summary | shortlog | log | commit | commitdiff | tree
raw | patch | inline | side by side (parent: f2032d4)
raw | patch | inline | side by side (parent: f2032d4)
author | Dave Airlie <airlied@redhat.com> | |
Wed, 19 Sep 2012 01:12:41 +0000 (11:12 +1000) | ||
committer | Dave Airlie <airlied@gmail.com> | |
Wed, 19 Sep 2012 09:59:26 +0000 (19:59 +1000) |
We hit this a lot with i915 and although we'd like to engineer things to hit
it a lot less, this commit at least makes it consume a few less cycles.
from something containing
movzwl 0x0(%rip),%r10d
to
add %r8,%rdx
I only noticed it while using perf to profile something else.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Dave Airlie <airlied@redhat.com>
it a lot less, this commit at least makes it consume a few less cycles.
from something containing
movzwl 0x0(%rip),%r10d
to
add %r8,%rdx
I only noticed it while using perf to profile something else.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Dave Airlie <airlied@redhat.com>
drivers/gpu/drm/drm_cache.c | patch | blob | history |
index 08758e061478eb77e2db77794d1d5e54c8625474..3dbc7f17eb1146a00cba823151dc2ea8f83ac2a1 100644 (file)
{
uint8_t *page_virtual;
unsigned int i;
+ const int size = boot_cpu_data.x86_clflush_size;
if (unlikely(page == NULL))
return;
page_virtual = kmap_atomic(page);
- for (i = 0; i < PAGE_SIZE; i += boot_cpu_data.x86_clflush_size)
+ for (i = 0; i < PAGE_SIZE; i += size)
clflush(page_virtual + i);
kunmap_atomic(page_virtual);
}