November 2006 - Wine-Devel - list.winehq.org

Bugs with new 'installer' keyword
by Dan Kegel Nov. 3, 2006

Nov. 3, 2006

I think the new 'installer' keyword is on every bug that needs it now. http://bugs.winehq.org/buglist.cgi?product=Wine&keywords_type=allwords&keyw… shows 94 downloadable apps with keyword 'installer'. Here's a breakdown by component: 24 wine-misc 17 wine-ole 15 wine-msi 7 wine-binary 6 wine-net 4 wine-x11 4 wine-msxml 3 wine-programs 3 wine-kernel 2 wine-user 2 wine-setupapi 2 wine-richedit 1 wine-shdocvw 1 Wine-Rebar 1 wine-msvcrt 1 wine-gdi-(printing) 1 wine-directx-d3d James is taking on the msi and setupapi ones, I think. Anyone else interested in plowing through these to see if we can resolve all the installer bugs before 1.0? - Dan

4 6

make test for 06/11/03 Wine
by Jozef Stefanka Nov. 3, 2006

Nov. 3, 2006

Hi ! I tested "make test" with today's (06/11/03) Wine checkouted from winehq.org CVS Testing system : Open SuSE 10.1 (i586), kernel : 2.6.16.13-4-default, glibc : 2.4-28 Failed tests : comctl32, kernel32, mshtml, user32 Logs are attached. Jozef

1 0

Re: ntdll: Add a stub implementation for RtlDllShutdownInProgress.
by Robert Shearman Nov. 3, 2006

Nov. 3, 2006

Mike McCormack wrote: > +NTSTATUS WINAPI RtlDllShutdownInProgress(void) > +{ > + FIXME("\n"); > + return STATUS_SUCCESS; > +} For undocumented functions like this, you should add a test case. -- Rob Shearman

2 3

mscoree: change WCHAR version[ ] in GetCORVersion to prevent a crash
by Louis. Lenders Nov. 3, 2006

Nov. 3, 2006

> >This will break the Acid Pro Media installer. It specifically checks >for the first value. I'm guessing your installer specifically checks >for the second value. I don't have a proper solution, just letting >you know that another installer will break. -- >James Hawkins Looks like it needs more testing then. will have another look later. Regards --------------------------------- Try the all-new Yahoo! Mail . "The New Version is radically easier to use" The Wall Street Journal

1 0

ntdll: Add a stub implementation for RtlDllShutdownInProgress
by Louis. Lenders Nov. 3, 2006

Nov. 3, 2006

a little test shows it returns some unknown status on my xp lap[top: 0x7c927f00, so likely not STATUS_SUCCESS. diff --git a/dlls/ntdll/tests/rtl.c b/dlls/ntdll/tests/rtl.c index 59871de..588e33e 100644 --- a/dlls/ntdll/tests/rtl.c +++ b/dlls/ntdll/tests/rtl.c @@ -66,6 +66,7 @@ static RTL_HANDLE * (WINAPI * pRtlAlloca static BOOLEAN (WINAPI * pRtlFreeHandle)(RTL_HANDLE_TABLE *, RTL_HANDLE *); static NTSTATUS (WINAPI *pRtlAllocateAndInitializeSid)(PSID_IDENTIFIER_AUTHORITY,BYTE,DWORD,DWORD,DWORD,DWORD,DWORD,DWORD,DWORD,DWORD,PSID*); static NTSTATUS (WINAPI *pRtlFreeSid)(PSID); +static NTSTATUS (WINAPI *pRtlDllShutdownInProgress)(void); #define LEN 16 static const char* src_src = "This is a test!"; /* 16 bytes long, incl NUL */ static ULONG src_aligned_block[4]; @@ -97,6 +98,7 @@ static void InitFunctionPtrs(void) pRtlFreeHandle = (void *)GetProcAddress(hntdll, "RtlFreeHandle"); pRtlAllocateAndInitializeSid = (void *)GetProcAddress(hntdll, "RtlAllocateAndInitializeSid"); pRtlFreeSid = (void *)GetProcAddress(hntdll, "RtlFreeSid"); + pRtlDllShutdownInProgress = (void *)GetProcAddress(hntdll, "RtlDllShutdownInProgress"); } strcpy((char*)src_aligned_block, src_src); ok(strlen(src) == 15, "Source must be 16 bytes long!\n"); @@ -848,6 +850,13 @@ static void test_RtlComputeCrc32(void) ok(crc == 0x40861dc2,"Expected 0x40861dc2, got %8x\n", crc); } +static void test_RtlDllShutdownInProgress(void) +{ + DWORD ret; + + ret = pRtlDllShutdownInProgress(); + ok(0x7c927f00 == ret,"Expected 0x7c927f00, got %08x\n", ret); +} typedef struct MY_HANDLE { @@ -933,4 +942,6 @@ START_TEST(rtl) test_HandleTables(); if (pRtlAllocateAndInitializeSid) test_RtlAllocateAndInitializeSid(); + if (pRtlDllShutdownInProgress) + test_RtlDllShutdownInProgress(); } Send instant messages to your online friends http://uk.messenger.yahoo.com

1 0

Re: configure.ac: Turn off the stack protector for the preloader. (resend)
by Alexandre Julliard Nov. 3, 2006

Nov. 3, 2006

Mike McCormack <mike(a)codeweavers.com> writes: > --- > configure.ac | 14 ++++++++++++++ > loader/Makefile.in | 1 + > 2 files changed, 15 insertions(+), 0 deletions(-) It's better to fix the preloader to setup the proper environment, like we already do for some other variants of that problem. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: shlexec: buffer work
by Alexandre Julliard Nov. 3, 2006

Nov. 3, 2006

Aric Stewart <aric(a)codeweavers.com> writes: > + len = ExpandEnvironmentStringsW(sei_tmp.lpFile, NULL, 0); > + if (len>0) > + { > + LPWSTR buf; > + buf = HeapAlloc(GetProcessHeap(),0,(len+1)*sizeof(WCHAR)); > + > + ExpandEnvironmentStringsW(sei_tmp.lpFile, buf, len+1); > + HeapFree(GetProcessHeap(), 0, wszApplicationName); > + dwApplicationNameLen = len+1; > + wszApplicationName = buf; > + > + sei_tmp.lpFile = wszApplicationName; > + } The indentation is all messed up, please fix your editor settings. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: winhlp32 implementation
by Kirill K. Smirnov Nov. 2, 2006

Nov. 2, 2006

Eric, there is a mess in winhelp and winhlp32 in "Context" and "Topic" concept: what winhelp calls "topic", winhlp32 calls "context", and what winhelp calls "Context" winhlp32 calls "Topic without CNT section". So, the buttons are messed too. I suspect there are other "features". We couldn't implement winhlp32 and winhelp in one box. I suggest to at least get rid of old winhelp interface in favor of winhlp32, and create symlink winhelp->winhlp32 in windows directory. Does it save the day? >IMO, we don't need two apps but we could make most apps believe both >winhelp & winhlp32 are present and put all effort into a single application >we should also reimplement winhelp based on richedit for the rendering >A+ richedit sound very good - select/copy very easy! (but riched32 is not complete yet) -- Kirill K. Smirnov

3 3

Re: comctl32: Don't free user32 objects with a gdi32 function in imagelist test.
by Alexandre Julliard Nov. 2, 2006

Nov. 2, 2006

Robert Shearman <rob(a)codeweavers.com> writes: > Use DestroyIcon for releasing the resources associated with created icons. That's probably correct, but it causes the test to fail: ../../../tools/runtest -q -P wine -M comctl32.dll -T ../../.. -p comctl32_test.exe.so imagelist.c && touch imagelist.ok imagelist.c:344: Test failed: icon 1 wasn't deleted imagelist.c:345: Test failed: icon 2 wasn't deleted imagelist.c:346: Test failed: icon 3 wasn't deleted imagelist.c:380: Test failed: icon 1 wasn't deleted imagelist.c:381: Test failed: icon 2 wasn't deleted imagelist.c:382: Test failed: icon 3 wasn't deleted make[2]: *** [imagelist.ok] Error 6 -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: comctl32: toolbar[7/8]: use the width of the current button instead of nButtonWidth to position the icon (fixes bug #3459)
by James Hawkins Nov. 1, 2006

Nov. 1, 2006

On 11/1/06, Mikołaj Zalewski <mikolaj(a)zalewski.pl> wrote: > BTNS_AUTOSIZE buttons can have a non-standard width. > You forgot the patch. -- James Hawkins

1 0