diff mbox

[v9,2/4] soc: mediatek: Init MT8173 scpsys driver earlier

Message ID 1476953798-23263-3-git-send-email-jamesjj.liao@mediatek.com (mailing list archive)
State New, archived
Headers show

Commit Message

James Liao Oct. 20, 2016, 8:56 a.m. UTC 
Some power domain comsumers may init before module_init.
So the power domain provider (scpsys) need to be initialized
earlier too.

Take an example for our IOMMU (M4U) and SMI. SMI is a bridge
between IOMMU and multimedia HW. SMI is responsible to
enable/disable iommu and help transfer data for each multimedia
HW. Both of them have to wait until the power and clocks are
enabled.

So scpsys driver should be initialized before SMI, and SMI should
be initialized before IOMMU, and then init IOMMU consumers
(display/vdec/venc/camera etc.).

IOMMU is subsys_init by default. So we need to init scpsys driver
before subsys_init.

Signed-off-by: James Liao <jamesjj.liao@mediatek.com>
Reviewed-by: Kevin Hilman <khilman@baylibre.com>
---
 drivers/soc/mediatek/mtk-scpsys.c | 19 ++++++++++++++++++-
 1 file changed, 18 insertions(+), 1 deletion(-)

Comments

Matthias Brugger Oct. 31, 2016, 12:08 a.m. UTC | #1 
Hi James,

On 10/20/2016 10:56 AM, James Liao wrote:
> Some power domain comsumers may init before module_init.
> So the power domain provider (scpsys) need to be initialized
> earlier too.
>
> Take an example for our IOMMU (M4U) and SMI. SMI is a bridge
> between IOMMU and multimedia HW. SMI is responsible to
> enable/disable iommu and help transfer data for each multimedia
> HW. Both of them have to wait until the power and clocks are
> enabled.
>
> So scpsys driver should be initialized before SMI, and SMI should
> be initialized before IOMMU, and then init IOMMU consumers
> (display/vdec/venc/camera etc.).
>
> IOMMU is subsys_init by default. So we need to init scpsys driver
> before subsys_init.
>
> Signed-off-by: James Liao <jamesjj.liao@mediatek.com>
> Reviewed-by: Kevin Hilman <khilman@baylibre.com>
> ---

I didn't applied this patch for now.
I answered you in v7 of this series [1]. I would prefer to see if we can 
fix that otherwise.

Would be great if you or Yong could provide some feedback.

Thanks,
Matthias

[1] https://patchwork.kernel.org/patch/9397405/

>  drivers/soc/mediatek/mtk-scpsys.c | 19 ++++++++++++++++++-
>  1 file changed, 18 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/soc/mediatek/mtk-scpsys.c b/drivers/soc/mediatek/mtk-scpsys.c
> index fa9ee69..dd7a07d 100644
> --- a/drivers/soc/mediatek/mtk-scpsys.c
> +++ b/drivers/soc/mediatek/mtk-scpsys.c
> @@ -613,4 +613,21 @@ static int scpsys_probe(struct platform_device *pdev)
>  		.of_match_table = of_match_ptr(of_scpsys_match_tbl),
>  	},
>  };
> -builtin_platform_driver(scpsys_drv);
> +
> +static int __init scpsys_drv_init(void)
> +{
> +	return platform_driver_register(&scpsys_drv);
> +}
> +
> +/*
> + * There are some Mediatek drivers which depend on the power domain driver need
> + * to probe in earlier initcall levels. So scpsys driver also need to probe
> + * earlier.
> + *
> + * IOMMU(M4U) and SMI drivers for example. SMI is a bridge between IOMMU and
> + * multimedia HW. IOMMU depends on SMI, and SMI is a power domain consumer,
> + * so the proper probe sequence should be scpsys -> SMI -> IOMMU driver.
> + * IOMMU drivers are initialized during subsys_init by default, so we need to
> + * move SMI and scpsys drivers to subsys_init or earlier init levels.
> + */
> +subsys_initcall(scpsys_drv_init);
>
James Liao Oct. 31, 2016, 2:08 a.m. UTC | #2 
On Mon, 2016-10-31 at 01:08 +0100, Matthias Brugger wrote:
> Hi James,
> 
> On 10/20/2016 10:56 AM, James Liao wrote:
> > Some power domain comsumers may init before module_init.
> > So the power domain provider (scpsys) need to be initialized
> > earlier too.
> >
> > Take an example for our IOMMU (M4U) and SMI. SMI is a bridge
> > between IOMMU and multimedia HW. SMI is responsible to
> > enable/disable iommu and help transfer data for each multimedia
> > HW. Both of them have to wait until the power and clocks are
> > enabled.
> >
> > So scpsys driver should be initialized before SMI, and SMI should
> > be initialized before IOMMU, and then init IOMMU consumers
> > (display/vdec/venc/camera etc.).
> >
> > IOMMU is subsys_init by default. So we need to init scpsys driver
> > before subsys_init.
> >
> > Signed-off-by: James Liao <jamesjj.liao@mediatek.com>
> > Reviewed-by: Kevin Hilman <khilman@baylibre.com>
> > ---
> 
> I didn't applied this patch for now.
> I answered you in v7 of this series [1]. I would prefer to see if we can 
> fix that otherwise.
> 
> Would be great if you or Yong could provide some feedback.
> 
> Thanks,
> Matthias
> 
> [1] https://patchwork.kernel.org/patch/9397405/
> 
> >  drivers/soc/mediatek/mtk-scpsys.c | 19 ++++++++++++++++++-
> >  1 file changed, 18 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/soc/mediatek/mtk-scpsys.c b/drivers/soc/mediatek/mtk-scpsys.c
> > index fa9ee69..dd7a07d 100644
> > --- a/drivers/soc/mediatek/mtk-scpsys.c
> > +++ b/drivers/soc/mediatek/mtk-scpsys.c
> > @@ -613,4 +613,21 @@ static int scpsys_probe(struct platform_device *pdev)
> >  		.of_match_table = of_match_ptr(of_scpsys_match_tbl),
> >  	},
> >  };
> > -builtin_platform_driver(scpsys_drv);
> > +
> > +static int __init scpsys_drv_init(void)
> > +{
> > +	return platform_driver_register(&scpsys_drv);
> > +}
> > +
> > +/*
> > + * There are some Mediatek drivers which depend on the power domain driver need
> > + * to probe in earlier initcall levels. So scpsys driver also need to probe
> > + * earlier.
> > + *
> > + * IOMMU(M4U) and SMI drivers for example. SMI is a bridge between IOMMU and
> > + * multimedia HW. IOMMU depends on SMI, and SMI is a power domain consumer,
> > + * so the proper probe sequence should be scpsys -> SMI -> IOMMU driver.
> > + * IOMMU drivers are initialized during subsys_init by default, so we need to
> > + * move SMI and scpsys drivers to subsys_init or earlier init levels.
> > + */
> > +subsys_initcall(scpsys_drv_init);
> >

Hi Matthias,

I got it, thanks.


Hi Yong,

Is this patch still needed on latest kernel for IOMMU driver? Or we have
other solutions for IOMMU driver init?


Best regards,

James
Honghui Zhang Nov. 15, 2016, 10:45 a.m. UTC | #3 
On Mon, 2016-10-31 at 10:08 +0800, James Liao wrote:
> On Mon, 2016-10-31 at 01:08 +0100, Matthias Brugger wrote:
> > Hi James,
> > 
> > On 10/20/2016 10:56 AM, James Liao wrote:
> > > Some power domain comsumers may init before module_init.
> > > So the power domain provider (scpsys) need to be initialized
> > > earlier too.
> > >
> > > Take an example for our IOMMU (M4U) and SMI. SMI is a bridge
> > > between IOMMU and multimedia HW. SMI is responsible to
> > > enable/disable iommu and help transfer data for each multimedia
> > > HW. Both of them have to wait until the power and clocks are
> > > enabled.
> > >
> > > So scpsys driver should be initialized before SMI, and SMI should
> > > be initialized before IOMMU, and then init IOMMU consumers
> > > (display/vdec/venc/camera etc.).
> > >
> > > IOMMU is subsys_init by default. So we need to init scpsys driver
> > > before subsys_init.
> > >
> > > Signed-off-by: James Liao <jamesjj.liao@mediatek.com>
> > > Reviewed-by: Kevin Hilman <khilman@baylibre.com>
> > > ---
> > 
> > I didn't applied this patch for now.
> > I answered you in v7 of this series [1]. I would prefer to see if we can 
> > fix that otherwise.
> > 
> > Would be great if you or Yong could provide some feedback.
> > 

Hi, Matthias,
  Yong had verified your propose and it seems didn't work[1].
But I'm looking at the recently merged device link patches[2], it maybe
could help with the probe sequence issue.

If we set the iommu client as the smi larb's device consumer of
device_link with the DL_DEV_PROBING flags, it will be wait for smi
larb's probe done. But I'm not test that solution yet, need to do a bit
more test to verify this.

thanks.

[1]
https://lists.linuxfoundation.org/pipermail/iommu/2016-July/018034.html
[2]https://www.mail-archive.com/linux-kernel@vger.kernel.org/msg1261311.html

> > Thanks,
> > Matthias
> > 
> > [1] https://patchwork.kernel.org/patch/9397405/
> > 
> > >  drivers/soc/mediatek/mtk-scpsys.c | 19 ++++++++++++++++++-
> > >  1 file changed, 18 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/soc/mediatek/mtk-scpsys.c b/drivers/soc/mediatek/mtk-scpsys.c
> > > index fa9ee69..dd7a07d 100644
> > > --- a/drivers/soc/mediatek/mtk-scpsys.c
> > > +++ b/drivers/soc/mediatek/mtk-scpsys.c
> > > @@ -613,4 +613,21 @@ static int scpsys_probe(struct platform_device *pdev)
> > >  		.of_match_table = of_match_ptr(of_scpsys_match_tbl),
> > >  	},

> _______________________________________________
> Linux-mediatek mailing list
> Linux-mediatek@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-mediatek
diff mbox

Patch

diff --git a/drivers/soc/mediatek/mtk-scpsys.c b/drivers/soc/mediatek/mtk-scpsys.c
index fa9ee69..dd7a07d 100644
--- a/drivers/soc/mediatek/mtk-scpsys.c
+++ b/drivers/soc/mediatek/mtk-scpsys.c
@@ -613,4 +613,21 @@  static int scpsys_probe(struct platform_device *pdev)
 		.of_match_table = of_match_ptr(of_scpsys_match_tbl),
 	},
 };
-builtin_platform_driver(scpsys_drv);
+
+static int __init scpsys_drv_init(void)
+{
+	return platform_driver_register(&scpsys_drv);
+}
+
+/*
+ * There are some Mediatek drivers which depend on the power domain driver need
+ * to probe in earlier initcall levels. So scpsys driver also need to probe
+ * earlier.
+ *
+ * IOMMU(M4U) and SMI drivers for example. SMI is a bridge between IOMMU and
+ * multimedia HW. IOMMU depends on SMI, and SMI is a power domain consumer,
+ * so the proper probe sequence should be scpsys -> SMI -> IOMMU driver.
+ * IOMMU drivers are initialized during subsys_init by default, so we need to
+ * move SMI and scpsys drivers to subsys_init or earlier init levels.
+ */
+subsys_initcall(scpsys_drv_init);