intel-gfx.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
From: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
To: Chris Wilson <chris@chris-wilson.co.uk>, intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 13/21] drm/i915: Check the CPU cached value of seqno after waking the waiter
Date: Mon, 6 Jun 2016 16:10:41 +0100	[thread overview]
Message-ID: <57559271.1050600@linux.intel.com> (raw)
In-Reply-To: <1464970133-29859-14-git-send-email-chris@chris-wilson.co.uk>


On 03/06/16 17:08, Chris Wilson wrote:
> If we have multiple waiters, we may find that many complete on the same
> wake up. If we first inspect the seqno from the CPU cache, we may reduce
> the number of heavyweight coherent seqno reads we require.
>
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
>   drivers/gpu/drm/i915/i915_drv.h | 14 ++++++++++----
>   1 file changed, 10 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
> index 4a71f4e9a97a..4ddb9ff319cb 100644
> --- a/drivers/gpu/drm/i915/i915_drv.h
> +++ b/drivers/gpu/drm/i915/i915_drv.h
> @@ -3918,6 +3918,12 @@ static inline bool __i915_request_irq_complete(struct drm_i915_gem_request *req)
>   {
>   	struct intel_engine_cs *engine = req->engine;
>
> +	/* Before we do the heavier coherent read of the seqno,
> +	 * check the value (hopefully) in the CPU cacheline.
> +	 */
> +	if (i915_gem_request_completed(req))
> +		return true;
> +
>   	/* Ensure our read of the seqno is coherent so that we
>   	 * do not "miss an interrupt" (i.e. if this is the last
>   	 * request and the seqno write from the GPU is not visible
> @@ -3929,11 +3935,11 @@ static inline bool __i915_request_irq_complete(struct drm_i915_gem_request *req)
>   	 * but it is easier and safer to do it every time the waiter
>   	 * is woken.
>   	 */
> -	if (engine->irq_seqno_barrier)
> +	if (engine->irq_seqno_barrier) {
>   		engine->irq_seqno_barrier(engine);
> -
> -	if (i915_gem_request_completed(req))
> -		return true;
> +		if (i915_gem_request_completed(req))
> +			return true;
> +	}
>
>   	/* We need to check whether any gpu reset happened in between
>   	 * the request being submitted and now. If a reset has occurred,
>

Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>

Regards,

Tvrtko
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

  reply	other threads:[~2016-06-06 15:10 UTC|newest]

Thread overview: 60+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-06-03 16:08 Breadcrumbs, again Chris Wilson
2016-06-03 16:08 ` [PATCH 01/21] drm/i915/shrinker: Flush active on objects before counting Chris Wilson
2016-06-03 16:08 ` [PATCH 02/21] drm/i915: Delay queuing hangcheck to wait-request Chris Wilson
2016-06-08  8:42   ` Daniel Vetter
2016-06-08  9:13     ` Chris Wilson
2016-06-03 16:08 ` [PATCH 03/21] drm/i915: Remove the dedicated hangcheck workqueue Chris Wilson
2016-06-06 12:52   ` Tvrtko Ursulin
2016-06-03 16:08 ` [PATCH 04/21] drm/i915: Make queueing the hangcheck work inline Chris Wilson
2016-06-03 16:08 ` [PATCH 05/21] drm/i915: Separate GPU hang waitqueue from advance Chris Wilson
2016-06-06 13:00   ` Tvrtko Ursulin
2016-06-07 12:11     ` Arun Siluvery
2016-06-03 16:08 ` [PATCH 06/21] drm/i915: Slaughter the thundering i915_wait_request herd Chris Wilson
2016-06-06 13:58   ` Tvrtko Ursulin
2016-06-03 16:08 ` [PATCH 07/21] drm/i915: Spin after waking up for an interrupt Chris Wilson
2016-06-06 14:39   ` Tvrtko Ursulin
2016-06-03 16:08 ` [PATCH 08/21] drm/i915: Use HWS for seqno tracking everywhere Chris Wilson
2016-06-06 14:55   ` Tvrtko Ursulin
2016-06-08  9:24     ` Chris Wilson
2016-06-03 16:08 ` [PATCH 09/21] drm/i915: Stop mapping the scratch page into CPU space Chris Wilson
2016-06-06 15:03   ` Tvrtko Ursulin
2016-06-03 16:08 ` [PATCH 10/21] drm/i915: Allocate scratch page from stolen Chris Wilson
2016-06-06 15:05   ` Tvrtko Ursulin
2016-06-03 16:08 ` [PATCH 11/21] drm/i915: Refactor scratch object allocation for gen2 w/a buffer Chris Wilson
2016-06-06 15:09   ` Tvrtko Ursulin
2016-06-08  9:27     ` Chris Wilson
2016-06-03 16:08 ` [PATCH 12/21] drm/i915: Add a delay between interrupt and inspecting the final seqno (ilk) Chris Wilson
2016-06-03 16:08 ` [PATCH 13/21] drm/i915: Check the CPU cached value of seqno after waking the waiter Chris Wilson
2016-06-06 15:10   ` Tvrtko Ursulin [this message]
2016-06-03 16:08 ` [PATCH 14/21] drm/i915: Only apply one barrier after a breadcrumb interrupt is posted Chris Wilson
2016-06-06 15:34   ` Tvrtko Ursulin
2016-06-08  9:35     ` Chris Wilson
2016-06-08  9:57       ` Tvrtko Ursulin
2016-06-03 16:08 ` [PATCH 15/21] drm/i915: Stop setting wraparound seqno on initialisation Chris Wilson
2016-06-08  8:54   ` Daniel Vetter
2016-06-03 16:08 ` [PATCH 16/21] drm/i915: Only query timestamp when measuring elapsed time Chris Wilson
2016-06-06 13:50   ` Tvrtko Ursulin
2016-06-03 16:08 ` [PATCH 17/21] drm/i915: Convert trace-irq to the breadcrumb waiter Chris Wilson
2016-06-07 12:04   ` Tvrtko Ursulin
2016-06-08  9:48     ` Chris Wilson
2016-06-08 10:16       ` Tvrtko Ursulin
2016-06-08 11:24         ` Chris Wilson
2016-06-08 11:47           ` Tvrtko Ursulin
2016-06-08 12:34             ` Chris Wilson
2016-06-08 12:44               ` Tvrtko Ursulin
2016-06-08 13:47                 ` Chris Wilson
2016-06-03 16:08 ` [PATCH 18/21] drm/i915: Embed signaling node into the GEM request Chris Wilson
2016-06-07 12:31   ` Tvrtko Ursulin
2016-06-08  9:54     ` Chris Wilson
2016-06-03 16:08 ` [PATCH 19/21] drm/i915: Move the get/put irq locking into the caller Chris Wilson
2016-06-07 12:46   ` Tvrtko Ursulin
2016-06-08 10:01     ` Chris Wilson
2016-06-08 10:18       ` Tvrtko Ursulin
2016-06-08 11:10         ` Chris Wilson
2016-06-08 11:49           ` Tvrtko Ursulin
2016-06-08 12:54             ` Chris Wilson
2016-06-03 16:08 ` [PATCH 20/21] drm/i915: Simplify enabling user-interrupts with L3-remapping Chris Wilson
2016-06-07 12:50   ` Tvrtko Ursulin
2016-06-03 16:08 ` [PATCH 21/21] drm/i915: Remove debug noise on detecting fault-injection of missed interrupts Chris Wilson
2016-06-07 12:51   ` Tvrtko Ursulin
2016-06-03 16:35 ` ✗ Ro.CI.BAT: failure for series starting with [01/21] drm/i915/shrinker: Flush active on objects before counting Patchwork

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=57559271.1050600@linux.intel.com \
    --to=tvrtko.ursulin@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).