[v2] drm: gem_cma_helper.c: Allow importing of contiguous scatterlists with nents > 1

Message ID 20171110133310.1225-1-Liviu.Dudau@arm.com
State New
Headers show
Series
  • [v2] drm: gem_cma_helper.c: Allow importing of contiguous scatterlists with nents > 1
Related show

Commit Message

Liviu Dudau Nov. 10, 2017, 1:33 p.m.
drm_gem_cma_prime_import_sg_table() will fail if the number of entries
in the sg_table > 1. However, you can have a device that uses an IOMMU
engine and can map a discontiguous buffer with multiple entries that
have consecutive sg_dma_addresses, effectively making it contiguous.
Allow for that scenario by testing the entries in the sg_table for
contiguous coverage.

Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Liviu Dudau <liviu.dudau@arm.com>
---

Laurent,

Thanks for the review! I would like to ask for one more favour: if you
are OK with this version, can you pull this patch through the drm-misc tree?

Many thanks,
Liviu

 drivers/gpu/drm/drm_gem_cma_helper.c | 22 ++++++++++++++++++++--
 include/drm/drm_gem_cma_helper.h     |  4 +++-
 2 files changed, 23 insertions(+), 3 deletions(-)

Comments

Liviu Dudau Nov. 15, 2017, 1:04 p.m. | #1
Hi,

On Sat, Nov 11, 2017 at 02:47:35PM +0200, Laurent Pinchart wrote:
> Hi Liviu,
> 
> Thank you for the patch.
> 
> On Friday, 10 November 2017 15:33:10 EET Liviu Dudau wrote:
> > drm_gem_cma_prime_import_sg_table() will fail if the number of entries
> > in the sg_table > 1. However, you can have a device that uses an IOMMU
> > engine and can map a discontiguous buffer with multiple entries that
> > have consecutive sg_dma_addresses, effectively making it contiguous.
> > Allow for that scenario by testing the entries in the sg_table for
> > contiguous coverage.
> > 
> > Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
> > Signed-off-by: Liviu Dudau <liviu.dudau@arm.com>
> > ---
> > 
> > Laurent,
> > 
> > Thanks for the review! I would like to ask for one more favour: if you
> > are OK with this version, can you pull this patch through the drm-misc tree?
> 
> I could, but I'd first need to set dim up, and I'm currently abroad with a bad 
> internet connection and a big deadline for the middle of next week (I know, 
> lots of excuses), so it's not very convenient for me at this time.

Any other drm-misc maintainers feeling helpful and willing to take this
patch in? Otherwise I can send it through the mali-dp tree if no one
objects.

Best regards,
Liviu

> 
> >  drivers/gpu/drm/drm_gem_cma_helper.c | 22 ++++++++++++++++++++--
> >  include/drm/drm_gem_cma_helper.h     |  4 +++-
> >  2 files changed, 23 insertions(+), 3 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/drm_gem_cma_helper.c
> > b/drivers/gpu/drm/drm_gem_cma_helper.c index 020e7668dfaba..43b179212052d
> > 100644
> > --- a/drivers/gpu/drm/drm_gem_cma_helper.c
> > +++ b/drivers/gpu/drm/drm_gem_cma_helper.c
> > @@ -482,8 +482,26 @@ drm_gem_cma_prime_import_sg_table(struct drm_device
> > *dev, {
> >  	struct drm_gem_cma_object *cma_obj;
> > 
> > -	if (sgt->nents != 1)
> > -		return ERR_PTR(-EINVAL);
> > +	if (sgt->nents != 1) {
> > +		/* check if the entries in the sg_table are contiguous */
> > +		dma_addr_t next_addr = sg_dma_address(sgt->sgl);
> > +		struct scatterlist *s;
> > +		unsigned int i;
> > +
> > +		for_each_sg(sgt->sgl, s, sgt->nents, i) {
> > +			/*
> > +			 * sg_dma_address(s) is only valid for entries
> > +			 * that have sg_dma_len(s) != 0
> > +			 */
> > +			if (!sg_dma_len(s))
> > +				continue;
> > +
> > +			if (sg_dma_address(s) != next_addr)
> > +				return ERR_PTR(-EINVAL);
> > +
> > +			next_addr = sg_dma_address(s) + sg_dma_len(s);
> > +		}
> > +	}
> > 
> >  	/* Create a CMA GEM buffer. */
> >  	cma_obj = __drm_gem_cma_create(dev, attach->dmabuf->size);
> > diff --git a/include/drm/drm_gem_cma_helper.h
> > b/include/drm/drm_gem_cma_helper.h index 58a739bf15f1f..214aa85adc8d5
> > 100644
> > --- a/include/drm/drm_gem_cma_helper.h
> > +++ b/include/drm/drm_gem_cma_helper.h
> > @@ -8,7 +8,9 @@
> >   * struct drm_gem_cma_object - GEM object backed by CMA memory allocations
> >   * @base: base GEM object
> >   * @paddr: physical address of the backing memory
> > - * @sgt: scatter/gather table for imported PRIME buffers
> > + * @sgt: scatter/gather table for imported PRIME buffers. The table can
> > have + *       more than one entry but they are guaranteed to have
> > contiguous + *       DMA addresses.
> >   * @vaddr: kernel virtual address of the backing memory
> >   */
> >  struct drm_gem_cma_object {
> 
> -- 
> Regards,
> 
> Laurent Pinchart
>

Patch

diff --git a/drivers/gpu/drm/drm_gem_cma_helper.c b/drivers/gpu/drm/drm_gem_cma_helper.c
index 020e7668dfaba..43b179212052d 100644
--- a/drivers/gpu/drm/drm_gem_cma_helper.c
+++ b/drivers/gpu/drm/drm_gem_cma_helper.c
@@ -482,8 +482,26 @@  drm_gem_cma_prime_import_sg_table(struct drm_device *dev,
 {
 	struct drm_gem_cma_object *cma_obj;
 
-	if (sgt->nents != 1)
-		return ERR_PTR(-EINVAL);
+	if (sgt->nents != 1) {
+		/* check if the entries in the sg_table are contiguous */
+		dma_addr_t next_addr = sg_dma_address(sgt->sgl);
+		struct scatterlist *s;
+		unsigned int i;
+
+		for_each_sg(sgt->sgl, s, sgt->nents, i) {
+			/*
+			 * sg_dma_address(s) is only valid for entries
+			 * that have sg_dma_len(s) != 0
+			 */
+			if (!sg_dma_len(s))
+				continue;
+
+			if (sg_dma_address(s) != next_addr)
+				return ERR_PTR(-EINVAL);
+
+			next_addr = sg_dma_address(s) + sg_dma_len(s);
+		}
+	}
 
 	/* Create a CMA GEM buffer. */
 	cma_obj = __drm_gem_cma_create(dev, attach->dmabuf->size);
diff --git a/include/drm/drm_gem_cma_helper.h b/include/drm/drm_gem_cma_helper.h
index 58a739bf15f1f..214aa85adc8d5 100644
--- a/include/drm/drm_gem_cma_helper.h
+++ b/include/drm/drm_gem_cma_helper.h
@@ -8,7 +8,9 @@ 
  * struct drm_gem_cma_object - GEM object backed by CMA memory allocations
  * @base: base GEM object
  * @paddr: physical address of the backing memory
- * @sgt: scatter/gather table for imported PRIME buffers
+ * @sgt: scatter/gather table for imported PRIME buffers. The table can have
+ *       more than one entry but they are guaranteed to have contiguous
+ *       DMA addresses.
  * @vaddr: kernel virtual address of the backing memory
  */
 struct drm_gem_cma_object {