[PATCH] drm/ttm: fix error handling if no BO can be swapped out v4

[PATCH] drm/ttm: fix error handling if no BO can be swapped out v4

[Christian König]

From: Shiwu Zhang <shiwu.zhang@amd.com>

In case that all pre-allocated BOs are busy, just continue to populate
BOs since likely half of system memory in total is still free.

v4 (chk): fix code moved to VMWGFX as well

Signed-off-by: Shiwu Zhang <shiwu.zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
---
 drivers/gpu/drm/ttm/ttm_device.c    | 2 +-
 drivers/gpu/drm/ttm/ttm_tt.c        | 2 ++
 drivers/gpu/drm/vmwgfx/ttm_memory.c | 2 +-
 3 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/ttm/ttm_device.c b/drivers/gpu/drm/ttm/ttm_device.c
index 1f2024164d72..5dc92b056f0a 100644
--- a/drivers/gpu/drm/ttm/ttm_device.c
+++ b/drivers/gpu/drm/ttm/ttm_device.c
@@ -112,7 +112,7 @@ int ttm_global_swapout(struct ttm_operation_ctx *ctx, gfp_t gfp_flags)
 {
 	struct ttm_global *glob = &ttm_glob;
 	struct ttm_device *bdev;
-	int ret = -EBUSY;
+	int ret = 0;
 
 	mutex_lock(&ttm_global_mutex);
 	list_for_each_entry(bdev, &glob->device_list, device_list) {
diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c
index 48c407cff112..539e0232cb3b 100644
--- a/drivers/gpu/drm/ttm/ttm_tt.c
+++ b/drivers/gpu/drm/ttm/ttm_tt.c
@@ -329,6 +329,8 @@ int ttm_tt_populate(struct ttm_device *bdev,
 	       ttm_dma32_pages_limit) {
 
 		ret = ttm_global_swapout(ctx, GFP_KERNEL);
+		if (ret == 0)
+			break;
 		if (ret < 0)
 			goto error;
 	}
diff --git a/drivers/gpu/drm/vmwgfx/ttm_memory.c b/drivers/gpu/drm/vmwgfx/ttm_memory.c
index 104b95a8c7a2..aeb0a22a2c34 100644
--- a/drivers/gpu/drm/vmwgfx/ttm_memory.c
+++ b/drivers/gpu/drm/vmwgfx/ttm_memory.c
@@ -280,7 +280,7 @@ static void ttm_shrink(struct ttm_mem_global *glob, bool from_wq,
 		spin_unlock(&glob->lock);
 		ret = ttm_global_swapout(ctx, GFP_KERNEL);
 		spin_lock(&glob->lock);
-		if (unlikely(ret < 0))
+		if (unlikely(ret <= 0))
 			break;
 	}
 
-- 
2.25.1

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH] drm/ttm: fix error handling if no BO can be swapped out v4

[Zack Rusin]

On 4/22/21 7:57 AM, Christian König wrote:
> From: Shiwu Zhang <shiwu.zhang@amd.com>
> 
> In case that all pre-allocated BOs are busy, just continue to populate
> BOs since likely half of system memory in total is still free.
> 
> v4 (chk): fix code moved to VMWGFX as well
> 
> Signed-off-by: Shiwu Zhang <shiwu.zhang@amd.com>
> Reviewed-by: Christian König <christian.koenig@amd.com>
> Signed-off-by: Christian König <christian.koenig@amd.com>
> ---
>   drivers/gpu/drm/ttm/ttm_device.c    | 2 +-
>   drivers/gpu/drm/ttm/ttm_tt.c        | 2 ++
>   drivers/gpu/drm/vmwgfx/ttm_memory.c | 2 +-
>   3 files changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/ttm/ttm_device.c b/drivers/gpu/drm/ttm/ttm_device.c
> index 1f2024164d72..5dc92b056f0a 100644
> --- a/drivers/gpu/drm/ttm/ttm_device.c
> +++ b/drivers/gpu/drm/ttm/ttm_device.c
> @@ -112,7 +112,7 @@ int ttm_global_swapout(struct ttm_operation_ctx *ctx, gfp_t gfp_flags)
>   {
>   	struct ttm_global *glob = &ttm_glob;
>   	struct ttm_device *bdev;
> -	int ret = -EBUSY;
> +	int ret = 0;

Looks good to me. Updating the comment above that function to note it returns either tge number of pages swapped out or an error might be a good idea because it's not obvious from the name.

Reviewed-by: Zack Rusin <zackr@vmware.com>
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel