wine-devel September 2015

wine-devel@winehq.org

73 participants
180 discussions

Start a nNew thread

Re: rsaenh/tests: Adjust length value to pass on Win 10 (try 2)
by Marvin 16 Sep '15

16 Sep '15

1 0

Re: rsaenh/tests: Adjust length value to pass on Win 10
by Marvin 16 Sep '15

16 Sep '15

1 0

The fg-xxx VMs now accessible on demand
by Francois Gouget 16 Sep '15

16 Sep '15

So as you may know I run WineTest in a variety of Windows configurations daily (28). The results can be seen on test.winehq.org but the problem is investigating the test failures because most of these configurations cannot be found on the Wine TestBot. I have three options to work around this: * As I mentionned before (1), if you're interested in a specific configuration I may be able to temporarily switch a WineTestBot VM to that configuration. * Another option is to send me an executable to run in the VM and I'll send back the results. * And the new option is that I can now give you VNC access to the VM. Access is open to any Wine developer (i.e. you got a patch accepted in Wine). To get access to the VM contact me on IRC or send me an email and give me the tag of the VM configuration you're interested in (e.g. fg-winxp-lusr). Then I'll start up the VM in the right configuration and send you the address to use to connect to the VM. Of course if I'm asleep or out you may have to wait a bit to get access. Also my uplink is not very fast so expect slow refreshes and some lag. But in the short test I've done it seems to be usable (at least to someone who's used to transatlantic VNC links with an elastic mouse pointer). And for reference here is a list of the available VM configurations. We may not care about some of them, for instance those running the earlier service pack revisions. Let me know if that's the case. Windows 2000 ------------ The French Windows 2000 Professional SP4 installed from an official Microsoft CD. Unless otherwise noted the tests are run in an administrator account. fg-win2000-4sp IE 5.00 SP2 fg-win2000-ie55 = 4sp + IE 5.5 SP2, MSXML3 fg-win2000-ie6 = ie55 + IE 6.0 SP1 fg-win2000-up1 = ie6 + all updates up to 2011/03/21 fg-win2000-uplast = up1 + DirectX 9.0c, WMP 9, .Net 1.1 & 2.0 fg-win2000-rusr = uplast but in a 'Restricted User' account fg-win2000-susr = uplast but in a 'Power User' account Windows XP ---------- I created this VM from the Compaq Windows XP partition of my (former) computer. So this started with a number of extra non-Microsoft applications (blasts from the past: Netscape 7, RealOne Player, etc.). On the one hand this is more representative of what Windows users normally have and on the other the configuration is not as 'straight out of the box' as the other VMs. That may not matter though. fg-winxp-1sp XP SP 1, IE 6.0 SP1, DirectX 8.1 fg-winxp-2sp = 1sp + XP SP2, IE 6.0, DirectX 9.0c, .Net 1.0 SP3 fg-winxp-2spie7 = 2sp + IE 7.0, WMP 10, .Net 1.1 fg-winxp-3sp = 2spie7 + XP SP3, WMP 11, .Net 2.0 SP2 & 3.0 SP2 & 3.5 SP1 fg-winxp-3spie8 = 3sp + IE 8 fg-winxp-last = 3spie8 + all updates up to 2011/09/08 fg-winxp-lusr = last but in a 'Limited User' account Windows 7 --------- Installed from an official English Windows 7 Ultimate CD. Unless otherwise noted the tests are run in an administrator account with Elevated Privileges. fg-win7u64-0sp Base install + French language pack (but still configured for English) fg-win7u64-1sp = 0sp + SP1 fg-win7u64-1spie9(64) = 1sp + IE 9.0, all language packs fg-win7u64-1spie9adm = 1spie9 but no 'Elevated Privileges' fg-win7u64-1spie9usr = 1spie9 but in a 'Standard user' account fg-win7u64-1spie9-ar = 1spie9 but in an Arabic locale fg-win7u64-1spie9-de = 1spie9 but in a German locale fg-win7u64-1spie9-es = 1spie9 but in a Spanish locale fg-win7u64-1spie9-fr = 1spie9 but in a French locale fg-win7u64-1spie9-he = 1spie9 but in a Hebrew locale fg-win7u64-1spie9-ja = 1spie9 but in a Japanese locale fg-win7u64-1spie9-ko = 1spie9 but in a Korean locale fg-win7u64-1spie9-mx = 1spie9 but in a mixed locale (1) https://www.winehq.org/pipermail/wine-devel/2015-September/109252.html -- Francois Gouget <fgouget(a)free.fr> http://fgouget.free.fr/ The nice thing about meditation is that it makes doing nothing quite respectable -- Paul Dean

1 0

Re: [PATCH 1/4] xaudio2: Implement IXAudio2SourceVoice frequency functions
by Chris Robinson 16 Sep '15

16 Sep '15

On 09/15/2015 10:25 AM, Andrew Eikum wrote: > @@ -801,14 +802,26 @@ static HRESULT WINAPI XA2SRC_SetFrequencyRatio(IXAudio2SourceVoice *iface, > float Ratio, UINT32 OperationSet) > { > XA2SourceImpl *This = impl_from_IXAudio2SourceVoice(iface); > + > TRACE("%p, %f, 0x%x\n", This, Ratio, OperationSet); > + > + EnterCriticalSection(&This->lock); > + > + This->freq_ratio = Ratio; > + This->freq = This->fmt->nSamplesPerSec * Ratio; > + > + LeaveCriticalSection(&This->lock); > + > return S_OK; ... > @@ -816,7 +829,16 @@ static HRESULT WINAPI XA2SRC_SetSourceSampleRate( > UINT32 NewSourceSampleRate) > { > XA2SourceImpl *This = impl_from_IXAudio2SourceVoice(iface); > + > TRACE("%p, %u\n", This, NewSourceSampleRate); > + > + EnterCriticalSection(&This->lock); > + > + This->fmt->nSamplesPerSec = NewSourceSampleRate; > + This->freq = NewSourceSampleRate * This->freq_ratio; > + > + LeaveCriticalSection(&This->lock); > + > return S_OK; > } ... > @@ -2989,7 +3012,7 @@ static BOOL xa2buffer_queue_period(XA2SourceImpl *src, XA2Buffer *buf, ALuint al > buf->offs_bytes += submit_bytes; > > alBufferData(al_buf, src->al_fmt, submit_buf, submit_bytes, > - src->fmt->nSamplesPerSec); > + src->freq); > > alSourceQueueBuffers(src->al_src, 1, &al_buf); > This doesn't strike me as correct. Assuming the SampleRate or FrequencyRatio can be altered in real-time, this will break the OpenAL buffer queue which needs to have the same sample rate on each buffer. SetFrequencyRatio looks like it'd map directly to the AL source's pitch property, while SetSourceSampleRate may be a bit trickier depending on when it's allowed to be called (it'd require stopping the source and clearing the queue, then rebuffering and requeueing the buffers to start where it left off). Also, I notice the OperationSet parameter isn't used. What is that for?

2 1

RFC: Hybrid (wineserver / process-local) native semaphores
by Daniel Santos 16 Sep '15

16 Sep '15

This patch set fixes most performance problems caused by ReleaseSemaphore() and WaitForSingle/MultipleObject(s) making server calls. One victim of this is Star Wars Battlefront (https://bugs.winehq.org/show_bug.cgi?id=29582), where the majority of the CPU time is spent on context switching while the program spams ReleaseSemaphore, WaitForSingle/MultipleObject(s) and GetForgroundWindow, the later of which I made a work-around hack for and that some players have been using. This patch set doubles performance of bug #29582. (When combined with GetForgroundWindow hack the problem is completely resolved.) The patch set works by having the server create a POSIX semaphore object and sharing the key to that object with the client process, enabling the client process to be able to implement ReleaseSemaphore and optimistic-case wait calls (where no blocking is reburied) without a server call. Blocking waits and any wait-multiple that cannot be resolved in the client process (e.g., bWaitAll=TRUE and objects include non-semaphores) is still handled by the server. (Implementing blocking wait calls on the client can yield some performance improvements because a context switch to another thread in the same program won't require swapping out the memory map & such, but I would expect this to be less significant.) However, upon further experimentation, I discovered that POSIX semaphores in glibc are actually implemented using a shared memory page, which may not be acceptable since a bad process can corrupt that page and potentially cause sem_* function calls in the server to fail as well as other client programs fail and/or deadlock. I am working on a System V adaptation, but I thought it would be a good idea to see feedback & comments now. Another problem is that this causes the threadpool test fails at line 1299, where the previous "release all semaphores and wait for callback" test is done in reverse order. I presume this is due to the nature of the linux scheduler being inconsistent with how Windows *happens* to schedule its threads. I have an idea for a fix for this already, but I will still have to dig deeper into it. The code is still in experimental quality (assert(0)s and such) and I've already re-worked the configure.ac stuff, I'm mostly concerned with feedback on the general scheme. Thanks! Daniel

5 46

Re: [PATCH 1/3] oledb32: Add DataConvert support of DBTYPE_VARIANT(NULL) to DBTYPE_DBDATE
by Nikolay Sivov 16 Sep '15

16 Sep '15

If it acts the same for all source variant types it looks like it'd be better to test them all and have this case handled once.

2 1

Re: [3/3] Add period to status bar, dialog box messages (try 2)
by Nikolay Sivov 16 Sep '15

16 Sep '15

On 15.09.2015 11:28, Florian Pelz wrote: > STRINGTABLE > { > - IDS_STATUS_DONE "Done" > + IDS_STATUS_DONE "Done." > IDS_STATUS_DOWNLOADINGFROM "Downloading from %s..." > } I don't think it's better really to have punctuation in status bars. Probably messages can live without it, unless they consist of multiple sentences, then last one needs termination too.

2 2

Re: [PATCH 1/2] user32: Implement semi-stub for FlashWindowEx. (resend)
by Marvin 16 Sep '15

16 Sep '15

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 https://testbot.winehq.org/JobDetails.pl?Key=16750 Your paranoid android. === w8 (32 bit win) === win.c:3417: Test failed: hwnd 000601B0 message 0118 === w864 (32 bit win) === win.c:3417: Test failed: hwnd 0003012C message 0118 === w1064 (32 bit win) === win.c:6658: Test failed: 0x80/0x1c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6680: Test failed: 0x80/0x1010000: window rect 0,0-1024,768 win.c:6658: Test failed: 0x80/0x81c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6658: Test failed: 0/0x41c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6658: Test failed: 0x40000/0x41c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6658: Test failed: 0x80/0x41c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6658: Test failed: 0x80/0x1c50000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6680: Test failed: 0x80/0x1050000: window rect 0,0-1024,768 win.c:6658: Test failed: 0x80/0x1c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6680: Test failed: 0x80/0x1410000: window rect 0,0-1024,768 === w1064 (64 bit win) === win.c:6658: Test failed: 0x80/0x1c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6680: Test failed: 0x80/0x1010000: window rect 0,0-1024,768 win.c:6658: Test failed: 0x80/0x81c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6658: Test failed: 0/0x41c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6658: Test failed: 0x40000/0x41c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6658: Test failed: 0x80/0x41c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6658: Test failed: 0x80/0x1c50000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6680: Test failed: 0x80/0x1050000: window rect 0,0-1024,768 win.c:6658: Test failed: 0x80/0x1c10000: window rect 0,0-1024,768 must be in 0,0-1024,728 win.c:6680: Test failed: 0x80/0x1410000: window rect 0,0-1024,768

1 0

Re: disabling wow64, testing 64-bit wine-mono P/invoke
by Hin-Tak Leung 16 Sep '15

16 Sep '15

------------------------------ On Wed, Sep 16, 2015 3:45 AM BST Vincent Povirk wrote: >> Is there any way of doing this with 64-bit wine and 64-bit wine-mono? > >Currently, Wine does not have the ability to load pure .NET PE images >in a 64-bit process. > >Ntdll needs to call _CorValidateImage during the loading process. >_CorValidateImage needs to convert the image, in memory, to the >correct architecture. > >It's complicated, and I don't have a solution. Actually you do :-). Using http://madewokherd.nfshost.com/omgsecret/mono-64bit-exe.tar.gz from https://bugs.winehq.org/show_bug.cgi?id=30518 It is a bit dated, but does the job. I was very confused by both of libmono-2.0-x86_64.dll libmono-2.0-x86.dll MonoPosixHelper-x86_64.dll MonoPosixHelper-x86.dll in c:/windows/mono/mono-2.0/bin/ In fact libmono-2.0-x86_64.dll MonoPosixHelper-x86_64.dll are never used, at the moment. When 64-bit mono loads a .NET PE image which then P/Invoke, it does so P/Invoke to 64-bit dlls.

2 1

Re: [PATCH 2/2] winex11.drv: Implement FlashWindow with _NET_WM_STATE_DEMANDS_ATTENTION. (try 2)
by Dmitry Timoshkov 16 Sep '15

16 Sep '15

Jactry Zeng <jzeng(a)codeweavers.com> wrote: > +void X11DRV_FlashWindow( HWND hwnd, BOOL bInvert ) > +{ > + struct x11drv_win_data *data = get_win_data( hwnd ); > + XEvent xev; > + > + if (!data) > + return; > + > + memset( &xev, 0, sizeof(xev) ); > + xev.type = ClientMessage; > + xev.xclient.window = data->whole_window; > + xev.xclient.message_type = x11drv_atom( _NET_WM_STATE ); > + xev.xclient.format = 32; > + xev.xclient.data.l[0] = 1; > + xev.xclient.data.l[1] = XInternAtom( data->display, "_NET_WM_STATE_DEMANDS_ATTENTION", > + False ); > + > + XSendEvent( data->display, DefaultRootWindow( data->display ), False, > + SubstructureNotifyMask, &xev ); > + release_win_data( data ); > +} This can't work, please have a look how XSendEvent is used in other parts of x11drv. -- Dmitry.

1 0

← Newer
1
...
5
6
7
8
9
10
11
...
18
Older →

Jump to page:

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 September 2015