From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
To: Daniel Vetter <daniel@ffwll.ch>
Cc: intel-gfx@lists.freedesktop.org, dri-devel@lists.freedesktop.org
Subject: Re: [PATCH 01/13] drm/atomic: Only update crtc->x/y if it's part of the state.
Date: Thu, 16 Jul 2015 11:38:18 +0200 [thread overview]
Message-ID: <55A77B8A.1090106@linux.intel.com> (raw)
In-Reply-To: <20150716092936.GG6223@phenom.ffwll.local>
Op 16-07-15 om 11:29 schreef Daniel Vetter:
> On Thu, Jul 16, 2015 at 11:17:29AM +0200, Maarten Lankhorst wrote:
>> Op 16-07-15 om 11:19 schreef Daniel Vetter:
>>> On Thu, Jul 16, 2015 at 10:59:14AM +0200, Maarten Lankhorst wrote:
>>>> Cc: dri-devel@lists.freedesktop.org
>>>> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
>>>> ---
>>>> drivers/gpu/drm/drm_atomic_helper.c | 7 +++++--
>>>> 1 file changed, 5 insertions(+), 2 deletions(-)
>>>>
>>>> diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c
>>>> index 0898afbc9e23..70e69904291d 100644
>>>> --- a/drivers/gpu/drm/drm_atomic_helper.c
>>>> +++ b/drivers/gpu/drm/drm_atomic_helper.c
>>>> @@ -667,8 +667,11 @@ drm_atomic_helper_update_legacy_modeset_state(struct drm_device *dev,
>>>> for_each_crtc_in_state(old_state, crtc, old_crtc_state, i) {
>>>> crtc->mode = crtc->state->mode;
>>>> crtc->enabled = crtc->state->enable;
>>>> - crtc->x = crtc->primary->state->src_x >> 16;
>>>> - crtc->y = crtc->primary->state->src_y >> 16;
>>>> +
>>>> + if (drm_atomic_get_existing_plane_state(old_state, crtc->primary)) {
>>>> + crtc->x = crtc->primary->state->src_x >> 16;
>>>> + crtc->y = crtc->primary->state->src_y >> 16;
>>>> + }
>>> What's the benefit here of only updating when something changed? The
>>> atomic state should be the master source so copying a few too many times
>>> shouldn't matter really.
>> Because you might not be holding plane lock, so primary->state may be garbage.
> Anyone who wants to touch primary plane must grab the crtc lock, so crtc
> lock would give you an implicit read lock. At least that's been my
> thinking, but it could be that the primary plane is used on some other
> crtc, and then this is indeed garbage.
This is only true if the plane is active. If there is none you can still update properties and
swap the plane state without locking the crtc.
> So maybe we need even more checks than what you propose:
>
> if (drm_atomic_get_existing_plane_state(old_state, crtc->primary) &&
> crtc->primary->state->crtc == crtc) {
> crtc->x = crtc->primary->state->src_x >> 16;
> crtc->y = crtc->primary->state->src_y >> 16;
> }
>
> I think a comment explaining this would help (or at least in the commit
> message!).
But the primary and cursor planes are not allowed to move between crtc's?
~Maarten
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2015-07-16 9:38 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <1437037166-9339-1-git-send-email-maarten.lankhorst@linux.intel.com>
2015-07-16 8:59 ` [PATCH 01/13] drm/atomic: Only update crtc->x/y if it's part of the state Maarten Lankhorst
2015-07-16 9:19 ` Daniel Vetter
2015-07-16 9:17 ` Maarten Lankhorst
2015-07-16 9:29 ` Daniel Vetter
2015-07-16 9:38 ` Maarten Lankhorst [this message]
2015-07-16 12:34 ` Daniel Vetter
2015-07-16 13:44 ` Maarten Lankhorst
2015-07-16 13:51 ` [PATCH v1.1 01/13] drm/atomic: Only update crtc->x/y if it's part of the state, v2 Maarten Lankhorst
2015-07-16 14:58 ` Daniel Vetter
2015-07-16 8:59 ` [PATCH 02/13] drm/atomic: Update legacy DPMS state during modesets Maarten Lankhorst
2015-07-16 9:19 ` Daniel Vetter
2015-07-16 9:24 ` Maarten Lankhorst
2015-07-16 9:31 ` Daniel Vetter
2015-07-27 11:04 ` [PATCH v1.1 02/13] drm/atomic: Update legacy DPMS state during modesets, v2 Maarten Lankhorst
2015-07-27 11:16 ` Daniel Vetter
2015-07-27 11:15 ` Maarten Lankhorst
2015-07-27 11:24 ` [PATCH v1.2 02/13] drm/atomic: Update legacy DPMS state during modesets, v3 Maarten Lankhorst
2015-07-27 13:56 ` 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=55A77B8A.1090106@linux.intel.com \
--to=maarten.lankhorst@linux.intel.com \
--cc=daniel@ffwll.ch \
--cc=dri-devel@lists.freedesktop.org \
--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