On Tue, Oct 18, 2016 at 10:33:11AM +0100, Tvrtko Ursulin wrote:
>
> On 17/10/2016 09:00, Chris Wilson wrote:
> >The scattergather list uses a 32bit size counter, we should avoid
> >exceeding it.
> >
> >v2: Also we should use unsigned int to match sg->length.
> >
> >Fixes: 871dfbd67d4e ("drm/i915: Allow compaction upto SWIOTLB max segment size")
> >Signed-off-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
> >Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxx> # v1
> >---
> > drivers/gpu/drm/i915/i915_gem.c | 6 +++---
> > 1 file changed, 3 insertions(+), 3 deletions(-)
> >
> >diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> >index 181bda2db587..a74ec10a5370 100644
> >--- a/drivers/gpu/drm/i915/i915_gem.c
> >+++ b/drivers/gpu/drm/i915/i915_gem.c
> >@@ -2205,7 +2205,7 @@ i915_gem_object_put_pages(struct drm_i915_gem_object *obj)
> > return 0;
> > }
> >-static unsigned long swiotlb_max_size(void)
> >+static unsigned int swiotlb_max_size(void)
> > {
> > #if IS_ENABLED(CONFIG_SWIOTLB)
> > return rounddown(swiotlb_nr_tbl() << IO_TLB_SHIFT, PAGE_SIZE);
> >@@ -2225,7 +2225,7 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
> > struct sgt_iter sgt_iter;
> > struct page *page;
> > unsigned long last_pfn = 0; /* suppress gcc warning */
> >- unsigned long max_segment;
> >+ unsigned int max_segment;
> > int ret;
> > gfp_t gfp;
> >@@ -2238,7 +2238,7 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
> > max_segment = swiotlb_max_size();
> > if (!max_segment)
> >- max_segment = obj->base.size;
> >+ max_segment = rounddown(UINT_MAX, PAGE_SIZE);
>
> rounddown looks like an overkill vs just UINT_MAX, no?
I think there is a subtle difference becuase we are couting in pages and
compare sg->length >= max_segment. If we left it at UINT_MAX, sg->length
would wrap to zero rather than start a new sg.
-Chris
--
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
