Bug 22537 - xorg freeze onf xorg 1.6.1.901 with refer to intel drivers [sony vaio sr19xn and dell mini 9]
Summary: xorg freeze onf xorg 1.6.1.901 with refer to intel drivers [sony vaio sr19xn ...
Status: RESOLVED WONTFIX
Alias: None
Product: xorg
Classification: Unclassified
Component: Server/General (show other bugs)
Version: unspecified
Hardware: x86 (IA32) Linux (All)
: high critical
Assignee: Keith Packard
QA Contact: Xorg Project Team
URL: https://bugs.gentoo.org/show_bug.cgi?...
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2009-06-29 11:10 UTC by Giovanni Pellerano
Modified: 2015-02-25 15:49 UTC (History)
8 users (show)

See Also:
i915 platform:
i915 features:

Attachments
gentoo patchset 1.6.1.901-r5 (2.55 KB, text/plain)
2009-06-29 11:12 UTC, Giovanni Pellerano 		no flags Details
gentoo patchset 1.6.1.901-r4 (1.22 KB, text/plain)
2009-06-29 11:20 UTC, Giovanni Pellerano 		no flags Details
View All

Description Giovanni Pellerano 2009-06-29 11:10:04 UTC 
Hi Gordon, Hi Jesse

I'm sorry to have bad news about the problem on the sony vaio sr19xn;
i think the freeze was never solved in the xorg tree.

I'm used xorg-server 1.6.1.901 on gentoo for some monts without problem with refer to the gentoo ebuild r1,r2,r3,r4 but now with the r5 I re-having the same freeze of some times ago on the Sony Vaio sr19xn.

now I've tested it also on a Dell Mini 9 and I've the same problem on it too.
reverting to the r4 solve the problem.

I'm attaching the list of the patchset included in the two gentoo ebuilds:

1.6.1.901-r4 1.6.1.901-r5 


here is the bug report i've filled for the gentoo guys: https://bugs.gentoo.org/show_bug.cgi?id=275852

thanks for all as usual.

regards

Giovanni Pellerano
Comment 1 Giovanni Pellerano 2009-06-29 11:12:34 UTC 
Created attachment 27225 [details]
gentoo patchset 1.6.1.901-r5
Comment 2 Giovanni Pellerano 2009-06-29 11:20:43 UTC 
Created attachment 27226 [details]
gentoo patchset 1.6.1.901-r4
Comment 3 Gordon Jin 2009-06-29 19:34:40 UTC 
So this sounds like an xserver regression. Could you figure out which patch in 1.6.1.901-r5 causes the regression? Thanks.
Comment 4 Giovanni Pellerano 2009-06-29 23:30:27 UTC 
sorry gordon at the moment i've problems to test it;
i will do; please keep me updated of the informations or test you need end if any other fill a similar bug report.

thanks
Comment 5 Priit Laes (irc: plaes) 2009-06-30 07:11:02 UTC 
(In reply to comment #3)
> So this sounds like an xserver regression. Could you figure out which patch in
> 1.6.1.901-r5 causes the regression? Thanks.
> 
> 
For me it was the following patch:

DRI2: Implement protocol for DRI2GetBuffersWithFormat

This change implements the protocol for DRI2GetBuffersWithFormat, but
the bulk of the differences are the changes to the extension / driver
interface to make this function work.  The old CreateBuffers and
DeleteBuffers routines are replaced with CreateBuffer and DeleteBuffer
(both singular).

This allows drivers to allocate buffers for a drawable one at a time.
As a result, 3D drivers can now allocate the (fake) front-buffer for a
window only when it is needed.  Since 3D drivers only ask for the
front-buffer on demand, the real front-buffer is always created.  This
allows CopyRegion impelemenations of SwapBuffers to continue working.
As with previous version of this code, if the client asks for the
front-buffer for a window, we instead give it the fake front-buffer.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kristian Høgsberg <krh@redhat.com>
Comment 6 Rémi Cardona 2009-06-30 07:17:56 UTC 
Oh and it seems only KMS is affected, UMS works fine.

Thanks
Comment 7 Julien Cristau 2009-06-30 07:20:10 UTC 
(In reply to comment #3)
> So this sounds like an xserver regression. Could you figure out which patch in
> 1.6.1.901-r5 causes the regression? Thanks.
> 
see http://lists.freedesktop.org/pipermail/xorg/2009-June/046267.html i think this is actually a driver bug.
Comment 8 Rémi Cardona 2009-06-30 10:49:46 UTC 
(In reply to comment #7)
> see http://lists.freedesktop.org/pipermail/xorg/2009-June/046267.html i think
> this is actually a driver bug.

Confirming, we did some searching with a user whose X wouldn't freeze but segfault.

Basically, [1] does not break the ABI like Julien said, but disables DRI2 on older drivers (2.7.1 in our case) or with drivers built with previous versions of the server.

However, with DRI2 disabled, the driver tries to initialize DRI1 which for some reason also fails. Hell then breaks loose and the driver segfaults later on in drm_intel_bo_alloc_for_render (IIRC, the function is from memory, I need to look it up).

Hope that helps

[1] http://cgit.freedesktop.org/xorg/xserver/commit/?h=server-1.6-branch&id=4fad615d689c61c6c3a000295a1fa755359737cb
Comment 9 Chris Wilson 2015-02-25 15:49:32 UTC 
We can safely say this is obsolete now, as these machines no longer even have UMS/DRI1.

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.