- wine-devel - winehq.org

Reload a DLL and _getptd returns zero
by Michael Ost 13 Jan '13

13 Jan '13

Hi list, Does anyone know why _getptd() calls would return zero on a DLL that has been reloaded? Does something happen to the TLS used by _getptd? At Muse Research we have been struggling with a deep bug for years where there is a crash _sometimes_ if you load a DLL a second time. Finally (with a big tip of the hat to Julien Pommier at PianoTeq!) we have a test case that demonstrates a (if not _the_) crash. The sequence is: 1. [main] start a thread 2. [main] load a library 3. [thread] call std::cout<< from the library 4. [main] free the library 5. [main] reload the library 6. [thread] call std::cout<< from the library -> crash! It looks like the crash happens because the MSVCRT function _getptd() is returning zero in step 6. This is dereferenced and crashes. In windows, there is no crash. The library in question is using static vc runtime linkage, so _getptd() is linked in and I don't know exactly what it is doing. But Wine's MSVCRT implementation of _getptd() gets its data from TlsGetValue. And I can see that just before the crash there is a TlsGetValue call that returns zero. Looking at TlsGetValue(), there must be something wrong with the values in TlsSlots. Maybe they persist for DLLs in Windows in a way they don't for Wine...? Or maybe Wine doesn't reinitialize them the same way when the DLL is reloaded? Any hints, thoughts? clues? Thanks! -- Michael Ost

4 4

Re: comctl32/listview: fix icon spacing calculation
by Nikolay Sivov 13 Jan '13

13 Jan '13

On 1/14/2013 00:00, Daniel Jelinski wrote: > -static DWORD LISTVIEW_SetIconSpacing(LISTVIEW_INFO *infoPtr, INT cx, INT cy) > +static DWORD LISTVIEW_SetIconSpacing(LISTVIEW_INFO *infoPtr, WORD cx, WORD cy) I don't really like it as WORD, what it really is a int value with special meaning reserved for -1. For 64bits and lParam == -1 you'll still get two -1 values. > + /* LPARAM = -1 - restore default processing */ > + ret = SendMessage(hwnd, LVM_SETICONSPACING, 0, MAKELPARAM(-1,-1)); > +todo_wine > + expect(100, LOWORD(ret)); > + expect(0xFFFF, HIWORD(ret)); > + > + /* NOTE: -1 is not the same as MAKELPARAM(-1,-1) in 64bit listview */ > + if(sizeof(LPARAM) == 8) > + { > + ret = SendMessage(hwnd, LVM_SETICONSPACING, 0, -1); > +todo_wine { > + expect(0xFFFF, LOWORD(ret)); > + expect(0xFFFF, HIWORD(ret)); > + } > + } That means you should just use lParam == -1. Also this message is supposed to use MAKELONG apparently, not sure if it makes a difference.

1 0

Re: comctl32/listview: fix icon spacing calculation
by Marvin 13 Jan '13

13 Jan '13

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=23897 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH] advapi32: add RegSetKeyValue{A,W}
by Nikolay Sivov 13 Jan '13

13 Jan '13

On 1/13/2013 09:27, danielfsantos(a)att.net wrote: > Adds RegSetKeyValue{A,W} to advapi32.dll > Refactors RegSetValue{A,W} to call RegSetKeyValue{A,W} > Fixes bug #32711 > --- > dlls/advapi32/advapi32.spec | 4 +- > dlls/advapi32/registry.c | 81 +++++++++++++++++++++++++++++++++---------- > 2 files changed, 64 insertions(+), 21 deletions(-) > > diff --git a/dlls/advapi32/advapi32.spec b/dlls/advapi32/advapi32.spec > index 21357fa..336518b 100644 > --- a/dlls/advapi32/advapi32.spec > +++ b/dlls/advapi32/advapi32.spec > @@ -645,8 +645,8 @@ > # @ stub RegSaveKeyExW > @ stdcall RegSaveKeyW(long ptr ptr) > @ stdcall RegSetKeySecurity(long long ptr) > -# @ stub RegSetKeyValueA > -# @ stub RegSetKeyValueW > +@ stdcall RegSetKeyValueA(long str str long ptr long) > +@ stdcall RegSetKeyValueW(long str str long ptr long) This should use wstr. > +LSTATUS WINAPI RegSetValueW( HKEY hkey, LPCWSTR name, DWORD type, LPCWSTR data, DWORD count ) > +{ > + return RegSetKeyValueW(hkey, name, NULL, type, data, count); > +} Don't use tabs please. And as I said in bug comment this needs tests.

1 0

Re: libwine: Use rpath-based install name and library references for libwine on Mac.
by Alexandre Julliard 12 Jan '13

12 Jan '13

Ken Thomases <ken(a)codeweavers.com> writes: > --- > configure.ac | 3 ++- > libs/wine/Makefile.in | 6 +++--- > 2 files changed, 5 insertions(+), 4 deletions(-) It doesn't seem to work when installed: wine: failed to initialize: dlopen(/usr/local/lib/wine/ntdll.dll.so, 258): image not found -- Alexandre Julliard julliard(a)winehq.org

2 4

Problem with editing FAQ and other pages at wiki
by Romanoff Alexei 12 Jan '13

12 Jan '13

Hello! I've tried editing wine FAQ and several other wiki pages, but failed. When I click on "Preview" or "Save" buttons, I always see error message like [Errno 31] Too many links: '/var/www/wine/wiki/data/pages/AlexeiRomanoff(2f)MoinEditorBackup' If you want to report a bug, please save this page and attach it to your bug report. A problem occurred in a Python script. Here is the sequence of function calls leading up to the error, in the order they occurred. 1. /usr/lib/python2.4/site-packages/MoinMoin/request.py in *run* (self=<MoinMoin.request.RequestModPy object>) 1. 1170 self.page.send_page(self, msg=msg) 2. 1171 else: 3. 1172 handler(self.page.page_name, self) 4. 1173 5. 1174 # generate page footer (actions that do not want this footer use - *handler* = <function do_edit> - *self* = <MoinMoin.request.RequestModPy object> - self.*page* = <MoinMoin.Page.Page instance> - self.page.*page_name* = u'FAQ' 2. /usr/lib/python2.4/site-packages/MoinMoin/wikiaction.py in *do_edit* (pagename=u'FAQ', request=<MoinMoin.request.RequestModPy object>) 1. 651 request.form.has_key('button_spellcheck') or 2. 652 request.form.has_key('button_newwords')): 3. 653 pg.sendEditor(preview=savetext, comment=comment) 4. 654 5. 655 # Preview with mode switch - *pg* = <MoinMoin.PageEditor.PageEditor instance> - pg.*sendEditor* = <bound method PageEditor.sendEditor of <MoinMoin.PageEditor.PageEditor instance>> - preview *undefined* - *savetext* = u"#pragma section-numbers 1\n#language en\n'''''Tran..., it can delete your entire hard drive.\nDeleted/\n" - *comment* = u'Fixed shell code example with invalid redirection of stderr to stdout instead of file' 3. /usr/lib/python2.4/site-packages/MoinMoin/PageEditor.py in *sendEditor* (self=<MoinMoin.PageEditor.PageEditor instance>, **kw={'comment': u'Fixed shell code example with invalid redirection of stderr to stdout instead of file', 'preview': u"#pragma section-numbers 1\n#language en\n'''''Tran..., it can delete your entire hard drive.\nDeleted/\n"}) 1. 284 # Make backup on previews - but not for new empty pages 2. 285 if preview and raw_body: 3. 286 self._make_backup(raw_body) 4. 287 5. 288 # Generate default content for new pages - *self* = <MoinMoin.PageEditor.PageEditor instance> - self.*_make_backup* = <bound method PageEditor._make_backup of <MoinMoin.PageEditor.PageEditor instance>> - *raw_body* = u"#pragma section-numbers 1\n#language en\n'''''Tran..., it can delete your entire hard drive.\nDeleted/\n" 4. /usr/lib/python2.4/site-packages/MoinMoin/PageEditor.py in *_make_backup* (self=<MoinMoin.PageEditor.PageEditor instance>, newtext=u"#pragma section-numbers 1\n#language en\n'''''Tran..., it can delete your entire hard drive.\nDeleted/\n", **kw={}) 1. 738 intro += _('## backup of page "%(pagename)s" submitted %(date)s') % { 2. 739 'pagename': pagename, 'date': date,} + u'\n' 3. 740 backuppage._write_file(intro + newtext) 4. 741 5. 742 return backuppage.url(self.request) - *backuppage* = <MoinMoin.PageEditor.PageEditor instance> - backuppage.*_write_file* = <bound method PageEditor._write_file of <MoinMoin.PageEditor.PageEditor instance>> - *intro* = u'#acl AlexeiRomanoff:read,write,delete All:\n\n' - *newtext* = u"#pragma section-numbers 1\n#language en\n'''''Tran..., it can delete your entire hard drive.\nDeleted/\n" 5. /usr/lib/python2.4/site-packages/MoinMoin/PageEditor.py in *_write_file* (self=<MoinMoin.PageEditor.PageEditor instance>, text=u'#acl AlexeiRomanoff:read,write,delete All:\n\n#pra..., it can delete your entire hard drive.\nDeleted/\n', action='SAVE', comment=u'', extra=u'') 1. 801 # The local log should be the standard edit log, not the 2. 802 # underlay copy log! 3. 803 pagelog = self.getPagePath('edit-log', use_underlay=0, isfile=1) 4. 804 llog = editlog.EditLog(self.request, filename=pagelog, 5. 805 uid_override=self.uid_override) - pagelog *undefined* - *self* = <MoinMoin.PageEditor.PageEditor instance> - self.*getPagePath* = <bound method PageEditor.getPagePath of <MoinMoin.PageEditor.PageEditor instance>> - use_underlay *undefined* - isfile *undefined* 6. /usr/lib/python2.4/site-packages/MoinMoin/Page.py in *getPagePath* (self=<MoinMoin.PageEditor.PageEditor instance>, *args=('edit-log',), **kw={'isfile': 1, 'use_underlay': 0}) 1. 348 """Return path to the page storage area.""" 2. 349 3. 350 return self.getPageStatus(*args, **kw)[1] 4. 351 5. 352 def split_title(self, request, force=0): - *self* = <MoinMoin.PageEditor.PageEditor instance> - self.*getPageStatus* = <bound method PageEditor.getPageStatus of <MoinMoin.PageEditor.PageEditor instance>> - *args* = ('edit-log',) - *kw* = {'isfile': 1, 'use_underlay': 0} 7. /usr/lib/python2.4/site-packages/MoinMoin/Page.py in *getPageStatus* (self=<MoinMoin.PageEditor.PageEditor instance>, *args=('edit-log',), **kw={'isfile': 1, 'use_underlay': 0}) 1. 342 dirname = fullpath 2. 343 if not os.path.exists(dirname): 3. 344 filesys.makeDirs(dirname) 4. 345 return underlay, fullpath 5. 346 - *global* *filesys* = <module 'MoinMoin.util.filesys' from '/usr/lib/python2.4/site-packages/MoinMoin/util/filesys.pyc'> - filesys.*makeDirs* = <function makedirs> - *dirname* = '/var/www/wine/wiki/data/pages/AlexeiRomanoff(2f)MoinEditorBackup' 8. /usr/lib/python2.4/site-packages/MoinMoin/util/filesys.py in *makedirs* (name='/var/www/wine/wiki/data/pages/AlexeiRomanoff(2f)MoinEditorBackup', mode=511) 1. 49 return 2. 50 try: 3. 51 os.mkdir(name, mode & config.umask) 4. 52 except OSError, err: 5. 53 if err.errno != errno.EEXIST: - *global* *os* = <module 'os' from '/usr/lib/python2.4/os.pyc'> - os.*mkdir* = <built-in function mkdir> - *name* = '/var/www/wine/wiki/data/pages/AlexeiRomanoff(2f)MoinEditorBackup' - *mode* = 511 - *global* *config* = <module 'MoinMoin.config' from '/usr/lib/python2.4/site-packages/MoinMoin/config.pyc'> - config.*umask* = 504 OSError [Errno 31] Too many links: '/var/www/wine/wiki/data/pages/AlexeiRomanoff(2f)MoinEditorBackup' - args = (31, 'Too many links') - errno = 31 - filename = '/var/www/wine/wiki/data/pages/AlexeiRomanoff(2f)MoinEditorBackup' - strerror = 'Too many links' It seems, something wrong happened at server side with MoinMoin. But how could be possible to contribute, if wiki isn't editable? I haven't found any specific mailing list or forum for wiki editors and decided to write here. -- Best regards, Alexei Romanoff software developer, Melesta

3 4

Re: Ken Thomases : libwine: Use rpath-based install name and library references for libwine on Mac.
by Charles Davis 11 Jan '13

11 Jan '13

On Jan 11, 2013, at 12:46 PM, Alexandre Julliard wrote: > Module: wine > Branch: master > Commit: f377591e98c490ca00e6479627832e41b2ad01b9 > URL: http://source.winehq.org/git/wine.git/?a=commit;h=f377591e98c490ca00e647962… > > Author: Ken Thomases <ken(a)codeweavers.com> > Date: Fri Jan 11 03:25:07 2013 -0600 > > libwine: Use rpath-based install name and library references for libwine on Mac. Huh? I thought this broke running installed Wine for you. Why did you commit this anyway? (Or is there another email message that I didn't see? Gmail has been acting wonky lately about delivering my email...) Chip

2 1

Re: [2/2] windowscodecs: Implement CreateBitmapFromHICON, based on a gdiplus implementation.
by Vincent Povirk 11 Jan '13

11 Jan '13

You appear to be leaking the IWICBitmapLock object. Also, it seems odd that you're using the buffer AND IWICBitmapLock. I assume the point of passing a buffer to BitmapImpl_Create (and wasting a bit of time/memory) was to avoid IWICBitmapLock.

2 1

Re: kernel32/tests : try3 added tests for copyfileex
by Dmitry Timoshkov 11 Jan '13

11 Jan '13

Patrick Rudolph <siro(a)das-labor.org> wrote: > + if (!pCopyFileExA) > + win_skip("CopyFileExA is not available\n"); Continuing at this point doesn't seem much useful. Also you need to use pCopyFileExA instead of direct CopyFileExA calls everywhere in the test, otherwise .exe won't even load under Windows due to dynamic linking errors. Sorry, I missed that in previous review. -- Dmitry.

1 0

Re: [2/2] windowscodecs: Implement CreateBitmapFromHICON, based on a gdiplus implementation. Resend.
by Vincent Povirk 11 Jan '13

11 Jan '13

I hadn't noticed you had resent this, and made comments on the earlier mail which still apply.

1 0

Re: kernel32/tests : try3 added tests for copyfileex
by Marvin 11 Jan '13

11 Jan '13

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=23881 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 1/2] d3d9: Return D3D_OK from the palette setters
by Henri Verbeet 11 Jan '13

11 Jan '13

On 11 January 2013 09:49, Stefan Dösinger <stefan(a)codeweavers.com> wrote: > - * Copyright 2006-2007, 2010 Stefan Dösinger for CodeWeavers > + * Copyright 2006-2008,2010-2011,2013 Stefan Dösinger for CodeWeavers I'm going to assume you're changing this because you checked it, but the formatting is a bit off. > + memset(&caps, 0, sizeof(caps)); That's redundant.

2 2

Re: [PATCH 2/2] d3d8: Return D3D_OK from the palette setters (try 2)
by Marvin 11 Jan '13

11 Jan '13

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=23877 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 5/5] d3d9/tests: Merge surface.c into device.c.
by Marvin 11 Jan '13

11 Jan '13

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=23875 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 4/5] d3d9/tests: Merge texture.c into device.c.
by Marvin 11 Jan '13

11 Jan '13

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=23874 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [2/2] windowscodecs: Implement CreateBitmapFromHICON, based on a gdiplus implementation. Resend.
by Marvin 11 Jan '13

11 Jan '13

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=23872 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [1/2] windowscodecs: Add a test for CreateBitmapFromHICON. Take 2.
by Marvin 11 Jan '13

11 Jan '13

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=23871 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 1/3] d3dx9_36: Implement ID3DXFileDataImpl_Lock and ID3DXFileDataImpl_Unlock.
by Rico Schüller 11 Jan '13

11 Jan '13

On 11.01.2013 09:40, Christian Costa wrote: > static HRESULT WINAPI ID3DXFileDataImpl_Unlock(ID3DXFileData *iface) > { > - FIXME("(%p)->(): stub\n", iface); > + TRACE("(%p)->()\n", iface); > > - return E_NOTIMPL; > + /* Nothing to do */ > + > + return S_OK; > } Hi Christian, I stumble over MSDN, it says in Remarks on http://msdn.microsoft.com/de-de/library/windows/desktop/bb205846%28v=vs.85%… ... from that sentence I'd think that you have to do some lock count. A test would probably be nice. Though, msdn has a bool as return value for the unlock function but in the header it's a HRESULT... Is msdn wrong here? Cheers Rico

2 1

Re: [1/2] windowscodecs: Add a test for CreateBitmapFromHICON.
by Nikolay Sivov 11 Jan '13

11 Jan '13

On 1/11/2013 14:39, Dmitry Timoshkov wrote: > + icon = CreateIconIndirect(&info); > + ok(icon != 0, "CreateIconIndirect failed\n"); > + DeleteObject(info.hbmColor); > + DeleteObject(info.hbmMask); > + > + hr = IWICImagingFactory_CreateBitmapFromHICON(factory, icon, &bitmap); > + ok(hr == S_OK, "CreateBitmapFromHICON error %#x\n", hr); > + > + IWICBitmap_GetPixelFormat(bitmap, &format); > + ok(IsEqualGUID(&format, &GUID_WICPixelFormat32bppBGRA), > + "unexpected pixel format %s\n", debugstr_guid(&format)); > + > + IWICBitmap_GetSize(bitmap, &width, &height); > + ok(hr == S_OK, "IWICBitmap_GetSize error %#x\n", hr); > + ok(width == 16, "expected 16, got %u\n", width); > + ok(height == 16, "expected 16, got %u\n", height); > + > + IWICBitmap_Release(bitmap); > +} Who is responsible for freeing icon handle? I guess it should be done here cause your implementation doesn't free it either.

3 4

Re: [PATCH 6/6] msvcrt: Added fwprintf tests
by Marvin 11 Jan '13

11 Jan '13

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=23870 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [1/2] windowscodecs: Add a test for CreateBitmapFromHICON.
by Marvin 11 Jan '13

11 Jan '13

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=23866 Your paranoid android. === WXPPROSP3 (32 bit bitmap) === bitmap.c:492: Test failed: 2: expected 255, got 192 bitmap.c:510: Test failed: 2: expected 255, got 192 === WVISTAADM (32 bit bitmap) === bitmap.c:492: Test failed: 2: expected 255, got 192 bitmap.c:510: Test failed: 2: expected 255, got 192 === W7PRO (32 bit bitmap) === bitmap.c:492: Test failed: 2: expected 255, got 192 bitmap.c:510: Test failed: 2: expected 255, got 192 === W7PROX64 (32 bit bitmap) === bitmap.c:492: Test failed: 2: expected 255, got 192 bitmap.c:510: Test failed: 2: expected 255, got 192 === TEST64_W7SP1 (32 bit bitmap) === bitmap.c:492: Test failed: 2: expected 255, got 192 bitmap.c:510: Test failed: 2: expected 255, got 192 === W7PROX64 (64 bit bitmap) === bitmap.c:492: Test failed: 2: expected 255, got 192 bitmap.c:510: Test failed: 2: expected 255, got 192 === TEST64_W7SP1 (64 bit bitmap) === bitmap.c:492: Test failed: 2: expected 255, got 192 bitmap.c:510: Test failed: 2: expected 255, got 192

1 0

Re: [2/2] windowscodecs: Implement CreateBitmapFromHICON, based on a gdiplus implementation.
by Marvin 11 Jan '13

11 Jan '13

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=23867 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 2/2] d3d8: Return D3D_OK from the palette setters
by Marvin 11 Jan '13

11 Jan '13

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=23865 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 1/2] d3d9: Return D3D_OK from the palette setters
by Marvin 11 Jan '13

11 Jan '13

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=23864 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: kernel32/tests : try2 added tests for copyfileex
by Dmitry Timoshkov 11 Jan '13

11 Jan '13

siro(a)das-labor.org wrote: > OK, but how do I know that the old behavior is broken ? WindowsXp behaves > like msdn quotes while newer windows don't. Do you always assume that the > latest implementation is correct ? Should wine just match the latest > test-results and ignore osversioninfo at all ? It's reasonable to assume that a later Windows implementation should be considered as the one to follow. Changing the reported Windows version in winecfg doesn't change an API behaviour (there are exceptions, but they don't apply in this case). -- Dmitry.

1 0

Re: kernel32/tests : try2 added tests for copyfileex
by Dmitry Timoshkov 11 Jan '13

11 Jan '13

Patrick Rudolph <siro(a)das-labor.org> wrote: > try2: > check the os version in case of PROGRESS_STOP > + ret = GetFileAttributesEx(dest_name, GetFileExInfoStandard, (void*)&fileInfo); > + > + /* Windows Vista and newer delete the file, while prior versions don't */ > + ok((!ret && info.dwMajorVersion > 5) || (ret && info.dwMajorVersion <= 5), > + "GetFileAttributesEx return-value wrong, got %u and MajorVersion is %u\n", ret, info.dwMajorVersion); Then you should mark old behaviour as broken() and don't check Windows version at all. -- Dmitry.

2 1

Re: d3dx9: Implement D3DXSHEvalConeLight (resend)
by Marvin 11 Jan '13

11 Jan '13

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=23862 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: user32: fix non-BOOL returns in BOOL functions
by Dmitry Timoshkov 11 Jan '13

11 Jan '13

Damjan Jovanovic <damjan.jov(a)gmail.com> wrote: > Changelog: > * user32: fix non-BOOL returns in BOOL functions GetClassInfo is special, it actually returns class atom on success, not just a boolean value. So it's better to leave it as it is now. -- Dmitry.

1 0

Re: [PATCH 2/4] kernel32/path : try2 implement copyfileexW progressroutine callback
by Dmitry Timoshkov 11 Jan '13

11 Jan '13

Patrick Rudolph <siro(a)das-labor.org> wrote: > try 2: > set progress routine pointer to NULL to disable progress reporting > test osversion and delete destination file on PROGRESS_STOP on vista > and newer This kind of version check is useless in Wine. -- Dmitry.

1 0

Re: [2/2] comctl32/listview: Add a special case for deselect all attempt when nothing is selected
by Marvin 10 Jan '13

10 Jan '13

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=23861 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: include: flesh out d3d11.idl
by Nikolay Sivov 10 Jan '13

10 Jan '13

On 1/10/2013 22:35, Austin English wrote: Some minor things. > +typedef enum D3D11_RESOURCE_DIMENSION > +{ > + D3D11_RESOURCE_DIMENSION_UNKNOWN = 0, > + D3D11_RESOURCE_DIMENSION_BUFFER = 1, > + D3D11_RESOURCE_DIMENSION_TEXTURE1D = 2, > + D3D11_RESOURCE_DIMENSION_TEXTURE2D = 3, > + D3D11_RESOURCE_DIMENSION_TEXTURE3D = 4, > +} D3D11_RESOURCE_DIMENSION; > + You could remove values in cases like that, result will be the same. > + D3D11_BIND_VERTEX_BUFFER = 0x00000001L, I think L suffix could be omitted. > +//FIXME: ...

1 0

Re: [PATCH 3/3] d3dx9: Use all 32 bits as mask.
by Matteo Bruni 10 Jan '13

10 Jan '13

2013/1/10 Rico Schüller <kgbricola(a)web.de>: > --- > dlls/d3dx9_36/texture.c | 6 +++--- > 1 Datei geändert, 3 Zeilen hinzugefügt(+), 3 Zeilen entfernt(- > > - DWORD i, v; > + DWORD i, v, mask32 = format->bits[c] == 32 ? -1 : ((1 << format->bits[c]) - 1); Usually we prefer to use ~0U instead of -1 with unsigned variables. Nice cleanups and fixes for larger pixel formats. Is this for Alan Wake?

2 1

Re: Truncate MIDI SysEx messages after termination byte
by Johannes Kroll 10 Jan '13

10 Jan '13

On Wed, 9 Jan 2013 16:10:18 +0100 Christian Costa <titan.costa(a)gmail.com> wrote: > 2013/1/9 Johannes Kroll <jkroll(a)lavabit.com> > > > On Wed, 09 Jan 2013 10:19:54 +0100 > > Christian Costa <titan.costa(a)gmail.com> wrote: > > > > > Le 09/01/2013 04:26, Johannes Kroll a écrit : > > > > Hi, > > > > > > > > On Wed, 09 Jan 2013 03:55:25 +0100 > > > > Christian Costa <titan.costa(a)gmail.com> wrote: > > > > > > > >> Hi, > > > >> > > > >> Please be consistent when using space. 1 space before and after > > > >> operators (<, ==, =, ...). > > > > spaces inserted. HTH. > > > > > > > > J. > > > That's better. You forgot some occurences in the for loop though. > > > > Is it OK if I leave the rest of the formatting to you? Any style is > > fine with me, whatever makes you happy and hopefully gets the patch > > included. > > > > Formatting only patch are not allowed. Clean patch and consistency in code > style help your patch get in. > Alexandre Julliard is the only judge not me. If he is ok with your patch > that's fine. > Anyway having the for loop consistent with rest should not be that hard. > > for(i= 0; i<lpMidiOutHdr->dwBufferLength; i++) The original patch was consistent with my coding style. To make assignments stick out visually, I use no space to the left of the assignment operator, one space to the right. On comparison operators, I use space either on both sides, or none. You will see this in all code I write. I attached another version of the patch. I added 3 more spaces and one line break in a long comment line. You or Alexandre can choose the version they like best. If you still see style issues, please consider changing the style to your liking. If that's not possible, the patch will have to be rejected. If anybody sees non-style-related issues with the patch, please let me know. Please keep any replies on the list(s). Have a nice day. Johannes

2 8

Re: [PATCH] d3dx9: Shift only as much as needed.
by Rico Schüller 10 Jan '13

10 Jan '13

Hi, please don't apply this patch. There are several other occurrences in the same file. I'll send an improved version. Cheers Rico On 10.01.2013 13:55, Rico Schüller wrote: > Hi, > > this patch removes useless shifts, which may result in wrong data when > the shift is bigger than 24. > > Cheers > Rico > > --- > dlls/d3dx9_36/texture.c | 15 ++++++++++++--- > 1 Datei geändert, 12 Zeilen hinzugefügt(+), 3 Zeilen entfernt(-) > > > > >

1 0

Re: [2/2] comctl32/listview: Use LVM_SETITEMSTATE directly in tests
by Marvin 10 Jan '13

10 Jan '13

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=23848 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH] ddraw: Don't discard buffers that haven't been used in draws (try 3)
by Marvin 10 Jan '13

10 Jan '13

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=23842 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 5/5] d3d9/tests: Merge query.c into device.c.
by Marvin 10 Jan '13

10 Jan '13

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=23841 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 4/5] d3d9/tests: Merge buffer.c into device.c.
by Marvin 10 Jan '13

10 Jan '13

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=23840 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 17/17] msvcrt: Added read tests on unicode files
by Marvin 10 Jan '13

10 Jan '13

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=23838 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 13/17] msvcrt: Added _setmode tests
by Marvin 10 Jan '13

10 Jan '13

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=23836 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 07/17] msvcrt: Added write tests on unicode files
by Marvin 10 Jan '13

10 Jan '13

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=23835 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 05/17] msvcrt: Added fopen_s test with ccs option
by Marvin 10 Jan '13

10 Jan '13

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=23834 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH 0/7] Series short description
by Alexandre Julliard 09 Jan '13

09 Jan '13

Eric Pouech <eric.pouech(a)orange.fr> writes: > The following series implements a new scheme for managing 'bare' consoles. > We now force every program started without a wine parent to be run under > wineconsole (transparently for the user). > This is done using a new backend in wineconsole called 'line'. This backend > only takes care of input; output is kept as a client side operation. > All terminal related code is moved from dlls/kernel32 to programs/wineconsole. > > This basically ensures that: > - Only wineconsole reads from standard input (and writes into input queue). > This fixes bug reported by Jason when $CONIN is opened with read access only. > - It also enforces that wineconsole always has control of the unix terminal, > hence ensuring that we properly reset the terminal settings upon exit. Running make test with this series hangs in the hlink.c test, there seems to be a wineconsole process that never terminates. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: [PATCH] crypt32: free the encoded msg (Coverity)
by Juan Lang 09 Jan '13

09 Jan '13

In case this was awaiting an ACK from me, this looks good. Thanks. --Juan

1 0

Re: usp10: Implement GPOS MarkToLigature Attachment Positioning Subtable
by Huw Davies 09 Jan '13

09 Jan '13

On Wed, Jan 09, 2013 at 11:36:51AM -0600, Aric Stewart wrote: > +typedef struct { > + WORD LigatureAnchor[1]; > +} GPOS_ComponentRecord; > + > +typedef struct { > + WORD ComponentCount; > + GPOS_ComponentRecord ComponentRecord[1]; > +} GPOS_LigatureAttach; > + > + for (i = 0; i < component_count && !offset; i++) > + { > + int k; > + for (k = 0; k < class_count && !offset; k++) > + offset = GET_BE_WORD(lt->ComponentRecord[i].LigatureAnchor[k]); > + } Hi Aric, This can't work as ComponentRecord is a variable-sized struct. Huw.

2 1

Re: [PATCH (try3)] ieframe: Strip 'file://' from file URLs in BEFORENAVIGATE2 callbacks
by Marvin 09 Jan '13

09 Jan '13

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=23822 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: [PATCH] d3d8/tests: Merge surface.c into device.c.
by Marvin 09 Jan '13

09 Jan '13

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=23820 Your paranoid android. === WINEBUILD (build) === Patch failed to apply

1 0

Re: aclocal: Use -Werror when checking compiler flags (try 2)
by Alexandre Julliard 09 Jan '13

09 Jan '13

Per Johansson <per(a)morth.org> writes: > ...if supported by compiler. > This makes clang exit with error on unsupported warnings. > --- > aclocal.m4 | 2 +- > configure.ac | 1 + > 2 files changed, 2 insertions(+), 1 deletion(-) > > diff --git a/aclocal.m4 b/aclocal.m4 > index 14cf916..a6d9f12 100644 > --- a/aclocal.m4 > +++ b/aclocal.m4 > @@ -74,7 +74,7 @@ AC_DEFUN([WINE_TRY_CFLAGS], > [AS_VAR_PUSHDEF([ac_var], ac_cv_cflags_[[$1]])dnl > AC_CACHE_CHECK([whether the compiler supports $1], ac_var, > [ac_wine_try_cflags_saved=$CFLAGS > -CFLAGS="$CFLAGS $1" > +CFLAGS="$CFLAGS $ac_wine_try_cflags_werror $1" > AC_LINK_IFELSE([AC_LANG_SOURCE([[int main(int argc, char **argv) { return 0; }]])], > [AS_VAR_SET(ac_var,yes)], [AS_VAR_SET(ac_var,no)]) aclocal macros should be stand-alone, they shouldn't depend on things done by the main configure. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [1/2] windowscodecs: Add a test for loading a .bmp file with BITMAPINFOHEADER.
by Vincent Povirk 09 Jan '13

09 Jan '13

Series looks fine to me. And yes, testbot has been broken for a while now..

1 0

Re: [PATCH 5/5] d3d8/tests: Merge surface.c into device.c.
by Alexandre Julliard 09 Jan '13

09 Jan '13

Henri Verbeet <hverbeet(a)codeweavers.com> writes: > --- > dlls/d3d8/tests/Makefile.in | 1 - > dlls/d3d8/tests/device.c | 704 ++++++++++++++++++++++++++++++++++++++++++- > dlls/d3d8/tests/surface.c | 653 --------------------------------------- > 3 files changed, 702 insertions(+), 656 deletions(-) > delete mode 100644 dlls/d3d8/tests/surface.c It doesn't work here: ../../../tools/runtest -q -P wine -M d3d8.dll -T ../../.. -p d3d8_test.exe.so device.c && touch device.ok device.c:4088: Test failed: Got unexpected hr 0x88760867. device.c:4091: Test failed: Got unexpected refcount 3, expected 4. device.c:4092: Test failed: Got unexpected ptr 0x15, expected 0x21fdb8. wine: Unhandled page fault on read access to 0x00000015 at address 0x684e3686 (thread 0009), starting debugger... -- Alexandre Julliard julliard(a)winehq.org

2 1

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

wine-devel