April 2009 - Wine-Devel - list.winehq.org

Re: Possible regression today's git wine-1.1.19-153-g55ad3fd
by Susan Cragin April 21, 2009

April 21, 2009

>2009/4/20 Reece Dunn <msclrhd(a)googlemail.com>: >> 2009/4/20 Susan Cragin <susancragin(a)earthlink.net>: >>>>Ohhh... got a nice crash after a re-boot. >>> >>> Never mind that. This looks more promising. At least, to me it does. The actctx error I get all the time, but I've never seen the other, and it happend right as the system froze up. >>> >>> fixme:actctx:parse_depend_manifests Could not find dependent assembly L"Microsoft.VC80.MFCLOC" (8.0.50608.0) >>> err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Broken pipe >>> err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Broken pipe >>> err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Broken pipe >>> err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Success >> >> Interesting. You are running Ubuntu studio 8.10 without pulseaudio, right? >> >> Have you updated the system between version runs (that may have >> reintroduced pulseaudio, for example)? >> >> Are you running any other sound applications (or the sound that is >> played after boot on Ubuntu)? >> >> Could you perform a git bisect to identify the guilty change? > >Doing some investigation, I have found >(http://alsa.opensrc.org/index.php/HowTo_Asynchronous_Playback) >""" >That is basically all there is to the callback. Note however, that >this example does not include error checking, which is a good idea of >course. Occasionally, you may have to restore from a buffer under-run, >in case the computer was caught up in some other process and didn't >get to calling your callback. Whenever this happens, the first >operation you perform on the device, in this case >snd_pcm_avail_update, will return -EPIPE (a 'Broken Pipe' error). You >can restore from this by calling the snd_pcm_prepare function again. >""" > >This may be what you are experiencing here. I'd need to look into this >further, but to be honest, I'm a little out of depth with the Windows >and Wine sound interfaces. > >- Reece Reece: I purged pulseaudio and checked that it had not been re-installed. I run Ubuntu 9.04 with all updates. I also use a test version of alsa-driver since my sound card is difficult to recognize but so far have had only great results -- no impact on wine at all. The problem I am experiencing manifests itself in the program freezing... could be the sound, could also be the buttons doing something odd. One button's on-off toggle switches suddenly with no input on my part. Regression test is below. Susan b958dfdc4769515c0ef56c401430dd3e3980aabf is first bad commit commit b958dfdc4769515c0ef56c401430dd3e3980aabf Author: Huw Davies <huw(a)codeweavers.com> Date: Thu Apr 16 11:51:45 2009 +0100 ole32: Change helper function to return the matching entry. :040000 040000 2964979dee4f353e1d565f7c20d8fdcc534f006d 8744cfee2e33db4799f58f102825103f565e18b9 M dlls

1 0

Re: dsound: Allow a special cbSize case in CreateSoundBuffer
by Maarten Lankhorst April 20, 2009

April 20, 2009

Hey Jesse, Jesse Allen schreef: > OK now do you have comments? I have test cases. > > On Sun, Apr 19, 2009 at 6:30 PM, Jesse Allen <the3dfxdude(a)gmail.com> wrote: > >> In CreateSoundBuffer, some apps pass sizeof(WAVEFORMATEXTENSIBLE) for >> the cbSize and we have to allow it, but only when the SubFormat is >> KSDATAFORMAT_SUBTYPE_PCM. >> >> Includes test cases which were verified on WinXP. >> Fixes bug #11716 >> If you want me to review a patch add me to cc please. The variable size_ok is unneeded here, you can rewrite that so that it will fall through if test fails. Apart from that it looks ok to me. Cheers, Maarten.

1 0

Re: Possible regression today's git wine-1.1.19-153-g55ad3fd
by Susan Cragin April 20, 2009

April 20, 2009

>Ohhh... got a nice crash after a re-boot. Never mind that. This looks more promising. At least, to me it does. The actctx error I get all the time, but I've never seen the other, and it happend right as the system froze up. fixme:actctx:parse_depend_manifests Could not find dependent assembly L"Microsoft.VC80.MFCLOC" (8.0.50608.0) err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Broken pipe err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Broken pipe err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Broken pipe err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Success

2 2

Janitor: list.h functions defined but not used
by Glenn L. McGrath April 20, 2009

April 20, 2009

Hi all, im new to the list, im interested in grinding away at some of the warnings wine generates... make depend was complaining about some static inline functions that it thinks arent used, this patch uses __attribute__ ((used)) to disable the warning. Glenn

7 7

Re: (Try2) wininet/tests: Fix HttpSendRequestW test on Win2k platform.
by Hans Leidekker April 20, 2009

April 20, 2009

On Monday 20 April 2009 20:14:14 Nicolas Le Cam wrote: > Win2k SP4 seems to be the only platform that follows what msdn says (see > http://msdn.microsoft.com/en-us/library/aa384247(VS.85).aspx) wininet is bundled with Internet Explorer, so it's more useful to look at the Explorer version than the Windows version. > Perhaps passing header's length to HttpSendRequest instead of adding a > broken expected value would be better. Let me know. No, the length parameter is the essence of this test. I wrote it because I saw IE7 call HttpSendRequestW like this which contradicts MSDN. The test shows that MSDN is outdated; on recent wininet HttpSendRequestW will calculate the header length for you. -Hans

2 1

Re: Possible regression today's git wine-1.1.19-153-g55ad3fd
by Susan Cragin April 20, 2009

April 20, 2009

>Dragon NaturallySpeaking freezes up today during training, without crashing. >I will do a regression test if needed, but suspect that the freeze may be already known and/or affect many programs. >Susan Ohhh... got a nice crash after a re-boot. fixme:gdiplus:GdipCreateHBITMAPFromBitmap stub fixme:actctx:parse_depend_manifests Could not find dependent assembly L"Microsoft.VC80.MFCLOC" (8.0.50608.0) err:ole:CoRevokeClassObject called from wrong apartment, should be called from 80000002c err:ole:CoRevokeClassObject called from wrong apartment, should be called from 80000002c err:ole:CoRevokeClassObject called from wrong apartment, should be called from 80000002c err:ole:CoRevokeClassObject called from wrong apartment, should be called from 80000002c err:ole:CoRevokeClassObject called from wrong apartment, should be called from 80000002c err:ole:CoRevokeClassObject called from wrong apartment, should be called from 80000002c err:ole:CoRevokeClassObject called from wrong apartment, should be called from 80000002c err:ole:CoRevokeClassObject called from wrong apartment, should be called from 80000002c err:ole:CoRevokeClassObject called from wrong apartment, should be called from 80000002c err:ntdll:RtlpWaitForCriticalSection section 0x7bc9c624 "loader.c: loader_section" wait timed out in thread 0040, blocked by 0042, retrying (60 sec) susan(a)ubuntu:~/.wine/drive_c/Program Files/Nuance/NaturallySpeaking10/Program$

1 0

Possible regression today's git wine-1.1.19-153-g55ad3fd
by Susan Cragin April 20, 2009

April 20, 2009

Dragon NaturallySpeaking freezes up today during training, without crashing. I will do a regression test if needed, but suspect that the freeze may be already known and/or affect many programs. The terminal output was singularly unhelpful. After deleting all the lines that I see every time often enough to have memorized them, I come up with the following, but none of it happened at the point of crash. Susan err:ole:local_server_thread ConnectNamedPipe failed with error 231 fixme:wininet:set_cookie Unknown additional option L"expires = Thu, 01-Jan-2099 00:00:00 GMT" fixme:winstation:GetUserObjectInformationW not supported index 4 err:ntdll:NtQueryInformationToken Unhandled Token Information class 12! fixme:mstask:MSTASK_ITaskScheduler_Enum 0x15bd40, 0x32f088: stub fixme:mstask:MSTASK_ITask_GetExitCode (0x155c18, 0x32f124): stub fixme:mstask:MSTASK_ITaskScheduler_Enum 0x15bd40, 0x32f088: stub fixme:mstask:MSTASK_ITask_GetExitCode (0x155c18, 0x32f124): stub fixme:mstask:MSTASK_ITaskScheduler_Enum 0x15bd40, 0x32f088: stub fixme:mstask:MSTASK_ITask_GetExitCode (0x155c18, 0x32f124): stub fixme:gdiplus:GdipSetLinePresetBlend not implemented fixme:keyboard:UnregisterHotKey (0x1008c,1): stub fixme:keyboard:RegisterHotKey (0x1008c,1,0x00000000,106): stub

1 0

Re: [3/3] msi: Add tests for importing a _SummaryInformation table. (resend)
by Alexandre Julliard April 20, 2009

April 20, 2009

Hans Leidekker <hans(a)codeweavers.com> writes: > diff --git a/dlls/msi/tests/db.c b/dlls/msi/tests/db.c > index b2c7d06..41510c4 100644 > --- a/dlls/msi/tests/db.c > +++ b/dlls/msi/tests/db.c This fails here: ../../../tools/runtest -q -P wine -M msi.dll -T ../../.. -p msi_test.exe.so db.c && touch db.ok db.c:1679: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1682: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1683: Test failed: Expected 14, got 0 db.c:1686: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1687: Test failed: Expected VT_I2, got 3341460 db.c:1688: Test failed: Expected 1252, got 1617263564 db.c:1692: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1693: Test failed: Expected VT_LPSTR, got 3341460 db.c:1694: Test failed: Expected 18, got 50 db.c:1695: Test failed: Expected "Installer Database", got db.c:1700: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1701: Test failed: Expected VT_LPSTR, got 3341460 db.c:1702: Test failed: Expected "Installer description", got db.c:1707: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1708: Test failed: Expected VT_LPSTR, got 3341460 db.c:1709: Test failed: Expected "WineHQ", got db.c:1714: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1715: Test failed: Expected VT_LPSTR, got 3341460 db.c:1716: Test failed: Expected "Installer", got db.c:1721: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1722: Test failed: Expected VT_LPSTR, got 3341460 db.c:1723: Test failed: Expected "Installer comments", got db.c:1728: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1729: Test failed: Expected VT_LPSTR, got 3341460 db.c:1730: Test failed: Expected "Intel;1033", got db.c:1735: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1736: Test failed: Expected VT_LPSTR, got 3341460 db.c:1737: Test failed: Expected "{12345678-1234-1234-1234-123456789012}", got db.c:1741: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1742: Test failed: Expected VT_FILETIME, got 3341460 db.c:1745: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1746: Test failed: Expected VT_FILETIME, got 3341460 db.c:1749: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1750: Test failed: Expected VT_I4, got 3341460 db.c:1751: Test failed: Expected 200, got 1617263564 db.c:1754: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1755: Test failed: Expected VT_I4, got 3341460 db.c:1756: Test failed: Expected 2, got 1617263564 db.c:1759: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1760: Test failed: Expected VT_I4, got 3341460 db.c:1761: Test failed: Expected 2, got 1617263564 db.c:1765: Test failed: Expected ERROR_SUCCESS, got 6 db.c:1766: Test failed: Expected VT_LPSTR, got 3341460 db.c:1767: Test failed: Expected "Vim", got make[2]: *** [db.ok] Error 44 -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: ntdll: add missing RtlReAllocateHeap valgrind hook, add tests
by Alexandre Julliard April 20, 2009

April 20, 2009

Dan Kegel <dank(a)kegel.com> writes: > More related to http://bugs.winehq.org/show_bug.cgi?id=18067 > This adds the missing valgrind hook, and adds tests that > show up the old problem if run under valgrind. The tests should go with the other existing heap tests in kernel32. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: advapi32: add a couple testcases for GetNamedSecurityInfo (try 3)
by Paul Vriens April 20, 2009

April 20, 2009

Austin English wrote: > Added win_skip for NT4, where it would crash (thanks Paul!). > > > > ------------------------------------------------------------------------ > > Discussed this one briefly with Austin. Still need a little bit of love. -- Cheers, Paul.

1 0