From: YoungJun Cho <yj44.cho@samsung.com>
To: Daniel Kurtz <djkurtz@chromium.org>
Cc: David Airlie <airlied@linux.ie>,
dri-devel <dri-devel@lists.freedesktop.org>,
devicetree@vger.kernel.org, Rob Herring <robh+dt@kernel.org>,
Pawel Moll <pawel.moll@arm.com>,
Mark Rutland <mark.rutland@arm.com>,
Ian Campbell <ijc+devicetree@hellion.org.uk>,
Kumar Gala <galak@codeaurora.org>,
linux-samsung-soc <linux-samsung-soc@vger.kernel.org>,
Thierry Reding <thierry.reding@gmail.com>,
Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
s.trumtrar@pengutronix.de,
Kyungmin Park <kyungmin.park@samsung.com>,
InKi Dae <inki.dae@samsung.com>,
Kukjin Kim <kgene.kim@samsung.com>,
Joonyoung Shim <jy0922.shim@samsung.com>,
Seung-Woo Kim <sw0312.kim@samsung.com>,
Andrzej Hajda <a.hajda@samsung.com>
Subject: Re: [PATCH v2 02/18] drm/exynos: use wait_event_timeout() for safety usage
Date: Wed, 21 May 2014 19:37:36 +0900 [thread overview]
Message-ID: <537C81F0.3060108@samsung.com> (raw)
In-Reply-To: <CAGS+omB_xgErG9u_0XB=UpJCrMhhkOu-Sjv_zupOn1V4i-V_-w@mail.gmail.com>
Hi Daniel,
On 05/21/2014 03:45 PM, Daniel Kurtz wrote:
> On Wed, May 21, 2014 at 2:28 PM, YoungJun Cho <yj44.cho@samsung.com> wrote:
>> Hi Daniel
>>
>>
>> On 05/21/2014 03:01 PM, Daniel Kurtz wrote:
>>>
>>> On Wed, May 21, 2014 at 12:42 PM, YoungJun Cho <yj44.cho@samsung.com>
>>> wrote:
>>>>
>>>> There could be the case that the page flip operation isn't finished
>>>> correctly
>>>> with some abnormal condition such as panel reset. So this patch replaces
>>>> wait_event() with wait_event_timeout() to avoid waiting for page flip
>>>> completion
>>>> infinitely.
>>>> And clears exynos_crtc->pending_flip in exynos_drm_crtc_page_flip() when
>>>> exynos_drm_crtc_mode_set_commit() is failed.
>>>>
>>>> Signed-off-by: YoungJun Cho <yj44.cho@samsung.com>
>>>> Acked-by: Inki Dae <inki.dae@samsung.com>
>>>> Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
>>>> ---
>>>> drivers/gpu/drm/exynos/exynos_drm_crtc.c | 7 +++++--
>>>> 1 file changed, 5 insertions(+), 2 deletions(-)
>>>>
>>>> diff --git a/drivers/gpu/drm/exynos/exynos_drm_crtc.c
>>>> b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
>>>> index 95c9435..3bf091d 100644
>>>> --- a/drivers/gpu/drm/exynos/exynos_drm_crtc.c
>>>> +++ b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
>>>> @@ -69,8 +69,10 @@ static void exynos_drm_crtc_dpms(struct drm_crtc
>>>> *crtc, int mode)
>>>>
>>>> if (mode > DRM_MODE_DPMS_ON) {
>>>> /* wait for the completion of page flip. */
>>>> - wait_event(exynos_crtc->pending_flip_queue,
>>>> - atomic_read(&exynos_crtc->pending_flip)
>>>> == 0);
>>>> + if (!wait_event_timeout(exynos_crtc->pending_flip_queue,
>>>> + !atomic_read(&exynos_crtc->pending_flip),
>>>> + HZ/20))
>>>> + atomic_set(&exynos_crtc->pending_flip, 0);
>>>
>>>
>>> I meant that changing this to wait_event_timeout() seems to be masking
>>> the original problem, that pending_flip wasn't being cleared.
>>
>>
>> Yes, I agree.
>>
>> The original purpose of this patch is to avoid lock-up during modetest (with
>> test_vsync) with command mode panel.
>> In MIPI DSI command mode interface, the display controller can not generate
>> VSYNC signal and uses TE signal instead which is generated by panel.
>> If there is an abnormal power off or reset condition, it is possible that
At first, I omitted the keyword "panel" at previous reply.
The abnormal power off or reset condition could be happened in panel
which generates TE signal.
>> the display controller misses the TE signal and it makes lock-up.
>> So I needed this patch.
>
> If flips are not completing on MIPI DSI, may I recommend that you
> start a timer when scheduling such a flip in the MIPI DSI driver that
> will expire after a timeout. In that timer's expiration handler, you
The MIPI DSI doesn't take care page flip operation at all.
And the exynos_drm_crtc_finish_page_flip() is called only by interrupt
handler of display controllers.
So there is no suitable place to set time-out for this.
> can then go through the normal process of completing a flip. This
> would make the timeout transparent to the generic exynos_drm_crtc
> layer, which can then do its normal flip complete processing: send
> pending vblank events, put vblank, clear pending_flip, etc.
> Does that make sense?
> Is it possible?
This function is exynos_drm_crtc_dpms() and my patch is for DPMS OFF case.
That means all page flip operations should be finished already to
control DPMS off and in almost every case, there is no problem.
But if there is an issue like this lock-up, I think it should escape
this condition.
Thank you.
Best regards YJ
>
> -Dan
>
>>> Now that you are now clearing pending_flip in the error path, you
>>> don't need the timeout, right?
>>>
>>
>> There might be my missing point. Would you explain more detail?
>>
>>
>> Thank you.
>> Best regards YJ
>>
>>> -Dan
>>>
>>>> drm_vblank_off(crtc->dev, exynos_crtc->pipe);
>>>> }
>>>>
>>>> @@ -259,6 +261,7 @@ static int exynos_drm_crtc_page_flip(struct drm_crtc
>>>> *crtc,
>>>> spin_lock_irq(&dev->event_lock);
>>>> drm_vblank_put(dev, exynos_crtc->pipe);
>>>> list_del(&event->base.link);
>>>> + atomic_set(&exynos_crtc->pending_flip, 0);
>>>> spin_unlock_irq(&dev->event_lock);
>>>>
>>>> goto out;
>>>> --
>>>> 1.7.9.5
>>>>
>>>
>>
>
next prev parent reply other threads:[~2014-05-21 10:37 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-05-21 4:42 [PATCH v2 00/18] drm/exynos: support MIPI DSI command mode display YoungJun Cho
2014-05-21 4:42 ` [PATCH v2 01/18] drm/exynos: dsi: move the EoT packets configuration point YoungJun Cho
2014-05-21 4:42 ` [PATCH v2 02/18] drm/exynos: use wait_event_timeout() for safety usage YoungJun Cho
2014-05-21 6:01 ` Daniel Kurtz
2014-05-21 6:28 ` YoungJun Cho
2014-05-21 6:45 ` Daniel Kurtz
2014-05-21 10:37 ` YoungJun Cho [this message]
2014-05-21 4:42 ` [PATCH v2 04/18] video: add command mode and command mode display timing YoungJun Cho
2014-05-21 11:02 ` Thierry Reding
2014-05-21 11:41 ` YoungJun Cho
2014-05-22 13:07 ` Inki Dae
2014-05-22 13:53 ` Thierry Reding
2014-05-26 10:14 ` Thierry Reding
2014-05-26 11:10 ` Heiko Stübner
2014-05-21 4:42 ` [PATCH v2 05/18] drm_modes: add command mode helpers YoungJun Cho
2014-05-21 4:42 ` [PATCH v2 06/18] ARM: dts: sysreg: add exynos5 compatible to DT bindings YoungJun Cho
2014-05-21 4:42 ` [PATCH v2 07/18] ARM: dts: samsung-fimd: add I80 specific properties YoungJun Cho
[not found] ` <1400647390-26590-1-git-send-email-yj44.cho-Sze3O3UU22JBDgjK7y7TUQ@public.gmane.org>
2014-05-21 4:42 ` [PATCH v2 03/18] ARM: dts: video: add command mode display timing DT bindings YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 08/18] drm/exynos: add TE handler to support MIPI DSI command mode interface YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 15/18] ARM: dts: exynos4: add system register node YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 16/18] ARM: dts: exynos5: add system register support YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 09/18] drm/exynos: dsi: add TE handler to support command mode interface YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 10/18] drm/exynos: fimd: support I80 interface YoungJun Cho
2014-05-26 9:00 ` Daniel Kurtz
2014-05-29 5:45 ` YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 11/18] ARM: dts: exynos_dsim: add exynos5420 compatible to DT bindings YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 12/18] drm/exynos: dsi: add driver data to support Exynos5420 YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 13/18] ARM: dts: s6e3fa0: add DT bindings YoungJun Cho
2014-05-26 13:41 ` Thierry Reding
2014-05-27 6:28 ` Andrzej Hajda
2014-05-27 7:53 ` Thierry Reding
2014-05-27 14:24 ` Inki Dae
2014-05-27 20:21 ` Thierry Reding
2014-05-28 4:50 ` Inki Dae
2014-05-28 6:44 ` Andrzej Hajda
2014-05-30 3:08 ` YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 14/18] drm/panel: add S6E3FA0 driver YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 17/18] ARM: dts: exynos5420: add mipi-phy node YoungJun Cho
2014-05-21 4:43 ` [PATCH v2 18/18] ARM: dts: exynos5420: add dsi node YoungJun Cho
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=537C81F0.3060108@samsung.com \
--to=yj44.cho@samsung.com \
--cc=a.hajda@samsung.com \
--cc=airlied@linux.ie \
--cc=devicetree@vger.kernel.org \
--cc=djkurtz@chromium.org \
--cc=dri-devel@lists.freedesktop.org \
--cc=galak@codeaurora.org \
--cc=ijc+devicetree@hellion.org.uk \
--cc=inki.dae@samsung.com \
--cc=jy0922.shim@samsung.com \
--cc=kgene.kim@samsung.com \
--cc=kyungmin.park@samsung.com \
--cc=laurent.pinchart@ideasonboard.com \
--cc=linux-samsung-soc@vger.kernel.org \
--cc=mark.rutland@arm.com \
--cc=pawel.moll@arm.com \
--cc=robh+dt@kernel.org \
--cc=s.trumtrar@pengutronix.de \
--cc=sw0312.kim@samsung.com \
--cc=thierry.reding@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.