[Bug 75279] New: XCloseDisplay() takes one minute around nouveau_dri.so, freezing Firefox startup

* [Bug 75279] New: XCloseDisplay() takes one minute around nouveau_dri.so, freezing Firefox startup
@ 2014-02-20 20:45 bugzilla-daemon
  2014-02-20 23:37 ` [Bug 75279] " bugzilla-daemon
  0 siblings, 1 reply; 2+ messages in thread
From: bugzilla-daemon @ 2014-02-20 20:45 UTC (permalink / raw)
  To: dri-devel

[-- Attachment #1.1: Type: text/plain, Size: 1852 bytes --]

https://bugs.freedesktop.org/show_bug.cgi?id=75279

          Priority: medium
            Bug ID: 75279
          Assignee: dri-devel@lists.freedesktop.org
           Summary: XCloseDisplay() takes one minute around
                    nouveau_dri.so, freezing Firefox startup
          Severity: normal
    Classification: Unclassified
                OS: Linux (All)
          Reporter: bjacob@mozilla.com
               URL: https://bugzilla.mozilla.org/show_bug.cgi?id=973192
          Hardware: x86-64 (AMD64)
            Status: NEW
           Version: XOrg CVS
         Component: DRM/other
           Product: DRI

Hi, I'm just forwarding to you a bug report originally made against Firefox:
https://bugzilla.mozilla.org/show_bug.cgi?id=973192

The symptom is Firefox freezing on startup. The cause (see around comment
39--42 on that bug) is that a XCloseDisplay() call takes about one minute to
return.

The original reporter of the bug has recorded a profile which shows time spent
in nouveau_dri.so (see comments 50, 51 on the mozilla bug)

If you want to see what the application code triggering this bug looks like,
it's all self-contained in this file:

http://hg.mozilla.org/mozilla-central/file/tip/toolkit/xre/glxtest.cpp

   228   glXMakeCurrent(dpy, None, nullptr);
   229   glXDestroyContext(dpy, context);
   230   XDestroyWindow(dpy, window);
   231   XFreeColormap(dpy, swa.colormap);
   232   XCloseDisplay(dpy);  // <--- this XCloseDisplay call takes 1 minute
   233   dlclose(libgl);

Is there some more information that you need, that isn't already on the mozilla
bug?

Is this application code doing something evil here?

Is this a known issue? Do you have a guess as to whether it might affect many
users (i.e., should we try working around it)?

-- 
You are receiving this mail because:
You are the assignee for the bug.

[-- Attachment #1.2: Type: text/html, Size: 3423 bytes --]

[-- Attachment #2: Type: text/plain, Size: 159 bytes --]

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 2+ messages in thread