2020-06-26 17:31:23

by Ralph Campbell

Subject: [PATCH] nouveau: fix page fault on device private memory

If system memory is migrated to device private memory and no GPU MMU
page table entry exists, the GPU will fault and call hmm_range_fault()
to get the PFN for the page. Since the .dev_private_owner pointer in
struct hmm_range is not set, hmm_range_fault returns an error which
results in the GPU program stopping with a fatal fault.
Fix this by setting .dev_private_owner appropriately.

Fixes: 08ddddda667b ("mm/hmm: check the device private page owner in hmm_range_fault()")
Signed-off-by: Ralph Campbell <[email protected]>
---

This is based on Linux-5.8.0-rc2 and is for Ben Skeggs nouveau tree.
It doesn't depend on any of the other nouveau/HMM changes I have
recently posted.

drivers/gpu/drm/nouveau/nouveau_svm.c | 1 +
1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/nouveau/nouveau_svm.c b/drivers/gpu/drm/nouveau/nouveau_svm.c
index ba9f9359c30e..6586d9d39874 100644
--- a/drivers/gpu/drm/nouveau/nouveau_svm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_svm.c
@@ -562,6 +562,7 @@ static int nouveau_range_fault(struct nouveau_svmm *svmm,
.end = notifier->notifier.interval_tree.last + 1,
.pfn_flags_mask = HMM_PFN_REQ_FAULT | HMM_PFN_REQ_WRITE,
.hmm_pfns = hmm_pfns,
+ .dev_private_owner = drm->dev,
};
struct mm_struct *mm = notifier->notifier.mm;
int ret;
--
2.20.1

2020-06-26 17:46:06

by Jason Gunthorpe

Subject: Re: [PATCH] nouveau: fix page fault on device private memory

On Fri, Jun 26, 2020 at 10:26:26AM -0700, Ralph Campbell wrote:
> If system memory is migrated to device private memory and no GPU MMU
> page table entry exists, the GPU will fault and call hmm_range_fault()
> to get the PFN for the page. Since the .dev_private_owner pointer in
> struct hmm_range is not set, hmm_range_fault returns an error which
> results in the GPU program stopping with a fatal fault.
> Fix this by setting .dev_private_owner appropriately.
>
> Fixes: 08ddddda667b ("mm/hmm: check the device private page owner in hmm_range_fault()")
> Signed-off-by: Ralph Campbell <[email protected]>
> ---
>
> This is based on Linux-5.8.0-rc2 and is for Ben Skeggs nouveau tree.
> It doesn't depend on any of the other nouveau/HMM changes I have
> recently posted.

Makes sense to me

Reviewed-by: Jason Gunthorpe <[email protected]>

Jason

2020-06-26 18:20:13

by John Hubbard

Subject: Re: [PATCH] nouveau: fix page fault on device private memory

On 2020-06-26 10:26, Ralph Campbell wrote:
> If system memory is migrated to device private memory and no GPU MMU
> page table entry exists, the GPU will fault and call hmm_range_fault()
> to get the PFN for the page. Since the .dev_private_owner pointer in
> struct hmm_range is not set, hmm_range_fault returns an error which
> results in the GPU program stopping with a fatal fault.
> Fix this by setting .dev_private_owner appropriately.
>
> Fixes: 08ddddda667b ("mm/hmm: check the device private page owner in hmm_range_fault()")
> Signed-off-by: Ralph Campbell <[email protected]>
> ---
>
> This is based on Linux-5.8.0-rc2 and is for Ben Skeggs nouveau tree.
> It doesn't depend on any of the other nouveau/HMM changes I have
> recently posted.

Maybe cc stable, seeing as how the problem affect Linux 5.7?

thanks,
--
John Hubbard
NVIDIA

>
> drivers/gpu/drm/nouveau/nouveau_svm.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/gpu/drm/nouveau/nouveau_svm.c b/drivers/gpu/drm/nouveau/nouveau_svm.c
> index ba9f9359c30e..6586d9d39874 100644
> --- a/drivers/gpu/drm/nouveau/nouveau_svm.c
> +++ b/drivers/gpu/drm/nouveau/nouveau_svm.c
> @@ -562,6 +562,7 @@ static int nouveau_range_fault(struct nouveau_svmm *svmm,
> .end = notifier->notifier.interval_tree.last + 1,
> .pfn_flags_mask = HMM_PFN_REQ_FAULT | HMM_PFN_REQ_WRITE,
> .hmm_pfns = hmm_pfns,
> + .dev_private_owner = drm->dev,
> };
> struct mm_struct *mm = notifier->notifier.mm;
> int ret;
>