Bug 2220 - RFE: Make glxgears a better GL client
Summary: RFE: Make glxgears a better GL client
Status: RESOLVED INVALID
Alias: None
Product: Mesa
Classification: Unclassified
Component: Demos (show other bugs)
Version: git
Hardware: All All
: high enhancement
Assignee: Roland Mainz
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on: 1426
Blocks:
  Show dependency treegraph
 
Reported: 2005-01-04 13:53 UTC by Roland Mainz
Modified: 2011-10-15 16:04 UTC (History)
2 users (show)

See Also:
i915 platform:
i915 features:


Attachments
Patch for 2005-01-04-trunk (13.25 KB, patch)
2005-01-04 14:02 UTC, Roland Mainz
Details | Splinter Review
Proposed patch to restore original behaviour (2.04 KB, patch)
2005-01-04 18:13 UTC, Roland Mainz
Details | Splinter Review
Backout attachment #1630, patch for 2005-03-07-trunk (13.63 KB, patch)
2005-03-07 11:37 UTC, Roland Mainz
Details | Splinter Review

Description Roland Mainz 2005-01-04 13:53:36 UTC
RFE: Make glxgears a better GL client.
Currently glxgears sends GL rendering instructions in an endless loop to the GL
engine without waiting for any response, making interactive usage of the
application quite problematic for GL implementations which have a large command
queue (in the case of one Win32 Xserver product it even rendered the whole
Xserver inaccessible as the GL engine allowed an nearly unlimited amount of
instructions queued - in that case it took around 5mins between a keypress and
the matching response).
Comment 1 Roland Mainz 2005-01-04 14:02:17 UTC
Created attachment 1630 [details] [review]
Patch for 2005-01-04-trunk
Comment 2 Roland Mainz 2005-01-04 14:06:09 UTC
Patch checked-in...

/cvs/xorg/xc/ChangeLog,v  <--  ChangeLog
new revision: 1.641; previous revision: 1.640
/cvs/xorg/xc/programs/glxgears/glxgears.c,v  <--  glxgears.c
new revision: 1.9; previous revision: 1.8
Mailing the commit message to xorg-commit@lists.freedesktop.org...

... marking bug as FIXED.
Comment 3 Adam Jackson 2005-01-04 14:30:58 UTC
please don't do this when using a direct context.
Comment 4 Roland Mainz 2005-01-04 18:13:01 UTC
Created attachment 1631 [details] [review]
Proposed patch to restore original behaviour

The patch turns off the usage of |glFinish()| (but still allows to turn it "on"
again via a command-line option) ...

ajax: Is the patch OK for you ?
Comment 5 Adam Jackson 2005-01-24 12:46:00 UTC
the issue here is that the software renderer is a pig.  changing glxgears is not
a solution.
Comment 6 Roland Mainz 2005-01-25 22:29:24 UTC
(In reply to comment #5)
> the issue here is that the software renderer is a pig.  changing glxgears is not
> a solution.

Agreed. I either backout the complete patch as-is or apply attachment #1631 [details] [review] ...
right now I am waiting for feedback from Sun staff about the original issue (but
if I am not getting feedback by friday I just rip the patch out... question is
then what we should do with the timing patch added a while ago - should we make
it optional or what ?). 
Comment 7 Roland Mainz 2005-03-07 11:29:55 UTC
OK... I'll back the junk out as there is no response from Sun yet... ;-(
Comment 8 Roland Mainz 2005-03-07 11:37:57 UTC
Created attachment 2035 [details] [review]
Backout attachment #1630 [details] [review], patch for 2005-03-07-trunk
Comment 9 Roland Mainz 2005-03-07 11:40:57 UTC
Attachment #1630 [details] backed-out...

/cvs/xorg/xc/ChangeLog,v  <--  xc/ChangeLog
new revision: 1.801; previous revision: 1.800
/cvs/xorg/xc/programs/glxgears/glxgears.c,v  <-- 
xc/programs/glxgears/glxgears.c
new revision: 1.10; previous revision: 1.9
Mailing the commit message to xorg-commit@lists.freedesktop.org...

... marking bug as INVALID.
Comment 10 Roland Mainz 2005-03-07 11:41:51 UTC
ajax:
Is there a bug which asks to get the issue fixed in the Xorg GLX implementation
?


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.