dri-devel.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
From: Haneen Mohammed <hamohammed.sa@gmail.com>
To: Chris Wilson <chris@chris-wilson.co.uk>
Cc: rodrigosiqueiramelo@gmail.com, dri-devel@lists.freedesktop.org
Subject: Re: [PATCH 1/4] drm/vkms: Add functions to map GEM backing storage
Date: Tue, 10 Jul 2018 18:12:11 +0300	[thread overview]
Message-ID: <20180710151211.GA4114@haneen-vb> (raw)
In-Reply-To: <153121035583.18500.12528342272398519561@skylake-alporthouse-com>

On Tue, Jul 10, 2018 at 09:12:36AM +0100, Chris Wilson wrote:
> Quoting Haneen Mohammed (2018-07-09 16:44:26)
> > +struct page **get_pages(struct vkms_gem_object *vkms_obj)
> > +{
> > +       struct drm_gem_object *gem_obj = &vkms_obj->gem;
> > +       struct page **pages = vkms_obj->pages;
> > +
> > +       if (!pages) {
> > +               mutex_lock(&vkms_obj->pages_lock);
> > +               pages = drm_gem_get_pages(gem_obj);
> > +               if (IS_ERR(pages)) {
> > +                       mutex_unlock(&vkms_obj->pages_lock);
> > +                       return pages;
> > +               }
> > +
> > +               vkms_obj->pages = pages;
> > +               mutex_unlock(&vkms_obj->pages_lock);
> 
> You have a race here with two callers setting pages. Trivially you fix
> it by checking if (!pages) again inside the lock, but the lock is
> superfluous in this case:
> 	if (!vkms_obj->pages) {
> 		srtuct pages **pages;
> 
> 		pages = drm_gem_get_pages(gem_obj);
> 		if (IS_ERR(pages))
> 			return pages;
> 		
> 		if (cmpxchg(&vkms_obj->pages, NULL, pages))
> 			put_pages(pages);
> 
> 		/* barrier() is implied */
> 	}
> 
> 	return vkms_obj->pages;
> -Chris

Thank you for the feedback!

- Haneen
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

  reply	other threads:[~2018-07-10 15:12 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-07-09 15:42 [PATCH 0/4] Add infrastructure needed for CRC support Haneen Mohammed
2018-07-09 15:44 ` [PATCH 1/4] drm/vkms: Add functions to map GEM backing storage Haneen Mohammed
2018-07-10  8:00   ` Daniel Vetter
2018-07-10  8:12   ` Chris Wilson
2018-07-10 15:12     ` Haneen Mohammed [this message]
2018-07-09 15:46 ` [PATCH 2/4] drm/vkms: map/unmap buffers in [prepare/cleanup]_fb hooks Haneen Mohammed
2018-07-10  8:04   ` Daniel Vetter
2018-07-09 15:47 ` [PATCH 3/4] drm/vkms: Add atomic_helper_check_plane_state Haneen Mohammed
2018-07-10  8:07   ` Daniel Vetter
2018-07-09 15:48 ` [PATCH 4/4] drm/vkms: subclass CRTC state Haneen Mohammed
2018-07-10  8:10   ` Daniel Vetter
2018-07-10 15:13     ` Haneen Mohammed

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=20180710151211.GA4114@haneen-vb \
    --to=hamohammed.sa@gmail.com \
    --cc=chris@chris-wilson.co.uk \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=rodrigosiqueiramelo@gmail.com \
    /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).