[PATCH v2 0/3] drm/ttm: Improve the TTM operation context gfp_retry_mayfail behaviour

[PATCH v2 0/3] drm/ttm: Improve the TTM operation context gfp_retry_mayfail behaviour

[Thomas Hellström]

Two small patches around the gfp_retry_mayfail behaviour, and
one patch to improve associated kerneldoc.

Patch 1 ensures we don't spam the kernel log on allocation failures.

Patch 2 extends the gfp_retry_mayfail to swap readback.

Patch 3 Improves on the associated kerneldocs.

v2:
- Add patch 3. (Sima).

Thomas Hellström (3):
  drm/ttm: Don't spam the log on buffer object backing store allocation
    failure
  drm/ttm: Avoid invoking the OOM killer when reading back swapped
    content
  drm/ttm: Update the struct ttm_operation_ctx kerneldoc

 drivers/gpu/drm/ttm/ttm_backup.c |  6 ++++--
 drivers/gpu/drm/ttm/ttm_pool.c   |  7 +++++--
 include/drm/ttm/ttm_backup.h     |  2 +-
 include/drm/ttm/ttm_bo.h         | 28 +++++++++++++++++++---------
 4 files changed, 29 insertions(+), 14 deletions(-)

-- 
2.53.0

[PATCH v2 1/3] drm/ttm: Don't spam the log on buffer object backing store allocation failure

[Thomas Hellström]

If the struct ttm_operation_ctx::gfp_retry_mayfail is true,
buffer object backing store allocation failures are expected to
silently fail with an error code to the caller. But currently an
elaborate warning is printed to the system log.

Don't spam the log in this way.

Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Simona Vetter <simona.vetter@ffwll.ch>
---
 drivers/gpu/drm/ttm/ttm_pool.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/ttm/ttm_pool.c b/drivers/gpu/drm/ttm/ttm_pool.c
index c0d95559197c..8fa9e09f6ee5 100644
--- a/drivers/gpu/drm/ttm/ttm_pool.c
+++ b/drivers/gpu/drm/ttm/ttm_pool.c
@@ -726,7 +726,7 @@ static int __ttm_pool_alloc(struct ttm_pool *pool, struct ttm_tt *tt,
 		gfp_flags |= __GFP_ZERO;
 
 	if (ctx->gfp_retry_mayfail)
-		gfp_flags |= __GFP_RETRY_MAYFAIL;
+		gfp_flags |= __GFP_RETRY_MAYFAIL | __GFP_NOWARN;
 
 	if (ttm_pool_uses_dma32(pool))
 		gfp_flags |= GFP_DMA32;
-- 
2.53.0

[PATCH v2 2/3] drm/ttm: Avoid invoking the OOM killer when reading back swapped content

[Thomas Hellström]

In situations where the system is very short on RAM, the shmem
readback from swap-space may invoke the OOM killer.

However, since this might be a recoverable situation where the caller
is indicating this by setting
struct ttm_operation_ctx::gfp_retry_mayfail to true, adjust the gfp
value used by the allocation accordingly.

Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Maarten Lankhorst <dev@lankhorst.se>
---
 drivers/gpu/drm/ttm/ttm_backup.c | 6 ++++--
 drivers/gpu/drm/ttm/ttm_pool.c   | 5 ++++-
 include/drm/ttm/ttm_backup.h     | 2 +-
 3 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/ttm/ttm_backup.c b/drivers/gpu/drm/ttm/ttm_backup.c
index 6bd4c123d94c..81df4cb5606b 100644
--- a/drivers/gpu/drm/ttm/ttm_backup.c
+++ b/drivers/gpu/drm/ttm/ttm_backup.c
@@ -44,18 +44,20 @@ void ttm_backup_drop(struct file *backup, pgoff_t handle)
  * @dst: The struct page to copy into.
  * @handle: The handle returned when the page was backed up.
  * @intr: Try to perform waits interruptible or at least killable.
+ * @additional_gfp: GFP mask to add to the default GFP mask if any.
  *
  * Return: 0 on success, Negative error code on failure, notably
  * -EINTR if @intr was set to true and a signal is pending.
  */
 int ttm_backup_copy_page(struct file *backup, struct page *dst,
-			 pgoff_t handle, bool intr)
+			 pgoff_t handle, bool intr, gfp_t additional_gfp)
 {
 	struct address_space *mapping = backup->f_mapping;
 	struct folio *from_folio;
 	pgoff_t idx = ttm_backup_handle_to_shmem_idx(handle);
 
-	from_folio = shmem_read_folio(mapping, idx);
+	from_folio = shmem_read_folio_gfp(mapping, idx, mapping_gfp_mask(mapping)
+					  | additional_gfp);
 	if (IS_ERR(from_folio))
 		return PTR_ERR(from_folio);
 
diff --git a/drivers/gpu/drm/ttm/ttm_pool.c b/drivers/gpu/drm/ttm/ttm_pool.c
index 8fa9e09f6ee5..aa41099c5ecf 100644
--- a/drivers/gpu/drm/ttm/ttm_pool.c
+++ b/drivers/gpu/drm/ttm/ttm_pool.c
@@ -530,6 +530,8 @@ static int ttm_pool_restore_commit(struct ttm_pool_tt_restore *restore,
 		p = first_page[i];
 		if (ttm_backup_page_ptr_is_handle(p)) {
 			unsigned long handle = ttm_backup_page_ptr_to_handle(p);
+			gfp_t additional_gfp = ctx->gfp_retry_mayfail ?
+				__GFP_RETRY_MAYFAIL | __GFP_NOWARN : 0;
 
 			if (IS_ENABLED(CONFIG_FAULT_INJECTION) && ctx->interruptible &&
 			    should_fail(&backup_fault_inject, 1)) {
@@ -543,7 +545,8 @@ static int ttm_pool_restore_commit(struct ttm_pool_tt_restore *restore,
 			}
 
 			ret = ttm_backup_copy_page(backup, restore->alloced_page + i,
-						   handle, ctx->interruptible);
+						   handle, ctx->interruptible,
+						   additional_gfp);
 			if (ret)
 				break;
 
diff --git a/include/drm/ttm/ttm_backup.h b/include/drm/ttm/ttm_backup.h
index c33cba111171..29b9c855af77 100644
--- a/include/drm/ttm/ttm_backup.h
+++ b/include/drm/ttm/ttm_backup.h
@@ -56,7 +56,7 @@ ttm_backup_page_ptr_to_handle(const struct page *page)
 void ttm_backup_drop(struct file *backup, pgoff_t handle);
 
 int ttm_backup_copy_page(struct file *backup, struct page *dst,
-			 pgoff_t handle, bool intr);
+			 pgoff_t handle, bool intr, gfp_t additional_gfp);
 
 s64
 ttm_backup_backup_page(struct file *backup, struct page *page,
-- 
2.53.0

[PATCH v2 3/3] drm/ttm: Update the struct ttm_operation_ctx kerneldoc

[Thomas Hellström]

Update the kerneldoc with a more elaborate description of some members,
including the gfp_retry_mayfail member. Use inline kerneldoc.

Suggested-by: Simona Vetter <simona.vetter@ffwll.ch>
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
---
 include/drm/ttm/ttm_bo.h | 28 +++++++++++++++++++---------
 1 file changed, 19 insertions(+), 9 deletions(-)

diff --git a/include/drm/ttm/ttm_bo.h b/include/drm/ttm/ttm_bo.h
index bca3a8849d47..8310bc3d55f9 100644
--- a/include/drm/ttm/ttm_bo.h
+++ b/include/drm/ttm/ttm_bo.h
@@ -167,24 +167,34 @@ struct ttm_bo_kmap_obj {
 /**
  * struct ttm_operation_ctx
  *
- * @interruptible: Sleep interruptible if sleeping.
- * @no_wait_gpu: Return immediately if the GPU is busy.
- * @gfp_retry_mayfail: Set the __GFP_RETRY_MAYFAIL when allocation pages.
- * @allow_res_evict: Allow eviction of reserved BOs. Can be used when multiple
- * BOs share the same reservation object.
- * faults. Should only be used by TTM internally.
- * @resv: Reservation object to allow reserved evictions with.
- * @bytes_moved: Statistics on how many bytes have been moved.
- *
  * Context for TTM operations like changing buffer placement or general memory
  * allocation.
  */
 struct ttm_operation_ctx {
+	/** @interruptible: Sleep interruptible if sleeping. */
 	bool interruptible;
+	/** @no_wait_gpu: Return immediately if the GPU is busy. */
 	bool no_wait_gpu;
+	/**
+	 * @gfp_retry_mayfail: Use __GFP_RETRY_MAYFAIL | __GFP_NOWARN
+	 * when allocation pages. This is to avoid invoking the OOM
+	 * killer when populating a buffer object, in order to
+	 * forward the error for it to be dealt with.
+	 */
 	bool gfp_retry_mayfail;
+	/**
+	 * @allow_res_evict: Allow eviction of reserved BOs. Can be used
+	 * when multiple BOs share the same reservation object @resv.
+	 */
 	bool allow_res_evict;
+	/**
+	 * @resv: Reservation object to be used together with
+	 * @allow_res_evict.
+	 */
 	struct dma_resv *resv;
+	/**
+	 * @bytes_moved: Statistics on how many bytes have been moved.
+	 */
 	uint64_t bytes_moved;
 };
 
-- 
2.53.0

Re: [PATCH v2 3/3] drm/ttm: Update the struct ttm_operation_ctx kerneldoc

[Maarten Lankhorst]

Reviewed-by: Maarten Lankhorst <dev@lankhorst.se>

Den 2026-03-17 kl. 15:18, skrev Thomas Hellström:
> Update the kerneldoc with a more elaborate description of some members,
> including the gfp_retry_mayfail member. Use inline kerneldoc.
> 
> Suggested-by: Simona Vetter <simona.vetter@ffwll.ch>
> Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
> ---
>  include/drm/ttm/ttm_bo.h | 28 +++++++++++++++++++---------
>  1 file changed, 19 insertions(+), 9 deletions(-)
> 
> diff --git a/include/drm/ttm/ttm_bo.h b/include/drm/ttm/ttm_bo.h
> index bca3a8849d47..8310bc3d55f9 100644
> --- a/include/drm/ttm/ttm_bo.h
> +++ b/include/drm/ttm/ttm_bo.h
> @@ -167,24 +167,34 @@ struct ttm_bo_kmap_obj {
>  /**
>   * struct ttm_operation_ctx
>   *
> - * @interruptible: Sleep interruptible if sleeping.
> - * @no_wait_gpu: Return immediately if the GPU is busy.
> - * @gfp_retry_mayfail: Set the __GFP_RETRY_MAYFAIL when allocation pages.
> - * @allow_res_evict: Allow eviction of reserved BOs. Can be used when multiple
> - * BOs share the same reservation object.
> - * faults. Should only be used by TTM internally.
> - * @resv: Reservation object to allow reserved evictions with.
> - * @bytes_moved: Statistics on how many bytes have been moved.
> - *
>   * Context for TTM operations like changing buffer placement or general memory
>   * allocation.
>   */
>  struct ttm_operation_ctx {
> +	/** @interruptible: Sleep interruptible if sleeping. */
>  	bool interruptible;
> +	/** @no_wait_gpu: Return immediately if the GPU is busy. */
>  	bool no_wait_gpu;
> +	/**
> +	 * @gfp_retry_mayfail: Use __GFP_RETRY_MAYFAIL | __GFP_NOWARN
> +	 * when allocation pages. This is to avoid invoking the OOM
> +	 * killer when populating a buffer object, in order to
> +	 * forward the error for it to be dealt with.
> +	 */
>  	bool gfp_retry_mayfail;
> +	/**
> +	 * @allow_res_evict: Allow eviction of reserved BOs. Can be used
> +	 * when multiple BOs share the same reservation object @resv.
> +	 */
>  	bool allow_res_evict;
> +	/**
> +	 * @resv: Reservation object to be used together with
> +	 * @allow_res_evict.
> +	 */
>  	struct dma_resv *resv;
> +	/**
> +	 * @bytes_moved: Statistics on how many bytes have been moved.
> +	 */
>  	uint64_t bytes_moved;
>  };
>  

Re: [PATCH v2 0/3] drm/ttm: Improve the TTM operation context gfp_retry_mayfail behaviour

[Thomas Hellström]

@Christian, Ack to merge these?

Thanks,
Thomas


On Tue, 2026-03-17 at 15:18 +0100, Thomas Hellström wrote:
> Two small patches around the gfp_retry_mayfail behaviour, and
> one patch to improve associated kerneldoc.
> 
> Patch 1 ensures we don't spam the kernel log on allocation failures.
> 
> Patch 2 extends the gfp_retry_mayfail to swap readback.
> 
> Patch 3 Improves on the associated kerneldocs.
> 
> v2:
> - Add patch 3. (Sima).
> 
> Thomas Hellström (3):
>   drm/ttm: Don't spam the log on buffer object backing store
> allocation
>     failure
>   drm/ttm: Avoid invoking the OOM killer when reading back swapped
>     content
>   drm/ttm: Update the struct ttm_operation_ctx kerneldoc
> 
>  drivers/gpu/drm/ttm/ttm_backup.c |  6 ++++--
>  drivers/gpu/drm/ttm/ttm_pool.c   |  7 +++++--
>  include/drm/ttm/ttm_backup.h     |  2 +-
>  include/drm/ttm/ttm_bo.h         | 28 +++++++++++++++++++---------
>  4 files changed, 29 insertions(+), 14 deletions(-)

Re: [PATCH v2 0/3] drm/ttm: Improve the TTM operation context gfp_retry_mayfail behaviour

[Thomas Hellström]

Hi!

Since Sima was OK with this series and I believe the changes are
completely non-controversial, I'll go ahead and merge this tomorrow
through drm-misc-next.

Thanks,
Thomas


On Wed, 2026-03-18 at 15:50 +0100, Thomas Hellström wrote:
> @Christian, Ack to merge these?
> 
> Thanks,
> Thomas
> 
> 
> On Tue, 2026-03-17 at 15:18 +0100, Thomas Hellström wrote:
> > Two small patches around the gfp_retry_mayfail behaviour, and
> > one patch to improve associated kerneldoc.
> > 
> > Patch 1 ensures we don't spam the kernel log on allocation
> > failures.
> > 
> > Patch 2 extends the gfp_retry_mayfail to swap readback.
> > 
> > Patch 3 Improves on the associated kerneldocs.
> > 
> > v2:
> > - Add patch 3. (Sima).
> > 
> > Thomas Hellström (3):
> >   drm/ttm: Don't spam the log on buffer object backing store
> > allocation
> >     failure
> >   drm/ttm: Avoid invoking the OOM killer when reading back swapped
> >     content
> >   drm/ttm: Update the struct ttm_operation_ctx kerneldoc
> > 
> >  drivers/gpu/drm/ttm/ttm_backup.c |  6 ++++--
> >  drivers/gpu/drm/ttm/ttm_pool.c   |  7 +++++--
> >  include/drm/ttm/ttm_backup.h     |  2 +-
> >  include/drm/ttm/ttm_bo.h         | 28 +++++++++++++++++++---------
> >  4 files changed, 29 insertions(+), 14 deletions(-)

Re: [PATCH v2 0/3] drm/ttm: Improve the TTM operation context gfp_retry_mayfail behaviour

[Christian König]

Sorry for the delay. Feel free to add my Acked-by.

Regards,
Christian.

On 3/19/26 20:30, Thomas Hellström wrote:
> Hi!
> 
> Since Sima was OK with this series and I believe the changes are
> completely non-controversial, I'll go ahead and merge this tomorrow
> through drm-misc-next.
> 
> Thanks,
> Thomas
> 
> 
> On Wed, 2026-03-18 at 15:50 +0100, Thomas Hellström wrote:
>> @Christian, Ack to merge these?
>>
>> Thanks,
>> Thomas
>>
>>
>> On Tue, 2026-03-17 at 15:18 +0100, Thomas Hellström wrote:
>>> Two small patches around the gfp_retry_mayfail behaviour, and
>>> one patch to improve associated kerneldoc.
>>>
>>> Patch 1 ensures we don't spam the kernel log on allocation
>>> failures.
>>>
>>> Patch 2 extends the gfp_retry_mayfail to swap readback.
>>>
>>> Patch 3 Improves on the associated kerneldocs.
>>>
>>> v2:
>>> - Add patch 3. (Sima).
>>>
>>> Thomas Hellström (3):
>>>   drm/ttm: Don't spam the log on buffer object backing store
>>> allocation
>>>     failure
>>>   drm/ttm: Avoid invoking the OOM killer when reading back swapped
>>>     content
>>>   drm/ttm: Update the struct ttm_operation_ctx kerneldoc
>>>
>>>  drivers/gpu/drm/ttm/ttm_backup.c |  6 ++++--
>>>  drivers/gpu/drm/ttm/ttm_pool.c   |  7 +++++--
>>>  include/drm/ttm/ttm_backup.h     |  2 +-
>>>  include/drm/ttm/ttm_bo.h         | 28 +++++++++++++++++++---------
>>>  4 files changed, 29 insertions(+), 14 deletions(-)

Re: [PATCH v2 0/3] drm/ttm: Improve the TTM operation context gfp_retry_mayfail behaviour

[Thomas Hellström]

On Fri, 2026-03-20 at 09:42 +0100, Christian König wrote:
> Sorry for the delay. Feel free to add my Acked-by.
> 
> Regards,
> Christian.

Thx!

/Thomas


> 
> On 3/19/26 20:30, Thomas Hellström wrote:
> > Hi!
> > 
> > Since Sima was OK with this series and I believe the changes are
> > completely non-controversial, I'll go ahead and merge this tomorrow
> > through drm-misc-next.
> > 
> > Thanks,
> > Thomas
> > 
> > 
> > On Wed, 2026-03-18 at 15:50 +0100, Thomas Hellström wrote:
> > > @Christian, Ack to merge these?
> > > 
> > > Thanks,
> > > Thomas
> > > 
> > > 
> > > On Tue, 2026-03-17 at 15:18 +0100, Thomas Hellström wrote:
> > > > Two small patches around the gfp_retry_mayfail behaviour, and
> > > > one patch to improve associated kerneldoc.
> > > > 
> > > > Patch 1 ensures we don't spam the kernel log on allocation
> > > > failures.
> > > > 
> > > > Patch 2 extends the gfp_retry_mayfail to swap readback.
> > > > 
> > > > Patch 3 Improves on the associated kerneldocs.
> > > > 
> > > > v2:
> > > > - Add patch 3. (Sima).
> > > > 
> > > > Thomas Hellström (3):
> > > >   drm/ttm: Don't spam the log on buffer object backing store
> > > > allocation
> > > >     failure
> > > >   drm/ttm: Avoid invoking the OOM killer when reading back
> > > > swapped
> > > >     content
> > > >   drm/ttm: Update the struct ttm_operation_ctx kerneldoc
> > > > 
> > > >  drivers/gpu/drm/ttm/ttm_backup.c |  6 ++++--
> > > >  drivers/gpu/drm/ttm/ttm_pool.c   |  7 +++++--
> > > >  include/drm/ttm/ttm_backup.h     |  2 +-
> > > >  include/drm/ttm/ttm_bo.h         | 28 +++++++++++++++++++-----
> > > > ----
> > > >  4 files changed, 29 insertions(+), 14 deletions(-)