[PATCH 04/23] RDMA: use dma_resv_wait() instead of extracting the fence

[PATCH 04/23] RDMA: use dma_resv_wait() instead of extracting the fence

[Christian König]

Use dma_resv_wait() instead of extracting the exclusive fence and
waiting on it manually.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Leon Romanovsky <leon@kernel.org>
Cc: Maor Gottlieb <maorg@nvidia.com>
Cc: Gal Pressman <galpress@amazon.com>
Cc: linux-media@vger.kernel.org
Cc: linaro-mm-sig@lists.linaro.org
---
 drivers/infiniband/core/umem_dmabuf.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/drivers/infiniband/core/umem_dmabuf.c b/drivers/infiniband/core/umem_dmabuf.c
index f0760741f281..d32cd7538835 100644
--- a/drivers/infiniband/core/umem_dmabuf.c
+++ b/drivers/infiniband/core/umem_dmabuf.c
@@ -16,7 +16,6 @@ int ib_umem_dmabuf_map_pages(struct ib_umem_dmabuf *umem_dmabuf)
 {
 	struct sg_table *sgt;
 	struct scatterlist *sg;
-	struct dma_fence *fence;
 	unsigned long start, end, cur = 0;
 	unsigned int nmap = 0;
 	int i;
@@ -68,11 +67,8 @@ int ib_umem_dmabuf_map_pages(struct ib_umem_dmabuf *umem_dmabuf)
 	 * may be not up-to-date. Wait for the exporter to finish
 	 * the migration.
 	 */
-	fence = dma_resv_excl_fence(umem_dmabuf->attach->dmabuf->resv);
-	if (fence)
-		return dma_fence_wait(fence, false);
-
-	return 0;
+	return dma_resv_wait_timeout(umem_dmabuf->attach->dmabuf->resv, false,
+				     false, MAX_SCHEDULE_TIMEOUT);
 }
 EXPORT_SYMBOL(ib_umem_dmabuf_map_pages);
 
-- 
2.25.1

Re: [PATCH 04/23] RDMA: use dma_resv_wait() instead of extracting the fence

[Daniel Vetter]

On Mon, Mar 21, 2022 at 02:58:37PM +0100, Christian König wrote:
> Use dma_resv_wait() instead of extracting the exclusive fence and
> waiting on it manually.
> 
> Signed-off-by: Christian König <christian.koenig@amd.com>
> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> Cc: Jason Gunthorpe <jgg@ziepe.ca>

Jason, can you ack this for merging through drm trees please?

Thanks, Daniel

> Cc: Leon Romanovsky <leon@kernel.org>
> Cc: Maor Gottlieb <maorg@nvidia.com>
> Cc: Gal Pressman <galpress@amazon.com>
> Cc: linux-media@vger.kernel.org
> Cc: linaro-mm-sig@lists.linaro.org
> ---
>  drivers/infiniband/core/umem_dmabuf.c | 8 ++------
>  1 file changed, 2 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/infiniband/core/umem_dmabuf.c b/drivers/infiniband/core/umem_dmabuf.c
> index f0760741f281..d32cd7538835 100644
> --- a/drivers/infiniband/core/umem_dmabuf.c
> +++ b/drivers/infiniband/core/umem_dmabuf.c
> @@ -16,7 +16,6 @@ int ib_umem_dmabuf_map_pages(struct ib_umem_dmabuf *umem_dmabuf)
>  {
>  	struct sg_table *sgt;
>  	struct scatterlist *sg;
> -	struct dma_fence *fence;
>  	unsigned long start, end, cur = 0;
>  	unsigned int nmap = 0;
>  	int i;
> @@ -68,11 +67,8 @@ int ib_umem_dmabuf_map_pages(struct ib_umem_dmabuf *umem_dmabuf)
>  	 * may be not up-to-date. Wait for the exporter to finish
>  	 * the migration.
>  	 */
> -	fence = dma_resv_excl_fence(umem_dmabuf->attach->dmabuf->resv);
> -	if (fence)
> -		return dma_fence_wait(fence, false);
> -
> -	return 0;
> +	return dma_resv_wait_timeout(umem_dmabuf->attach->dmabuf->resv, false,
> +				     false, MAX_SCHEDULE_TIMEOUT);
>  }
>  EXPORT_SYMBOL(ib_umem_dmabuf_map_pages);
>  
> -- 
> 2.25.1
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

Re: [PATCH 04/23] RDMA: use dma_resv_wait() instead of extracting the fence

[Jason Gunthorpe]

On Wed, Mar 23, 2022 at 02:22:01PM +0100, Daniel Vetter wrote:
> On Mon, Mar 21, 2022 at 02:58:37PM +0100, Christian König wrote:
> > Use dma_resv_wait() instead of extracting the exclusive fence and
> > waiting on it manually.
> > 
> > Signed-off-by: Christian König <christian.koenig@amd.com>
> > Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > Cc: Jason Gunthorpe <jgg@ziepe.ca>
> 
> Jason, can you ack this for merging through drm trees please?

Sure, it looks trivial, but I didn't see the whole series:

Acked-by: Jason Gunthorpe <jgg@nvidia.com>

Jason

Re: [PATCH 04/23] RDMA: use dma_resv_wait() instead of extracting the fence

[Daniel Vetter]

On Wed, 23 Mar 2022 at 17:32, Jason Gunthorpe <jgg@ziepe.ca> wrote:
>
> On Wed, Mar 23, 2022 at 02:22:01PM +0100, Daniel Vetter wrote:
> > On Mon, Mar 21, 2022 at 02:58:37PM +0100, Christian König wrote:
> > > Use dma_resv_wait() instead of extracting the exclusive fence and
> > > waiting on it manually.
> > >
> > > Signed-off-by: Christian König <christian.koenig@amd.com>
> > > Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > > Cc: Jason Gunthorpe <jgg@ziepe.ca>
> >
> > Jason, can you ack this for merging through drm trees please?
>
> Sure, it looks trivial, but I didn't see the whole series:
>
> Acked-by: Jason Gunthorpe <jgg@nvidia.com>

The entire series reworks how dma_resv stores fences (and what exactly
the mean), which is why we need to get users away from some of these
low-level accessors and towards functions at a slightly higher level.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch