January 2003 - Wine-Devel - list.winehq.org

Re: MciWave: some small cleanups
by Francois Gouget Jan. 11, 2003

Jan. 11, 2003

On Sat, 11 Jan 2003, Eric Pouech wrote: > see changelog for the details [...] > Name: mciwave > ChangeLog: > License: X11 Actually the changelog is a bit short... -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ Linux: Because rebooting is for adding new hardware

2 1

Re: [D3D8 - 1] resend of Jason Texture work + many fixes
by Raphaël Junqueira Jan. 11, 2003

Jan. 11, 2003

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Saturday 11 January 2003 01:27, Raphaël Junqueira wrote: > Hi again, I forgot to say with this patch now: - - i have run successfully the NVIDIA Effects Browser 3.0 (http://developer.nvidia.com/view.asp?IO=NVEffectsBrowser_3) with many effects as Aniso (http://developer.nvidia.com/view.asp?IO=Anisotropic_Lighting) or membrane_lighting (http://developer.nvidia.com/view.asp?IO=membrane_lighting) in pure vertex shader software emulation - - Warcraf3 Demo begin to run - - SplinterCell Demo now complain about stencil support anyone have other Dx8 games demos for testing (simple if possible) ? > Thanks > Raphael -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.1 (GNU/Linux) iD8DBQE+H2hKp7NA3AmQTU4RAvalAJ9v746sFf+RiEGzC/Vuv1NBnprAsQCaAwdQ WiIc/yV8HbZe0iw3CwJoFuE= =He8Q -----END PGP SIGNATURE-----

2 2

C++ and winedbg: testers wanted
by Eric Pouech Jan. 11, 2003

Jan. 11, 2003

as some folks have already noticed, loading ELF modules with C++ compilation units gave a zillion of traces while parsing debug info in winedbg this patch should silence this type of output it won't however add the C++ ABI to the debugger: heavy C++ debugging will require gdb (and also DwarfII debug info; stabs is so broken in this respect) on the C++ front, what should be possible is: - every class should be seen as a regular structures - non virtual methods will be available as functions (of course, this will be seen as a regular parameter, and functions will be seen as undecorated names... this may recall to some of us the old good days of cfront ;-) anyway, as gcc has some discrepencies it its stabs output in its version, I'd like to get some feedback of people loading C++ info into winedbg I already tested gcc 3.2 and 2.96, but results from 2.95, 3.0 and 3.3 would be appreciated TIA A+ -- Eric Pouech Name: wd_stabs ChangeLog: - added experimental parsing for C++ code (but winedbg doesn't support C++ ABI) - now loading symbols for included files License: X11 GenDate: 2003/01/11 17:14:34 UTC ModifiedFiles: programs/winedbg/stabs.c AddedFiles: =================================================================== RCS file: /home/cvs/cvsroot/wine/wine/programs/winedbg/stabs.c,v retrieving revision 1.3 diff -u -u -r1.3 stabs.c --- programs/winedbg/stabs.c 29 Sep 2002 18:35:31 -0000 1.3 +++ programs/winedbg/stabs.c 11 Jan 2003 17:14:30 -0000 @@ -31,6 +31,8 @@ * available (hopefully) from http:\\sources.redhat.com\gdb\onlinedocs */ +#define WITH_CPP_STABS_PARSING 1 + #include "config.h" #include <sys/types.h> @@ -269,13 +271,13 @@ if (**x=='(') { (*x)++; /* '(' */ - filenr=strtol(*x,x,10); /* <int> */ + filenr=strtol(*x,x,10); /* <int> */ (*x)++; /* ',' */ - subnr=strtol(*x,x,10); /* <int> */ + subnr=strtol(*x,x,10); /* <int> */ (*x)++; /* ')' */ } else { filenr = 0; - subnr = strtol(*x,x,10); /* <int> */ + subnr = strtol(*x,x,10); /* <int> */ } return DEBUG_FileSubNr2StabEnum(filenr,subnr); } @@ -343,14 +345,58 @@ if (*ptd->ptr++ != ';') return -1; /* ';' */ return 0; } +#ifdef WITH_CPP_STABS_PARSING +static inline int DEBUG_PTS_ReadMethodInfo(struct ParseTypedefData* ptd, + struct datatype* dt) +{ + char* tmp; + + if (*ptd->ptr++ == ':') + { + if (!(tmp = strchr(ptd->ptr, ';'))) return -1; + ptd->ptr = tmp + 1; + + } + if (!(*ptd->ptr >= '0' && *ptd->ptr <= '9')) return -1; + ptd->ptr++; + if (!(ptd->ptr[0] >= 'A' && *ptd->ptr <= 'D')) return -1; + ptd->ptr++; + if (*ptd->ptr++ != '.') return -1; + if (*ptd->ptr != ';') + { + while (*ptd->ptr) + { + /* some GCC versions (2.96 for example) really spit out wrongly defined + * stabs... in some cases, the methods are not correctly handled + * so just swallow the garbage until either .; (end of method) + * or ;; (end of struct) are found + */ + if (ptd->ptr[0] == '.' && ptd->ptr[1] == ';') + { + ptd->ptr++; + break; + } + if (ptd->ptr[0] == ';' && ptd->ptr[1] == ';') + break; + ptd->ptr++; + } + } + + return 0; +} +#endif -static inline int DEBUG_PTS_ReadAggregate(struct ParseTypedefData* ptd, struct datatype* sdt) +static inline int DEBUG_PTS_ReadAggregate(struct ParseTypedefData* ptd, + struct datatype* sdt) { int sz, ofs; char* last; struct datatype* adt; int idx; int doadd; +#ifdef WITH_CPP_STABS_PARSING + BOOL inMethod; +#endif sz = strtol(ptd->ptr, &last, 10); if (last == ptd->ptr) return -1; @@ -366,29 +412,67 @@ while (*ptd->ptr != ';') { /* agg_name : type ',' <int:offset> ',' <int:size> */ idx = ptd->idx; + if (DEBUG_PTS_ReadID(ptd) == -1) return -1; /* Ref. TSDF R2.130 Section 7.4. When the field name is a method name * it is followed by two colons rather than one. */ - if (*ptd->ptr == ':') ptd->ptr++; + if (*ptd->ptr == ':') + { + ptd->ptr++; +#ifdef WITH_CPP_STABS_PARSING + { + char* psemic = strchr(ptd->ptr, ';'); + char* psharp = strchr(ptd->ptr, '#'); + if (!psharp || psharp > psemic) + { + struct datatype* dt = NULL; + /* FIXME: hack... this shall be a ctor */ + if (DEBUG_PTS_ReadTypedef(ptd, NULL, &dt) == -1) return -1; + if (DEBUG_PTS_ReadMethodInfo(ptd, dt) == -1) return -1; + if (*ptd->ptr++ != ';') return -1; + ptd->idx = idx; + continue; + } + } + inMethod = TRUE; + } + else + { + inMethod = FALSE; +#endif + } + if (DEBUG_PTS_ReadTypedef(ptd, NULL, &adt) == -1) return -1; if (!adt) return -1; - if (*ptd->ptr++ != ',') return -1; - if (DEBUG_PTS_ReadNum(ptd, &ofs) == -1) return -1; - if (*ptd->ptr++ != ',') return -1; - if (DEBUG_PTS_ReadNum(ptd, &sz) == -1) return -1; - if (*ptd->ptr++ != ';') return -1; + if (*ptd->ptr != ',') + { +#ifdef WITH_CPP_STABS_PARSING + if (!inMethod) +#endif + return -1; + ptd->ptr++; + } + else + { + ptd->ptr++; + if (DEBUG_PTS_ReadNum(ptd, &ofs) == -1) return -1; + if (*ptd->ptr++ != ',') return -1; + if (DEBUG_PTS_ReadNum(ptd, &sz) == -1) return -1; + if (*ptd->ptr++ != ';') return -1; - if (doadd) DEBUG_AddStructElement(sdt, ptd->buf + idx, adt, ofs, sz); + if (doadd) DEBUG_AddStructElement(sdt, ptd->buf + idx, adt, ofs, sz); + } ptd->idx = idx; } ptd->ptr++; /* ';' */ return 0; } -static inline int DEBUG_PTS_ReadEnum(struct ParseTypedefData* ptd, struct datatype* edt) +static inline int DEBUG_PTS_ReadEnum(struct ParseTypedefData* ptd, + struct datatype* edt) { int ofs; int idx; @@ -405,7 +489,8 @@ return 0; } -static inline int DEBUG_PTS_ReadArray(struct ParseTypedefData* ptd, struct datatype* adt) +static inline int DEBUG_PTS_ReadArray(struct ParseTypedefData* ptd, + struct datatype* adt) { int lo, hi; struct datatype* rdt; @@ -463,14 +548,17 @@ /* then the real definitions */ switch (*ptd->ptr++) { case '*': +#ifdef WITH_CPP_STABS_PARSING + case '&': +#endif new_dt = DEBUG_NewDataType(DT_POINTER, NULL); if (DEBUG_PTS_ReadTypedef(ptd, NULL, &ref_dt) == -1) return -1; DEBUG_SetPointerType(new_dt, ref_dt); break; - case 'k': /* 'const' modifier */ - case 'B': /* 'volatile' modifier */ - /* just kinda ignore the modifier, I guess -gmt */ - if (DEBUG_PTS_ReadTypedef(ptd, NULL, &new_dt) == -1) return -1; + case 'k': /* 'const' modifier */ + case 'B': /* 'volatile' modifier */ + /* just kinda ignore the modifier, I guess -gmt */ + if (DEBUG_PTS_ReadTypedef(ptd, NULL, &new_dt) == -1) return -1; break; case '(': ptd->ptr--; @@ -588,7 +676,8 @@ case 31: basic = DT_BASIC_LONGLONGINT; break; case 32: basic = DT_BASIC_ULONGLONGINT; break; default: - DEBUG_Printf(DBG_CHN_MESG, "Unsupported integral type (%d/%d)\n", lo, sz); + DEBUG_Printf(DBG_CHN_MESG, + "Unsupported integral type (%d/%d)\n", lo, sz); return -1; } if (!(new_dt = DEBUG_GetBasicType(basic))) { @@ -598,6 +687,35 @@ if (*ptd->ptr++ != ';') return -1; } break; +#ifdef WITH_CPP_STABS_PARSING + case '#': + new_dt = DEBUG_NewDataType(DT_FUNC, NULL); + if (*ptd->ptr == '#') + { + ptd->ptr++; + if (DEBUG_PTS_ReadTypedef(ptd, NULL, &ref_dt) == -1) return -1; + DEBUG_SetPointerType(new_dt, ref_dt); + if (*ptd->ptr++ != ';') return -1; + + } + else + { + struct datatype* cls_dt; + struct datatype* pmt_dt; + + if (DEBUG_PTS_ReadTypedef(ptd, NULL, &cls_dt) == -1) return -1; + if (*ptd->ptr++ != ',') return -1; + if (DEBUG_PTS_ReadTypedef(ptd, NULL, &ref_dt) == -1) return -1; + DEBUG_SetPointerType(new_dt, ref_dt); + while (*ptd->ptr++ == ',') + { + if (DEBUG_PTS_ReadTypedef(ptd, NULL, &pmt_dt) == -1) return -1; + } + } + if (DEBUG_PTS_ReadMethodInfo(ptd, new_dt) == -1) return -1; + + break; +#endif default: DEBUG_Printf(DBG_CHN_MESG, "Unknown type '%c'\n", ptd->ptr[-1]); return -1; @@ -647,11 +765,15 @@ if ((ptd.ptr = strchr(ptr, ':'))) { ptd.ptr++; if (*ptd.ptr != '(') ptd.ptr++; +#ifdef WITH_CPP_STABS_PARSING + /* most of type definitions take one char, except Tt */ + if (*ptd.ptr != '(') ptd.ptr++; +#endif ret = DEBUG_PTS_ReadTypedef(&ptd, typename, &dt); } if (ret == -1 || *ptd.ptr) { - DEBUG_Printf(DBG_CHN_MESG, "failure on %s at %s\n", ptr, ptd.ptr); + DEBUG_Printf(DBG_CHN_MESG, "Failure on %s at %s\n", ptr, ptd.ptr); return FALSE; } @@ -729,7 +851,7 @@ strtabinc = 0; stabbuff[0] = '\0'; - for(i=0; i < nstab; i++, stab_ptr++ ) + for (i = 0; i < nstab; i++, stab_ptr++) { ptr = strs + (unsigned int) stab_ptr->n_un.n_name; if( ptr[strlen(ptr) - 1] == '\\' ) @@ -944,7 +1066,7 @@ } #endif - if( *ptr == '\0' ) + if( *ptr == '\0' ) /* end of N_SO file */ { /* * Nuke old path. @@ -1004,7 +1127,7 @@ stabbuff[0] = '\0'; #if 0 - DEBUG_Printf(DBG_CHN_MESG, "%d %x %s\n", stab_ptr->n_type, + DEBUG_Printf(DBG_CHN_MESG, "0x%02x %x %s\n", stab_ptr->n_type, (unsigned int) stab_ptr->n_value, strs + (unsigned int) stab_ptr->n_un.n_name); #endif @@ -1044,7 +1167,7 @@ nsym = symtab->sh_size / sizeof(*symp); strp = (char *) (addr + strtab->sh_offset); - for(i=0; i < nsym; i++, symp++) + for (i = 0; i < nsym; i++, symp++) { /* * Ignore certain types of entries which really aren't of that much @@ -1165,7 +1288,7 @@ } if (stabsect == -1 || stabstrsect == -1) { - DEBUG_Printf(DBG_CHN_WARN, "no .stab section\n"); + DEBUG_Printf(DBG_CHN_WARN, "No .stab section\n"); goto leave; } @@ -1181,7 +1304,7 @@ dil = DIL_LOADED; } else { dil = DIL_ERROR; - DEBUG_Printf(DBG_CHN_WARN, "bad stabs\n"); + DEBUG_Printf(DBG_CHN_WARN, "Couldn't read correctly read stabs\n"); goto leave; } @@ -1350,7 +1473,9 @@ if (filename == NULL) return DIL_ERROR; if (DEBUG_FindModuleByName(filename, DMT_ELF)) return DIL_LOADED; +#ifndef WITH_CPP_STABS_PARSING if (strstr (filename, "libstdc++")) return DIL_ERROR; /* We know we can't do it */ +#endif dil = DEBUG_ProcessElfFile(filename, load_offset, dyn_addr); /* if relative pathname, try some absolute base dirs */ @@ -1383,6 +1508,7 @@ for (lm_addr = (u_long)dbg_hdr->r_map; lm_addr; lm_addr = (u_long)lm.l_next) { if (!DEBUG_READ_MEM_VERBOSE((void*)lm_addr, &lm, sizeof(lm))) return FALSE; + if (lm.l_addr != 0 && DEBUG_READ_MEM_VERBOSE((void*)lm.l_addr, &ehdr, sizeof(ehdr)) && ehdr.e_type == ET_DYN && /* only look at dynamic modules */

1 0

Re: patch for "GetFileTitleW: illegal call to GetFileTitleA"
by Dmitry Timoshkov Jan. 11, 2003

Jan. 11, 2003

"liu spider" <liuspider(a)yahoo.com> wrote: > this is a patch to get rid of W->A calls: > GetFileTitleW: illegal call to GetFileTitleA > + ret = GetFileTitleW(lpwstrFile, lpwstrTitle, cbBuf); > + HeapFree(GetProcessHeap(), 0, lpwstrTitle); > + HeapFree(GetProcessHeap(), 0, lpwstrFile); > + return ret; > +} It would be a good idea to translate the result from unicode to ANSI. One more suggestion: RtlCreateUnicodeStringFromAsciiz is supposed to reduce trouble of conversion ASCIIZ strings to unicode. -- Dmitry.

2 1

Some ddraw / GDI palette problem
by Lionel Ulmer Jan. 11, 2003

Jan. 11, 2003

Hi all, In 'old' D3D games, a lot of textures are created paletted... And each of these texture has its own DDraw palette object. And thus each has some GDI memory allocated. This means that after some time, we run out of GDI Heap space due to all these palettes created there : trace:ddraw:Main_DirectDraw_CreatePalette (0x4038ff48)->(00000044,0x406b1340,0x406b1978,(nil)) trace:palette:CreatePalette entries=256 err:local:LOCAL_GetBlock not enough space in GDI heap 01b7 for 1052 bytes Now that problem is fixed in one game (it was due to some bad reference counting is some D3D code fixed in patch 79), but if ever it happens in yet another game, is it possible to increase this space ? Lionel -- Lionel Ulmer - http://www.bbrox.org/

2 1

Re: Help with /?? stuff
by John K. Hohm Jan. 11, 2003

Jan. 11, 2003

> > Alberto Massari wrote: > > I guess Wine is converting the NULL into \?? to be able to store it as > > a string. Shachar Shemesh wrote: > You guessed wrong. The leading \?? is there under both Wine and Windows > 2000, and in both cases there is ALSO a NULL. As neither Wine nor > Windows 2000 then accepts that as a valid path, I'll just remove it. I > was just wondering whether anyone knows why Windows 2000 puts that extra > stuff in (it's obvious why wine does that - to be compatible). I'm not sure about \??, but see this excerpt from MSDN Re CreateFile: lpFileName [in] Pointer to a null-terminated string that specifies the name of the object to create or open. In the ANSI version of this function, the name is limited to MAX_PATH characters. To extend this limit to 32,767 wide characters, call the Unicode version of the function and prepend "\\?\" to the path. For more information, see Naming a File. Windows 95/98/Me: This string must not exceed MAX_PATH characters.

3 2

Re: PATCH - Check/Define for vsnprintf/_vsnprintf
by Dimitrie O. Paun Jan. 11, 2003

Jan. 11, 2003

On January 10, 2003 07:47 pm, Steven Edwards wrote: > Changelog: (porting fix for Mingw/Windows) > check for vsnprintf and _vsnprintf > define vsnprintf with _vsnprintf Alexandre, why don't we link with oldnames.lib in Windows, and get rid of all these tests? -- Dimi.

4 4

shlwapi.dll from Wine regression tests on Windows 2000 SP3
by Steven Edwards Jan. 11, 2003

Jan. 11, 2003

I did a compare of windows shlwapi with WINEs shlwapi. With Windows 2000 SP3 shlwapi.dll we get - C:\Documents and Settings\Administrator\Desktop\test>shlwapi_test.exe clist clist: 236 tests executed, 0 marked as todo, 0 failures. C:\Documents and Settings\Administrator\Desktop\test>shlwapi_test.exe generated generated: 25 tests executed, 0 marked as todo, 0 failures. C:\Documents and Settings\Administrator\Desktop\test>shlwapi_test.exe path path.c:92: Test failed: Expected ?query=x&return=y, but got query=x&return=y path: 33 tests executed, 0 marked as todo, 1 failure. C:\Documents and Settings\Administrator\Desktop\test>shlwapi_test.exe shreg shreg.c:171: Test failed: (44,43) shreg.c:183: Test failed: (2) shreg.c:195: Test failed: (14,43) shreg.c:196: Test failed: (2) shreg.c:204: Test failed: Expected ERROR_MORE_DATA shreg.c:205: Test failed: () shreg.c:206: Test failed: () shreg.c:207: Test failed: (1,43) shreg.c:218: Test failed: (44,43) shreg.c:262: Test failed: didn't open copy shreg: 40 tests executed, 0 marked as todo, 10 failures. C:\Documents and Settings\Administrator\Desktop\test> With the current WINE shlwapi on Windows 2000 we get C:\Documents and Settings\Administrator\Desktop\test>shlwapi_test.exe clist clist: 236 tests executed, 0 marked as todo, 0 failures. C:\Documents and Settings\Administrator\Desktop\test>shlwapi_test.exe generated generated: 25 tests executed, 0 marked as todo, 0 failures. C:\Documents and Settings\Administrator\Desktop\test>shlwapi_test.exe path fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:81: Test failed: UrlGetPartA didn't return S_OK fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:83: Test failed: UrlGetPartW didn't return S_OK path.c:87: Test failed: Strings didn't match between ascii and unicode UrlGetPar t! path.c:92: Test failed: Expected localhost, but got x☺� fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:81: Test failed: UrlGetPartA didn't return S_OK fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:83: Test failed: UrlGetPartW didn't return S_OK path.c:87: Test failed: Strings didn't match between ascii and unicode UrlGetPar t! path.c:92: Test failed: Expected 21, but got x☺� fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:81: Test failed: UrlGetPartA didn't return S_OK fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:83: Test failed: UrlGetPartW didn't return S_OK path.c:87: Test failed: Strings didn't match between ascii and unicode UrlGetPar t! path.c:92: Test failed: Expected foo, but got x☺� fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:81: Test failed: UrlGetPartA didn't return S_OK fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:83: Test failed: UrlGetPartW didn't return S_OK path.c:87: Test failed: Strings didn't match between ascii and unicode UrlGetPar t! path.c:92: Test failed: Expected bar, but got x☺� fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:81: Test failed: UrlGetPartA didn't return S_OK fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:83: Test failed: UrlGetPartW didn't return S_OK path.c:87: Test failed: Strings didn't match between ascii and unicode UrlGetPar t! path.c:92: Test failed: Expected http, but got x☺� fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:81: Test failed: UrlGetPartA didn't return S_OK fixme:shell:URL_ParseUrl failed to parse L"http://foo:bar(a)localhost:21/internal. php?query=x&return=y" path.c:83: Test failed: UrlGetPartW didn't return S_OK path.c:87: Test failed: Strings didn't match between ascii and unicode UrlGetPar t! path.c:92: Test failed: Expected ?query=x&return=y, but got x☺� path: 33 tests executed, 0 marked as todo, 24 failures. C:\Documents and Settings\Administrator\Desktop\test> C:\Documents and Settings\Administrator\Desktop\test>shlwapi_test.exe shreg shreg.c:171: Test failed: (44,43) shreg.c:195: Test failed: (44,43) shreg.c:205: Test failed: () shreg.c:206: Test failed: () shreg.c:207: Test failed: (44,43) shreg.c:218: Test failed: (44,43) shreg.c:262: Test failed: didn't open copy shreg: 37 tests executed, 0 marked as todo, 7 failures. C:\Documents and Settings\Administrator\Desktop\test> __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com

1 0

dx8 Patches - Question
by Ann and Jason Edmeades Jan. 11, 2003

Jan. 11, 2003

Just a question, I noticed some of the dx8 patches are not applied - Is it due to the Chrismas backlog where Alexander took a very well deserved break, or because they have slipped through the cracks (or are there problems with them)? Will it help to resubmit them based off the current cvs since there were about 4 chained together each based off the previous one? Jason

3 3

wine/ole error on Mingw32/MSYS
by Steven Edwards Jan. 11, 2003

Jan. 11, 2003

Platform is a older MSYS and GCC (3.1 something I think). I'm still getting settled in from getting tied-down so I cant test this on the latest mingw-current. only on dial-up here for now =(. Anyway any idea about this? Thanks Steven $ make gcc -c -I. -I. -I../include -I../include -DWINE_NOWINSOCK -Wall -mpreferred-sta ck-boundary=2 -gstabs+ -D__WINESRC__ -D_REENTRANT -o dx8guid.o dx8guid.c In file included from dx8guid.c:26: ../include/winuser.h:2396: unnamed fields of type other than struct or union are not allowed ../include/winuser.h:2409: unnamed fields of type other than struct or union are not allowed make: *** [dx8guid.o] Error 1 __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com

3 2

Re: msvcrt: stubs for chsize, dup, dup2
by Steven Edwards Jan. 11, 2003

Jan. 11, 2003

--- "Dimitrie O. Paun" <dpaun(a)rogers.com> wrote: > These are needed by the inline functions in the header. > > ChangeLog > Create stubs for chsize, dup, and dup2. I think chsize and ftruncate do the same thing so is there anyway to forward the call for chsize to ftruncate on platforms that have it? I added a configure check for this a while back because windows doesnt have ftruncate and makedep needs it. __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com

1 0

[RFC] msvcrt: defines break C++ code
by Dimitrie O. Paun Jan. 10, 2003

Jan. 10, 2003

Folks, We have a lot of code in msvcrt headers like so: #define umask _umask #define unlink _unlink #define write _write This breaks C++ that define a write method in a header, and then implement it like so: #include <io.h> void MyClass::write(...) { write(...); } I suggest we turn those defines into inlines, like this: inline int write(int fd, const void* ptr, unsigned int size) { return _write(fd, ptr, size); } Any other solutions? -- Dimi.

4 6

Windows 2000 conformance test update
by Dave Miller Jan. 10, 2003

Jan. 10, 2003

Here are three updates for the status page. I think user32/sysparams passed for me previously. C:\winetests>kernel32_test.exe path path.c:514: Test failed: GetLongPathNameA: wrong return code, 97 instead of 42 path.c:881:TMP=C:\DOCUME~1\davidm\LOCALS~1\Temp path.c:892:TMP=C:\WINNT path.c:902:TMP=C:\ path.c:912:TMP=C: path: 1730 tests executed, 0 marked as todo, 1 failure. C:\winetests>urlmon_test generated generated: 4 tests executed, 0 marked as todo, 0 failures. C:\winetests>user32_test sysparams sysparams.c:184:testing SPI_{GET,SET}BEEP sysparams.c:306:testing SPI_{GET,SET}MOUSE sysparams.c:374:testing SPI_{GET,SET}BORDER sysparams.c:170: Test failed: Wrong value in registry: subKey=Control Panel\Desktop, valName=BorderWidth, testValue=1, value= sysparams.c:170: Test failed: Wrong value in registry: subKey=Control Panel\Desktop, valName=BorderWidth, testValue=0, value= sysparams.c:139: Test failed: Missed a message: change_counter=2 sysparams.c:170: Test failed: Wrong value in registry: subKey=Control Panel\Desktop, valName=BorderWidth, testValue=7, value= sysparams.c:170: Test failed: Wrong value in registry: subKey=Control Panel\Desktop, valName=BorderWidth, testValue=20, value= sysparams.c:406:testing SPI_{GET,SET}KEYBOARDSPEED sysparams.c:139: Test failed: Missed a message: change_counter=2 sysparams.c:440:testing SPI_ICONHORIZONTALSPACING sysparams.c:170: Test failed: Wrong value in registry: subKey=Control Panel\Desktop, valName=IconSpacing, testValue=101, value= sysparams.c:170: Test failed: Wrong value in registry: subKey=Control Panel\Desktop, valName=IconSpacing, testValue=32, value= sysparams.c:480:testing SPI_{GET,SET}SCREENSAVETIMEOUT sysparams.c:515:testing SPI_{GET,SET}SCREENSAVEACTIVE sysparams.c:555:testing SPI_{GET,SET}KEYBOARDDELAY sysparams.c:590:testing SPI_ICONVERTICALSPACING sysparams.c:170: Test failed: Wrong value in registry: subKey=Control Panel\Desktop, valName=IconVerticalSpacing, testValue=101, value= sysparams.c:170: Test failed: Wrong value in registry: subKey=Control Panel\Desktop, valName=IconVerticalSpacing, testValue=32, value= sysparams.c:635:testing SPI_{GET,SET}ICONTITLEWRAP sysparams.c:170: Test failed: Wrong value in registry: subKey=Control Panel\Desktop, valName=IconTitleWrap, testValue=1, value= sysparams.c:170: Test failed: Wrong value in registry: subKey=Control Panel\Desktop, valName=IconTitleWrap, testValue=0, value= sysparams.c:670:testing SPI_{GET,SET}MENUDROPALIGNMENT sysparams.c:707:testing SPI_SETDOUBLECLKWIDTH sysparams.c:738:testing SPI_SETDOUBLECLKHEIGHT sysparams.c:770:testing SPI_SETDOUBLECLICKTIME sysparams.c:817:testing SPI_SETMOUSEBUTTONSWAP sysparams.c:880:testing SPI_{GET,SET}DRAGFULLWINDOWS sysparams.c:915:testing SPI_{GET,SET}WORKAREA sysparams.c:953:testing SPI_{GET,SET}SHOWSOUNDS sysparams.c:997:testing SPI_{GET,SET}DESKWALLPAPER sysparams: 318 tests executed, 0 marked as todo, 12 failures.

1 0

Re: wine/dlls/comcat regsvr.c comcat_main.c
by Dimitrie O. Paun Jan. 10, 2003

Jan. 10, 2003

On January 10, 2003 12:48 am, John K. Hohm wrote: > + CLSID const *clsid; /* NULL for end of list */ [...] > + { NULL } /* list terminator */ Why not terminate the list with a NULL pointer to the struct instead, it's simpler and more idiomatic... -- Dimi.

2 1

Re: Wine, Mozilla & IE ?
by Francois Gouget Jan. 10, 2003

Jan. 10, 2003

On Fri, 10 Jan 2003, Dan Kegel wrote: [...] > I don't know the particulars, but I do know that g++'s > ABI only very recently stabilized. ^^^^^^^^^^^^^^^^^^^^^^^^ Isn't that a contradiction in terms? [...] > (There was one little ABI tweak between 3.2 and 3.3, > and there may be more later, but they probably > won't break many apps.) That too seems incompatible with something declared 'stable'. Not that I have anything against C++, I like OO programming. But these g++ compatibility issues have been like a cold shower. <rant class="my opinion only"> >From the point of vue of an ISV, distributing a Wine binary is already enough trouble with the C library changing their internal implementation (which of course they are absolutely free to do since it's *internal*) and FreeType breaking backwards compatibility in minor releases, but if ISVs also have to deal with g++ changing their ABI on a regular basis... Well, I guess ISVs could always compile these libraries with Visual C++ and ship them as Windows dlls and thus benefit from a stable and Wine compatible ABI... </rant> -- Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ tcA thgirypoC muinelliM latigiD eht detaloiv tsuj evah uoY

3 4

Readfile on a duplicated pipe hangs
by Uwe Bonnes Jan. 10, 2003

Jan. 10, 2003

Hallo, Marcus sent me patches that made Altera quartus install. Thanks. Now running the package hangs in the way it communicates with created processes. Here is what happens in my understanding: M: Masterprocess C: Childprocess M: CreatePipe -> handle_read + handle_write M: DuplicateHandle handle_read -> handle_read_dup M: CloseHandle (handle_read) M: Call CreateProcessA with handle_write as hstderr M: Return Createprocess M: CloseHandle (handle_write) M: Call Readfile (handle_read_dup) C: initializes, executes and exit M: Return Readfile retval=00000001 M: Endless loop of Call Readfile (handle_read_dup) retval=00000001 MSDN says, that ReadFile should fail on an anonymous pipe, if the othe end has close. With following patch, the application starts up. Feedback welcome. Index: wine/server/pipe.c =================================================================== RCS file: /home/wine/wine/server/pipe.c,v retrieving revision 1.24 diff -u -r1.24 pipe.c --- wine/server/pipe.c 30 May 2002 20:12:58 -0000 1.24 +++ wine/server/pipe.c 10 Jan 2003 20:28:43 -0000 @@ -141,6 +141,13 @@ static int pipe_get_info( struct object *obj, struct get_file_info_reply *reply, int *flags ) { + struct pipe *pipe = (struct pipe *)obj; + assert( obj->ops == &pipe_ops ); + if (!pipe->other) + { + set_error( STATUS_PIPE_BROKEN ); + return -1; + } if (reply) { reply->type = FILE_TYPE_PIPE; -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

1 0

Re: Wine, Mozilla & IE ?
by Dan Kegel Jan. 10, 2003

Jan. 10, 2003

Dan Kegel wrote: > With gcc 3.2 and later, shared libraries > produced by g++ will have a stable ABI, > and different machines (with different versions > of gcc >= 3.2) will produce compatible shared libraries. I forgot to mention: this also means that you can mix and match shared libraries produced by *other* c++ compilers, not just gcc! See http://www.linuxbase.org/spec/refspecs/ http://www.codesourcery.com/cxx-abi/ http://www.armdevzone.com/EABI/CppABI.html http://gcc.gnu.org/gcc-3.2/c++-abi.html http://www.intel.com/software/products/compilers/clin/clinux.htm and, for a bit of solaris history wrt c++ abis, http://forte.sun.com/s1scc/articles/CC_abi/CC_abi.html -- Dan Kegel Linux User #78045 http://www.kegel.com

1 0

Re: Clean up OSS_Wave{In,Out}Init
by Eric Pouech Jan. 10, 2003

Jan. 10, 2003

Francois Gouget wrote: > > This modifies OSS_Wave{In,Out}Init to report the exact capabilities of > the OSS driver. The new code does not try to use SNDCTL_DSP_SETFMTS > since some (old) cards may support 16bit or stereo audio at some sample > rates but not other (higher) rates. It now also reports the 48kHz and > 96kHz formats if they are supported. given the susceptibility of some OSS implementations, I'd suggest reimplementing the test loop as: for fmt for channels for rates if open succeeds 1/ flag format as supported 2/ close device endif endfor endfor endfor this would also put us out of danger if the device doesn't accept changing its mode on the fly A+ -- Eric Pouech

1 0

Re: widl patch
by Greg Turner Jan. 10, 2003

Jan. 10, 2003

On Wednesday 08 January 2003 12:31 pm, Ove Kaaven wrote: > Feel free to recompile the .idl files after applying this patch... > it's not necessary, but sooner or later someone is going to try what > I'm currently trying... to compile MIDL-generated proxy/stub code > with Wine headers. Indeed, I am doing this now with my RPC test. I'm not sure we are targeting the same stuff, however. You can see my work-in-progress in the K patchseries, which only has one (rejected) "official" patch: K01. Actually, what I'm really trying to do is mix and match. I want the .h header to be generated by widl, and the _s.c and _c.c stubs to be generated by MIDL. wire_marshall should be a barrel of laughs ;) Maybe after I get this batch working I'll whip up some tests for 'ya that use it. I encounter the following gaps: o widl doesn't generate forward declarations for MIDL_user_allocate and MIDL_user_free, should be trivial to fix. o widl doesn't parse the implicit_handle stuff in the .acf, which cause MIDL to put an extern declaration in for the handle. o RpcTry/Except/etc. Ugh, what a $&^@$*% train wreck. More-or-less unfixable, and not widl's fault anyhow. ATM I may hack around this using some evil, incorrect, and, frankly, downright offensive macros. Long term, of course, I'm interested in seeing exception handling souce-compatibility in wine... perhaps, winegcc, or some cousin, could parse VC++ __try & family and generate wine-compatible __TRY & family to match? Given the freedom to parse/generate, regardless of the implementation specifics, it becomes a surmountable problem IMO. Otherwise, it works perfectly. Good work! Any chance either of the first two points will be fixed soon? This would leave me with only one inelegant hack (two if calling MIDL counts), and an otherwise tolerable build process. Thanks, -- gmt "It does not take a majority to prevail ... but rather an irate, tireless minority, keen on setting brushfires of freedom in the minds of men." --Samuel Adams, Patriot

3 5

Kaleidagraph demo
by Jos van den Oever Jan. 10, 2003

Jan. 10, 2003

Hello Wine developers, The one program holding back linux adoption in my lab is the application Kaleidgraph. Kaleidagraph is a graphing application that is avaibable for Windows and Macintosh. I've been trying to use Kaleidgraph with wine (wine-20021219) without a windows installation on the computer. The installation works pretty well: Kaleigraph is installed, only the installer does not exit cleanly but not until it asks about displaying the readme file. Running Kaleidagraph does give quite a few problems. The problems seem to arise from bugs and not missing features in wine. One bug for example is, that almost all menu entries are listed twice ("File Edit Gallery ... Windows Edit Gallery ... Windows Help"). I'd like to help in making this program run by testing with the CVS version and submitting bugs. However, I think it would be easier if someone more experienced looks at the program first and irons out the biggest problems, because they are instantly obvious when you try to run the program with wine. There is a demo version of Kaleidagraph that can be downloaded for free from www.kaleidagraph.com. Best regards, Jos

2 3

Possible bug?
by Mike Hearn Jan. 10, 2003

Jan. 10, 2003

Hi, I think this code in menu.c, BOOL WINAPI SetMenuItemInfoA(HMENU hmenu, UINT item, BOOL bypos, const MENUITEMINFOA *lpmii) { if ((lpmii->fType & (MF_HILITE|MF_POPUP)) || (lpmii->fState)) { /* QuickTime does pass invalid data into SetMenuItemInfo. * do some of the checks Windows does. */ WARN("Bad masks for type (0x%08x) or state (0x%08x)\n", lpmii->fType,lpmii->fState ); return FALSE; } .... is buggy, because it checks the type and state even when fMask doesn't indicate that these fields are being used. It was added by Marcus Meissner at LinuxTag in 2001 which is why I'm confused, the bug I'm playing with/learning how to debug with is a regression and didn't exist a few months ago, yet that code has been in Wine for some time. Could somebody more experienced than me please look over the description of the MENUITEMINFO struct: http://msdn.microsoft.com/library/default.asp?url=/library/en-us/winui/WinU… and the SetMenuItemInfo call: http://msdn.microsoft.com/library/default.asp?url=/library/en-us/winui/WinU… .. and tell me if I'm reading this situation correctly? I don't think it should be returning false like this when fMask is 6, ie MIIM_ID | MIIM_SUBMENU. It'd only take a minute, promise :) -- Mike Hearn <m.hearn(a)signal.qinetiq.com> QinetiQ - Malvern Technology Center

1 0

Re: Re: Re: html browser for wine (khtml)
by fenix＠club-internet.fr Jan. 10, 2003

Jan. 10, 2003

Hi again, ----Message d'origine---- >Sujet: Re: Re: html browser for wine (khtml) >De: Mike Hearn <m.hearn(a)signal.qinetiq.com> >A: fenix(a)club-internet.fr >Copie à: m.hearn(a)signal.dera.gov.uk, hetz(a)witch.dyndns.org, >Date: 10 Jan 2003 09:47:41 +0000 > >> As i know, khtml support most of the IE-isms (for compatibility, try to actve "IE support" in konqueror). >> And for ActiveX, it must be simply supported by a khtml plugin (as crossover plugin do) > >If only it were that simple. Rather ironically the thing that led to me Wine in the first place was that >I needed to use features of an ActiveX control that were only available under IE. Same problem here :) > Although it would also (in theory) >work in Mozilla or Konq as well, in practice they didn't support rich enough plugin interfaces for host scripting >merge (where the plugin hooks into the browsers script interpreter and partially merges them together). Hence I now >sometimes need to use IE under Linux. http://webcvs.kde.org/cgi-bin/cvsweb.cgi/kdelibs/kparts/browserextension.h?… -> LiveConnectExtension is for scripting Interface (kjs and JAVA support use it so it will be sufficient for ActiveX/ActiveScript) (I have played with it a long time ago for a mini-basic scripting support) >There are lots of IEisms that I doubt Konq replicates. ActiveX controls are sometimes a lot more complex than just >displaying some controls on a page, they can actually integrate with IE to quite an impressive extent. maybe but all ActiveX i have seen are only embeded plugins or code who: - sometimes modify html rendering (ex: embeding directx panel, ...) - do some windows calls And i don't see what you cannot do with konq plugins. >For instance: > >- DirectX filters hmm, problem here, DirectShow and DirectMedia are not supported by wine yet ( it's at end on my todo list :) ). >- DHTML Behaviours if you talk by unsupported specific IE-DHTML behaviors by kthml ... we have to see if we can easily extend khtml parser without forking kthml code (i haven't see how parser interact with plugins) Maybe asking khtml people if its possible integrating more IE-isms compat in kthml ... >- ActiveScript (needs a vbscript interpreter as well as a JS interpreter) >- ActiveX data binding >- .NET integration .NET integration is difficult because it can use SOAP and somes XMLs stupid things :( >and so on. I expect upstream khtml could get 80% of the way there, but the remaining 20% would require tight >integration with Wine and some extensive code changes. >-- >Mike Hearn <m.hearn(a)signal.qinetiq.com> >QinetiQ - Malvern Technology Center > Raphael

1 0

Re: Re: html browser for wine (khtml)
by fenix＠club-internet.fr Jan. 10, 2003

Jan. 10, 2003

Hi, ----Message d'origine---- >Sujet: Re: html browser for wine (khtml) >De: Mike Hearn <m.hearn(a)signal.qinetiq.com> >A: hetz(a)witch.dyndns.org >Copie à: dpaun(a)rogers.com, wine-devel(a)winehq.com >Date: 10 Jan 2003 09:14:54 +0000 <snip> >About duplication - I think it's probably most likely that the best course would be >to copy the KHTML source into the Wine tree, because I'd expect modifications would >be needed in order to support all the IEisms that will crop up in regular usage of >the WebBrowser control. KHTML out of the box won't, and never will support ActiveX, >IE specific markup or any of the other proprietary stuff MS threw in for fun during >the browser wars. The KDE team would rightly reject such patches from being integrated >into upstream. As i know, khtml support most of the IE-isms (for compatibility, try to actve "IE support" in konqueror). And for ActiveX, it must be simply supported by a khtml plugin (as crossover plugin do) >thanks -mike >-- >Mike Hearn <m.hearn(a)signal.qinetiq.com> >QinetiQ - Malvern Technology Center Raphael

2 1

html browser for wine (khtml)
by Hetz Ben Hamo Jan. 10, 2003

Jan. 10, 2003

Hi, Yesterday, Apple released their web browser - Safari. The license is LGPL and the code is based on KDE's KHTML rendering engine. Whats the difference? now you don't need QT, so license wise - it's kosher ;) Thanks, Hetz

5 9

Re: getting started
by Dimitrie O. Paun Jan. 10, 2003

Jan. 10, 2003

On January 10, 2003 01:14 am, Glen Kaukola wrote: > All they have is C++ code. So maybe I should go with my > original plan of compiling stuff with visual studio and seeing if it > runs under wine. Or do you have any other suggestions? Compiling stuff with visual studio is a first step. But it should be very simple to do a Makefile for the examples, how many files do you have in one example? Here is an example Makefile that you can use under mingw: (let me know if it doesn't work, I just typed it inline here, no testing) >--------------- cut here ------------------------< CC = gcc # change this to winegcc under Wine CXX = g++ # change this to wineg++ under Wine RC = windres # change this to wrc under Wine # There six lines will change between examples TARGET = sample1.exe #name of the executable SRCS = file1.cpp file2.cpp ... RSRC = res1.rc res2.rc ... # If you need to add defines, add them to CPPFLAGS # CPPFLAGS, INCDIRS, LIBDIRS, and LIBS may change between examples CPPFLAGS = # add stuff here like -D_WIN32_IE=0x0400 CFLAGS = -mno-cygwin -W -Wall -g -gstabs+ -fno-exceptions -fno-rtti LDFLAGS = -mno-cygwin -mwindows INCDIRS = # add stuff there like -I../include LIBDIRS = # add stuff like -L../lib LIBS = -lcomctl32 # we'll need to add more libs in here for sure OBJ = $(SRCS:.c=.o) $(RSRC:.c=.o) %.o : %.rc $(RC) $(CPPFLAGS) $< $@ %.o : %.cpp $(CXX) $(CFLAGS) $(CPPFLAGS) $(INCDIRS) -c -o $@ $< %.o : %.c $(CC) $(CFLAGS) $(CPPFLAGS) $(INCDIRS) -c -o $@ $< $(TARGET): $(OBJS) $(CXX) $(LDFLAGS) $(LIBDIRS) $(LIBS) -o $@ $(OBJS) # A few nice targets all: $(TARGET) clean: rm -f $(TARGET) $(OBJS) >--------------- cut here ------------------------< It's not hard, just look in the .dsp project to see what you need to put in these variables. -- Dimi.

2 1