Re: Fwd: Please revert wrong kernel OOPS patch in -69 !

[Luis Henriques <luis.henriques@canonical.com>]

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
>