On Sat, 24 Oct 2015, Gerald Pfeifer wrote:
> This avoids potential shift overflows. I only should be necessary
> for D3DVSD_TOKENTYPESHIFT, but it makes sense to have this consistent.
Correction, this per se does not avoid shift overflows, I sent an
older patch. It still may make sense to apply (please advise), but
we'd need a different change in addition. I'll have a look.
Gerald
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=17842
Your paranoid android.
=== w1064 (32 bit inputprocessor) ===
inputprocessor.c:791: Test failed: Deactivation came unexpectedly
Hi Matteo,
On 29/10/15 08:27, Matteo Bruni wrote:
> Signed-off-by: Matteo Bruni <mbruni(a)codeweavers.com>
> ---
> This avoids a crash on exit with World of Warships when the OS version
> is set to Windows 7.
>
> dlls/msctf/threadmgr.c | 145 ++++++++++++++++++++++++++++---------------------
> include/msctf.idl | 14 +++++
> 2 files changed, 96 insertions(+), 63 deletions(-)
Can you please provide tests?
Best Regards
Alistair Leslie-Hughes
On Tue, Oct 20, 2015 at 12:15:22PM +0900, Alexandre Julliard wrote:
> Andrew Eikum <aeikum(a)codeweavers.com> writes:
>
> > Hm, it works here whether PA is running or not. What's your setup? PA
> > is obviously installed. Is it running? What version do you have
> > installed? Can you give me a +devenum,+mmdevapi,+pulse,+alsa,+tid,+seh
> > log?
>
> Version 7.0, it's running AFAICT. Log is attached. Note that this is
> with only the first patch applied, it also crashes (but not always which
> is why I missed it at first). Other sound tests crash as well.
>
This was seen in my patchset, and not in wine-staging, because we
unload the pulse driver after it returns Low priority which is lower
than winealsa's Preferred priority. Winepulse misbehaved when it was
unloaded. In wine-staging, the Pulse driver is always Preferred and so
would never be unloaded.
I believe the crash happened because the DLL would be unloaded while
the PA mainloop thread was still running. DllMain does call
pa_mainloop_quit, but doesn't wait for the thread to actually exit.
The mainloop thread would occasionally trigger a callback after the
module has been unloaded, before acknowledging the quit status. This
is why you didn't always see the crash.
This has been remedied in the latest patchset. We now wait for the
thread to quit in DllMain, see patch 3/5.
Andrew
On 10/29/15 15:50, Daniel Lehman wrote:
> +int __cdecl _Thrd_join(_Thrd_t thr, int *unk)
> +{
> + DWORD ret;
> +
> + TRACE("(%p %u %p)\n", thr.hnd, thr.id, unk);
> + ret = WaitForSingleObject(thr.hnd, INFINITE);
> + if (unk)
> + GetExitCodeThread(thr.hnd, (DWORD *)unk);
Please rename the unk parameter. You already know that it's a thread
exit code.
You should probably also exit earlier if WaitForSingleObject fails. It
doesn't really make sense call GetExitCodeThread in this case. Probably
there's also no need to close the handle if WaitForSingleObject fails.
Thanks,
Piotr
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=17823
Your paranoid android.
=== build (build) ===
Patch failed to apply
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=17817
Your paranoid android.
=== build (build) ===
Patch failed to apply
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=17816
Your paranoid android.
=== build (build) ===
Patch failed to apply