From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
To: Chris Wilson <chris@chris-wilson.co.uk>, intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 12/24] drm/i915: Store a persistent reference for an object in the execbuffer cache
Date: Fri, 19 May 2017 12:37:00 +0300 [thread overview]
Message-ID: <1495186620.3819.14.camel@linux.intel.com> (raw)
In-Reply-To: <20170518094638.5469-12-chris@chris-wilson.co.uk>
On to, 2017-05-18 at 10:46 +0100, Chris Wilson wrote:
> If we take a reference to the object/vma when it is first used in an
> execbuf, we can keep that reference until the object's file-local handle
> is closed. Thereby saving a frequent ref/unref pair.
>
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
The benchmark this gives improvements on might make sense to add to the
commit message.
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Regards, Joonas
--
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2017-05-19 9:37 UTC|newest]
Thread overview: 47+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-05-18 9:46 [PATCH 01/24] drm/i915/selftests: Pretend to be a gfx pci device Chris Wilson
2017-05-18 9:46 ` [PATCH 02/24] drm/i915: Mark CPU cache as dirty on every transition for CPU writes Chris Wilson
2017-05-18 9:46 ` [PATCH 03/24] drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty Chris Wilson
2017-05-19 10:22 ` Chris Wilson
2017-05-22 20:39 ` Dongwon Kim
2017-05-18 9:46 ` [PATCH 04/24] drm/i915: Reinstate reservation_object zapping for batch_pool objects Chris Wilson
2017-05-18 9:46 ` [PATCH 05/24] drm/i915: Amalgamate execbuffer parameter structures Chris Wilson
2017-05-18 9:46 ` [PATCH 06/24] drm/i915: Use vma->exec_entry as our double-entry placeholder Chris Wilson
2017-05-18 9:46 ` [PATCH 07/24] drm/i915: Split vma exec_link/evict_link Chris Wilson
2017-05-18 9:46 ` [PATCH 08/24] drm/i915: Store a direct lookup from object handle to vma Chris Wilson
2017-05-18 9:46 ` [PATCH 09/24] drm/i915: Pass vma to relocate entry Chris Wilson
2017-05-18 9:46 ` [PATCH 10/24] drm/i915: Disable EXEC_OBJECT_ASYNC when doing relocations Chris Wilson
2017-05-19 9:05 ` Joonas Lahtinen
2017-05-18 9:46 ` [PATCH 11/24] drm/i915: Eliminate lots of iterations over the execobjects array Chris Wilson
2017-05-18 9:46 ` [PATCH 12/24] drm/i915: Store a persistent reference for an object in the execbuffer cache Chris Wilson
2017-05-19 9:37 ` Joonas Lahtinen [this message]
2017-05-18 9:46 ` [PATCH 13/24] drm/i915: First try the previous execbuffer location Chris Wilson
2017-05-18 9:46 ` [PATCH 14/24] drm/i915: Wait upon userptr get-user-pages within execbuffer Chris Wilson
2017-05-18 9:46 ` [PATCH 15/24] drm/i915: Allow execbuffer to use the first object as the batch Chris Wilson
2017-05-18 9:46 ` [PATCH 16/24] drm/i915: Async GPU relocation processing Chris Wilson
2017-05-18 9:46 ` [PATCH 17/24] drm/i915: Stash a pointer to the obj's resv in the vma Chris Wilson
2017-05-19 9:19 ` Joonas Lahtinen
2017-05-18 9:46 ` [PATCH 18/24] drm/i915: Convert execbuf to use struct-of-array packing for critical fields Chris Wilson
2017-05-19 9:25 ` Chris Wilson
2017-05-18 9:46 ` [PATCH 19/24] drm/i915/scheduler: Support user-defined priorities Chris Wilson
2017-08-02 21:55 ` Jason Ekstrand
2017-05-18 9:46 ` [PATCH 20/24] drm/i915: Group all the global context information together Chris Wilson
2017-05-19 9:35 ` Joonas Lahtinen
2017-05-18 9:46 ` [PATCH 21/24] drm/i915: Allow contexts to be unreferenced locklessly Chris Wilson
2017-05-18 9:46 ` [PATCH 22/24] drm/i915: Enable rcu-only context lookups Chris Wilson
2017-05-19 10:36 ` Joonas Lahtinen
2017-05-18 9:46 ` [PATCH 23/24] drm/i915: Keep a recent cache of freed contexts objects for reuse Chris Wilson
2017-05-18 13:52 ` Tvrtko Ursulin
2017-05-18 14:19 ` Chris Wilson
2017-05-19 9:09 ` Tvrtko Ursulin
2017-05-22 10:51 ` Tvrtko Ursulin
2017-05-22 11:06 ` Chris Wilson
2017-05-18 9:46 ` [PATCH 24/24] RFC drm/i915: Expose a PMU interface for perf queries Chris Wilson
2017-05-18 23:48 ` Dmitry Rogozhkin
2017-05-19 8:01 ` Chris Wilson
2017-05-19 14:54 ` Dmitry Rogozhkin
2017-06-15 11:17 ` Tvrtko Ursulin
2017-05-18 12:31 ` ✓ Fi.CI.BAT: success for series starting with [01/24] drm/i915/selftests: Pretend to be a gfx pci device Patchwork
2017-05-18 13:35 ` [PATCH 01/24] " Tvrtko Ursulin
2017-05-18 14:46 ` Chris Wilson
2017-05-19 9:02 ` Joonas Lahtinen
2017-05-19 9:10 ` Chris Wilson
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1495186620.3819.14.camel@linux.intel.com \
--to=joonas.lahtinen@linux.intel.com \
--cc=chris@chris-wilson.co.uk \
--cc=intel-gfx@lists.freedesktop.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.