* Re: Fwd: Please revert wrong kernel OOPS patch in -69 !
2016-04-07 13:05 ` Fwd: Please revert wrong kernel OOPS patch in -69 ! Nicolai Hähnle
@ 2016-04-07 13:31 ` Luis Henriques
2016-04-18 1:34 ` Greg Kroah-Hartman
1 sibling, 0 replies; 3+ messages in thread
From: Luis Henriques @ 2016-04-07 13:31 UTC (permalink / raw)
To: Nicolai Hähnle
Cc: Taras Prokopenko, christian.koenig, stable, Kamal Mostafa,
Sasha Levin, Jiri Slaby, Greg Kroah-Hartman
On Thu, Apr 07, 2016 at 08:05:53AM -0500, Nicolai H�hnle wrote:
> Hmm, I was under the impression from an email I got that this was already
> reverted from 3.16. Clearly, communication is hard.
>
> Let me re-iterate, to be sure: kernels 3.17 and older
>
> 1. need to revert the backport of upstream commit
> f6ff4f67cdf8455d0a4226eeeaf5af17c37d05eb (I think most have done that
> already) and
>
> 2. should apply the corrected backport which I am attaching yet again (seems
> like when I sent this patch around two weeks ago, it wasn't picked up).
>
Thanks a lot for reiterating this, Nicolai.
So, you're absolutely right: this *was* reverted in the 3.16 kernel,
in version 3.16.7-ckt26 (and afaik from any other stable trees).
Unfortunately, version 3.16.7-ckt25 was actually released *before*
this regression was identified.
I'm not sure which other stable kernels were actually released with
this regression, but there was definitely a window of time where this
could have happen.
Cheers,
--
Lu�s
> Thanks,
> Nicolai
>
> On 07.04.2016 04:15, Taras Prokopenko wrote:
> >
> >---------- Forwarded message ----------
> >From: *Taras Prokopenko* <taras.prokopenko@gmail.com
> ><mailto:taras.prokopenko@gmail.com>>
> >Date: 2016-04-07 12:04 GMT+03:00
> >Subject: Please revert wrong kernel OOPS patch in -69 !
> >To: luis.henriques@canonical.com <mailto:luis.henriques@canonical.com>
> >Cc: jslaby@suse.cz <mailto:jslaby@suse.cz>
> >
> >
> >Please test your changes! Lost 1/2 a day to figure out whats causing my
> >pc to hang.
> >
> >OopsText:
> > BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
> > IP: [<ffffffffc03b83ed>] radeon_fence_ref+0xd/0x50 [radeon]
> > PGD 0
> > Oops: 0002 [#1] SMP
> >
> >
> >https://lists.ubuntu.com/archives/kernel-team/2016-February/072148.html
> >
> >Damn
> >
> From ad94965f69c2681832f64473d28c23ae71b6e52f Mon Sep 17 00:00:00 2001
> From: =?UTF-8?q?Nicolai=20H=C3=A4hnle?= <nicolai.haehnle@amd.com>
> Date: Tue, 15 Mar 2016 12:56:45 -0500
> Subject: [PATCH] drm/radeon: hold reference to fences in radeon_sa_bo_new
> (3.17 and older)
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
>
> [Backport of upstream commit f6ff4f67cdf8455d0a4226eeeaf5af17c37d05eb, with
> an additional NULL pointer guard that is required for kernels 3.17 and older.
>
> To be precise, any kernel that does *not* have commit 954605ca3 "drm/radeon:
> use common fence implementation for fences, v4" requires this additional
> NULL pointer guard.]
>
> An arbitrary amount of time can pass between spin_unlock and
> radeon_fence_wait_any, so we need to ensure that nobody frees the
> fences from under us.
>
> Based on the analogous fix for amdgpu.
>
> Signed-off-by: Nicolai H??hnle <nicolai.haehnle@amd.com>
> Reviewed-by: Christian K??nig <christian.koenig@amd.com> (v1 + fix)
> Tested-by: Lutz Euler <lutz.euler@freenet.de>
> Cc: stable@vger.kernel.org
> ---
> drivers/gpu/drm/radeon/radeon_sa.c | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/drivers/gpu/drm/radeon/radeon_sa.c b/drivers/gpu/drm/radeon/radeon_sa.c
> index f0bac68..8962411 100644
> --- a/drivers/gpu/drm/radeon/radeon_sa.c
> +++ b/drivers/gpu/drm/radeon/radeon_sa.c
> @@ -349,8 +349,15 @@ int radeon_sa_bo_new(struct radeon_device *rdev,
> /* see if we can skip over some allocations */
> } while (radeon_sa_bo_next_hole(sa_manager, fences, tries));
>
> + for (i = 0; i < RADEON_NUM_RINGS; ++i) {
> + if (fences[i])
> + radeon_fence_ref(fences[i]);
> + }
> +
> spin_unlock(&sa_manager->wq.lock);
> r = radeon_fence_wait_any(rdev, fences, false);
> + for (i = 0; i < RADEON_NUM_RINGS; ++i)
> + radeon_fence_unref(&fences[i]);
> spin_lock(&sa_manager->wq.lock);
> /* if we have nothing to wait for block */
> if (r == -ENOENT && block) {
> --
> 2.5.0
>
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: Fwd: Please revert wrong kernel OOPS patch in -69 !
2016-04-07 13:05 ` Fwd: Please revert wrong kernel OOPS patch in -69 ! Nicolai Hähnle
2016-04-07 13:31 ` Luis Henriques
@ 2016-04-18 1:34 ` Greg Kroah-Hartman
1 sibling, 0 replies; 3+ messages in thread
From: Greg Kroah-Hartman @ 2016-04-18 1:34 UTC (permalink / raw)
To: Nicolai Hähnle
Cc: Taras Prokopenko, christian.koenig, Luis Henriques, stable,
Kamal Mostafa, Sasha Levin, Jiri Slaby
On Thu, Apr 07, 2016 at 08:05:53AM -0500, Nicolai H�hnle wrote:
> Hmm, I was under the impression from an email I got that this was already
> reverted from 3.16. Clearly, communication is hard.
>
> Let me re-iterate, to be sure: kernels 3.17 and older
>
> 1. need to revert the backport of upstream commit
> f6ff4f67cdf8455d0a4226eeeaf5af17c37d05eb (I think most have done that
> already) and
>
> 2. should apply the corrected backport which I am attaching yet again (seems
> like when I sent this patch around two weeks ago, it wasn't picked up).
Thanks, I've fixed this up now for the 3.14-stable tree, sorry for the
mess.
greg k-h
^ permalink raw reply [flat|nested] 3+ messages in thread