The mandriva nvidia drivers package maintainer just submitted a patch from mandriva's tree to upstream that fixes an ordering issue in the Kernel Compatibility Layer of nvidia's drivers: http://www.nvnews.net/vbulletin/showthread.php?t=152065 This issue seems to affect all versions of the nvidia drivers dating back to 190.53 in the portage tree. What is sad is that this was pointed out on multiple occasions in bug 301318, even before the issue entered the tree, but nothing has been done to fix it. The most recent mention of it is in comment 22 in bug 301318: http://bugs.gentoo.org/show_bug.cgi?id=301318#c22 I believe that this is causing a major problem on my system where it will infrequently cause compositing to fail at boot. Reproducible: Sometimes Steps to Reproduce: See my thread at the nvnews.net forums: http://www.nvnews.net/vbulletin/showthread.php?t=151685 Actual Results: Compositing fails randomly (less than 1/10 of the time) at boot in all drivers, but the properly patched 190.53 drivers. Expected Results: Compositing should always work.
I'm aware of this on the other bug you created. I'm awaiting a response from NVIDIA. *** This bug has been marked as a duplicate of bug 301318 ***
Is there any chance that this could be made a blocker bug for bugs 301318 and 321061? This is an issue common to several versions of the Nvidia drivers, so it probably deserves its own bug, such that any bug reports for specific versions of the nvidia drivers could be blocked by this.
(In reply to comment #2) > Is there any chance that this could be made a blocker bug for bugs 301318 and > 321061? This is an issue common to several versions of the Nvidia drivers, so > it probably deserves its own bug, such that any bug reports for specific > versions of the nvidia drivers could be blocked by this. > As explained by zander on the forums, since 195.x.y the calls expand out to the same value so its a non-issue.
(In reply to comment #3) > (In reply to comment #2) > > Is there any chance that this could be made a blocker bug for bugs 301318 and > > 321061? This is an issue common to several versions of the Nvidia drivers, so > > it probably deserves its own bug, such that any bug reports for specific > > versions of the nvidia drivers could be blocked by this. > > > > As explained by zander on the forums, since 195.x.y the calls expand out to the > same value so its a non-issue. > The NULL is in the 7th parameter in Nvidia's version when it should be in the 5th parameter. That is not quite the same thing. Do you have a link to zander's post?