(HCTX); + if (context->context.lcPktData & PK_STATUS) + rc +=sizeof(UINT); + if (context->context.lcPktData & PK_TIME) + rc += sizeof(LONG); + if (context->context.lcPktData & PK_CHANGED) + rc += sizeof(WTPKT); + if (context->context.lcPktData & PK_SERIAL_NUMBER) + rc += sizeof(UINT); + if (context->context.lcPktData & PK_CURSOR) + rc += sizeof(UINT); + if (context->context.lcPktData & PK_BUTTONS) + rc += sizeof(DWORD); + if (context->context.lcPktData & PK_X) + rc += sizeof(DWORD); + if (context->context.lcPktData & PK_Y) + rc += sizeof(DWORD); + if (context->context.lcPktData & PK_Z) + rc += sizeof(DWORD); + if (context->context.lcPktData & PK_NORMAL_PRESSURE) + rc += sizeof(UINT); + if (context->context.lcPktData & PK_TANGENT_PRESSURE) + rc += sizeof(UINT); + if (context->context.lcPktData & PK_ORIENTATION) + rc += sizeof(ORIENTATION); + if (context->context.lcPktData & PK_ROTATION) + rc += sizeof(ROTATION); - rc *= n; - memset(lpPkt,0,rc); + rc *= n; + memset(lpPkt,0,rc); + } } @@ -856,7 +871,14 @@ */ int WINAPI WTPacketsGet(HCTX hCtx, int cMaxPkts, LPVOID lpPkts) { - int limit; + /* + Rob 19-6-2003 + WTPacketsGet fails with lpPkts set to NULL. + TO DO: Handle null condition. + Null condition flushes queue (Of Max packets????) + */ + int limit = 0; + int maxPktsToGet; LPOPENCONTEXT context; LPVOID ptr = lpPkts; @@ -872,15 +894,20 @@ return 0; } - for(limit = 0; limit < cMaxPkts && limit < context->PacketsQueued; limit++) + maxPktsToGet = min(cMaxPkts,context->PacketsQueued); + if(lpPkts != NULL) + { + for(limit = 0; limit < maxPktsToGet; limit++) ptr=TABLET_CopyPacketData(context ,ptr, &context->PacketQueue[limit]); - - if (limit < context->PacketsQueued) + } + + if (maxPktsToGet < context->PacketsQueued) { - memcpy(context->PacketQueue, &context->PacketQueue[limit], - (context->QueueSize - (limit))*sizeof(WTPACKET)); + /* TO DO: Wrong function mem areas overlap. should use memmove*/ + memmove(context->PacketQueue, &context->PacketQueue[maxPktsToGet], + (context->QueueSize - (maxPktsToGet))*sizeof(WTPACKET)); } - context->PacketsQueued -= limit; + context->PacketsQueued -= maxPktsToGet; LeaveCriticalSection(&csTablet); TRACE("Copied %i packets\n",limit); @@ -911,7 +938,7 @@ if ((rc+1) < context->QueueSize) { - memcpy(context->PacketQueue, &context->PacketQueue[rc+1], + memmove(context->PacketQueue, &context->PacketQueue[rc+1], (context->QueueSize - (rc+1))*sizeof(WTPACKET)); } context->PacketsQueued -= (rc+1); @@ -1080,7 +1107,7 @@ for (limit = 0; limit < cMaxPkts && limit < context->PacketsQueued; limit++) ptr = TABLET_CopyPacketData(context ,ptr, &context->PacketQueue[limit]); - + LeaveCriticalSection(&csTablet); TRACE("Copied %i packets\n",limit); return limit; @@ -1099,7 +1126,7 @@ UINT num = 0; TRACE("(%p, %u, %u, %d, %p, %p)\n", - hCtx, wBegin, wEnd, cMaxPkts, lpPkts, lpNPkts); + hCtx, wBegin, wEnd, cMaxPkts, lpPkts, lpNPkts); context = TABLET_FindOpenContext(hCtx); @@ -1130,7 +1157,7 @@ /* remove read packets */ if ((end+1) < context->PacketsQueued) - memcpy( &context->PacketQueue[bgn], &context->PacketQueue[end+1], + memmove( &context->PacketQueue[bgn], &context->PacketQueue[end+1], (context->PacketsQueued - ((end-bgn)+1)) * sizeof (WTPACKET)); context->PacketsQueued -= ((end-bgn)+1);

2 2

Wine Status DLL's #2
by Tom 26 Jun '03

26 Jun '03

Here is #2 any one have any comments ? Tom

3 4

Re: Mac OS X Support : Asm syntax support
by Marcus Meissner 26 Jun '03

26 Jun '03

On Wed, Jun 25, 2003 at 02:32:29PM +0200, Pierre d'Herbemont wrote: > Hi, > > Here is a patch which should add support for Mac OS X ppc asm syntax, > and some Mach-O quirk (like text section). > I hope it will be ok. It mainly relay on Macro which are stored in > build.h (any better place?). > > Cheers, > > Pierre > > ChangeLog: > - Add support for Mac OS X assembler syntax This needs 2 places fixed: > +# define LD_SECTION_TEXT ".section \".text\"" Should be # define LD_SECTION_TEXT ".section \\\".text\\\"" and: > +# define PPC_LOW(mem,index) "(" mem ")@l(##index)" should be: # define PPC_LOW(mem,index) "(" mem ")@l(" #index ")" Ciao, Marcus

2 1