[PATCH] drm/amdgpu: fix check in fbdev init

[PATCH] drm/amdgpu: fix check in fbdev init

[Alex Deucher]

The new vkms virtual display code is atomic so there is
no need to call drm_helper_disable_unused_functions()
when it is enabled.  Doing so can result in a segfault.
When the driver switched from the old virtual display code
to the new atomic virtual display code, it was missed that
we enable virtual display unconditionally under SR-IOV
so the checks here missed that case.  Add the missing
check for SR-IOV.

There is no equivalent of this patch for Linus' tree
because the relevant code no longer exists.  This patch
is only relevant to kernels 5.15 and 5.16.

Fixes: 84ec374bd580 ("drm/amdgpu: create amdgpu_vkms (v4)")
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 5.15.x
Cc: hgoffin@amazon.com
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
index cd0acbea75da..d58ab9deb028 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
@@ -341,7 +341,8 @@ int amdgpu_fbdev_init(struct amdgpu_device *adev)
 	}
 
 	/* disable all the possible outputs/crtcs before entering KMS mode */
-	if (!amdgpu_device_has_dc_support(adev) && !amdgpu_virtual_display)
+	if (!amdgpu_device_has_dc_support(adev) && !amdgpu_virtual_display &&
+	    !amdgpu_sriov_vf(adev))
 		drm_helper_disable_unused_functions(adev_to_drm(adev));
 
 	drm_fb_helper_initial_config(&rfbdev->helper, bpp_sel);
-- 
2.35.3

Re: [PATCH] drm/amdgpu: fix check in fbdev init

[Alex Deucher]

On Tue, Jul 19, 2022 at 2:57 PM Alex Deucher <alexander.deucher@amd.com> wrote:
>
> The new vkms virtual display code is atomic so there is
> no need to call drm_helper_disable_unused_functions()
> when it is enabled.  Doing so can result in a segfault.
> When the driver switched from the old virtual display code
> to the new atomic virtual display code, it was missed that
> we enable virtual display unconditionally under SR-IOV
> so the checks here missed that case.  Add the missing
> check for SR-IOV.
>
> There is no equivalent of this patch for Linus' tree
> because the relevant code no longer exists.  This patch
> is only relevant to kernels 5.15 and 5.16.
>
> Fixes: 84ec374bd580 ("drm/amdgpu: create amdgpu_vkms (v4)")
> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
> Cc: stable@vger.kernel.org # 5.15.x
> Cc: hgoffin@amazon.com

Hi Greg,

Is there any chance this can get applied?  It fixes a regression on
5.15 and 5.16.

Thanks,

Alex

> ---
>  drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
> index cd0acbea75da..d58ab9deb028 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
> @@ -341,7 +341,8 @@ int amdgpu_fbdev_init(struct amdgpu_device *adev)
>         }
>
>         /* disable all the possible outputs/crtcs before entering KMS mode */
> -       if (!amdgpu_device_has_dc_support(adev) && !amdgpu_virtual_display)
> +       if (!amdgpu_device_has_dc_support(adev) && !amdgpu_virtual_display &&
> +           !amdgpu_sriov_vf(adev))
>                 drm_helper_disable_unused_functions(adev_to_drm(adev));
>
>         drm_fb_helper_initial_config(&rfbdev->helper, bpp_sel);
> --
> 2.35.3
>

Re: [PATCH] drm/amdgpu: fix check in fbdev init

[Greg KH]

On Wed, Aug 10, 2022 at 11:28:18AM -0400, Alex Deucher wrote:
> On Tue, Jul 19, 2022 at 2:57 PM Alex Deucher <alexander.deucher@amd.com> wrote:
> >
> > The new vkms virtual display code is atomic so there is
> > no need to call drm_helper_disable_unused_functions()
> > when it is enabled.  Doing so can result in a segfault.
> > When the driver switched from the old virtual display code
> > to the new atomic virtual display code, it was missed that
> > we enable virtual display unconditionally under SR-IOV
> > so the checks here missed that case.  Add the missing
> > check for SR-IOV.
> >
> > There is no equivalent of this patch for Linus' tree
> > because the relevant code no longer exists.  This patch
> > is only relevant to kernels 5.15 and 5.16.
> >
> > Fixes: 84ec374bd580 ("drm/amdgpu: create amdgpu_vkms (v4)")
> > Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
> > Cc: stable@vger.kernel.org # 5.15.x
> > Cc: hgoffin@amazon.com
> 
> Hi Greg,
> 
> Is there any chance this can get applied?  It fixes a regression on
> 5.15 and 5.16.

Ah, missed this as it was not obvious that this was a not-upstream
commit at all, sorry.

I'll dig it out of lore.kernel.org and queue it up for the next round of
releases, but note, this is our "busy time" with many patches marked for
stable.

Oh and 5.16 is long end-of-life, nothing anyone can do there, and no one
should be using that kernel version anymore, so no issues there.

thanks,

greg k-h

Re: [PATCH] drm/amdgpu: fix check in fbdev init

[Alex Deucher]

On Wed, Aug 10, 2022 at 11:38 AM Greg KH <gregkh@linuxfoundation.org> wrote:
>
> On Wed, Aug 10, 2022 at 11:28:18AM -0400, Alex Deucher wrote:
> > On Tue, Jul 19, 2022 at 2:57 PM Alex Deucher <alexander.deucher@amd.com> wrote:
> > >
> > > The new vkms virtual display code is atomic so there is
> > > no need to call drm_helper_disable_unused_functions()
> > > when it is enabled.  Doing so can result in a segfault.
> > > When the driver switched from the old virtual display code
> > > to the new atomic virtual display code, it was missed that
> > > we enable virtual display unconditionally under SR-IOV
> > > so the checks here missed that case.  Add the missing
> > > check for SR-IOV.
> > >
> > > There is no equivalent of this patch for Linus' tree
> > > because the relevant code no longer exists.  This patch
> > > is only relevant to kernels 5.15 and 5.16.
> > >
> > > Fixes: 84ec374bd580 ("drm/amdgpu: create amdgpu_vkms (v4)")
> > > Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
> > > Cc: stable@vger.kernel.org # 5.15.x
> > > Cc: hgoffin@amazon.com
> >
> > Hi Greg,
> >
> > Is there any chance this can get applied?  It fixes a regression on
> > 5.15 and 5.16.
>
> Ah, missed this as it was not obvious that this was a not-upstream
> commit at all, sorry.
>
> I'll dig it out of lore.kernel.org and queue it up for the next round of
> releases, but note, this is our "busy time" with many patches marked for
> stable.
>
> Oh and 5.16 is long end-of-life, nothing anyone can do there, and no one
> should be using that kernel version anymore, so no issues there.

Thanks Greg.  Much appreciated.

Alex

>
> thanks,
>
> greg k-h

Re: [PATCH] drm/amdgpu: fix check in fbdev init

[Greg KH]

On Wed, Aug 10, 2022 at 11:39:39AM -0400, Alex Deucher wrote:
> On Wed, Aug 10, 2022 at 11:38 AM Greg KH <gregkh@linuxfoundation.org> wrote:
> >
> > On Wed, Aug 10, 2022 at 11:28:18AM -0400, Alex Deucher wrote:
> > > On Tue, Jul 19, 2022 at 2:57 PM Alex Deucher <alexander.deucher@amd.com> wrote:
> > > >
> > > > The new vkms virtual display code is atomic so there is
> > > > no need to call drm_helper_disable_unused_functions()
> > > > when it is enabled.  Doing so can result in a segfault.
> > > > When the driver switched from the old virtual display code
> > > > to the new atomic virtual display code, it was missed that
> > > > we enable virtual display unconditionally under SR-IOV
> > > > so the checks here missed that case.  Add the missing
> > > > check for SR-IOV.
> > > >
> > > > There is no equivalent of this patch for Linus' tree
> > > > because the relevant code no longer exists.  This patch
> > > > is only relevant to kernels 5.15 and 5.16.
> > > >
> > > > Fixes: 84ec374bd580 ("drm/amdgpu: create amdgpu_vkms (v4)")
> > > > Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
> > > > Cc: stable@vger.kernel.org # 5.15.x
> > > > Cc: hgoffin@amazon.com
> > >
> > > Hi Greg,
> > >
> > > Is there any chance this can get applied?  It fixes a regression on
> > > 5.15 and 5.16.
> >
> > Ah, missed this as it was not obvious that this was a not-upstream
> > commit at all, sorry.
> >
> > I'll dig it out of lore.kernel.org and queue it up for the next round of
> > releases, but note, this is our "busy time" with many patches marked for
> > stable.
> >
> > Oh and 5.16 is long end-of-life, nothing anyone can do there, and no one
> > should be using that kernel version anymore, so no issues there.
> 
> Thanks Greg.  Much appreciated.

Sorry for the delay, now queued up.

greg k-h