Re: [PATCH 1/2] drm/i915/dmabuf: fix sg_table handling in map_dma_buf

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 28/10/2022 14:55, Ruhl, Michael J wrote:
-----Original Message-----
From: Intel-gfx <intel-gfx-bounces@xxxxxxxxxxxxxxxxxxxxx> On Behalf Of
Matthew Auld
Sent: Thursday, October 27, 2022 11:27 AM
To: intel-gfx@xxxxxxxxxxxxxxxxxxxxx
Subject:  [PATCH 1/2] drm/i915/dmabuf: fix sg_table handling in
map_dma_buf

We need to iterate over the original entries here for the sg_table,
pulling out the struct page for each one, to be remapped. However
currently this incorrectly iterates over the final dma mapped entries,
which is likely just one gigantic sg entry if the iommu is enabled,
leading to us only mapping the first struct page (and any physically
contiguous pages following it), even if there is potentially lots more
data to follow.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/7306
Signed-off-by: Matthew Auld <matthew.auld@xxxxxxxxx>
Cc: Lionel Landwerlin <lionel.g.landwerlin@xxxxxxxxx>
Cc: Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxxxxxxxx>
Cc: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx>
---
drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
b/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
index 07eee1c09aaf..05ebbdfd3b3b 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
@@ -40,13 +40,13 @@ static struct sg_table *i915_gem_map_dma_buf(struct
dma_buf_attachment *attachme
		goto err;
	}

-	ret = sg_alloc_table(st, obj->mm.pages->nents, GFP_KERNEL);
+	ret = sg_alloc_table(st, obj->mm.pages->orig_nents, GFP_KERNEL);
	if (ret)
		goto err_free;

	src = obj->mm.pages->sgl;
	dst = st->sgl;
-	for (i = 0; i < obj->mm.pages->nents; i++) {
+	for (i = 0; i < obj->mm.pages->orig_nents; i++) {

This really should use the for_each_sg() macro.

I proposed a clean up patch a while back that looked like this:

diff --git a/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
index e2cdc2640c08..ccc5d46aa749 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c
@@ -7,6 +7,7 @@
  #include <linux/dma-buf.h>
  #include <linux/highmem.h>
  #include <linux/dma-resv.h>
+#include <linux/scatterlist.h>
#include "gem/i915_gem_dmabuf.h"
  #include "i915_drv.h"
@@ -41,12 +42,10 @@ static struct sg_table *i915_gem_map_dma_buf(struct dma_buf_attachment *attach,
         if (ret)
                 goto err_free;
- src = obj->mm.pages->sgl;
         dst = sgt->sgl;
-       for (i = 0; i < obj->mm.pages->nents; i++) {
+       for_each_sg(obj->mm.pages->sgl, src, obj->mm.pages->nents, i) {
                 sg_set_page(dst, sg_page(src), src->length, 0);
                 dst = sg_next(dst);
-               src = sg_next(src);
         }

If you are updating the for loop, this might be a reasonable update as well.

Ok, but such cleanups should normally be a separate patch. I'll grab your series and bolt that onto this one.


Mike

		sg_set_page(dst, sg_page(src), src->length, 0);
		dst = sg_next(dst);
		src = sg_next(src);
--
2.37.3




[Index of Archives]     [AMD Graphics]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux