[PATCH 00/12] Prepare various SH/R Mobile/Car drivers for CCF migration

[PATCH 00/12] Prepare various SH/R Mobile/Car drivers for CCF migration

[Laurent Pinchart]

Hello,

This patch series, based on v3.12-rc7, prepares various Renesas SH-Mobile,
R-Mobile and R-Car drivers for migration to CCF by adding clock prepare and
unprepare support.

The patches are pretty straightforward. Most of the drivers called
clk_enable and clk_disable in sleepable context, I've thus just converted them
to clk_prepare_enable and clk_disable_unprepare.

The clocksource drivers were slightly more complex to handle as clk_enable and
clk_disable were called in non-sleepable context. As the clocksource framework
architecture doesn't provide any sleepable callback in which clocks could be
prepared and unprepared, I've added clk_prepare and clk_unprepare calls in the
probe and suspend/resume handlers (clocksources can't be removed so the remove
handler doesn't need a clk_unprepare call).

I'd like to get all these patches merged in v3.14. As they will need to go
through their respective subsystems' trees, I would appreciate if all
maintainers involved could notify me when they merge patches from this series
in their tree to help me tracking the merge status. I don't plan to send pull
requests individually for these patches, and I will repost patches
individually if changes are requested during review.

Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Cc: linux-kernel@vger.kernel.org
Cc: Tejun Heo <tj@kernel.org>
Cc: linux-ide@vger.kernel.org
Cc: David Airlie <airlied@linux.ie>
Cc: dri-devel@lists.freedesktop.org
Cc: Wolfram Sang <wsa@the-dreams.de>
Cc: linux-i2c@vger.kernel.org
Cc: Chris Ball <cjb@laptop.org>
Cc: Guennadi Liakhovetski <g.liakhovetski+renesas@gmail.com>
Cc: linux-mmc@vger.kernel.org
Cc: Felipe Balbi <balbi@ti.com>
Cc: linux-usb@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com>
Cc: Tomi Valkeinen <tomi.valkeinen@ti.com>
Cc: linux-fbdev@vger.kernel.org

Laurent Pinchart (12):
  clocksource: sh_cmt: Add clk_prepare/unprepare support
  clocksource: sh_mtu2: Add clk_prepare/unprepare support
  clocksource: sh_tmu: Add clk_prepare/unprepare support
  sata_rcar: Convert to clk_prepare/unprepare
  drm: shmob_drm: Convert to clk_prepare/unprepare
  i2c: sh_mobile: Convert to clk_prepare/unprepare
  mmc: sh_mmcif: Convert to clk_prepare/unprepare
  mmc: sh_mobile_sdhi: Convert to clk_prepare/unprepare
  usb: gadget: r8a66597-udc: Convert to clk_prepare/unprepare
  usb: r8a66597-hcd: Convert to clk_prepare/unprepare
  fbdev: shmobile-hdmi: Convert to clk_prepare/unprepare
  fbdev: shmobile-lcdcfb: Convert to clk_prepare/unprepare

 drivers/ata/sata_rcar.c                   | 10 +++++-----
 drivers/clocksource/sh_cmt.c              | 20 ++++++++++++++++----
 drivers/clocksource/sh_mtu2.c             | 16 ++++++++++++++--
 drivers/clocksource/sh_tmu.c              | 20 +++++++++++++++++---
 drivers/gpu/drm/shmobile/shmob_drm_crtc.c |  4 ++--
 drivers/i2c/busses/i2c-sh_mobile.c        |  8 ++++----
 drivers/mmc/host/sh_mmcif.c               | 12 ++++++------
 drivers/mmc/host/sh_mobile_sdhi.c         |  4 ++--
 drivers/usb/gadget/r8a66597-udc.c         |  6 +++---
 drivers/usb/host/r8a66597-hcd.c           |  4 ++--
 drivers/video/sh_mobile_hdmi.c            |  6 +++---
 drivers/video/sh_mobile_lcdcfb.c          |  4 ++--
 12 files changed, 76 insertions(+), 38 deletions(-)

-- 
Regards,

Laurent Pinchart

[PATCH 05/12] drm: shmob_drm: Convert to clk_prepare/unprepare

[Laurent Pinchart]

Turn clk_enable() and clk_disable() calls into clk_prepare_enable() and
clk_disable_unprepare() to get ready for the migration to the common
clock framework.

Cc: David Airlie <airlied@linux.ie>
Cc: dri-devel@lists.freedesktop.org
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/shmobile/shmob_drm_crtc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
index 54bad98..562f9a4 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
@@ -40,7 +40,7 @@
 static void shmob_drm_clk_on(struct shmob_drm_device *sdev)
 {
 	if (sdev->clock)
-		clk_enable(sdev->clock);
+		clk_prepare_enable(sdev->clock);
 #if 0
 	if (sdev->meram_dev && sdev->meram_dev->pdev)
 		pm_runtime_get_sync(&sdev->meram_dev->pdev->dev);
@@ -54,7 +54,7 @@ static void shmob_drm_clk_off(struct shmob_drm_device *sdev)
 		pm_runtime_put_sync(&sdev->meram_dev->pdev->dev);
 #endif
 	if (sdev->clock)
-		clk_disable(sdev->clock);
+		clk_disable_unprepare(sdev->clock);
 }
 
 /* -----------------------------------------------------------------------------
-- 
1.8.1.5

Re: [PATCH 05/12] drm: shmob_drm: Convert to clk_prepare/unprepare

[Laurent Pinchart]

Hi Dave,

Could you please pick this patch up ?

On Monday 28 October 2013 23:49:22 Laurent Pinchart wrote:
> Turn clk_enable() and clk_disable() calls into clk_prepare_enable() and
> clk_disable_unprepare() to get ready for the migration to the common
> clock framework.
> 
> Cc: David Airlie <airlied@linux.ie>
> Cc: dri-devel@lists.freedesktop.org
> Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> ---
>  drivers/gpu/drm/shmobile/shmob_drm_crtc.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
> b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c index 54bad98..562f9a4 100644
> --- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
> +++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
> @@ -40,7 +40,7 @@
>  static void shmob_drm_clk_on(struct shmob_drm_device *sdev)
>  {
>  	if (sdev->clock)
> -		clk_enable(sdev->clock);
> +		clk_prepare_enable(sdev->clock);
>  #if 0
>  	if (sdev->meram_dev && sdev->meram_dev->pdev)
>  		pm_runtime_get_sync(&sdev->meram_dev->pdev->dev);
> @@ -54,7 +54,7 @@ static void shmob_drm_clk_off(struct shmob_drm_device
> *sdev) pm_runtime_put_sync(&sdev->meram_dev->pdev->dev);
>  #endif
>  	if (sdev->clock)
> -		clk_disable(sdev->clock);
> +		clk_disable_unprepare(sdev->clock);
>  }
> 
>  /*
> ---------------------------------------------------------------------------
-- 
Regards,

Laurent Pinchart

Re: [PATCH 05/12] drm: shmob_drm: Convert to clk_prepare/unprepare

[Thierry Reding]

[-- Attachment #1.1: Type: text/plain, Size: 1938 bytes --]

On Sat, Nov 09, 2013 at 01:51:04PM +0100, Laurent Pinchart wrote:
> Hi Dave,
> 
> Could you please pick this patch up ?
> 
> On Monday 28 October 2013 23:49:22 Laurent Pinchart wrote:
> > Turn clk_enable() and clk_disable() calls into clk_prepare_enable() and
> > clk_disable_unprepare() to get ready for the migration to the common
> > clock framework.
> > 
> > Cc: David Airlie <airlied@linux.ie>
> > Cc: dri-devel@lists.freedesktop.org
> > Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > ---
> >  drivers/gpu/drm/shmobile/shmob_drm_crtc.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
> > b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c index 54bad98..562f9a4 100644
> > --- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
> > +++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
> > @@ -40,7 +40,7 @@
> >  static void shmob_drm_clk_on(struct shmob_drm_device *sdev)
> >  {
> >  	if (sdev->clock)
> > -		clk_enable(sdev->clock);
> > +		clk_prepare_enable(sdev->clock);

Sorry for jumping in so late, but shouldn't this be split into two
separate calls, clk_prepare() in .probe() and clk_enable() here?

Also note that both clk_prepare() and clk_enable() (and therefore
clk_prepare_enable() as well) can fail, so you should really check
the return values here.

> >  #if 0
> >  	if (sdev->meram_dev && sdev->meram_dev->pdev)
> >  		pm_runtime_get_sync(&sdev->meram_dev->pdev->dev);
> > @@ -54,7 +54,7 @@ static void shmob_drm_clk_off(struct shmob_drm_device
> > *sdev) pm_runtime_put_sync(&sdev->meram_dev->pdev->dev);
> >  #endif
> >  	if (sdev->clock)
> > -		clk_disable(sdev->clock);
> > +		clk_disable_unprepare(sdev->clock);

Similarily I'd expect this to be clk_disable() only, with the
clk_unprepare() in .remove(). Or perhaps there's a very good reason to
do both here?

Thierry

[-- Attachment #1.2: Type: application/pgp-signature, Size: 836 bytes --]

[-- Attachment #2: Type: text/plain, Size: 159 bytes --]

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

Re: [PATCH 05/12] drm: shmob_drm: Convert to clk_prepare/unprepare

[Laurent Pinchart]

[-- Attachment #1.1: Type: text/plain, Size: 3016 bytes --]

Hi Thierry,

On Monday 11 November 2013 09:55:24 Thierry Reding wrote:
> On Sat, Nov 09, 2013 at 01:51:04PM +0100, Laurent Pinchart wrote:
> > Hi Dave,
> > 
> > Could you please pick this patch up ?
> > 
> > On Monday 28 October 2013 23:49:22 Laurent Pinchart wrote:
> > > Turn clk_enable() and clk_disable() calls into clk_prepare_enable() and
> > > clk_disable_unprepare() to get ready for the migration to the common
> > > clock framework.
> > > 
> > > Cc: David Airlie <airlied@linux.ie>
> > > Cc: dri-devel@lists.freedesktop.org
> > > Signed-off-by: Laurent Pinchart
> > > <laurent.pinchart+renesas@ideasonboard.com>
> > > ---
> > > 
> > >  drivers/gpu/drm/shmobile/shmob_drm_crtc.c | 4 ++--
> > >  1 file changed, 2 insertions(+), 2 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
> > > b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c index 54bad98..562f9a4
> > > 100644
> > > --- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
> > > +++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
> > > @@ -40,7 +40,7 @@
> > >  static void shmob_drm_clk_on(struct shmob_drm_device *sdev)
> > >  {
> > >  	if (sdev->clock)
> > > -		clk_enable(sdev->clock);
> > > +		clk_prepare_enable(sdev->clock);
> 
> Sorry for jumping in so late, but shouldn't this be split into two
> separate calls, clk_prepare() in .probe() and clk_enable() here?

The clock prepare and enable operations are split to allow clock 
implementations to sleep. Clocks should be kept disable whenever possible, the 
clk_enable() and clk_disable() calls should be as close as possible to the 
time range during which the clock needs to be enabled. This means that those 
calls might happen in a context where sleeping isn't allowed. If a clock 
implementation needs to sleep to enable the clock (by performing an I2C access 
for instance), that operation should be performed at prepare time.

>From a clock user point of view, both clk_prepare() and clk_enable() should be 
called as late as possible. If clk_enable() needs to be called in an atomic 
context clk_prepare() must be called earlier, in a non-atomic context(). 
Otherwise there'e no point in splitting the two calls.

> Also note that both clk_prepare() and clk_enable() (and therefore
> clk_prepare_enable() as well) can fail, so you should really check
> the return values here.

Yes, that's a good point. I'd like to fix that in a separate patch in order to 
avoid delaying this one.

> > >  #if 0
> > >  	if (sdev->meram_dev && sdev->meram_dev->pdev)
> > >  		pm_runtime_get_sync(&sdev->meram_dev->pdev->dev);
> > > @@ -54,7 +54,7 @@ static void shmob_drm_clk_off(struct shmob_drm_device
> > > *sdev)
> > > 	pm_runtime_put_sync(&sdev->meram_dev->pdev->dev);
> > >  #endif
> > >  	if (sdev->clock)
> > > -		clk_disable(sdev->clock);
> > > +		clk_disable_unprepare(sdev->clock);
> 
> Similarily I'd expect this to be clk_disable() only, with the
> clk_unprepare() in .remove(). Or perhaps there's a very good reason to
> do both here?

-- 
Regards,

Laurent Pinchart

[-- Attachment #1.2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

[-- Attachment #2: Type: text/plain, Size: 159 bytes --]

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