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

RE: question about copyright and code, is this legal?
by Patrik Stridvall 09 Oct '03

09 Oct '03

> On Fri, 3 Oct 2003 23:07, Patrik Stridvall wrote: > > [correct summary of the law deleted] > > > As far as Wine is concerned though I can't think of > > any Windows API that: > > 1. Returns pure facts (not having any unique expression) > > 2. Would fulfill the requirement of needed a significant > > effort to collect to collect the fact. > > > > So no I wouldn't worry about it as far as Wine is concerned. > > Unfortunately it appears there may be one: LCMapString. If > all the things we > have discovered about the tables are correct (and I am by no > means assuming > that they are), then the Microsoft sort key table meets the > industrious > collection requirements and returns facts in a way that could not be > duplicated without duplicating the table (it returns the > table entry in > literal form, so it couldn't even be represented as a > different table). The significant effort requirement is for EACH fact seen by ITSELF. For example if a make a database with the home and work addresses and telephone numbers for every Wine contributor EACH fact seen by ITSELF will take significant effort (at least on average) since they are in most cases are not readily available. In your example above anybody knowledgable in some language, say Japanese, can easily determine the return values for inputs involving their language for LCString AFAICS. Futhermore the example I made above is what the law was designed to protect so it is doubtful any court would even attempt to extend it to your example. The intent of the lawmakers is VERY important here in Sweden. > This is what is (generally) returned by LCMapString. The > actual table is > included in an external file under NT (sortkey.nls), so one > option (likely > the best option) would be to implement LCMapString using a > standard table > from unicode.org, then provide a facility so that people in > the US and Canada > can (at their own risk of course) use the Microsoft versions > relying on > Feist. In addition I doubt any court in Sweden would even try the case since the facts in questions was intentionally published by Microsoft in the US where the collection of facts was not protected because of Feist and thus legally importable from there...

2 1

european cvs mirror problems
by Ivan Leo Murray-Smith 09 Oct '03

09 Oct '03

I'm getting this message cvs update -PAd cvs [update aborted]: connect to rhlx01.fht-esslingen.de(129.143.116.10):2401 failed: Connection refused will this be fixed?

1 0

Re: Byteswapping
by Warren_Baird＠cimmetry.com 09 Oct '03

09 Oct '03

> The best is to use the Rtl*ByteSwap functions from winternl.h. That is certainly the best solution --- I hadn't noticed that those had been added... I'll try redoing our swapping code with them and make sure it all works... Warren

1 0

Re: [janatorial] HeapReAlloc
by Dimitrie O. Paun 09 Oct '03

09 Oct '03

On Wed, 8 Oct 2003, Oleg Prokhorov wrote: > Change-log: > fix wine code depended on That's cool, but we need to somehow manage this. There are almost 300 occurences of memory reallocation, and we need to somehow keep track of what was reviwed/fixed etc., otherwise we risk missing quite a few... We should start with a comprehensive list of things that need fixing... -- Dimi.

4 3

imagehlp - MS_VC/Mingw porting
by Steven Edwards 09 Oct '03

09 Oct '03

The part about the imagehlp version in the header doesnt have to be in there I just figured we might as well add it while we are trying to be the same. Changelog: Bring imagehlp (mostly) inline with the PSDK and Mingw Index: wine/dlls/imagehlp/debug.c =================================================================== RCS file: /home/wine/wine/dlls/imagehlp/debug.c,v retrieving revision 1.11 diff -u -r1.11 debug.c --- wine/dlls/imagehlp/debug.c 5 Sep 2003 23:08:37 -0000 1.11 +++ wine/dlls/imagehlp/debug.c 9 Oct 2003 01:01:39 -0000 @@ -75,7 +75,7 @@ */ BOOL WINAPI StackWalk( DWORD MachineType, HANDLE hProcess, HANDLE hThread, - PSTACKFRAME StackFrame, PVOID ContextRecord, + LPSTACKFRAME StackFrame, LPVOID ContextRecord, PREAD_PROCESS_MEMORY_ROUTINE ReadMemoryRoutine, PFUNCTION_TABLE_ACCESS_ROUTINE FunctionTableAccessRoutine, PGET_MODULE_BASE_ROUTINE GetModuleBaseRoutine, Index: wine/dlls/imagehlp/imagehlp_main.c =================================================================== RCS file: /home/wine/wine/dlls/imagehlp/imagehlp_main.c,v retrieving revision 1.15 diff -u -r1.15 imagehlp_main.c --- wine/dlls/imagehlp/imagehlp_main.c 5 Sep 2003 23:08:37 -0000 1.15 +++ wine/dlls/imagehlp/imagehlp_main.c 9 Oct 2003 01:01:39 -0000 @@ -58,7 +58,7 @@ /*********************************************************************** * ImagehlpApiVersion (IMAGEHLP.@) */ -PAPI_VERSION WINAPI ImagehlpApiVersion() +LPAPI_VERSION WINAPI ImagehlpApiVersion(VOID) { return &IMAGEHLP_ApiVersion; } @@ -66,7 +66,7 @@ /*********************************************************************** * ImagehlpApiVersionEx (IMAGEHLP.@) */ -PAPI_VERSION WINAPI ImagehlpApiVersionEx(PAPI_VERSION AppVersion) +LPAPI_VERSION WINAPI ImagehlpApiVersionEx(LPAPI_VERSION AppVersion) { if(!AppVersion) return NULL; Index: wine/include/imagehlp.h =================================================================== RCS file: /home/wine/wine/include/imagehlp.h,v retrieving revision 1.19 diff -u -r1.19 imagehlp.h --- wine/include/imagehlp.h 5 Sep 2003 23:15:44 -0000 1.19 +++ wine/include/imagehlp.h 9 Oct 2003 01:01:51 -0000 @@ -25,6 +25,8 @@ extern "C" { #endif /* defined(__cplusplus) */ +#define API_VERSION_NUMBER 7 /* 7 is the default */ + /*********************************************************************** * Types */ @@ -168,7 +170,7 @@ USHORT MinorVersion; USHORT Revision; USHORT Reserved; -} API_VERSION, *PAPI_VERSION; +} API_VERSION, *LPAPI_VERSION; typedef struct _IMAGE_DEBUG_INFORMATION { LIST_ENTRY List; @@ -414,11 +416,11 @@ BOOL WINAPI ImageUnload( PLOADED_IMAGE LoadedImage ); -PAPI_VERSION WINAPI ImagehlpApiVersion( +LPAPI_VERSION WINAPI ImagehlpApiVersion( void ); -PAPI_VERSION WINAPI ImagehlpApiVersionEx( - PAPI_VERSION AppVersion +LPAPI_VERSION WINAPI ImagehlpApiVersionEx( + LPAPI_VERSION AppVersion ); BOOL WINAPI MakeSureDirectoryPathExists( LPCSTR DirPath __________________________________ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com

1 0

Protected mode interrupt problem
by Mike McCormack 08 Oct '03

08 Oct '03

Hi All, I've run into an interrupt related problem with Quicken 2002's installer. Two 16bit programs launch, one after the other. Each program hooks INT 0, however since our interrupt tables are global to all tasks, the second task retrieves the interrupt vector that the first task set. The first task exits, and restores the original INT 0 vector. When the second task exits, tries to restore the INT 0 vector set by the first task, which now has an invalid selector, so it crashes. After a bit of discussion and writing a test program, we discovered that it appears that some interrupt vectors are task local, and some are global. The local ones appear to be stored in the TDB (task descriptor block), in offsets 0x26 - 0x4a. The attached patch fixes this, though we're not sure that it is correct for DOS programs. I have enclosed a test program that shows the problem. Mike Index: include/task.h =================================================================== RCS file: /cvstrees/crossover/office/wine/include/task.h,v retrieving revision 1.1.1.5 diff -u -r1.1.1.5 task.h --- include/task.h 12 Sep 2003 18:37:36 -0000 1.1.1.5 +++ include/task.h 7 Oct 2003 06:29:43 -0000 @@ -98,13 +98,13 @@ FARPROC16 sighandler WINE_PACKED; /* 26 Signal handler */ FARPROC16 userhandler WINE_PACKED; /* 2a USER signal handler */ FARPROC16 discardhandler WINE_PACKED; /* 2e Handler for GlobalNotify() */ - DWORD int0 WINE_PACKED; /* 32 int 0 (divide by 0) handler */ - DWORD int2 WINE_PACKED; /* 36 int 2 (NMI) handler */ - DWORD int4 WINE_PACKED; /* 3a int 4 (INTO) handler */ - DWORD int6 WINE_PACKED; /* 3e int 6 (invalid opc) handler */ - DWORD int7 WINE_PACKED; /* 42 int 7 (coprocessor) handler */ - DWORD int3e WINE_PACKED; /* 46 int 3e (80x87 emu) handler */ - DWORD int75 WINE_PACKED; /* 4a int 75 (80x87 error) handler */ + FARPROC16 int0 WINE_PACKED; /* 32 int 0 (divide by 0) handler */ + FARPROC16 int2 WINE_PACKED; /* 36 int 2 (NMI) handler */ + FARPROC16 int4 WINE_PACKED; /* 3a int 4 (INTO) handler */ + FARPROC16 int6 WINE_PACKED; /* 3e int 6 (invalid opc) handler */ + FARPROC16 int7 WINE_PACKED; /* 42 int 7 (coprocessor) handler */ + FARPROC16 int3e WINE_PACKED; /* 46 int 3e (80x87 emu) handler */ + FARPROC16 int75 WINE_PACKED; /* 4a int 75 (80x87 error) handler */ DWORD compat_flags WINE_PACKED; /* 4e Compatibility flags */ BYTE unused4[2]; /* 52 */ struct _TEB *teb; /* 54 Pointer to thread database */ Index: dlls/winedos/interrupts.c =================================================================== RCS file: /cvstrees/crossover/office/wine/dlls/winedos/interrupts.c,v retrieving revision 1.1.1.8 diff -u -r1.1.1.8 interrupts.c --- dlls/winedos/interrupts.c 12 Sep 2003 18:37:31 -0000 1.1.1.8 +++ dlls/winedos/interrupts.c 7 Oct 2003 06:29:47 -0000 @@ -25,6 +25,7 @@ #include "wine/winbase16.h" #include "thread.h" +#include "task.h" WINE_DEFAULT_DEBUG_CHANNEL(int); WINE_DECLARE_DEBUG_CHANNEL(relay); @@ -541,9 +542,42 @@ */ FARPROC16 DOSVM_GetPMHandler16( BYTE intnum ) { + TDB *pTask; + FARPROC16 proc = 0; + + pTask = TASK_GetCurrent(); + if (pTask) + { + switch( intnum ) + { + case 0x00: + proc = pTask->int0; + break; + case 0x02: + proc = pTask->int2; + break; + case 0x04: + proc = pTask->int4; + break; + case 0x06: + proc = pTask->int6; + break; + case 0x07: + proc = pTask->int7; + break; + case 0x3e: + proc = pTask->int3e; + break; + case 0x75: + proc = pTask->int75; + break; + } + if( proc ) + return proc; + } if (!DOSVM_Vectors16[intnum]) { - FARPROC16 proc = (FARPROC16)MAKESEGPTR( DOSVM_dpmi_segments->int16_sel, + proc = (FARPROC16)MAKESEGPTR( DOSVM_dpmi_segments->int16_sel, DOSVM_STUB_PM16 * intnum ); DOSVM_Vectors16[intnum] = proc; } @@ -558,9 +592,41 @@ */ void DOSVM_SetPMHandler16( BYTE intnum, FARPROC16 handler ) { + TDB *pTask; + TRACE("Set protected mode interrupt vector %02x <- %04x:%04x\n", intnum, HIWORD(handler), LOWORD(handler) ); - DOSVM_Vectors16[intnum] = handler; + + pTask = TASK_GetCurrent(); + if (!pTask) + return; + switch( intnum ) + { + case 0x00: + pTask->int0 = handler; + break; + case 0x02: + pTask->int2 = handler; + break; + case 0x04: + pTask->int4 = handler; + break; + case 0x06: + pTask->int6 = handler; + break; + case 0x07: + pTask->int7 = handler; + break; + case 0x3e: + pTask->int3e = handler; + break; + case 0x75: + pTask->int75 = handler; + break; + default: + DOSVM_Vectors16[intnum] = handler; + break; + } } #include <stdio.h> #include <windows.h> void myintfn(void) { } int PASCAL WinMain(HANDLE hInstance, HANDLE hPrevInstance, LPSTR lpCmdLine, int nCmdShow) { int old_seg, old_ofs, new_seg, new_ofs; char str[16]; /* get the old interrupt handler for int 0 */ _asm { push bx push es mov ax,0x3502 int 21h mov old_seg, es mov old_ofs, bx pop es pop bx } /* print it in a message box */ sprintf(str, "old %04x:%04x", old_seg, old_ofs ); MessageBox( 0, str, "Int 02", MB_OK); /* set the new interrupt handler */ _asm { push ds push dx mov ax,0x2502 push cs pop ds mov dx, offset myintfn int 21h pop dx pop ds } if( !hPrevInstance) WinExec("taskint.exe",SW_SHOW); /* fetch it back again */ _asm { push bx push es mov ax,0x3502 int 21h mov new_seg, es mov new_ofs, bx pop es pop bx } sprintf(str, "new %04x:%04x", new_seg, new_ofs ); MessageBox( 0, str, "Int 02", MB_OK); /* now set the original back again */ _asm { push ds push dx mov ax,0x2502 mov ds, old_seg mov dx, old_ofs int 21h pop dx pop ds } return 0; }

1 0

Re: patch for implementation of EMF driver SetPixel metarecord.
by Dimitrie O. Paun 08 Oct '03

08 Oct '03

On Wed, 8 Oct 2003 Dave_Belanger(a)cimmetry.com wrote: This patches are so hard to review. Not only are they not inlined, but they are attached as Application/OCTET-STREAM. This means I have to type about 8 keystrokes to open them in pine. Please, pretty please, try to inline them. If this doesn't wokr (have you tried?), at least attach them as text/plain. Thanks. -- Dimi. P.S. Inlining them is _far_ supperior to attachment. First, we can review them without any additional action. Second, we can simply press "Reply" and quote the code we want to comment on. This is very important.

1 0

Theme configuration options
by Kevin Koltzau 08 Oct '03

08 Oct '03

I've come to a point in the development of uxtheme where configuration options will be needed, and am looking for feedback on how this should be done. I currently forsee the following options needed: EnableTheme ="Y|N" ThemeMethod = "XP|GTK|KDE|..." GTK, KDE, etc are conceivable, but wont be in initial implementation, will default to XP ThemeName="name of theme" for example "Luna" if ThemeMethod=="XP" and the default XP theme is installed These should probably go into the [Tweak.Layout] section (possibly only enabled with a "WineLook"="WinXP") and of course needs the ability for application specific settings The theme API also has a function EnableTheming(BOOL) which under XP turns theming on/off for all applications immediately (all applications are redrawn), and it is persistant across reboots. I havent checked up on if I can update the EnableTheme config option from code, or if I would need to possibly store this value elsewhere in the registry Another possible option would simply be to only apply that call to the current running app, and not make it persistant..but that would deviate from how windows handles it Thanks Kevin

5 9

Byteswapping
by Warren_Baird＠cimmetry.com 08 Oct '03

08 Oct '03

Hi all, There are some changes I'd like to submit as patches, but I'm not sure the best way to go about it. We need to add some byteswapping code in a few places to make things happier on big-endian machines. The problem is that the byteswap macros only exist inside wine/tools/wrc right now - we've added a new header containing the byteswap macros in wine/include, and so they are just duplicated in wine/include and wine/tools/wrc - but this is hardly the ideal solution. The best solution I can think of at the moment is to put the byteswap macros into wine/include/wine/port.h, which is one of the few main wine header that is included in the wrc code - but then we'll have to include config.h and port.h in at least one place they aren't currently included : dlls/ole32/storage32.c Does anyone have any better suggestions? Thanks, Warren

2 1

Problem with NONAMELESSUNION on MS_VC and Mingw
by Steven Edwards 08 Oct '03

08 Oct '03

Hello, I have been having problems compiling on Mingw and MS_VC + PSDK code that uses #define NONAMELESSUNION. This is not the first place I have ran in to this problem. When I try to build libwine I am getting the following error: loader.c: In function `fixup_resources': loader.c:193: error: union has no member named `s3' loader.c:194: error: union has no member named `s3' loader.c: At top level: loader.c:170: warning: `fixup_imports' defined but not used make: *** [loader.o] Error 1 I had the same problem on MS_VC with the PSDK but I cant test/submut another report on this right now as my broadband at home on my dev box with the PSDK is still out. Can anyone else with the PSDK check? Thanks Steven __________________________________ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com

1 0