intel-gfx.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
From: Daniel Vetter <daniel@ffwll.ch>
To: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Andy Whitcroft <apw@canonical.com>,
	Daniel Vetter <daniel.vetter@ffwll.ch>,
	intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 8/8] drm/i915: Fix tiling corruption from pipelined fencing
Date: Sat, 19 Mar 2011 23:17:56 +0100	[thread overview]
Message-ID: <20110319221756.GA16343@viiv.ffwll.ch> (raw)
In-Reply-To: <1300435330-15934-9-git-send-email-chris@chris-wilson.co.uk>

On Fri, Mar 18, 2011 at 08:02:10AM +0000, Chris Wilson wrote:
> ... even though it was disabled. A mistake in the handling of fence reuse
> caused us to skip the vital delay of waiting for the object to finish
> rendering before changing the register.

Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Nice catch and good simplification of the code-flow. One nitpick about a
possible further cleanup below.

> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index d4bf061..c5dfb59 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -2581,8 +2581,23 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj,
>  		reg = &dev_priv->fence_regs[obj->fence_reg];
>  		list_move_tail(&reg->lru_list, &dev_priv->mm.fence_list);
>  
> -		if (!obj->fenced_gpu_access && !obj->last_fenced_seqno)
> -			pipelined = NULL;
> +		if (obj->tiling_changed) {
> +			ret = i915_gem_object_flush_fence(obj, pipelined);
> +			if (ret)
> +				return ret;
> +
> +			if (!obj->fenced_gpu_access && !obj->last_fenced_seqno)
> +				pipelined = NULL;
> +
> +			if (pipelined) {
> +				reg->setup_seqno =
> +					i915_gem_next_request_seqno(pipelined);
> +				obj->last_fenced_seqno = reg->setup_seqno;
> +				obj->last_fenced_ring = pipelined;
> +			}
> +
> +			goto update;

I think we could move the update label 3 lines up, which would make the
above if(pipelined) clause unnecessary. Maybe even drop the goto and
extract the tail of get_fence as a function of its own.

> +		}
>  
>  		if (!pipelined) {
>  			if (reg->setup_seqno) {
> @@ -2601,31 +2616,6 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj,
>  			ret = i915_gem_object_flush_fence(obj, pipelined);
>  			if (ret)
>  				return ret;
> -		} else if (obj->tiling_changed) {
> -			if (obj->fenced_gpu_access) {
> -				if (obj->base.write_domain & I915_GEM_GPU_DOMAINS) {
> -					ret = i915_gem_flush_ring(obj->ring,
> -								  0, obj->base.write_domain);
> -					if (ret)
> -						return ret;
> -				}
> -
> -				obj->fenced_gpu_access = false;
> -			}
> -		}
> -
> -		if (!obj->fenced_gpu_access && !obj->last_fenced_seqno)
> -			pipelined = NULL;
> -		BUG_ON(!pipelined && reg->setup_seqno);
> -
> -		if (obj->tiling_changed) {
> -			if (pipelined) {
> -				reg->setup_seqno =
> -					i915_gem_next_request_seqno(pipelined);
> -				obj->last_fenced_seqno = reg->setup_seqno;
> -				obj->last_fenced_ring = pipelined;
> -			}
> -			goto update;
>  		}
>  
>  		return 0;
> -- 
> 1.7.2.3
> 

-- 
Daniel Vetter
Mail: daniel@ffwll.ch
Mobile: +41 (0)79 365 57 48

  reply	other threads:[~2011-03-19 22:18 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-03-18  8:02 Patch queue for -fixes Chris Wilson
2011-03-18  8:02 ` [PATCH 1/8] drm/i915: Remove surplus POSTING_READs before wait_for_vblank Chris Wilson
2011-03-18 14:42   ` Keith Packard
2011-03-18 16:23   ` Jesse Barnes
2011-03-18  8:02 ` [PATCH 2/8] drm/i915: skip redundant operations whilst enabling pipes and planes Chris Wilson
2011-03-18 14:46   ` Keith Packard
2011-03-18 16:25   ` Jesse Barnes
2011-03-18 17:37     ` Chris Wilson
2011-03-18 17:41       ` Jesse Barnes
2011-03-18  8:02 ` [PATCH 3/8] drm/i915: Flush the plane enable using the DSPCNTR latch Chris Wilson
2011-03-18 15:09   ` Keith Packard
2011-03-18 16:26   ` Jesse Barnes
2011-03-18 17:42     ` Chris Wilson
2011-03-18  8:02 ` [PATCH 4/8] drm/i915: Only wait for vblank after pipe enabling on gen2 Chris Wilson
2011-03-18 15:10   ` Keith Packard
2011-03-18 16:27   ` Jesse Barnes
2011-03-18  8:02 ` [PATCH 5/8] drm/i915: Prevent racy removal of request from client list Chris Wilson
2011-03-18  8:02 ` [PATCH 6/8] drm: Fix use-after-free in drm_gem_vm_close() Chris Wilson
2011-03-18 15:15   ` [Intel-gfx] " Keith Packard
2011-03-18  8:02 ` [PATCH 7/8] drm/i915: Re-enable self-refresh Chris Wilson
2011-03-18 15:22   ` Keith Packard
2011-03-18 17:48     ` Chris Wilson
2011-03-18 18:12     ` Eric Anholt
2011-03-18 19:08       ` Chris Wilson
2011-03-18  8:02 ` [PATCH 8/8] drm/i915: Fix tiling corruption from pipelined fencing Chris Wilson
2011-03-19 22:17   ` Daniel Vetter [this message]
2011-03-19 22:25     ` Chris Wilson
2011-03-19 23:12       ` Daniel Vetter

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=20110319221756.GA16343@viiv.ffwll.ch \
    --to=daniel@ffwll.ch \
    --cc=apw@canonical.com \
    --cc=chris@chris-wilson.co.uk \
    --cc=daniel.vetter@ffwll.ch \
    --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).