[PATCH v2 00/21] DRM: fix struct sg_table nents vs. orig_nents misuse

[PATCH v2 00/21] DRM: fix struct sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

Dear All,

During the Exynos DRM GEM rework and fixing the issues in the 
drm_prime_sg_to_page_addr_arrays() function [1] I've noticed that most
drivers in DRM framework incorrectly use nents and orig_nents entries of
the struct sg_table.

In case of the most DMA-mapping implementations exchanging those two
entries or using nents for all loops on the scatterlist is harmless,
because they both have the same value. There exists however a DMA-mapping
implementations, for which such incorrect usage breaks things. The nents
returned by dma_map_sg() might be lower than the nents passed as its
parameter and this is perfectly fine. DMA framework or IOMMU is allowed
to join consecutive chunks while mapping if such operation is supported
by the underlying HW (bus, bridge, IOMMU, etc). Example of the case
where dma_map_sg() might return 1 'DMA' chunk for the 4 'physical' pages
is described here [2]

The DMA-mapping framework documentation [3] states that dma_map_sg()
returns the numer of the created entries in the DMA address space.
However the subsequent calls to dma_sync_sg_for_{device,cpu} and
dma_unmap_sg must be called with the original number of entries passed to
dma_map_sg. The common pattern in DRM drivers were to assign the
dma_map_sg() return value to sg_table->nents and use that value for
the subsequent calls to dma_sync_sg_* or dma_unmap_sg functions. Also
the code iterated over nents times to access the pages stored in the
processed scatterlist, while it should use orig_nents as the numer of
the page entries.

I've tried to identify all such incorrect usage of sg_table->nents and
this is a result of my research. It looks that the incorrect pattern has
been copied over the many drivers mainly in the DRM subsystem. Too bad in
most cases it even worked correctly if the system used simple,
linear DMA-mapping implementation, for which swapping nents and
orig_nents doesn't make any difference.

The biggest TODO is DRM/i915 driver and I don't feel brave enough to fix
it fully. The driver creatively uses sg_table->orig_nents to store the
size of the allocate scatterlist and ignores the number of the entries
returned by dma_map_sg function. In this patchset I only fixed the
sg_table objects exported by dmabuf related functions. I hope that I
didn't break anything there.

As a follow-up for this patchset I will prepare a common
dma_{map,sync,unmap}_sgtable() wrappers, which will use a proper sg_table
entries and call respective DMA-mapping functions. I hope this will help
to avoid issue like this in the future.

Patches are based on top of Linux next-20200504.

Best regards,
Marek Szyprowski


References:

[1] https://lkml.org/lkml/2020/3/27/555 
[2] https://lkml.org/lkml/2020/3/29/65
[3] Documentation/DMA-API-HOWTO.txt


Changelog:

v2:
- dropped most of the changes to drm/i915
- added fixes for rcar-du, xen, media and ion
- fixed a few issues pointed by kbuild test robot
- added wide cc: list for each patch

v1: https://lore.kernel.org/linux-iommu/c01c9766-9778-fd1f-f36e-2dc7bd376ba4@arm.com/T/#m879a727e4e46b5479ef8603994b1a006fb2aa337
- initial version


Patch summary:

Marek Szyprowski (21):
  drm: core: fix sg_table nents vs. orig_nents misuse
  drm: amdgpu: fix sg_table nents vs. orig_nents misuse
  drm: armada: fix sg_table nents vs. orig_nents misuse
  drm: etnaviv: fix sg_table nents vs. orig_nents misuse
  drm: exynos: fix sg_table nents vs. orig_nents misuse
  drm: i915: fix sg_table nents vs. orig_nents misuse for dmabuf objects
  drm: lima: fix sg_table nents vs. orig_nents misuse
  drm: msm: fix sg_table nents vs. orig_nents misuse
  drm: panfrost: fix sg_table nents vs. orig_nents misuse
  drm: radeon: fix sg_table nents vs. orig_nents misuse
  drm: rockchip: fix sg_table nents vs. orig_nents misuse
  drm: tegra: fix sg_table nents vs. orig_nents misuse
  drm: virtio: fix sg_table nents vs. orig_nents misuse
  drm: vmwgfx: fix sg_table nents vs. orig_nents misuse
  drm: xen: fix sg_table nents vs. orig_nents misuse
  drm: host1x: fix sg_table nents vs. orig_nents misuse
  drm: rcar-du: fix sg_table nents vs. orig_nents misuse
  xen: gntdev: fix sg_table nents vs. orig_nents misuse
  dmabuf: fix sg_table nents vs. orig_nents misuse
  media: pci: fix common ALSA DMA-mapping related code
  staging: ion: fix sg_table nents vs. orig_nents misuse

 drivers/dma-buf/heaps/heap-helpers.c             |  7 ++++---
 drivers/dma-buf/udmabuf.c                        |  5 +++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c      |  7 ++++---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c          |  8 ++++----
 drivers/gpu/drm/armada/armada_gem.c              | 14 ++++++++-----
 drivers/gpu/drm/drm_cache.c                      |  2 +-
 drivers/gpu/drm/drm_gem_shmem_helper.c           |  7 ++++---
 drivers/gpu/drm/drm_prime.c                      |  9 +++++----
 drivers/gpu/drm/etnaviv/etnaviv_gem.c            | 10 ++++++----
 drivers/gpu/drm/exynos/exynos_drm_g2d.c          |  7 ++++---
 drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c       |  9 +++++----
 drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.c |  5 +++--
 drivers/gpu/drm/lima/lima_gem.c                  |  4 ++--
 drivers/gpu/drm/msm/msm_gem.c                    |  8 ++++----
 drivers/gpu/drm/msm/msm_iommu.c                  |  3 ++-
 drivers/gpu/drm/panfrost/panfrost_gem.c          |  3 ++-
 drivers/gpu/drm/panfrost/panfrost_mmu.c          |  4 +++-
 drivers/gpu/drm/radeon/radeon_ttm.c              | 11 ++++++-----
 drivers/gpu/drm/rockchip/rockchip_drm_gem.c      | 15 +++++++-------
 drivers/gpu/drm/tegra/gem.c                      | 25 ++++++++++++------------
 drivers/gpu/drm/tegra/plane.c                    | 13 ++++++------
 drivers/gpu/drm/virtio/virtgpu_object.c          | 11 ++++++-----
 drivers/gpu/drm/virtio/virtgpu_vq.c              |  8 ++++----
 drivers/gpu/drm/vmwgfx/vmwgfx_ttm_buffer.c       |  6 +++---
 drivers/gpu/drm/xen/xen_drm_front_gem.c          |  2 +-
 drivers/gpu/host1x/job.c                         | 17 ++++++++--------
 drivers/media/pci/cx23885/cx23885-alsa.c         |  2 +-
 drivers/media/pci/cx25821/cx25821-alsa.c         |  2 +-
 drivers/media/pci/cx88/cx88-alsa.c               |  2 +-
 drivers/media/pci/saa7134/saa7134-alsa.c         |  2 +-
 drivers/media/platform/vsp1/vsp1_drm.c           |  7 ++++---
 drivers/staging/android/ion/ion.c                | 17 ++++++++--------
 drivers/staging/android/ion/ion_heap.c           |  6 +++---
 drivers/staging/android/ion/ion_system_heap.c    |  2 +-
 drivers/xen/gntdev-dmabuf.c                      | 10 ++++++----
 35 files changed, 149 insertions(+), 121 deletions(-)

-- 
1.9.1

Re: [PATCH v2 00/21] DRM: fix struct sg_table nents vs. orig_nents misuse

[Christoph Hellwig]

On Mon, May 04, 2020 at 02:50:17PM +0200, Marek Szyprowski wrote:
> v2:
> - dropped most of the changes to drm/i915
> - added fixes for rcar-du, xen, media and ion
> - fixed a few issues pointed by kbuild test robot
> - added wide cc: list for each patch

Didn't you plan to add dma_map_sgbuf and co helper?

Re: [PATCH v2 00/21] DRM: fix struct sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

Hi Christoph,

On 04.05.2020 14:52, Christoph Hellwig wrote:
> On Mon, May 04, 2020 at 02:50:17PM +0200, Marek Szyprowski wrote:
>> v2:
>> - dropped most of the changes to drm/i915
>> - added fixes for rcar-du, xen, media and ion
>> - fixed a few issues pointed by kbuild test robot
>> - added wide cc: list for each patch
> Didn't you plan to add dma_map_sgbuf and co helper?

Yes, I have a followup patches for that, but I wanted to fix the 
existing code in the first step. Then I wanted to send a wrappers and 
their application. Do you think I should do everything at once, in one 
patchset?

Best regards
-- 
Marek Szyprowski, PhD
Samsung R&D Institute Poland

Re: [PATCH v2 00/21] DRM: fix struct sg_table nents vs. orig_nents misuse

[Christoph Hellwig]

On Mon, May 04, 2020 at 03:05:30PM +0200, Marek Szyprowski wrote:
> Hi Christoph,
> 
> On 04.05.2020 14:52, Christoph Hellwig wrote:
> > On Mon, May 04, 2020 at 02:50:17PM +0200, Marek Szyprowski wrote:
> >> v2:
> >> - dropped most of the changes to drm/i915
> >> - added fixes for rcar-du, xen, media and ion
> >> - fixed a few issues pointed by kbuild test robot
> >> - added wide cc: list for each patch
> > Didn't you plan to add dma_map_sgbuf and co helper?
> 
> Yes, I have a followup patches for that, but I wanted to fix the 
> existing code in the first step. Then I wanted to send a wrappers and 
> their application. Do you think I should do everything at once, in one 
> patchset?

That would be my preference.  The helpers should be mostly trivial
wrappers, so they can easily backported, and they force passing of the
correct parameters.  So I don't really see a need to fix up all the 20+
places up first just to convert them to the proper API a little later.

[PATCH v2 01/21] drm: core: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
 drivers/gpu/drm/drm_cache.c            | 2 +-
 drivers/gpu/drm/drm_gem_shmem_helper.c | 7 ++++---
 drivers/gpu/drm/drm_prime.c            | 9 +++++----
 3 files changed, 10 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/drm_cache.c b/drivers/gpu/drm/drm_cache.c
index 03e01b0..63bd497 100644
--- a/drivers/gpu/drm/drm_cache.c
+++ b/drivers/gpu/drm/drm_cache.c
@@ -127,7 +127,7 @@ static void drm_cache_flush_clflush(struct page *pages[],
 		struct sg_page_iter sg_iter;
 
 		mb(); /*CLFLUSH is ordered only by using memory barriers*/
-		for_each_sg_page(st->sgl, &sg_iter, st->nents, 0)
+		for_each_sg_page(st->sgl, &sg_iter, st->orig_nents, 0)
 			drm_clflush_page(sg_page_iter_page(&sg_iter));
 		mb(); /*Make sure that all cache line entry is flushed*/
 
diff --git a/drivers/gpu/drm/drm_gem_shmem_helper.c b/drivers/gpu/drm/drm_gem_shmem_helper.c
index df31e57..f47caa7 100644
--- a/drivers/gpu/drm/drm_gem_shmem_helper.c
+++ b/drivers/gpu/drm/drm_gem_shmem_helper.c
@@ -118,7 +118,7 @@ void drm_gem_shmem_free_object(struct drm_gem_object *obj)
 	} else {
 		if (shmem->sgt) {
 			dma_unmap_sg(obj->dev->dev, shmem->sgt->sgl,
-				     shmem->sgt->nents, DMA_BIDIRECTIONAL);
+				     shmem->sgt->orig_nents, DMA_BIDIRECTIONAL);
 			sg_free_table(shmem->sgt);
 			kfree(shmem->sgt);
 		}
@@ -396,7 +396,7 @@ void drm_gem_shmem_purge_locked(struct drm_gem_object *obj)
 	WARN_ON(!drm_gem_shmem_is_purgeable(shmem));
 
 	dma_unmap_sg(obj->dev->dev, shmem->sgt->sgl,
-		     shmem->sgt->nents, DMA_BIDIRECTIONAL);
+		     shmem->sgt->orig_nents, DMA_BIDIRECTIONAL);
 	sg_free_table(shmem->sgt);
 	kfree(shmem->sgt);
 	shmem->sgt = NULL;
@@ -623,7 +623,8 @@ struct sg_table *drm_gem_shmem_get_pages_sgt(struct drm_gem_object *obj)
 		goto err_put_pages;
 	}
 	/* Map the pages for use by the h/w. */
-	dma_map_sg(obj->dev->dev, sgt->sgl, sgt->nents, DMA_BIDIRECTIONAL);
+	sgt->nents = dma_map_sg(obj->dev->dev, sgt->sgl, sgt->orig_nents,
+				DMA_BIDIRECTIONAL);
 
 	shmem->sgt = sgt;
 
diff --git a/drivers/gpu/drm/drm_prime.c b/drivers/gpu/drm/drm_prime.c
index 282774e..f3e2d2a 100644
--- a/drivers/gpu/drm/drm_prime.c
+++ b/drivers/gpu/drm/drm_prime.c
@@ -626,8 +626,9 @@ struct sg_table *drm_gem_map_dma_buf(struct dma_buf_attachment *attach,
 	else
 		sgt = obj->dev->driver->gem_prime_get_sg_table(obj);
 
-	if (!dma_map_sg_attrs(attach->dev, sgt->sgl, sgt->nents, dir,
-			      DMA_ATTR_SKIP_CPU_SYNC)) {
+	sgt->nents = dma_map_sg_attrs(attach->dev, sgt->sgl, sgt->orig_nents,
+				      dir, DMA_ATTR_SKIP_CPU_SYNC);
+	if (!sgt->nents) {
 		sg_free_table(sgt);
 		kfree(sgt);
 		sgt = ERR_PTR(-ENOMEM);
@@ -652,7 +653,7 @@ void drm_gem_unmap_dma_buf(struct dma_buf_attachment *attach,
 	if (!sgt)
 		return;
 
-	dma_unmap_sg_attrs(attach->dev, sgt->sgl, sgt->nents, dir,
+	dma_unmap_sg_attrs(attach->dev, sgt->sgl, sgt->orig_nents, dir,
 			   DMA_ATTR_SKIP_CPU_SYNC);
 	sg_free_table(sgt);
 	kfree(sgt);
@@ -975,7 +976,7 @@ int drm_prime_sg_to_page_addr_arrays(struct sg_table *sgt, struct page **pages,
 	 */
 	page_index = 0;
 	dma_index = 0;
-	for_each_sg(sgt->sgl, sg, sgt->nents, count) {
+	for_each_sg(sgt->sgl, sg, sgt->orig_nents, count) {
 		page_len = sg->length;
 		page = sg_page(sg);
 		dma_len = sg_dma_len(sg);
-- 
1.9.1

[PATCH v2 02/21] drm: amdgpu: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c | 7 ++++---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c     | 8 ++++----
 2 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
index 43d8ed7..4df813e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
@@ -307,8 +307,9 @@ static struct sg_table *amdgpu_dma_buf_map(struct dma_buf_attachment *attach,
 		if (IS_ERR(sgt))
 			return sgt;
 
-		if (!dma_map_sg_attrs(attach->dev, sgt->sgl, sgt->nents, dir,
-				      DMA_ATTR_SKIP_CPU_SYNC))
+		sgt->nents = dma_map_sg_attrs(attach->dev, sgt->sgl, sgt->orig_nents,
+					      dir, DMA_ATTR_SKIP_CPU_SYNC);
+		if (!sgt->nents)
 			goto error_free;
 		break;
 
@@ -349,7 +350,7 @@ static void amdgpu_dma_buf_unmap(struct dma_buf_attachment *attach,
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
 
 	if (sgt->sgl->page_link) {
-		dma_unmap_sg(attach->dev, sgt->sgl, sgt->nents, dir);
+		dma_unmap_sg(attach->dev, sgt->sgl, sgt->orig_nents, dir);
 		sg_free_table(sgt);
 		kfree(sgt);
 	} else {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index eff1f73..1f8c507 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -1043,7 +1043,6 @@ static int amdgpu_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(ttm->bdev);
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
-	unsigned nents;
 	int r;
 
 	int write = !(gtt->userflags & AMDGPU_GEM_USERPTR_READONLY);
@@ -1059,8 +1058,9 @@ static int amdgpu_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 
 	/* Map SG to device */
 	r = -ENOMEM;
-	nents = dma_map_sg(adev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
-	if (nents == 0)
+	ttm->sg->nents = dma_map_sg(adev->dev, ttm->sg->sgl,
+				    ttm->sg->orig_nents, direction);
+	if (ttm->sg->nents == 0)
 		goto release_sg;
 
 	/* convert SG to linear array of pages and dma addresses */
@@ -1091,7 +1091,7 @@ static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 		return;
 
 	/* unmap the pages mapped to the device */
-	dma_unmap_sg(adev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
+	dma_unmap_sg(adev->dev, ttm->sg->sgl, ttm->sg->orig_nents, direction);
 
 	sg_free_table(ttm->sg);
 
-- 
1.9.1

[PATCH v2 03/21] drm: armada: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/armada/armada_gem.c | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/armada/armada_gem.c b/drivers/gpu/drm/armada/armada_gem.c
index 976685f..749647f 100644
--- a/drivers/gpu/drm/armada/armada_gem.c
+++ b/drivers/gpu/drm/armada/armada_gem.c
@@ -407,8 +407,10 @@ int armada_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 			sg_set_page(sg, page, PAGE_SIZE, 0);
 		}
 
-		if (dma_map_sg(attach->dev, sgt->sgl, sgt->nents, dir) == 0) {
-			num = sgt->nents;
+		sgt->nents = dma_map_sg(attach->dev, sgt->sgl, sgt->orig_nents,
+					dir);
+		if (sgt->nents == 0) {
+			num = sgt->orig_nents;
 			goto release;
 		}
 	} else if (dobj->page) {
@@ -418,7 +420,9 @@ int armada_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 
 		sg_set_page(sgt->sgl, dobj->page, dobj->obj.size, 0);
 
-		if (dma_map_sg(attach->dev, sgt->sgl, sgt->nents, dir) == 0)
+		sgt->nents = dma_map_sg(attach->dev, sgt->sgl, sgt->orig_nents,
+					dir);
+		if (sgt->nents == 0)
 			goto free_table;
 	} else if (dobj->linear) {
 		/* Single contiguous physical region - no struct page */
@@ -449,11 +453,11 @@ static void armada_gem_prime_unmap_dma_buf(struct dma_buf_attachment *attach,
 	int i;
 
 	if (!dobj->linear)
-		dma_unmap_sg(attach->dev, sgt->sgl, sgt->nents, dir);
+		dma_unmap_sg(attach->dev, sgt->sgl, sgt->orig_nents, dir);
 
 	if (dobj->obj.filp) {
 		struct scatterlist *sg;
-		for_each_sg(sgt->sgl, sg, sgt->nents, i)
+		for_each_sg(sgt->sgl, sg, sgt->orig_nents, i)
 			put_page(sg_page(sg));
 	}
 
-- 
1.9.1

[PATCH v2 04/21] drm: etnaviv: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/etnaviv/etnaviv_gem.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.c b/drivers/gpu/drm/etnaviv/etnaviv_gem.c
index dc9ef30..a224a97 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.c
@@ -27,7 +27,8 @@ static void etnaviv_gem_scatter_map(struct etnaviv_gem_object *etnaviv_obj)
 	 * because display controller, GPU, etc. are not coherent.
 	 */
 	if (etnaviv_obj->flags & ETNA_BO_CACHE_MASK)
-		dma_map_sg(dev->dev, sgt->sgl, sgt->nents, DMA_BIDIRECTIONAL);
+		sgt->nents = dma_map_sg(dev->dev, sgt->sgl, sgt->orig_nents,
+					DMA_BIDIRECTIONAL);
 }
 
 static void etnaviv_gem_scatterlist_unmap(struct etnaviv_gem_object *etnaviv_obj)
@@ -51,7 +52,8 @@ static void etnaviv_gem_scatterlist_unmap(struct etnaviv_gem_object *etnaviv_obj
 	 * discard those writes.
 	 */
 	if (etnaviv_obj->flags & ETNA_BO_CACHE_MASK)
-		dma_unmap_sg(dev->dev, sgt->sgl, sgt->nents, DMA_BIDIRECTIONAL);
+		dma_unmap_sg(dev->dev, sgt->sgl, sgt->orig_nents,
+			     DMA_BIDIRECTIONAL);
 }
 
 /* called with etnaviv_obj->lock held */
@@ -405,7 +407,7 @@ int etnaviv_gem_cpu_prep(struct drm_gem_object *obj, u32 op,
 
 	if (etnaviv_obj->flags & ETNA_BO_CACHED) {
 		dma_sync_sg_for_cpu(dev->dev, etnaviv_obj->sgt->sgl,
-				    etnaviv_obj->sgt->nents,
+				    etnaviv_obj->sgt->orig_nents,
 				    etnaviv_op_to_dma_dir(op));
 		etnaviv_obj->last_cpu_prep_op = op;
 	}
@@ -422,7 +424,7 @@ int etnaviv_gem_cpu_fini(struct drm_gem_object *obj)
 		/* fini without a prep is almost certainly a userspace error */
 		WARN_ON(etnaviv_obj->last_cpu_prep_op == 0);
 		dma_sync_sg_for_device(dev->dev, etnaviv_obj->sgt->sgl,
-			etnaviv_obj->sgt->nents,
+			etnaviv_obj->sgt->orig_nents,
 			etnaviv_op_to_dma_dir(etnaviv_obj->last_cpu_prep_op));
 		etnaviv_obj->last_cpu_prep_op = 0;
 	}
-- 
1.9.1

[PATCH v2 05/21] drm: exynos: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/exynos/exynos_drm_g2d.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/exynos/exynos_drm_g2d.c b/drivers/gpu/drm/exynos/exynos_drm_g2d.c
index fcee33a..f995b0c 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_g2d.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_g2d.c
@@ -396,7 +396,7 @@ static void g2d_userptr_put_dma_addr(struct g2d_data *g2d,
 
 out:
 	dma_unmap_sg(to_dma_dev(g2d->drm_dev), g2d_userptr->sgt->sgl,
-			g2d_userptr->sgt->nents, DMA_BIDIRECTIONAL);
+		     g2d_userptr->sgt->orig_nents, DMA_BIDIRECTIONAL);
 
 	pages = frame_vector_pages(g2d_userptr->vec);
 	if (!IS_ERR(pages)) {
@@ -511,8 +511,9 @@ static dma_addr_t *g2d_userptr_get_dma_addr(struct g2d_data *g2d,
 
 	g2d_userptr->sgt = sgt;
 
-	if (!dma_map_sg(to_dma_dev(g2d->drm_dev), sgt->sgl, sgt->nents,
-				DMA_BIDIRECTIONAL)) {
+	sgt->nents = dma_map_sg(to_dma_dev(g2d->drm_dev), sgt->sgl,
+				sgt->orig_nents, DMA_BIDIRECTIONAL);
+	if (!sgt->nents) {
 		DRM_DEV_ERROR(g2d->dev, "failed to map sgt with dma region.\n");
 		ret = -ENOMEM;
 		goto err_sg_free_table;
-- 
1.9.1

[PATCH v2 06/21] drm: i915: fix sg_table nents vs. orig_nents misuse for dmabuf objects

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h.

This driver creatively uses sg_table->orig_nents to store the size of the
allocate scatterlist and ignores the number of the entries returned by
dma_map_sg function. The sg_table->orig_nents is (mis)used to properly
free the (over)allocated scatterlist.

This patch only fixes the sg_table->nents entries in the sg_table objects
exported by the dmabuf related functions, so the other drivers, which
might share buffers with i915 could rely on the nents and orig_nents
values.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c       | 9 +++++----
 drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.c | 5 +++--
 2 files changed, 8 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
index 7db5a79..98159df 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
@@ -48,9 +48,10 @@ static struct sg_table *i915_gem_map_dma_buf(struct dma_buf_attachment *attachme
 		src = sg_next(src);
 	}
 
-	if (!dma_map_sg_attrs(attachment->dev,
-			      st->sgl, st->nents, dir,
-			      DMA_ATTR_SKIP_CPU_SYNC)) {
+	st->nents = dma_map_sg_attrs(attachment->dev,
+				     st->sgl, st->orig_nents, dir,
+				     DMA_ATTR_SKIP_CPU_SYNC);
+	if (!st->nents) {
 		ret = -ENOMEM;
 		goto err_free_sg;
 	}
@@ -74,7 +75,7 @@ static void i915_gem_unmap_dma_buf(struct dma_buf_attachment *attachment,
 	struct drm_i915_gem_object *obj = dma_buf_to_obj(attachment->dmabuf);
 
 	dma_unmap_sg_attrs(attachment->dev,
-			   sg->sgl, sg->nents, dir,
+			   sg->sgl, sg->orig_nents, dir,
 			   DMA_ATTR_SKIP_CPU_SYNC);
 	sg_free_table(sg);
 	kfree(sg);
diff --git a/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.c b/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.c
index debaf7b..5723525 100644
--- a/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.c
+++ b/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.c
@@ -28,7 +28,8 @@ static struct sg_table *mock_map_dma_buf(struct dma_buf_attachment *attachment,
 		sg = sg_next(sg);
 	}
 
-	if (!dma_map_sg(attachment->dev, st->sgl, st->nents, dir)) {
+	st->nents = dma_map_sg(attachment->dev, st->sgl, st->orig_nents, dir);
+	if (!st->nents) {
 		err = -ENOMEM;
 		goto err_st;
 	}
@@ -46,7 +47,7 @@ static void mock_unmap_dma_buf(struct dma_buf_attachment *attachment,
 			       struct sg_table *st,
 			       enum dma_data_direction dir)
 {
-	dma_unmap_sg(attachment->dev, st->sgl, st->nents, dir);
+	dma_unmap_sg(attachment->dev, st->sgl, st->orig_nents, dir);
 	sg_free_table(st);
 	kfree(st);
 }
-- 
1.9.1

[PATCH v2 07/21] drm: lima: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/lima/lima_gem.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/lima/lima_gem.c b/drivers/gpu/drm/lima/lima_gem.c
index 5404e0d..3edd2ff 100644
--- a/drivers/gpu/drm/lima/lima_gem.c
+++ b/drivers/gpu/drm/lima/lima_gem.c
@@ -70,7 +70,7 @@ int lima_heap_alloc(struct lima_bo *bo, struct lima_vm *vm)
 
 	if (bo->base.sgt) {
 		dma_unmap_sg(dev, bo->base.sgt->sgl,
-			     bo->base.sgt->nents, DMA_BIDIRECTIONAL);
+			     bo->base.sgt->orig_nents, DMA_BIDIRECTIONAL);
 		sg_free_table(bo->base.sgt);
 	} else {
 		bo->base.sgt = kmalloc(sizeof(*bo->base.sgt), GFP_KERNEL);
@@ -80,7 +80,7 @@ int lima_heap_alloc(struct lima_bo *bo, struct lima_vm *vm)
 		}
 	}
 
-	dma_map_sg(dev, sgt.sgl, sgt.nents, DMA_BIDIRECTIONAL);
+	sgt.nents = dma_map_sg(dev, sgt.sgl, sgt.orig_nents, DMA_BIDIRECTIONAL);
 
 	*bo->base.sgt = sgt;
 
-- 
1.9.1

[PATCH v2 08/21] drm: msm: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/msm/msm_gem.c   | 8 ++++----
 drivers/gpu/drm/msm/msm_iommu.c | 3 ++-
 2 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c
index 5a6a79f..54c3bbb 100644
--- a/drivers/gpu/drm/msm/msm_gem.c
+++ b/drivers/gpu/drm/msm/msm_gem.c
@@ -54,10 +54,10 @@ static void sync_for_device(struct msm_gem_object *msm_obj)
 
 	if (get_dma_ops(dev) && IS_ENABLED(CONFIG_ARM64)) {
 		dma_sync_sg_for_device(dev, msm_obj->sgt->sgl,
-			msm_obj->sgt->nents, DMA_BIDIRECTIONAL);
+			msm_obj->sgt->orig_nents, DMA_BIDIRECTIONAL);
 	} else {
 		dma_map_sg(dev, msm_obj->sgt->sgl,
-			msm_obj->sgt->nents, DMA_BIDIRECTIONAL);
+			msm_obj->sgt->orig_nents, DMA_BIDIRECTIONAL);
 	}
 }
 
@@ -67,10 +67,10 @@ static void sync_for_cpu(struct msm_gem_object *msm_obj)
 
 	if (get_dma_ops(dev) && IS_ENABLED(CONFIG_ARM64)) {
 		dma_sync_sg_for_cpu(dev, msm_obj->sgt->sgl,
-			msm_obj->sgt->nents, DMA_BIDIRECTIONAL);
+			msm_obj->sgt->orig_nents, DMA_BIDIRECTIONAL);
 	} else {
 		dma_unmap_sg(dev, msm_obj->sgt->sgl,
-			msm_obj->sgt->nents, DMA_BIDIRECTIONAL);
+			msm_obj->sgt->orig_nents, DMA_BIDIRECTIONAL);
 	}
 }
 
diff --git a/drivers/gpu/drm/msm/msm_iommu.c b/drivers/gpu/drm/msm/msm_iommu.c
index ad58cfe..b0ca084 100644
--- a/drivers/gpu/drm/msm/msm_iommu.c
+++ b/drivers/gpu/drm/msm/msm_iommu.c
@@ -43,7 +43,8 @@ static int msm_iommu_map(struct msm_mmu *mmu, uint64_t iova,
 	struct msm_iommu *iommu = to_msm_iommu(mmu);
 	size_t ret;
 
-	ret = iommu_map_sg(iommu->domain, iova, sgt->sgl, sgt->nents, prot);
+	ret = iommu_map_sg(iommu->domain, iova, sgt->sgl, sgt->orig_nents,
+			   prot);
 	WARN_ON(!ret);
 
 	return (ret == len) ? 0 : -EINVAL;
-- 
1.9.1

[PATCH v2 09/21] drm: panfrost: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/panfrost/panfrost_gem.c | 3 ++-
 drivers/gpu/drm/panfrost/panfrost_mmu.c | 4 +++-
 2 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/panfrost/panfrost_gem.c b/drivers/gpu/drm/panfrost/panfrost_gem.c
index 17b654e..22fec7c 100644
--- a/drivers/gpu/drm/panfrost/panfrost_gem.c
+++ b/drivers/gpu/drm/panfrost/panfrost_gem.c
@@ -42,7 +42,8 @@ static void panfrost_gem_free_object(struct drm_gem_object *obj)
 		for (i = 0; i < n_sgt; i++) {
 			if (bo->sgts[i].sgl) {
 				dma_unmap_sg(pfdev->dev, bo->sgts[i].sgl,
-					     bo->sgts[i].nents, DMA_BIDIRECTIONAL);
+					     bo->sgts[i].orig_nents,
+					     DMA_BIDIRECTIONAL);
 				sg_free_table(&bo->sgts[i]);
 			}
 		}
diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c b/drivers/gpu/drm/panfrost/panfrost_mmu.c
index ed28aeb..2d9b1f9 100644
--- a/drivers/gpu/drm/panfrost/panfrost_mmu.c
+++ b/drivers/gpu/drm/panfrost/panfrost_mmu.c
@@ -517,7 +517,9 @@ static int panfrost_mmu_map_fault_addr(struct panfrost_device *pfdev, int as,
 	if (ret)
 		goto err_pages;
 
-	if (!dma_map_sg(pfdev->dev, sgt->sgl, sgt->nents, DMA_BIDIRECTIONAL)) {
+	sgt->nents = dma_map_sg(pfdev->dev, sgt->sgl, sgt->orig_nents,
+				DMA_BIDIRECTIONAL);
+	if (!sgt->nents) {
 		ret = -EINVAL;
 		goto err_map;
 	}
-- 
1.9.1

Re: [PATCH v2 09/21] drm: panfrost: fix sg_table nents vs. orig_nents misuse

[Steven Price]

On 04/05/2020 13:53, Marek Szyprowski wrote:
> The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
> numer of the created entries in the DMA address space. However the
> subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
> called with the original number of entries passed to dma_map_sg. The
> sg_table->nents in turn holds the result of the dma_map_sg call as stated
> in include/linux/scatterlist.h. Adapt the code to obey those rules.

I find this commit message a bit confusing, but AFAICT the problem with 
the Panfrost code is really in mmu_map_sg() where we don't have the 
return value from dma_map_sg() and the for_each_sg() loop could (in 
theory) run off the end of the list.

The fix seems correct - store the return where it's meant to be (nents) 
and make sure when unmapping we use the original (orig_nents). So you 
might also consider adding:

Fixes: f3ba91228e8e ("drm/panfrost: Add initial panfrost driver")

Even better would be the wrappers you mention in the cover letter! ;)

Reviewed-by: Steven Price <steven.price@arm.com>

> 
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---
> For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
> vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
> ---
>   drivers/gpu/drm/panfrost/panfrost_gem.c | 3 ++-
>   drivers/gpu/drm/panfrost/panfrost_mmu.c | 4 +++-
>   2 files changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/panfrost/panfrost_gem.c b/drivers/gpu/drm/panfrost/panfrost_gem.c
> index 17b654e..22fec7c 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_gem.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_gem.c
> @@ -42,7 +42,8 @@ static void panfrost_gem_free_object(struct drm_gem_object *obj)
>   		for (i = 0; i < n_sgt; i++) {
>   			if (bo->sgts[i].sgl) {
>   				dma_unmap_sg(pfdev->dev, bo->sgts[i].sgl,
> -					     bo->sgts[i].nents, DMA_BIDIRECTIONAL);
> +					     bo->sgts[i].orig_nents,
> +					     DMA_BIDIRECTIONAL);
>   				sg_free_table(&bo->sgts[i]);
>   			}
>   		}
> diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c b/drivers/gpu/drm/panfrost/panfrost_mmu.c
> index ed28aeb..2d9b1f9 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_mmu.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_mmu.c
> @@ -517,7 +517,9 @@ static int panfrost_mmu_map_fault_addr(struct panfrost_device *pfdev, int as,
>   	if (ret)
>   		goto err_pages;
>   
> -	if (!dma_map_sg(pfdev->dev, sgt->sgl, sgt->nents, DMA_BIDIRECTIONAL)) {
> +	sgt->nents = dma_map_sg(pfdev->dev, sgt->sgl, sgt->orig_nents,
> +				DMA_BIDIRECTIONAL);
> +	if (!sgt->nents) {
>   		ret = -EINVAL;
>   		goto err_map;
>   	}
> 

[PATCH v2 10/21] drm: radeon: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/radeon/radeon_ttm.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 5d50c9e..f8275c8 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -481,7 +481,7 @@ static int radeon_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 {
 	struct radeon_device *rdev = radeon_get_rdev(ttm->bdev);
 	struct radeon_ttm_tt *gtt = (void *)ttm;
-	unsigned pinned = 0, nents;
+	unsigned pinned = 0;
 	int r;
 
 	int write = !(gtt->userflags & RADEON_GEM_USERPTR_READONLY);
@@ -522,8 +522,9 @@ static int radeon_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 		goto release_sg;
 
 	r = -ENOMEM;
-	nents = dma_map_sg(rdev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
-	if (nents == 0)
+	ttm->sg->nents = dma_map_sg(rdev->dev, ttm->sg->sgl,
+				    ttm->sg->orig_nents, direction);
+	if (ttm->sg->nents == 0)
 		goto release_sg;
 
 	drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages,
@@ -554,9 +555,9 @@ static void radeon_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 		return;
 
 	/* free the sg table and pages again */
-	dma_unmap_sg(rdev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
+	dma_unmap_sg(rdev->dev, ttm->sg->sgl, ttm->sg->orig_nents, direction);
 
-	for_each_sg_page(ttm->sg->sgl, &sg_iter, ttm->sg->nents, 0) {
+	for_each_sg_page(ttm->sg->sgl, &sg_iter, ttm->sg->orig_nents, 0) {
 		struct page *page = sg_page_iter_page(&sg_iter);
 		if (!(gtt->userflags & RADEON_GEM_USERPTR_READONLY))
 			set_page_dirty(page);
-- 
1.9.1

[PATCH v2 11/21] drm: rockchip: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/rockchip/rockchip_drm_gem.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_gem.c b/drivers/gpu/drm/rockchip/rockchip_drm_gem.c
index 0d18846..a024c71 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_gem.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_gem.c
@@ -37,7 +37,7 @@ static int rockchip_gem_iommu_map(struct rockchip_gem_object *rk_obj)
 	rk_obj->dma_addr = rk_obj->mm.start;
 
 	ret = iommu_map_sg(private->domain, rk_obj->dma_addr, rk_obj->sgt->sgl,
-			   rk_obj->sgt->nents, prot);
+			   rk_obj->sgt->orig_nents, prot);
 	if (ret < rk_obj->base.size) {
 		DRM_ERROR("failed to map buffer: size=%zd request_size=%zd\n",
 			  ret, rk_obj->base.size);
@@ -98,11 +98,11 @@ static int rockchip_gem_get_pages(struct rockchip_gem_object *rk_obj)
 	 * TODO: Replace this by drm_clflush_sg() once it can be implemented
 	 * without relying on symbols that are not exported.
 	 */
-	for_each_sg(rk_obj->sgt->sgl, s, rk_obj->sgt->nents, i)
+	for_each_sg(rk_obj->sgt->sgl, s, rk_obj->sgt->orig_nents, i)
 		sg_dma_address(s) = sg_phys(s);
 
-	dma_sync_sg_for_device(drm->dev, rk_obj->sgt->sgl, rk_obj->sgt->nents,
-			       DMA_TO_DEVICE);
+	dma_sync_sg_for_device(drm->dev, rk_obj->sgt->sgl,
+			       rk_obj->sgt->orig_nents, DMA_TO_DEVICE);
 
 	return 0;
 
@@ -351,7 +351,8 @@ void rockchip_gem_free_object(struct drm_gem_object *obj)
 			rockchip_gem_iommu_unmap(rk_obj);
 		} else {
 			dma_unmap_sg(drm->dev, rk_obj->sgt->sgl,
-				     rk_obj->sgt->nents, DMA_BIDIRECTIONAL);
+				     rk_obj->sgt->orig_nents,
+				     DMA_BIDIRECTIONAL);
 		}
 		drm_prime_gem_destroy(obj, rk_obj->sgt);
 	} else {
@@ -493,14 +494,14 @@ static unsigned long rockchip_sg_get_contiguous_size(struct sg_table *sgt,
 			struct sg_table *sg,
 			struct rockchip_gem_object *rk_obj)
 {
-	int count = dma_map_sg(drm->dev, sg->sgl, sg->nents,
+	int count = dma_map_sg(drm->dev, sg->sgl, sg->orig_nents,
 			       DMA_BIDIRECTIONAL);
 	if (!count)
 		return -EINVAL;
 
 	if (rockchip_sg_get_contiguous_size(sg, count) < attach->dmabuf->size) {
 		DRM_ERROR("failed to map sg_table to contiguous linear address.\n");
-		dma_unmap_sg(drm->dev, sg->sgl, sg->nents,
+		dma_unmap_sg(drm->dev, sg->sgl, sg->orig_nents,
 			     DMA_BIDIRECTIONAL);
 		return -EINVAL;
 	}
-- 
1.9.1

[PATCH v2 12/21] drm: tegra: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/tegra/gem.c   | 25 +++++++++++++------------
 drivers/gpu/drm/tegra/plane.c | 13 +++++++------
 2 files changed, 20 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c
index 6237681..5710ab4 100644
--- a/drivers/gpu/drm/tegra/gem.c
+++ b/drivers/gpu/drm/tegra/gem.c
@@ -98,8 +98,8 @@ static struct sg_table *tegra_bo_pin(struct device *dev, struct host1x_bo *bo,
 		 * the SG table needs to be copied to avoid overwriting any
 		 * other potential users of the original SG table.
 		 */
-		err = sg_alloc_table_from_sg(sgt, obj->sgt->sgl, obj->sgt->nents,
-					     GFP_KERNEL);
+		err = sg_alloc_table_from_sg(sgt, obj->sgt->sgl,
+					     obj->sgt->orig_nents, GFP_KERNEL);
 		if (err < 0)
 			goto free;
 	} else {
@@ -197,7 +197,7 @@ static int tegra_bo_iommu_map(struct tegra_drm *tegra, struct tegra_bo *bo)
 	bo->iova = bo->mm->start;
 
 	bo->size = iommu_map_sg(tegra->domain, bo->iova, bo->sgt->sgl,
-				bo->sgt->nents, prot);
+				bo->sgt->orig_nents, prot);
 	if (!bo->size) {
 		dev_err(tegra->drm->dev, "failed to map buffer\n");
 		err = -ENOMEM;
@@ -264,7 +264,7 @@ static struct tegra_bo *tegra_bo_alloc_object(struct drm_device *drm,
 static void tegra_bo_free(struct drm_device *drm, struct tegra_bo *bo)
 {
 	if (bo->pages) {
-		dma_unmap_sg(drm->dev, bo->sgt->sgl, bo->sgt->nents,
+		dma_unmap_sg(drm->dev, bo->sgt->sgl, bo->sgt->orig_nents,
 			     DMA_FROM_DEVICE);
 		drm_gem_put_pages(&bo->gem, bo->pages, true, true);
 		sg_free_table(bo->sgt);
@@ -290,9 +290,9 @@ static int tegra_bo_get_pages(struct drm_device *drm, struct tegra_bo *bo)
 		goto put_pages;
 	}
 
-	err = dma_map_sg(drm->dev, bo->sgt->sgl, bo->sgt->nents,
-			 DMA_FROM_DEVICE);
-	if (err == 0) {
+	bo->sgt->nents = dma_map_sg(drm->dev, bo->sgt->sgl, bo->sgt->orig_nents,
+				DMA_FROM_DEVICE);
+	if (bo->sgt->nents == 0) {
 		err = -EFAULT;
 		goto free_sgt;
 	}
@@ -571,7 +571,8 @@ int tegra_drm_mmap(struct file *file, struct vm_area_struct *vma)
 			goto free;
 	}
 
-	if (dma_map_sg(attach->dev, sgt->sgl, sgt->nents, dir) == 0)
+	sgt->nents = dma_map_sg(attach->dev, sgt->sgl, sgt->orig_nents, dir);
+	if (sgt->nents == 0)
 		goto free;
 
 	return sgt;
@@ -590,7 +591,7 @@ static void tegra_gem_prime_unmap_dma_buf(struct dma_buf_attachment *attach,
 	struct tegra_bo *bo = to_tegra_bo(gem);
 
 	if (bo->pages)
-		dma_unmap_sg(attach->dev, sgt->sgl, sgt->nents, dir);
+		dma_unmap_sg(attach->dev, sgt->sgl, sgt->orig_nents, dir);
 
 	sg_free_table(sgt);
 	kfree(sgt);
@@ -609,7 +610,7 @@ static int tegra_gem_prime_begin_cpu_access(struct dma_buf *buf,
 	struct drm_device *drm = gem->dev;
 
 	if (bo->pages)
-		dma_sync_sg_for_cpu(drm->dev, bo->sgt->sgl, bo->sgt->nents,
+		dma_sync_sg_for_cpu(drm->dev, bo->sgt->sgl, bo->sgt->orig_nents,
 				    DMA_FROM_DEVICE);
 
 	return 0;
@@ -623,8 +624,8 @@ static int tegra_gem_prime_end_cpu_access(struct dma_buf *buf,
 	struct drm_device *drm = gem->dev;
 
 	if (bo->pages)
-		dma_sync_sg_for_device(drm->dev, bo->sgt->sgl, bo->sgt->nents,
-				       DMA_TO_DEVICE);
+		dma_sync_sg_for_device(drm->dev, bo->sgt->sgl,
+				       bo->sgt->orig_nents, DMA_TO_DEVICE);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/tegra/plane.c b/drivers/gpu/drm/tegra/plane.c
index 9ccfb56..3982bf8 100644
--- a/drivers/gpu/drm/tegra/plane.c
+++ b/drivers/gpu/drm/tegra/plane.c
@@ -130,9 +130,10 @@ static int tegra_dc_pin(struct tegra_dc *dc, struct tegra_plane_state *state)
 		}
 
 		if (sgt) {
-			err = dma_map_sg(dc->dev, sgt->sgl, sgt->nents,
-					 DMA_TO_DEVICE);
-			if (err == 0) {
+			sgt->nents = dma_map_sg(dc->dev, sgt->sgl,
+						sgt->orig_nents,
+						DMA_TO_DEVICE);
+			if (sgt->nents == 0) {
 				err = -ENOMEM;
 				goto unpin;
 			}
@@ -143,7 +144,7 @@ static int tegra_dc_pin(struct tegra_dc *dc, struct tegra_plane_state *state)
 			 * map its SG table to a single contiguous chunk of
 			 * I/O virtual memory.
 			 */
-			if (err > 1) {
+			if (sgt->nents > 1) {
 				err = -EINVAL;
 				goto unpin;
 			}
@@ -165,7 +166,7 @@ static int tegra_dc_pin(struct tegra_dc *dc, struct tegra_plane_state *state)
 		struct sg_table *sgt = state->sgt[i];
 
 		if (sgt)
-			dma_unmap_sg(dc->dev, sgt->sgl, sgt->nents,
+			dma_unmap_sg(dc->dev, sgt->sgl, sgt->orig_nents,
 				     DMA_TO_DEVICE);
 
 		host1x_bo_unpin(dc->dev, &bo->base, sgt);
@@ -185,7 +186,7 @@ static void tegra_dc_unpin(struct tegra_dc *dc, struct tegra_plane_state *state)
 		struct sg_table *sgt = state->sgt[i];
 
 		if (sgt)
-			dma_unmap_sg(dc->dev, sgt->sgl, sgt->nents,
+			dma_unmap_sg(dc->dev, sgt->sgl, sgt->orig_nents,
 				     DMA_TO_DEVICE);
 
 		host1x_bo_unpin(dc->dev, &bo->base, sgt);
-- 
1.9.1

[PATCH v2 13/21] drm: virtio: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/virtio/virtgpu_object.c | 11 ++++++-----
 drivers/gpu/drm/virtio/virtgpu_vq.c     |  8 ++++----
 2 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/virtio/virtgpu_object.c b/drivers/gpu/drm/virtio/virtgpu_object.c
index 6ccbd01..12f6bee 100644
--- a/drivers/gpu/drm/virtio/virtgpu_object.c
+++ b/drivers/gpu/drm/virtio/virtgpu_object.c
@@ -73,7 +73,8 @@ void virtio_gpu_cleanup_object(struct virtio_gpu_object *bo)
 		if (shmem->pages) {
 			if (shmem->mapped) {
 				dma_unmap_sg(vgdev->vdev->dev.parent,
-					     shmem->pages->sgl, shmem->mapped,
+					     shmem->pages->sgl,
+					     shmem->pages->orig_nents,
 					     DMA_TO_DEVICE);
 				shmem->mapped = 0;
 			}
@@ -157,13 +158,13 @@ static int virtio_gpu_object_shmem_init(struct virtio_gpu_device *vgdev,
 	}
 
 	if (use_dma_api) {
-		shmem->mapped = dma_map_sg(vgdev->vdev->dev.parent,
+		shmem->pages->nents = dma_map_sg(vgdev->vdev->dev.parent,
 					   shmem->pages->sgl,
-					   shmem->pages->nents,
+					   shmem->pages->orig_nents,
 					   DMA_TO_DEVICE);
-		*nents = shmem->mapped;
+		*nents = shmem->mapped = shmem->pages->nents;
 	} else {
-		*nents = shmem->pages->nents;
+		*nents = shmem->pages->orig_nents;
 	}
 
 	*ents = kmalloc_array(*nents, sizeof(struct virtio_gpu_mem_entry),
diff --git a/drivers/gpu/drm/virtio/virtgpu_vq.c b/drivers/gpu/drm/virtio/virtgpu_vq.c
index 9e663a5..661325b 100644
--- a/drivers/gpu/drm/virtio/virtgpu_vq.c
+++ b/drivers/gpu/drm/virtio/virtgpu_vq.c
@@ -604,8 +604,8 @@ void virtio_gpu_cmd_transfer_to_host_2d(struct virtio_gpu_device *vgdev,
 
 	if (use_dma_api)
 		dma_sync_sg_for_device(vgdev->vdev->dev.parent,
-				       shmem->pages->sgl, shmem->pages->nents,
-				       DMA_TO_DEVICE);
+				       shmem->pages->sgl,
+				       shmem->pages->orig_nents, DMA_TO_DEVICE);
 
 	cmd_p = virtio_gpu_alloc_cmd(vgdev, &vbuf, sizeof(*cmd_p));
 	memset(cmd_p, 0, sizeof(*cmd_p));
@@ -1020,8 +1020,8 @@ void virtio_gpu_cmd_transfer_to_host_3d(struct virtio_gpu_device *vgdev,
 
 	if (use_dma_api)
 		dma_sync_sg_for_device(vgdev->vdev->dev.parent,
-				       shmem->pages->sgl, shmem->pages->nents,
-				       DMA_TO_DEVICE);
+				       shmem->pages->sgl,
+				       shmem->pages->orig_nents, DMA_TO_DEVICE);
 
 	cmd_p = virtio_gpu_alloc_cmd(vgdev, &vbuf, sizeof(*cmd_p));
 	memset(cmd_p, 0, sizeof(*cmd_p));
-- 
1.9.1

[PATCH v2 14/21] drm: vmwgfx: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/vmwgfx/vmwgfx_ttm_buffer.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_buffer.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_buffer.c
index bf0bc46..a5fd128 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_buffer.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_buffer.c
@@ -362,7 +362,7 @@ static void vmw_ttm_unmap_from_dma(struct vmw_ttm_tt *vmw_tt)
 {
 	struct device *dev = vmw_tt->dev_priv->dev->dev;
 
-	dma_unmap_sg(dev, vmw_tt->sgt.sgl, vmw_tt->sgt.nents,
+	dma_unmap_sg(dev, vmw_tt->sgt.sgl, vmw_tt->sgt.orig_nents,
 		DMA_BIDIRECTIONAL);
 	vmw_tt->sgt.nents = vmw_tt->sgt.orig_nents;
 }
@@ -449,10 +449,10 @@ static int vmw_ttm_map_dma(struct vmw_ttm_tt *vmw_tt)
 		if (unlikely(ret != 0))
 			goto out_sg_alloc_fail;
 
-		if (vsgt->num_pages > vmw_tt->sgt.nents) {
+		if (vsgt->num_pages > vmw_tt->sgt.orig_nents) {
 			uint64_t over_alloc =
 				sgl_size * (vsgt->num_pages -
-					    vmw_tt->sgt.nents);
+					    vmw_tt->sgt.orig_nents);
 
 			ttm_mem_global_free(glob, over_alloc);
 			vmw_tt->sg_alloc_size -= over_alloc;
-- 
1.9.1

[PATCH v2 15/21] drm: xen: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/drm/xen/xen_drm_front_gem.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/xen/xen_drm_front_gem.c b/drivers/gpu/drm/xen/xen_drm_front_gem.c
index f0b85e0..ba4bdc5 100644
--- a/drivers/gpu/drm/xen/xen_drm_front_gem.c
+++ b/drivers/gpu/drm/xen/xen_drm_front_gem.c
@@ -215,7 +215,7 @@ struct drm_gem_object *
 		return ERR_PTR(ret);
 
 	DRM_DEBUG("Imported buffer of size %zu with nents %u\n",
-		  size, sgt->nents);
+		  size, sgt->orig_nents);
 
 	return &xen_obj->base;
 }
-- 
1.9.1

[PATCH v2 16/21] drm: host1x: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/gpu/host1x/job.c | 17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c
index a10643a..3ea185e 100644
--- a/drivers/gpu/host1x/job.c
+++ b/drivers/gpu/host1x/job.c
@@ -166,8 +166,9 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 				goto unpin;
 			}
 
-			err = dma_map_sg(dev, sgt->sgl, sgt->nents, dir);
-			if (!err) {
+			sgt->nents = dma_map_sg(dev, sgt->sgl, sgt->orig_nents,
+						dir);
+			if (!sgt->nents) {
 				err = -ENOMEM;
 				goto unpin;
 			}
@@ -217,7 +218,7 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 		}
 
 		if (!IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) && host->domain) {
-			for_each_sg(sgt->sgl, sg, sgt->nents, j)
+			for_each_sg(sgt->sgl, sg, sgt->orig_nents, j)
 				gather_size += sg->length;
 			gather_size = iova_align(&host->iova, gather_size);
 
@@ -231,7 +232,7 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 
 			err = iommu_map_sg(host->domain,
 					iova_dma_addr(&host->iova, alloc),
-					sgt->sgl, sgt->nents, IOMMU_READ);
+					sgt->sgl, sgt->orig_nents, IOMMU_READ);
 			if (err == 0) {
 				__free_iova(&host->iova, alloc);
 				err = -EINVAL;
@@ -241,9 +242,9 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 			job->unpins[job->num_unpins].size = gather_size;
 			phys_addr = iova_dma_addr(&host->iova, alloc);
 		} else if (sgt) {
-			err = dma_map_sg(host->dev, sgt->sgl, sgt->nents,
-					 DMA_TO_DEVICE);
-			if (!err) {
+			sgt->nents = dma_map_sg(host->dev, sgt->sgl,
+						sgt->orig_nents, DMA_TO_DEVICE);
+			if (!sgt->nents) {
 				err = -ENOMEM;
 				goto unpin;
 			}
@@ -647,7 +648,7 @@ void host1x_job_unpin(struct host1x_job *job)
 		}
 
 		if (unpin->dev && sgt)
-			dma_unmap_sg(unpin->dev, sgt->sgl, sgt->nents,
+			dma_unmap_sg(unpin->dev, sgt->sgl, sgt->orig_nents,
 				     unpin->dir);
 
 		host1x_bo_unpin(dev, unpin->bo, sgt);
-- 
1.9.1

[PATCH v2 17/21] drm: rcar-du: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/media/platform/vsp1/vsp1_drm.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/media/platform/vsp1/vsp1_drm.c b/drivers/media/platform/vsp1/vsp1_drm.c
index a4a45d6..b54a30f 100644
--- a/drivers/media/platform/vsp1/vsp1_drm.c
+++ b/drivers/media/platform/vsp1/vsp1_drm.c
@@ -912,8 +912,9 @@ int vsp1_du_map_sg(struct device *dev, struct sg_table *sgt)
 	 * skip cache sync. This will need to be revisited when support for
 	 * non-coherent buffers will be added to the DU driver.
 	 */
-	return dma_map_sg_attrs(vsp1->bus_master, sgt->sgl, sgt->nents,
-				DMA_TO_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
+	sgt->nents = dma_map_sg_attrs(vsp1->bus_master, sgt->sgl,
+			sgt->orig_nents, DMA_TO_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
+	return sgt->nents;
 }
 EXPORT_SYMBOL_GPL(vsp1_du_map_sg);
 
@@ -921,7 +922,7 @@ void vsp1_du_unmap_sg(struct device *dev, struct sg_table *sgt)
 {
 	struct vsp1_device *vsp1 = dev_get_drvdata(dev);
 
-	dma_unmap_sg_attrs(vsp1->bus_master, sgt->sgl, sgt->nents,
+	dma_unmap_sg_attrs(vsp1->bus_master, sgt->sgl, sgt->orig_nents,
 			   DMA_TO_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
 }
 EXPORT_SYMBOL_GPL(vsp1_du_unmap_sg);
-- 
1.9.1

[PATCH v2 18/21] xen: gntdev: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/xen/gntdev-dmabuf.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/xen/gntdev-dmabuf.c b/drivers/xen/gntdev-dmabuf.c
index 75d3bb9..ed749fd 100644
--- a/drivers/xen/gntdev-dmabuf.c
+++ b/drivers/xen/gntdev-dmabuf.c
@@ -248,7 +248,7 @@ static void dmabuf_exp_ops_detach(struct dma_buf *dma_buf,
 		if (sgt) {
 			if (gntdev_dmabuf_attach->dir != DMA_NONE)
 				dma_unmap_sg_attrs(attach->dev, sgt->sgl,
-						   sgt->nents,
+						   sgt->orig_nents,
 						   gntdev_dmabuf_attach->dir,
 						   DMA_ATTR_SKIP_CPU_SYNC);
 			sg_free_table(sgt);
@@ -288,8 +288,10 @@ static void dmabuf_exp_ops_detach(struct dma_buf *dma_buf,
 	sgt = dmabuf_pages_to_sgt(gntdev_dmabuf->pages,
 				  gntdev_dmabuf->nr_pages);
 	if (!IS_ERR(sgt)) {
-		if (!dma_map_sg_attrs(attach->dev, sgt->sgl, sgt->nents, dir,
-				      DMA_ATTR_SKIP_CPU_SYNC)) {
+		sgt->nents = dma_map_sg_attrs(attach->dev, sgt->sgl,
+					      sgt->orig_nents, dir,
+					      DMA_ATTR_SKIP_CPU_SYNC);
+		if (!sgt->nents) {
 			sg_free_table(sgt);
 			kfree(sgt);
 			sgt = ERR_PTR(-ENOMEM);
@@ -625,7 +627,7 @@ static struct gntdev_dmabuf *dmabuf_imp_alloc_storage(int count)
 
 	/* Now convert sgt to array of pages and check for page validity. */
 	i = 0;
-	for_each_sg_page(sgt->sgl, &sg_iter, sgt->nents, 0) {
+	for_each_sg_page(sgt->sgl, &sg_iter, sgt->orig_nents, 0) {
 		struct page *page = sg_page_iter_page(&sg_iter);
 		/*
 		 * Check if page is valid: this can happen if we are given
-- 
1.9.1

[PATCH v2 19/21] dmabuf: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/dma-buf/heaps/heap-helpers.c | 7 ++++---
 drivers/dma-buf/udmabuf.c            | 5 +++--
 2 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/drivers/dma-buf/heaps/heap-helpers.c b/drivers/dma-buf/heaps/heap-helpers.c
index 9f964ca..b923863 100644
--- a/drivers/dma-buf/heaps/heap-helpers.c
+++ b/drivers/dma-buf/heaps/heap-helpers.c
@@ -144,8 +144,9 @@ struct sg_table *dma_heap_map_dma_buf(struct dma_buf_attachment *attachment,
 
 	table = &a->table;
 
-	if (!dma_map_sg(attachment->dev, table->sgl, table->nents,
-			direction))
+	table->nents = dma_map_sg(attachment->dev, table->sgl,
+				  table->orig_nents, direction);
+	if (!table->nents)
 		table = ERR_PTR(-ENOMEM);
 	return table;
 }
@@ -154,7 +155,7 @@ static void dma_heap_unmap_dma_buf(struct dma_buf_attachment *attachment,
 				   struct sg_table *table,
 				   enum dma_data_direction direction)
 {
-	dma_unmap_sg(attachment->dev, table->sgl, table->nents, direction);
+	dma_unmap_sg(attachment->dev, table->sgl, table->orig_nents, direction);
 }
 
 static vm_fault_t dma_heap_vm_fault(struct vm_fault *vmf)
diff --git a/drivers/dma-buf/udmabuf.c b/drivers/dma-buf/udmabuf.c
index acb26c6..ea0cf71 100644
--- a/drivers/dma-buf/udmabuf.c
+++ b/drivers/dma-buf/udmabuf.c
@@ -63,7 +63,8 @@ static struct sg_table *get_sg_table(struct device *dev, struct dma_buf *buf,
 					GFP_KERNEL);
 	if (ret < 0)
 		goto err;
-	if (!dma_map_sg(dev, sg->sgl, sg->nents, direction)) {
+	sg->nents = dma_map_sg(dev, sg->sgl, sg->orig_nents, direction);
+	if (!sg->nents) {
 		ret = -EINVAL;
 		goto err;
 	}
@@ -78,7 +79,7 @@ static struct sg_table *get_sg_table(struct device *dev, struct dma_buf *buf,
 static void put_sg_table(struct device *dev, struct sg_table *sg,
 			 enum dma_data_direction direction)
 {
-	dma_unmap_sg(dev, sg->sgl, sg->nents, direction);
+	dma_unmap_sg(dev, sg->sgl, sg->orig_nents, direction);
 	sg_free_table(sg);
 	kfree(sg);
 }
-- 
1.9.1

[PATCH v2 20/21] media: pci: fix common ALSA DMA-mapping related code

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/media/pci/cx23885/cx23885-alsa.c | 2 +-
 drivers/media/pci/cx25821/cx25821-alsa.c | 2 +-
 drivers/media/pci/cx88/cx88-alsa.c       | 2 +-
 drivers/media/pci/saa7134/saa7134-alsa.c | 2 +-
 4 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/media/pci/cx23885/cx23885-alsa.c b/drivers/media/pci/cx23885/cx23885-alsa.c
index df44ed7..3f366e4 100644
--- a/drivers/media/pci/cx23885/cx23885-alsa.c
+++ b/drivers/media/pci/cx23885/cx23885-alsa.c
@@ -129,7 +129,7 @@ static int cx23885_alsa_dma_unmap(struct cx23885_audio_dev *dev)
 	if (!buf->sglen)
 		return 0;
 
-	dma_unmap_sg(&dev->pci->dev, buf->sglist, buf->sglen, PCI_DMA_FROMDEVICE);
+	dma_unmap_sg(&dev->pci->dev, buf->sglist, buf->nr_pages, PCI_DMA_FROMDEVICE);
 	buf->sglen = 0;
 	return 0;
 }
diff --git a/drivers/media/pci/cx25821/cx25821-alsa.c b/drivers/media/pci/cx25821/cx25821-alsa.c
index 3016164..c40304d 100644
--- a/drivers/media/pci/cx25821/cx25821-alsa.c
+++ b/drivers/media/pci/cx25821/cx25821-alsa.c
@@ -193,7 +193,7 @@ static int cx25821_alsa_dma_unmap(struct cx25821_audio_dev *dev)
 	if (!buf->sglen)
 		return 0;
 
-	dma_unmap_sg(&dev->pci->dev, buf->sglist, buf->sglen, PCI_DMA_FROMDEVICE);
+	dma_unmap_sg(&dev->pci->dev, buf->sglist, buf->nr_pages, PCI_DMA_FROMDEVICE);
 	buf->sglen = 0;
 	return 0;
 }
diff --git a/drivers/media/pci/cx88/cx88-alsa.c b/drivers/media/pci/cx88/cx88-alsa.c
index 7d7acee..3c6fe6c 100644
--- a/drivers/media/pci/cx88/cx88-alsa.c
+++ b/drivers/media/pci/cx88/cx88-alsa.c
@@ -332,7 +332,7 @@ static int cx88_alsa_dma_unmap(struct cx88_audio_dev *dev)
 	if (!buf->sglen)
 		return 0;
 
-	dma_unmap_sg(&dev->pci->dev, buf->sglist, buf->sglen,
+	dma_unmap_sg(&dev->pci->dev, buf->sglist, buf->nr_pages,
 		     PCI_DMA_FROMDEVICE);
 	buf->sglen = 0;
 	return 0;
diff --git a/drivers/media/pci/saa7134/saa7134-alsa.c b/drivers/media/pci/saa7134/saa7134-alsa.c
index 544ca57..398c47f 100644
--- a/drivers/media/pci/saa7134/saa7134-alsa.c
+++ b/drivers/media/pci/saa7134/saa7134-alsa.c
@@ -313,7 +313,7 @@ static int saa7134_alsa_dma_unmap(struct saa7134_dev *dev)
 	if (!dma->sglen)
 		return 0;
 
-	dma_unmap_sg(&dev->pci->dev, dma->sglist, dma->sglen, PCI_DMA_FROMDEVICE);
+	dma_unmap_sg(&dev->pci->dev, dma->sglist, dma->nr_pages, PCI_DMA_FROMDEVICE);
 	dma->sglen = 0;
 	return 0;
 }
-- 
1.9.1

[PATCH v2 21/21] staging: ion: fix sg_table nents vs. orig_nents misuse

[Marek Szyprowski]

The Documentation/DMA-API-HOWTO.txt states that dma_map_sg returns the
numer of the created entries in the DMA address space. However the
subsequent calls to dma_sync_sg_for_{device,cpu} and dma_unmap_sg must be
called with the original number of entries passed to dma_map_sg. The
sg_table->nents in turn holds the result of the dma_map_sg call as stated
in include/linux/scatterlist.h. Adapt the code to obey those rules.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
For more information, see '[PATCH v2 00/21] DRM: fix struct sg_table nents
vs. orig_nents misuse' thread: https://lkml.org/lkml/2020/5/4/373
---
 drivers/staging/android/ion/ion.c             | 17 +++++++++--------
 drivers/staging/android/ion/ion_heap.c        |  6 +++---
 drivers/staging/android/ion/ion_system_heap.c |  2 +-
 3 files changed, 13 insertions(+), 12 deletions(-)

diff --git a/drivers/staging/android/ion/ion.c b/drivers/staging/android/ion/ion.c
index 38b51ea..b14170c 100644
--- a/drivers/staging/android/ion/ion.c
+++ b/drivers/staging/android/ion/ion.c
@@ -147,14 +147,14 @@ static struct sg_table *dup_sg_table(struct sg_table *table)
 	if (!new_table)
 		return ERR_PTR(-ENOMEM);
 
-	ret = sg_alloc_table(new_table, table->nents, GFP_KERNEL);
+	ret = sg_alloc_table(new_table, table->orig_nents, GFP_KERNEL);
 	if (ret) {
 		kfree(new_table);
 		return ERR_PTR(-ENOMEM);
 	}
 
 	new_sg = new_table->sgl;
-	for_each_sg(table->sgl, sg, table->nents, i) {
+	for_each_sg(table->sgl, sg, table->orig_nents, i) {
 		memcpy(new_sg, sg, sizeof(*sg));
 		new_sg->dma_address = 0;
 		new_sg = sg_next(new_sg);
@@ -227,8 +227,9 @@ static struct sg_table *ion_map_dma_buf(struct dma_buf_attachment *attachment,
 
 	table = a->table;
 
-	if (!dma_map_sg(attachment->dev, table->sgl, table->nents,
-			direction))
+	table->nents = dma_map_sg(attachment->dev, table->sgl,
+				  table->orig_nents, direction);
+	if (!table->nents)
 		return ERR_PTR(-ENOMEM);
 
 	return table;
@@ -238,7 +239,7 @@ static void ion_unmap_dma_buf(struct dma_buf_attachment *attachment,
 			      struct sg_table *table,
 			      enum dma_data_direction direction)
 {
-	dma_unmap_sg(attachment->dev, table->sgl, table->nents, direction);
+	dma_unmap_sg(attachment->dev, table->sgl, table->orig_nents, direction);
 }
 
 static int ion_mmap(struct dma_buf *dmabuf, struct vm_area_struct *vma)
@@ -297,7 +298,7 @@ static int ion_dma_buf_begin_cpu_access(struct dma_buf *dmabuf,
 
 	mutex_lock(&buffer->lock);
 	list_for_each_entry(a, &buffer->attachments, list) {
-		dma_sync_sg_for_cpu(a->dev, a->table->sgl, a->table->nents,
+		dma_sync_sg_for_cpu(a->dev, a->table->sgl, a->table->orig_nents,
 				    direction);
 	}
 
@@ -320,8 +321,8 @@ static int ion_dma_buf_end_cpu_access(struct dma_buf *dmabuf,
 
 	mutex_lock(&buffer->lock);
 	list_for_each_entry(a, &buffer->attachments, list) {
-		dma_sync_sg_for_device(a->dev, a->table->sgl, a->table->nents,
-				       direction);
+		dma_sync_sg_for_device(a->dev, a->table->sgl,
+				       a->table->orig_nents, direction);
 	}
 	mutex_unlock(&buffer->lock);
 
diff --git a/drivers/staging/android/ion/ion_heap.c b/drivers/staging/android/ion/ion_heap.c
index 0755b11..f2f7ca7 100644
--- a/drivers/staging/android/ion/ion_heap.c
+++ b/drivers/staging/android/ion/ion_heap.c
@@ -38,7 +38,7 @@ void *ion_heap_map_kernel(struct ion_heap *heap,
 	else
 		pgprot = pgprot_writecombine(PAGE_KERNEL);
 
-	for_each_sg(table->sgl, sg, table->nents, i) {
+	for_each_sg(table->sgl, sg, table->orig_nents, i) {
 		int npages_this_entry = PAGE_ALIGN(sg->length) / PAGE_SIZE;
 		struct page *page = sg_page(sg);
 
@@ -71,7 +71,7 @@ int ion_heap_map_user(struct ion_heap *heap, struct ion_buffer *buffer,
 	int i;
 	int ret;
 
-	for_each_sg(table->sgl, sg, table->nents, i) {
+	for_each_sg(table->sgl, sg, table->orig_nents, i) {
 		struct page *page = sg_page(sg);
 		unsigned long remainder = vma->vm_end - addr;
 		unsigned long len = sg->length;
@@ -142,7 +142,7 @@ int ion_heap_buffer_zero(struct ion_buffer *buffer)
 	else
 		pgprot = pgprot_writecombine(PAGE_KERNEL);
 
-	return ion_heap_sglist_zero(table->sgl, table->nents, pgprot);
+	return ion_heap_sglist_zero(table->sgl, table->orig_nents, pgprot);
 }
 
 int ion_heap_pages_zero(struct page *page, size_t size, pgprot_t pgprot)
diff --git a/drivers/staging/android/ion/ion_system_heap.c b/drivers/staging/android/ion/ion_system_heap.c
index b83a1d1..34f6e81 100644
--- a/drivers/staging/android/ion/ion_system_heap.c
+++ b/drivers/staging/android/ion/ion_system_heap.c
@@ -162,7 +162,7 @@ static void ion_system_heap_free(struct ion_buffer *buffer)
 	if (!(buffer->private_flags & ION_PRIV_FLAG_SHRINKER_FREE))
 		ion_heap_buffer_zero(buffer);
 
-	for_each_sg(table->sgl, sg, table->nents, i)
+	for_each_sg(table->sgl, sg, table->orig_nents, i)
 		free_buffer_page(sys_heap, buffer, sg_page(sg));
 	sg_free_table(table);
 	kfree(table);
-- 
1.9.1