Seems extending of any PIN_XXX flags to 64b has no sense for all other cases.
But also in the code there is mismatching of 64 and 32b flags/data usage (32b flags may be put to 64b arguments).
Seems that shouldn't cause any issues anymore, but if you think that it should be fixed - please, let me know - would like to complete that case.
On Fri, Nov 2, 2018 at 5:47 PM Sergii Romantsov <sergii.romantsov@xxxxxxxxx> wrote:
Operating with gtt-addresses has to be done with 64b variables.
CC: Kenneth Graunke <kenneth@xxxxxxxxxxxxx>
CC: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
Fixes: a363bb2cd0e2 (i965: Allocate VMA in userspace for full-PPGTT systems.)
Fixes: c7c6e46f913b (drm/i915: Convert execbuf to use struct-of-array packing for critical fields)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108619
Signed-off-by: Sergii Romantsov <sergii.romantsov@xxxxxxxxxxxxxxx>
---
drivers/gpu/drm/i915/i915_gem_execbuffer.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 3f0c612..1973b79 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -611,7 +611,7 @@ static int eb_reserve_vma(const struct i915_execbuffer *eb,
if (exec_flags & EXEC_OBJECT_PINNED) {
pin_flags |= entry->offset | PIN_OFFSET_FIXED;
- pin_flags &= ~PIN_NONBLOCK; /* force overlapping checks */
+ pin_flags &= ~(u64)PIN_NONBLOCK; /* force overlapping checks */
} else if (exec_flags & __EXEC_OBJECT_NEEDS_BIAS) {
pin_flags |= BATCH_OFFSET_BIAS | PIN_OFFSET_BIAS;
}
--
2.7.4
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Sergii Romantsov
_______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
