April 2005 - Wine-Devel - list.winehq.org

Hexdumping function?
by Benjamin Cutler April 16, 2005

April 16, 2005

I'm writing a conformance test for my implementation of powrprof.dll that I've been working on, and something I'd like to do in the test is dump the contents of some structs to the screen without trying to make it pretty. I cobbled together the following function, but I'd like to know if there's something like this already so I don't reinvent the wheel: void hexdump(void *p, int len) { int i; long *src = (long *)p; len /= 4; for(i = 0; i < len; i++) { printf("%08lX", src[i]); } }

2 1

Exception in wcstombs_sbcs
by Paul Vriens April 16, 2005

April 16, 2005

Hi, when I run the latest Process Explorer (Version 9.0.3) I sometimes get an exception: fixme:ntdll:NtQuerySymbolicLinkObject (0x777e2ce4,0x777ebc28,0x12e4f878) stub wine: Unhandled exception (thread 000b), starting debugger... WineDbg starting on pid 0xa Unhandled exception: page fault on read access to 0x12e5001a in 32-bit code (0x1014df10). In 32 bit mode. Register dump: CS:0073 SS:007b DS:007b ES:007b FS:003b GS:0033 EIP:1014df10 ESP:12e4f7bc EBP:12e4f7d0 EFLAGS:00210202( - 00 - -RI1) EAX:00000000 EBX:1023d978 ECX:77e2dae0 EDX:101fcb40 ESI:12e4fffc EDI:101fdd40 Stack dump: 0x12e4f7bc: 00001bb0 101fcb40 777e2ce4 00001bb0 0x12e4f7cc: 00000000 12e4f804 777c802e 1023d3fc 0x12e4f7dc: 12e4f8fc 00001830 77e2d760 00001bb0 0x12e4f7ec: 00001bb0 00000000 00000000 777e2ce4 0x12e4f7fc: 12e4f87c 12e4f8f0 12e4f834 777c7e5b 0x12e4f80c: 77e2d760 00001bb0 00000000 12e4f8fc Backtrace: =>1 0x1014df10 wcstombs_sbcs(table=0x1023d3fc, src=0x12e4f8fc, srclen=0x1830, dst=0x77e2d760, dstlen=0x1bb0) [wctomb.c:161] in libwine_unicode.so.1 (0x12e4f7d0) 2 0x777c802e RtlUnicodeToMultiByteN(dst=0x77e2d760, dstlen=0x1bb0, reslen=0x0, src=0x12e4f8fc, srclen=0x3760) [/data/install/linux/wine-src/wine/dlls/ntdll/rtlstr.c:800] in ntdll (0x12e4f804) 3 0x777c7e5b RtlUnicodeStringToAnsiString(ansi=0x12e4f87c, uni=0x12e4f8f0, doalloc=0x1) [/data/install/linux/wine-src/wine/dlls/ntdll/rtlstr.c:696] in ntdll (0x12e4f834) 4 0x004195ff in procexp (+0x195ff) (0x00000043) 5 0x00000000 (0x00000000) 0x1014df10 wcstombs_sbcs+0x98 [wctomb.c:161] in libwine_unicode.so.1: movw 0x1e(%esi),%dx The information above is from an older (Mar-01-05) version of wine but CVS has the same problem. Is there a way we can trace the stuff in libs/unicode ? This exception occurs not every time (I guess 3 out of 5 runs) and the only similarity is the exception is in wcstombs_sbcs. Cheers, Paul.

2 1

Re: Is this an acceptable fix? If so I'll fix the rest of the file.
by Dmitry Timoshkov April 16, 2005

April 16, 2005

"Jakob Eriksson" <jakov(a)vmlinux.org> wrote: > @@ -1109,7 +1109,7 @@ > bufdesc.dwSize=sizeof(bufdesc); > bufdesc.dwFlags=DSBCAPS_PRIMARYBUFFER|DSBCAPS_CTRL3D; > rc=IDirectSound_CreateSoundBuffer(dso,&bufdesc,&primary,NULL); > - ok(rc==DS_OK && primary!=NULL,"IDirectSound_CreateSoundBuffer() failed " > + ok(rc==E_INVALIDARG || (rc==DS_OK && primary!=NULL),"IDirectSound_CreateSoundBuffer() failed " > "to create a 3D primary buffer: %s\n",DXGetErrorString8(rc)); I'm afraid you need to find what exactly IDirectSound_CreateSoundBuffer doesn't like in the passed arguments. This might be DSBCAPS_CTRL3D flag. -- Dmitry.

1 0

Re: advapi32 registry test, fail gracefully when no permissions
by James Hawkins April 16, 2005

April 16, 2005

On 4/14/05, Jakob Eriksson <jakov(a)vmlinux.org> wrote: > > ERROR_PRIVILEGE_NOT_HELD returned for NT4 here: > > http://test.winehq.org/data/200504141000/nt4_IDWASEMPTY/advapi32:registry.t… > > > Index: dlls/advapi32/tests/registry.c > =================================================================== > RCS file: /home/wine/wine/dlls/advapi32/tests/registry.c,v > retrieving revision 1.18 > diff -u -r1.18 registry.c > --- dlls/advapi32/tests/registry.c 13 Apr 2005 14:40:58 -0000 1.18 > +++ dlls/advapi32/tests/registry.c 14 Apr 2005 18:35:54 -0000 > @@ -357,6 +357,7 @@ > DWORD ret; > > ret = RegSaveKey(hkey_main, "saved_key", NULL); > + if (ERROR_PRIVILEGE_NOT_HELD == ret) return; > ok(ret == ERROR_SUCCESS, "expected ERROR_SUCCESS, got %ld\n", ret); > } > > @@ -366,6 +367,7 @@ > HKEY hkHandle; > > ret = RegLoadKey(HKEY_LOCAL_MACHINE, "Test", "saved_key"); > + if (ERROR_PRIVILEGE_NOT_HELD == ret) return; > ok(ret == ERROR_SUCCESS, "expected ERROR_SUCCESS, got %ld\n", ret); > > ret = RegOpenKey(HKEY_LOCAL_MACHINE, "Test", &hkHandle); I have a better fix for this to send in. As the test stands now, it doesn't pass on any windows machines because you have to manually set the backup and restore privileges. The patch I have sets these permissions then unsets them after the tests are run. -- James Hawkins

1 0

winiet/ftp.c question about a strncpy elimination
by Peter Berg Larsen April 15, 2005

April 15, 2005

What is the correct behaviour in FtpGetCurrentDirectory if the buffer parameter is to small? http://msdn.microsoft.com/library/default.asp?url=/library/en-us/wininet/wi… Index: dlls/wininet/ftp.c =================================================================== RCS file: /home/wine/wine/dlls/wininet/ftp.c,v retrieving revision 1.53 diff -u -r1.53 ftp.c --- dlls/wininet/ftp.c 28 Mar 2005 14:58:52 -0000 1.53 +++ dlls/wininet/ftp.c 15 Apr 2005 20:15:04 -0000 @@ -813,6 +813,13 @@ * FTP_FtpGetCurrentDirectoryA (Internal) * * Retrieves the current directory + * + * Parameters + * + * lpwfs [I] Handle + * lpszCurrentDirectory [O] The absolute current path with '\0' termination. + * lpwdCurrentDirectory [I] The size of lpszCurrentDirectory. + * [O] The length of the absolute current path in lpszCurrentDirectory. * * RETURNS * TRUE on success @@ -864,11 +871,19 @@ } len = lastpos - firstpos - 1; - strncpyW(lpszCurrentDirectory, &lpszResponseBuffer[firstpos+1], - len < *lpdwCurrentDirectory ? len : *lpdwCurrentDirectory); - HeapFree(GetProcessHeap(), 0, lpszResponseBuffer); - *lpdwCurrentDirectory = len; - bSuccess = TRUE; + if (len >= *lpdwCurrentDirectory) + { + *lpdwCurrentDirectory = len + 1; + FTP_SetResponseError(ERROR_INTERNET_BAD_OPTION_LENGTH); + } + else + { + *lpdwCurrentDirectory = len; + memcpy(lpszCurrentDirectory, &lpszResponseBuffer[firstpos+1], len*sizeof(WCHAR)); + lpszCurrentDirectory[len] = '\0'; + bSuccess = TRUE; + } + HeapFree(GetProcessHeap(), 0, lpszResponseBuffer); } else FTP_SetResponseError(nResCode);

1 0

Re: TrashCan Day 3 - Use FreeDesktop TrashCan
by Steven Edwards April 15, 2005

April 15, 2005

Hi, --- Diego 'Flameeyes' Petten� <flameeyes(a)users.berlios.de> wrote: > Well actually KDE 3.4 uses the trash specs. Was the ones before which doesn't > use it. Ok great. I only have 3.3 on my devel system. I will update and check it out. Thanks Steven __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/

1 0

TrashCan Day 3 - Use FreeDesktop TrashCan
by Steven Edwards April 15, 2005

April 15, 2005

Hi, Yes I know this does not look that nice atm. Its still a work in progress but you get the idea of where I am going with it. With this patch any SHFileOperation with the FOF_ALLOWUNDO flag will move files to the users ~/.Trash. There is a draft spec on www.freedesktop.org on how the Trash is supposed to work but currently KDE does something totally different and GNOME is not compliant so I am not sure how I am going to deal with that yet. Thanks Steven __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/

2 1

Re: USB status ...
by Hiji April 15, 2005

April 15, 2005

--- Lars Segerlund <lars.segerlund(a)comsys.se> wrote: > > What is the status of wines USB support ? > > I have an IDE for a developement board and I really > would like to flash it from > Linux, ( I don't have windows), and this uses USB > to talk to the board. > Is there anything done yet or am I on my own ? > > Can someone also giva a hint on the feasibiliyt, > complexity of managing a USB > implementation for wine ? I am rather good at linux > device drivers and such. > > / regards, Lars Segerlund. There's not much I can say in the way of development, but as a user, getting USB support (beyond what's supported in the kernel) would be a HUGE leap of improvement for Wine. I think others have said before that there are a lot of apps out there that will run on Wine, but are useless because they depend on their hardware counterpart. (In the music world, Guitarport comes to mind.) Hiji __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

1 0

script for quickly finding large functions
by Joris Huizer April 15, 2005

April 15, 2005

Hello, I made this script to find large functions; I think it may be usefull to add at the janitorial page where the note on large functions was placed; It assumes no things are between functions, so it gives some incorrect matches; also, it accidentilly reads some if() and switch() statements, but in some (most) cases those are actually in large functions if they show up high. I hope this is usefull, it did find some interesting large functions like HRESULT WINAPI IDirectMusicContainerImpl_IPersistStream_Load (dlls/dmloader/container.c) which has like 7 levels of nested switch statements regards, Joris BEGIN {line_no = 0; file="";} { if ($1 != file) { if (line_no != 0 && index(file,".c") != 0) { printf("?: %s %s\n",file,var); } line_no = 0; file = $1 } if (line_no != 0 && index(file,".c") != 0) { printf("%6d: %s %s\n",($2 - line_no),file,var); } line_no = $2 var = ""; for (i = 3; i <= NF; i++) { var = sprintf("%s%s ",var,$i); } }

1 0

Re: Updown: keep track of WS_DISABLED
by Peter Berg Larsen April 15, 2005

April 15, 2005

diff -u -p -r1.66 updown.c --- dlls/comctl32/updown.c 25 Mar 2005 10:27:11 -0000 1.66 +++ dlls/comctl32/updown.c 14 Apr 2005 12:44:29 -0000 @@ -775,6 +775,8 @@ static LRESULT WINAPI UpDownWindowProc(H break; case WM_ENABLE: + infoPtr->dwStyle &= ~WS_DISABLED; + infoPtr->dwStyle |= (wParam ? 0 : WS_DISABLED); if (infoPtr->dwStyle & WS_DISABLED) UPDOWN_CancelMode (infoPtr); InvalidateRect (infoPtr->Self, NULL, FALSE); break; I had to look twice at this - I think it would added cleaner with: if (wParam) infoPtr->dwStyle &= ~WS_DISABLED; else infoPtr->dwStyle |= WS_DISABLED; or less clean infoPtr->dwStyle &= ~WS_DISABLED; if (!wParam) infoPtr->dwStyle |= WS_DISABLED; or looking at the next diff line: if (wParam) infoPtr->dwStyle &= ~WS_DISABLED; else { infoPtr->dwStyle |= WS_DISABLED; UPDOWN_CancelMode(infoPtr); } Peter

1 0