This is the mail archive of the cygwin-xfree mailing list for the Cygwin XFree86 project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Windows 7 Aero mode issue again


On 26/10/2010 23:58, Eliot Moss wrote:
Ok, here's a first email about *some* symptoms ...

If I run the latest installed XWin, or this one:
ftp://cygwin.com/pub/cygwinx/XWin.20101026-git-6105a1d4e1e137f0.exe.bz2

without -resize and then "sleep" and unsleep the
laptop, X does not paint anything but the cursor
and keyboard events do not seem to go to the
(invisible) windows. I have to kill the server
and all the X jobs. I will now proceed to the
other tests you suggested ...

Hmmm... I don't think you've mentioned this before.


I assume this is a regression from 1.8.0 (before -resize was added), as we used to survive a hibernate cycle?

You're probably right about earlier version being
sensitive to whether -resize was given. I had
forgotten about that dependency of behavior.

On 27/10/2010 00:15, Eliot Moss wrote: > Ok ... the 20101026 snapshot works properly with -resize and > -engine 1 set: DWM does not go away or get complained about > if I sleep then unsleep. Yay!

Good.

Just to point out this isn't specific to the snapshot, though, -engine 1 should workaround the problem in all cases.

I guess this confirms that it is a problem with our use of directdraw and a hibernate cycle. (as the workaround forces XWin to use the shadow GDI drawing engine rather than the shadow directdraw drawing engine)

However, I can't see what is wrong with XWin's use of ddraw. We never try to draw directly to the primary surface (which is what causes DWM to be shut down), so why it thinks we are during a hibernate or resume is obscure.

It would probably be a good idea to test with some other ddraw application over a hibernate, to confirm this is a problem with XWin and not with the intel video drivers, but I'm not sure what other ddraw application is similar enough to make this a valid test (The simple tests in dxdiag probably wouldn't prove anything)

> I did notice one little thing: its log file seems to go to
> /var/log/XWin.0.log rather than /var/log/xwin/XWin.0.log.

I hadn't updated the script I use to build snapshots. Thanks for pointing this out, I've now corrected that.

> I include the -logverbose 3 output for you, and the stderr.

Sorry, I was a little unclear. If you are still interested in pursing this, a log without -engine 1 and with -logverbose 3 would be most helpful. If you could annotate it to show at approximately what timestamp the suspend and resume occurred, that would be even better.

--
Jon TURNEY
Volunteer Cygwin/X X Server maintainer

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://x.cygwin.com/docs/
FAQ:                   http://x.cygwin.com/docs/faq/


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]