December 2010 - Wine-Devel - list.winehq.org

wine's configure says "gst/gstpad.h usability... No"
by wylda＠volny.cz Dec. 12, 2010

Dec. 12, 2010

Hi guys, i'm overworked so wanted to relax a bit with wine :-D and noticed that Debian Squeeze says: gstreamer-0.10 development files not found, gstreamer support disabled but "dpkg -l | grep -i gst": ii libgstreamer-plugins-base0.10-0 0.10.30-1 ii libgstreamer0.10-0 0.10.30-1 ii libgstreamer0.10-0-dbg 0.10.30-1 ii libgstreamer0.10-dev 0.10.30-1 also "find / -iname gstpad.h": /usr/include/gstreamer-0.10/gst/gstpad.h Thanks for any tips to enjoy your new release. Adding config.log if that helps somehow. W.

5 5

Re: shdocvw: Fix testfailure on some NT4 machines
by Marvin Dec. 12, 2010

Dec. 12, 2010

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=7562 Your paranoid android. === WVISTAADM (32 bit shdocvw) === Failure running script in VM: The specified guest user must be logged in interactively to perform this operation

1 0

Re: [PATCH 5/6] mshtml: Added IHTMLWindow2::open test.
by Paul Vriens Dec. 12, 2010

Dec. 12, 2010

Hi Jacek, On 06/14/2010 02:20 AM, Jacek Caban wrote: > --- > dlls/mshtml/tests/htmldoc.c | 100 > ++++++++++++++++++++++++++++++++++++++++++- > 1 files changed, 98 insertions(+), 2 deletions(-) > Some of these tests fail on IE6 SP1: http://test.winehq.org/data/tests/mshtml:htmldoc.html What would be the best approach to fix these? We can add a broken() for the 3577 todo_wine 3578 ok(new_window == NULL, "new_window != NULL\n"); but what about the 3582 todo_wine 3583 CHECK_CALLED(EvaluateNewWindow); introduce a CHECK_CALLED_BROKEN also in this file? -- Cheers, Paul.

2 1

Please ban spammer from the forum
by Rosanne DiMesio Dec. 12, 2010

Dec. 12, 2010

Dan or Jeremy-- Please ban user hary from the forum. Right now he's posting spam as fast as I'm deleting it. -- Rosanne DiMesio <dimesio(a)earthlink.net>

1 0

Re: Wine: Use GCC's -Wlogical-op if supported.
by Alexandre Julliard Dec. 11, 2010

Dec. 11, 2010

Gerald Pfeifer <gerald(a)pfeifer.com> writes: > In the past this has found a dozen or two real issues and lead to some > simplifications. Now the tree is clean in that regard, so we can make > this a default. I still see a ton of warnings here (gcc 4.4.5), mostly because of the glibc string macros. -- Alexandre Julliard julliard(a)winehq.org

2 3

Re: ntdll: Change NtQuerySystemInformation to return 100ns ticks for SYSTEM_PROCESSOR_PERFORMANCE_INFORMATION info class (try 3)
by Juan Lang Dec. 11, 2010

Dec. 11, 2010

Hi Louis, +#define USER_HZ_TO_TICKSPERSEC(x) (ULONGLONG)10000000 * x / user_hz It's poor style to assume in a macro that user_hz exists. Pass it as a macro parameter instead. Also, this is a question rather than a suggestion for a fix: + long user_hz = sysconf(_SC_CLK_TCK); Can this value change during the lifetime of a program? I expect not, but I could be mistaken. If not, why not make it static and initialize it only once? --Juan

1 0

wine-1.3.8 regression for the curses backend to wineconsole
by Alan W. Irwin Dec. 11, 2010

Dec. 11, 2010

The command wineconsole --backend=curses cmd hangs with the following error message for 1.3.8: err:ntdll:RtlpWaitForCriticalSection section 0x7ee17dec "?" wait timed out in thread 0065, blocked by 0000, retrying (60 sec) err:ntdll:RtlpWaitForCriticalSection section 0x7ee17dec "?" wait timed out in thread 004d, blocked by 0000, retrying (60 sec) There are no such issues for the default wineconsole backend, but I have always preferred using the curses backend because cut and paste was available for it. That backend worked fine (according to my notes) for wine-1.1.41. I built wine-1.3.8 on my Debian testing box a week ago, but saved the output of all build steps. The configure stage had the following curses-related messages: software(a)raven> grep curses configure.out checking curses.h usability... yes checking curses.h presence... yes checking for curses.h... yes checking ncurses.h usability... yes checking ncurses.h presence... yes checking for ncurses.h... yes checking for -lncurses... libncurses.so.5 These seem fine, and all curses-related output of the "make depend", "make", and "make install" steps seem fine as well. I am generally pretty happy with wine-1.3.8 except for this one curses backend regression. Can others confirm this issue? Is there some workaround such as using a different version of ncurses? Here are the curses-related packages I have installed at the moment on Debian testing: irwin(a)raven> dpkg --list |grep curse ii lib32ncurses5 5.7+20100313-4 shared libraries for terminal handling (32-bit) ii lib32ncurses5-dev 5.7+20100313-4 developer's libraries for ncurses (32-bit) ii libncurses5 5.7+20100313-4 shared libraries for terminal handling ii libncurses5-dev 5.7+20100313-4 developer's libraries and docs for ncurses ii libncursesw5 5.7+20100313-4 shared libraries for terminal handling (wide character support) ii ncurses-base 5.7+20100313-4 basic terminal type definitions ii ncurses-bin 5.7+20100313-4 terminal-related programs and man pages Alan __________________________ Alan W. Irwin Astronomical research affiliation with Department of Physics and Astronomy, University of Victoria (astrowww.phys.uvic.ca). Programming affiliations with the FreeEOS equation-of-state implementation for stellar interiors (freeeos.sf.net) PLplot scientific plotting software package (plplot.org) the libLASi project (unifont.org/lasi) the Loads of Linux Links project (loll.sf.net) and the Linux Brochure Project (lbproject.sf.net) __________________________ Linux-powered Science __________________________

3 4

Re: [PATCH] ntdll: Handle executable file mappings from noexec filesystems
by Alexandre Julliard Dec. 11, 2010

Dec. 11, 2010

Marcus Meissner <meissner(a)suse.de> writes: > Hi, > > regression from my previous patch, which forced EXEC permissions on > some mappings, which in turn gets EPERM when the filesystem is > "noexec" mounted. I don't think that's a good idea. If the filesystem is no exec we should report it and fail. -- Alexandre Julliard julliard(a)winehq.org

4 6

Re: msi/tests: Fix some test failures on Wow64.
by Marvin Dec. 11, 2010

Dec. 11, 2010

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=7542 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 07/10] mshtml: Added QueryService(SID_SBindHost) implementation
by Marvin Dec. 10, 2010

Dec. 10, 2010

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=7538 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

winedbg --gdb shows multiple segfaults during any app startup
by Ilya Basin Dec. 10, 2010

Dec. 10, 2010

In both 1.3.8 and git winedbg --gdb or plain gdb catches several segfaults, but if I ignore them, the program works. $ CC="ccache gcc" ./configure CFLAGS="-g -O0" make ./wine winedbg.exe --gdb notepad Wine-gdb> cont Continuing. Program received signal SIGSEGV, Segmentation fault. 0x7e942cf4 in create_alpha_bitmap (color=0x33c, mask=0x338, src_info=0x1319a8, color_bits=0x7ebf411c) at cursoricon.c:664 664 ptr[0] = ptr[0] * alpha / 255; Wine-gdb> Continuing. ... Program exited normally. Wine-gdb> quit

2 1

None
by paulo lesgaz Dec. 10, 2010

Dec. 10, 2010

http://webpage.lv/page.php

1 0

Re: ntdll: Change NtQuerySystemInformation to return 100ns ticks for SYSTEM_PROCESSOR_PERFORMANCE_INFORMATION info class (try 2)
by Alexandre Julliard Dec. 10, 2010

Dec. 10, 2010

Louis Lenders <xerox_xerox2000(a)yahoo.co.uk> writes: > diff --git a/dlls/ntdll/nt.c b/dlls/ntdll/nt.c > index 7ad56e4..d9f0398 100644 > --- a/dlls/ntdll/nt.c > +++ b/dlls/ntdll/nt.c > @@ -64,6 +64,8 @@ > #include <mach/vm_map.h> > #endif > > +#define USER_HZ_TO_TICKSPERSEC(x) (ULONGLONG)10000000 * x / sysconf(_SC_CLK_TCK) Don't hide sysconf in a macro, call it just once where needed. -- Alexandre Julliard julliard(a)winehq.org

1 0

Cross-DLL toolbar bitmaps
by Erich Hoover Dec. 10, 2010

Dec. 10, 2010

I was looking into adding better bitmaps for the toolbar controls in hhctrl.ocx (now that we have fancy alpha transparency support) and I wanted to make sure that the bitmaps used matched everywhere else in Wine. Looking into how this works currently, I noticed that keeping all the toolbar bitmaps in sync is currently done manually. For example, dlls/shell32/idb_tb_large.bmp uses older icons for backward and forward where dlls/shdocvw/ietoolbar.bmp and dlls/comctl32/idb_std_large.bmp use the new icons. So, my thought was that it might be desirable to split apart the individual icons, move them all to one centralized location, and then institute a makefile rule for chaining the icons together into toolbar bitmaps. Any other thoughts? Does this sound reasonable and worth pursuing? Erich Hoover ehoover(a)mines.edu

2 3

64-bit down to 11 failures !!
by Austin English Dec. 10, 2010

Dec. 10, 2010

2 2

Re: riched20/tests: Mark behaviour at 120 dpi as broken
by Marvin Dec. 9, 2010

Dec. 9, 2010

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=7522 Your paranoid android. === W7PROX64 (32 bit editor) === editor.c:4702: Test failed: test paste: strcmp = 1, text='testing paste editor.c:4720: Test failed: test paste: strcmp = 1 editor.c:4732: Test failed: test paste: strcmp = -1 editor.c:4762: Test failed: test paste: strcmp = -1, actual = 'Ts' editor.c:5977: Test failed: Cursor is at 0 instead of 4 editor.c:5983: Test failed: Cursor is at 0 instead of 9 editor.c:5989: Test failed: Cursor is at 0 instead of 4 editor.c:6003: Test failed: Cursor is at 8 instead of 9 editor.c:6011: Test failed: Cursor is at 11 instead of 9 editor.c:6022: Test failed: Cursor is at 0 instead of 8 editor.c:6047: Test failed: Cursor is at 0 instead of 8 === W7PROX64 (64 bit editor) === editor.c:4702: Test failed: test paste: strcmp = 1, text='testing paste editor.c:4720: Test failed: test paste: strcmp = 1 editor.c:4732: Test failed: test paste: strcmp = -1 editor.c:4762: Test failed: test paste: strcmp = -1, actual = 'Ts' editor.c:5977: Test failed: Cursor is at 0 instead of 4 editor.c:5983: Test failed: Cursor is at 0 instead of 9 editor.c:5989: Test failed: Cursor is at 0 instead of 4 editor.c:6003: Test failed: Cursor is at 8 instead of 9 editor.c:6011: Test failed: Cursor is at 11 instead of 9 editor.c:6022: Test failed: Cursor is at 0 instead of 8 editor.c:6047: Test failed: Cursor is at 0 instead of 8

1 0

Re: riched20/tests: Mark behaviour at 120 dpi as broken
by James Mckenzie Dec. 9, 2010

Dec. 9, 2010

Andre: My computer is set up for 125 dpi (Large Fonts on WindowsXP). Should we be testing for this as well? This should give different results than what is stated here. Maybe there is a better way to test this than using a fixed value of 39 in the ok comparision. James McKenzie -----Original Message----- >From: André Hentschel <nerv(a)dawncrow.de> >Sent: Dec 9, 2010 11:38 AM >To: wine-patches <wine-patches(a)winehq.org> >Subject: riched20/tests: Mark behaviour at 120 dpi as broken > >comparison to 96 dpi can be found at http://test.winehq.org/data/23906816d87795c363f0199f33ff70f2732c6ab5/index_… >--- > dlls/riched20/tests/editor.c | 4 +++- > 1 files changed, 3 insertions(+), 1 deletions(-) > >diff --git a/dlls/riched20/tests/editor.c b/dlls/riched20/tests/editor.c >index efd2023..5321d07 100644 >--- a/dlls/riched20/tests/editor.c >+++ b/dlls/riched20/tests/editor.c >@@ -6082,7 +6082,9 @@ static void test_EM_CHARFROMPOS(void) > point.x = 1000; > point.y = 40; > result = SendMessage(hwnd, EM_CHARFROMPOS, 0, (LPARAM)&point); >- todo_wine ok(result == 39, "expected character index of 39 but got %d\n", result); >+ todo_wine ok(result == 39, >+ broken(result == 34) /* 120 dpi */, >+ "expected character index of 39 but got %d\n", result); > > point.x = 1000; > point.y = -1; >-- > >Best Regards, André Hentschel > >

1 0

Trace/breakpoint trap using VS2008 remote debugger
by Joe Dluzen Dec. 9, 2010

Dec. 9, 2010

Hi all, I've got the Visual Studio Remote Debugger working with breakpoints, but not with forcing the process to pause. It immediately exits with "Trace/breakpoint trap." I've tried looking up the error, but it seems that no one has this specific set of circumstances. Any ideas? Thanks, Joe

2 2

Re: [1/2] msi: Don't try to install or remove ODBC drivers if the component is disabled.
by Joris Huizer Dec. 9, 2010

Dec. 9, 2010

Hello, I noticed in this patch, exactly the same lines are added in all the functions; wouldn't it be better to add a(n) (inline) function isComponentEnabled(rec,package) to reduce the repetition? HTH, Joris

2 1

Re: [1/2] winetest: Add a trace in the log when skipping a subtest.
by Alexandre Julliard Dec. 9, 2010

Dec. 9, 2010

Francois Gouget <fgouget(a)free.fr> writes: > I'm not happy with the situation when we skip a whole test executable. > One issue is that in such a case we don't even extract the executable > right now, thus making it impossible to print the subtest name. So we'd > need a different skip line format for that case. Or we could decide to > extract all the files and query their subtest names anyway and only > disable actually running the tests. The problem with this is that one of the reasons for skipping a test is that loading the test binary crashes the system, so loading it to get the subtest names wouldn't help. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH 5/5] mshtml: Added GUID_CUSTOM_CONFIRMOBJECTSAFETY with CONFIRMSAFETYACTION_LOADOBJECT flag tests
by Marvin Dec. 9, 2010

Dec. 9, 2010

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=7515 Your paranoid android. === WXPPROSP3 (32 bit script) === script.c:1462: Test failed: expected AXSetInterfaceSafetyOptions_IDispatch_data script.c:1393: Test failed: policy = 0, expected 3 script.c:1407: Test failed: expected AXSetInterfaceSafetyOptions_IDispatch_data script.c:1407: Test failed: expected AXSetInterfaceSafetyOptions_IDispatch_data script.c:1407: Test failed: expected AXSetInterfaceSafetyOptions_IDispatch_data script.c:1586: Test failed: expected AXSetInterfaceSafetyOptions_IDispatch_data script.c:1587: Test failed: policy = 0, expected 3 === WVISTAADM (32 bit script) === script.c:1462: Test failed: expected AXSetInterfaceSafetyOptions_IDispatch_data script.c:1393: Test failed: policy = 0, expected 3 script.c:1407: Test failed: expected AXSetInterfaceSafetyOptions_IDispatch_data script.c:1407: Test failed: expected AXSetInterfaceSafetyOptions_IDispatch_data script.c:1407: Test failed: expected AXSetInterfaceSafetyOptions_IDispatch_data script.c:1586: Test failed: expected AXSetInterfaceSafetyOptions_IDispatch_data script.c:1587: Test failed: policy = 0, expected 3

1 0

Re: [2/2] ntdll: Implement missing 64 bit shifts (try 2)
by Alexandre Julliard Dec. 9, 2010

Dec. 9, 2010

Mike Gibson <mike.gibson(a)storagecraft.com> writes: > @@ -144,8 +144,23 @@ LONGLONG WINAPI RtlLargeIntegerShiftRight( LONGLONG a, INT count ) > */ > LONGLONG WINAPI RtlLargeIntegerArithmeticShift( LONGLONG a, INT count ) > { > - /* FIXME: gcc does arithmetic shift here, but it may not be true on all platforms */ > - return a >> count; > +#ifdef HAVE_SIMPLE_ARITHMETIC_RIGHT_SHIFT > + return a >> count; > +#else > + /* create mask to extend the sign bit */ > + LONGLONG mask; > + mask = ( LONGLONG ) 1 << 63; > + mask &= a; > + mask >>= count; That's an arithmetic shift, you are only moving the problem. Also your calling convention for the _allsh* functions isn't correct, as a simple test case demonstrates. Do you really have an app that uses this? -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: dsound: Standardize the COM usage in duplex.c.
by Alexandre Julliard Dec. 9, 2010

Dec. 9, 2010

Michael Stefaniuc <mstefani(a)redhat.de> writes: > --- > dlls/dsound/duplex.c | 86 ++++++++++++++++++++++++++++++------------------- > 1 files changed, 53 insertions(+), 33 deletions(-) > > diff --git a/dlls/dsound/duplex.c b/dlls/dsound/duplex.c > index 4a1fbd2..2fcf64c 100644 > --- a/dlls/dsound/duplex.c > +++ b/dlls/dsound/duplex.c > @@ -44,7 +44,7 @@ WINE_DEFAULT_DEBUG_CHANNEL(dsound); > typedef struct IDirectSoundFullDuplexImpl > { > /* IUnknown fields */ > - const IDirectSoundFullDuplexVtbl *lpVtbl; > + IDirectSoundFullDuplex IDirectSoundFullDuplex_iface; > LONG ref; > > /* IDirectSoundFullDuplexImpl fields */ > @@ -57,23 +57,43 @@ typedef struct IDirectSoundFullDuplexImpl > } IDirectSoundFullDuplexImpl; > > typedef struct IDirectSoundFullDuplex_IUnknown { > - const IUnknownVtbl *lpVtbl; > + IUnknown IUnknown_iface; > LONG ref; > IDirectSoundFullDuplexImpl *pdsfd; > } IDirectSoundFullDuplex_IUnknown; > > typedef struct IDirectSoundFullDuplex_IDirectSound8 { > - const IDirectSound8Vtbl *lpVtbl; > + IDirectSound8 IDirectSound8_iface; > LONG ref; > IDirectSoundFullDuplexImpl *pdsfd; > } IDirectSoundFullDuplex_IDirectSound8; > > typedef struct IDirectSoundFullDuplex_IDirectSoundCapture { > - const IDirectSoundCaptureVtbl *lpVtbl; > + IDirectSoundCapture IDirectSoundCapture_iface; > LONG ref; > IDirectSoundFullDuplexImpl *pdsfd; > } IDirectSoundFullDuplex_IDirectSoundCapture; All these objects are redundant, they should be merged. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: Outlook-friendly stub for CreatePrivateObjectSecurity
by James Mckenzie Dec. 8, 2010

Dec. 8, 2010

Nick Sukharev <nicksukharev(a)gmail.com> wrote: > >Tab characters fixed. You need to resubmit as [Try 2] if you make a correction in the future in a new message, please. James McKenzie

1 0

Re: Outlook-friendly stub for CreatePrivateObjectSecurity
by Dmitry Timoshkov Dec. 8, 2010

Dec. 8, 2010

Nick Sukharev <nicksukharev(a)gmail.com> wrote: > + if (!InitializeSecurityDescriptor(*NewDescriptor, SECURITY_DESCRIPTOR_REVISION)) { > + HeapFree(GetProcessHeap(), 0, *NewDescriptor); > + return FALSE; > + } You are mixing tabs and spaces. -- Dmitry.

1 0