Bug#687442: [Backport] Use DRM/KMS subsystem from Linux 3.4 in wheezy
Le samedi 05 janvier 2013 à 17:25 +0100, Julien Cristau a écrit :
> On Fri, Jan 4, 2013 at 17:13:35 +0100, Julien Cristau wrote:
>
> > http://people.debian.org/~jcristau/linux-image-3.2.0-4.drm-686-pae_3.2.35-3~jcristau.1_i386.deb
> > sha1sum 17980ee5bafc40bb4e7bf42576ce40ac81c75833
> >
> > http://people.debian.org/~jcristau/linux-image-3.2.0-4.drm-amd64_3.2.35-3~jcristau.1_amd64.deb
> > sha1sum f6711fe6d0d924aab82ec82fe1a86102a69a8c32
Hi Julien et al.
I tested with an Asus UX31A (Ivybridge Mobile GT2, eDP1 display), using
2D and 3D drivers from Wheezy (respectively versions 2.19.0-6 and
8.0.5-3).
On the DE side, I used GNOME Shell and Cinnamon (both need 3D
acceleration).
On the kernel side, I used:
i915.i915_enable_rc6={0, 1, 3} → no issue (don't know if it worth to
test RC6pp)
i915.semaphores=1 → ditto
In order to stress the GPU, I played some video bits in 1080p mostly
using VAAPI, played video games (native and emulated), and the usual
glxgear to see if all went well, and that's a *success*, the kernel is
pretty stable (contrary to the "raw" 3.2 series which causes some
freezes on this system, but that's another story I need to
investigate ;-)).
Voilà, to me this kernel is in really good shape, so thank you Julien
for backporting this stuff!
TODO:
- Testing VGA and HDMI outputs
- Testing my old GPUs to prevent regressions (Radeon X800XL and Nvidia
7600GS)
- Testing SNA (?)
- Advertising about this kernel (I found it haphazardly by diving in the
BTS)
> >
> Added
> http://people.debian.org/~jcristau/linux-image-3.2.0-4.drm-486_3.2.35-3~jcristau.1_i386.deb
> sha1sum 76f67adbed522f8e42216f8595c8b5d1699af9ec
>
> Cheers,
> Julien
Regards,
Vincent
PS: By the way if Greg KH tags a new longterm kernel, will there be a
chance to see the DRM subsystem backported in Wheezy or that's
definitely too invasive and we'll stay with 3.4 (I don't have any issue
with that, I'm just curious).
Reply to: