public inbox for intel-gfx@lists.freedesktop.org
 help / color / mirror / Atom feed
From: Ankitprasad Sharma <ankitprasad.r.sharma@intel.com>
To: Chris Wilson <chris@chris-wilson.co.uk>
Cc: intel-gfx@lists.freedesktop.org, akash.goel@intel.com
Subject: Re: [PATCH 1/3] drm/i915: Support for pre-populating the object with system pages
Date: Thu, 27 Aug 2015 11:18:15 +0530	[thread overview]
Message-ID: <1440654495.26385.5.camel@ankitprasad-desktop> (raw)
In-Reply-To: <20150824123554.GD25712@nuc-i3427.alporthouse.com>

On Mon, 2015-08-24 at 13:35 +0100, Chris Wilson wrote:
> On Mon, Aug 24, 2015 at 05:28:14PM +0530, ankitprasad.r.sharma@intel.com wrote:
> > +static int
> > +__i915_gem_object_get_pages(struct drm_i915_gem_object *obj)
> > +{
> > +	const struct drm_i915_gem_object_ops *ops = obj->ops;
> > +	int ret;
> > +
> > +	WARN_ON(obj->pages);
> > +
> > +	if (obj->madv != I915_MADV_WILLNEED) {
> > +		DRM_DEBUG("Attempting to obtain a purgeable object\n");
> > +		return -EFAULT;
> > +	}
> > +
> > +	BUG_ON(obj->pages_pin_count);
> 
> Put the parameter checking into i915_gem_object_get_pages(). The __i915
> version is only allowed from strict contexts and we can place the burden
> of being correct on the caller.
> 
> > +	ret = ops->get_pages(obj);
> > +	if (ret)
> > +		return ret;
> > +
> > +	obj->get_page.sg = obj->pages->sgl;
> > +	obj->get_page.last = 0;
> > +
> > +	return 0;
> > +}
> > +
> >  /* Ensure that the associated pages are gathered from the backing storage
> >   * and pinned into our object. i915_gem_object_get_pages() may be called
> >   * multiple times before they are released by a single call to
> > @@ -2339,28 +2377,17 @@ int
> >  i915_gem_object_get_pages(struct drm_i915_gem_object *obj)
> >  {
> >  	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
> > -	const struct drm_i915_gem_object_ops *ops = obj->ops;
> >  	int ret;
> >  
> >  	if (obj->pages)
> >  		return 0;
> >  
> > -	if (obj->madv != I915_MADV_WILLNEED) {
> > -		DRM_DEBUG("Attempting to obtain a purgeable object\n");
> > -		return -EFAULT;
> > -	}
> > -
> > -	BUG_ON(obj->pages_pin_count);
> > -
> > -	ret = ops->get_pages(obj);
> > +	ret = __i915_gem_object_get_pages(obj);
> >  	if (ret)
> >  		return ret;
> >  
> >  	list_add_tail(&obj->global_list, &dev_priv->mm.unbound_list);
> 
> I am tempted to say this should be in a new
> 
> __i915_gem_object_get_pages__tail_locked()
> 
> so that we don't have to hunt down users if we ever need to modify the
> global lists.
Could not get you here. 
is it just to add list_add_tail in a separate function
__i915_gem_object_get_pages__tail_locked(), or a new variant of
__i915_gem_object_get_pages() that will also do the link list insertion

Thanks,
Ankit

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

  reply	other threads:[~2015-08-27  6:06 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-08-24 11:58 [PATCH 0/3] Reduce the time for which 'struct_mutex' is held ankitprasad.r.sharma
2015-08-24 11:58 ` [PATCH 1/3] drm/i915: Support for pre-populating the object with system pages ankitprasad.r.sharma
2015-08-24 12:35   ` Chris Wilson
2015-08-27  5:48     ` Ankitprasad Sharma [this message]
2015-08-27  7:45       ` Chris Wilson
2015-08-25 10:51   ` Siluvery, Arun
2015-08-27  6:34     ` Ankitprasad Sharma
2015-08-27  7:43       ` Chris Wilson
2015-08-24 11:58 ` [PATCH 2/3] drm/i915: Support for the clflush of pre-populated pages ankitprasad.r.sharma
2015-08-24 12:43   ` Chris Wilson
2015-08-24 11:58 ` [PATCH 3/3] drm/i915: Only move to the CPU write domain if keeping the GTT pages ankitprasad.r.sharma
2015-08-24 12:46   ` Chris Wilson
2015-08-26  9:25 ` [PATCH 0/3] Reduce the time for which 'struct_mutex' is held Daniel Vetter
2015-08-27 10:48   ` Ankitprasad Sharma
  -- strict thread matches above, loose matches on Subject: below --
2015-08-27 11:04 [PATCH v2 " ankitprasad.r.sharma
2015-08-27 11:04 ` [PATCH 1/3] drm/i915: Support for pre-populating the object with system pages ankitprasad.r.sharma

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=1440654495.26385.5.camel@ankitprasad-desktop \
    --to=ankitprasad.r.sharma@intel.com \
    --cc=akash.goel@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