drm/i915: Move object backing storage manipulation to its own locking

Browse source

Break the allocation of the backing storage away from struct_mutex into
a per-object lock. This allows parallel page allocation, provided we can
do so outside of struct_mutex (i.e. set-domain-ioctl, pwrite, GTT
fault), i.e. before execbuf! The increased cost of the atomic counters
are hidden behind i915_vma_pin() for the typical case of execbuf, i.e.
as the object is typically bound between execbufs, the page_pin_count is
static. The cost will be felt around set-domain and pwrite, but offset
by the improvement from reduced struct_mutex contention.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161028125858.23563-14-chris@chris-wilson.co.uk

...

This commit is contained in:

Chris Wilson 2016-10-28 13:58:37 +01:00

parent 03ac84f183

commit 1233e2db19

5 changed files with 114 additions and 78 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download patch file Download diff file Expand all files Collapse all files

36

drivers/gpu/drm/i915/i915_drv.h

View file

93

drivers/gpu/drm/i915/i915_gem.c

View file

51

drivers/gpu/drm/i915/i915_gem_shrinker.c

View file

2

drivers/gpu/drm/i915/i915_gem_tiling.c

View file

10

drivers/gpu/drm/i915/i915_gem_userptr.c

View file