drm/ttm: fix the tt_populated check in ttm_tt_destroy()

Commit Message

Ben Skeggs Sept. 17, 2013, 4:21 a.m. UTC

From: Ben Skeggs <[e-mail]>

After a vmalloc failure in ttm_dma_tt_alloc_page_directory(),
ttm_dma_tt_init() will call ttm_tt_destroy() to cleanup, and end up
inside the driver's unpopulate() hook when populate() has never yet
been called.

On nouveau, the first issue to be hit because of this is that
dma_address[] may be a NULL pointer.  After working around this,
ttm_pool_unpopulate() may potentially hit the same issue with
the pages[] array.

It seems to make more sense to avoid calling unpopulate on already
unpopulated TTMs than to add checks to all the implementations.

Signed-off-by: Ben Skeggs <[e-mail]>
Cc: Jerome Glisse <[e-mail]>
Cc: Thomas Hellstrom <[e-mail]>
---
 drivers/gpu/drm/ttm/ttm_tt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Thomas Hellstrom Sept. 17, 2013, 6:40 a.m. UTC | #1

On 09/17/2013 06:21 AM, Ben Skeggs wrote:
> From: Ben Skeggs <[e-mail]>
>
> After a vmalloc failure in ttm_dma_tt_alloc_page_directory(),
> ttm_dma_tt_init() will call ttm_tt_destroy() to cleanup, and end up
> inside the driver's unpopulate() hook when populate() has never yet
> been called.
>
> On nouveau, the first issue to be hit because of this is that
> dma_address[] may be a NULL pointer.  After working around this,
> ttm_pool_unpopulate() may potentially hit the same issue with
> the pages[] array.
>
> It seems to make more sense to avoid calling unpopulate on already
> unpopulated TTMs than to add checks to all the implementations.
>
> Signed-off-by: Ben Skeggs <[e-mail]>
> Cc: Jerome Glisse <[e-mail]>
> Cc: Thomas Hellstrom <[e-mail]>
> ---
>   drivers/gpu/drm/ttm/ttm_tt.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c
> index 5e93a52..210d503 100644
> --- a/drivers/gpu/drm/ttm/ttm_tt.c
> +++ b/drivers/gpu/drm/ttm/ttm_tt.c
> @@ -170,7 +170,7 @@ void ttm_tt_destroy(struct ttm_tt *ttm)
>   		ttm_tt_unbind(ttm);
>   	}
>   
> -	if (likely(ttm->pages != NULL)) {
> +	if (ttm->state == tt_unbound) {
>   		ttm->bdev->driver->ttm_tt_unpopulate(ttm);
>   	}
>   
Looks good to me.

Reviewed-by: Thomas Hellstrom <[e-mail]>

Patch

diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c
index 5e93a52..210d503 100644
--- a/drivers/gpu/drm/ttm/ttm_tt.c
+++ b/drivers/gpu/drm/ttm/ttm_tt.c
@@ -170,7 +170,7 @@  void ttm_tt_destroy(struct ttm_tt *ttm)
 		ttm_tt_unbind(ttm);
 	}
 
-	if (likely(ttm->pages != NULL)) {
+	if (ttm->state == tt_unbound) {
 		ttm->bdev->driver->ttm_tt_unpopulate(ttm);
 	}