Re: [PATCH 6/6] drm/amdkfd: fix boot failure when iommu is disabled in Picasso.

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

On Fri, Dec 10, 2021 at 09:46:08AM -0500, James Zhu wrote:
> 
> On 2021-12-10 9:35 a.m., Greg Kroah-Hartman wrote:
> > On Fri, Dec 10, 2021 at 09:14:30AM -0500, James Zhu wrote:
> > > On 2021-12-10 8:33 a.m., Greg Kroah-Hartman wrote:
> > > > On Thu, Dec 09, 2021 at 05:09:56PM -0500, James Zhu wrote:
> > > > > From: Yifan Zhang <yifan1.zhang@amd.com>
> > > > > 
> > > > > commit afd18180c07026f94a80ff024acef5f4159084a4 upstream.
> > > > > 
> > > > > When IOMMU disabled in sbios and kfd in iommuv2 path, iommuv2
> > > > > init will fail. But this failure should not block amdgpu driver init.
> > > > > 
> > > > > Reported-by: youling <youling257@gmail.com>
> > > > > Tested-by: youling <youling257@gmail.com>
> > > > > Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
> > > > > Reviewed-by: James Zhu <James.Zhu@amd.com>
> > > > > Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
> > > > > Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > > > > Signed-off-by: James Zhu <James.Zhu@amd.com>
> > > > > ---
> > > > >    drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 ----
> > > > >    drivers/gpu/drm/amd/amdkfd/kfd_device.c    | 3 +++
> > > > >    2 files changed, 3 insertions(+), 4 deletions(-)
> > > > > 
> > > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> > > > > index 488e574f5da1..f262c4e7a48a 100644
> > > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> > > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> > > > > @@ -2255,10 +2255,6 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
> > > > >    		amdgpu_xgmi_add_device(adev);
> > > > >    	amdgpu_amdkfd_device_init(adev);
> > > > > -	r = amdgpu_amdkfd_resume_iommu(adev);
> > > > > -	if (r)
> > > > > -		goto init_failed;
> > > > > -
> > > > >    	amdgpu_fru_get_product_info(adev);
> > > > >    init_failed:
> > > > > diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
> > > > > index 1204dae85797..b35f0af71f00 100644
> > > > > --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
> > > > > +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
> > > > > @@ -751,6 +751,9 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd,
> > > > >    	kfd_cwsr_init(kfd);
> > > > > +	if (kgd2kfd_resume_iommu(kfd))
> > > > > +		goto device_iommu_error;
> > > > > +
> > > > >    	if (kfd_resume(kfd))
> > > > >    		goto kfd_resume_error;
> > > > > -- 
> > > > > 2.25.1
> > > > > 
> > > > Like I said last time, do not change the backport unless you HAVE to.
> > > > You did it here again for no good reason :(
> > > [JZ] Yes, I should add more explanation next time.
> > > 
> > > Backport conflict fix to remove  svm_migrate_init((struct amdgpu_device
> > > *)kfd->kgd);
> > > 
> > > new AMD svm feature has not been added for 5.10 So it is safe to remove it.
> > No, I am talking about the fact that you fixed up a coding style fix in
> > this backport that is not in the original commit in Linus's tree.
> 
> [JZ] I see. this fix is not necessary. Do you want me to send v2 with
> 
> this unnecessary coding style fix dropping for backport?
> 

I took what was in Linus's tree already.  Please verify that what I
applied to the queue still works.

thanks,

greg k-h