Re: [PATCH 2/2] mtd: nand: sunxi: add reset line support

Re: [PATCH 2/2] mtd: nand: sunxi: add reset line support

[Boris Brezillon]

+Philipp

On Sun, 19 Jun 2016 19:37:39 +0800
Icenowy Zheng <icenowy@aosc.xyz> wrote:

> The NAND controller on some sun8i chips needs its reset line to be deasserted
> before they can enter working state. This commit added the reset line process
> to the driver.
> 
> Signed-off-by: Icenowy Zheng <icenowy@aosc.xyz>
> ---
>  drivers/mtd/nand/sunxi_nand.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
> 
> diff --git a/drivers/mtd/nand/sunxi_nand.c b/drivers/mtd/nand/sunxi_nand.c
> index a83a690..1502748 100644
> --- a/drivers/mtd/nand/sunxi_nand.c
> +++ b/drivers/mtd/nand/sunxi_nand.c
> @@ -39,6 +39,7 @@
>  #include <linux/gpio.h>
>  #include <linux/interrupt.h>
>  #include <linux/iopoll.h>
> +#include <linux/reset.h>
>  
>  #define NFC_REG_CTL		0x0000
>  #define NFC_REG_ST		0x0004
> @@ -269,6 +270,7 @@ struct sunxi_nfc {
>  	void __iomem *regs;
>  	struct clk *ahb_clk;
>  	struct clk *mod_clk;
> +	struct reset_control *reset;
>  	unsigned long assigned_cs;
>  	unsigned long clk_rate;
>  	struct list_head chips;
> @@ -1871,6 +1873,18 @@ static int sunxi_nfc_probe(struct platform_device *pdev)
>  	if (ret)
>  		goto out_ahb_clk_unprepare;
>  
> +	nfc->reset = devm_reset_control_get_optional(dev, "ahb");
> +	if (PTR_ERR(nfc->reset) == -EPROBE_DEFER)
> +		return PTR_ERR(nfc->reset);

Actually you should test for != -ENOENT, because all error codes except
this one should stop the ->probe().

BTW, this devm_reset_control_get_optional() is really weird. While most
_optional() methods return NULL when the element is not defined in the
DT, this one returns -ENOTENT, which makes it impossible to
differentiate a real error from a undefined reset line (which is a
valid case for _optional()).

Philipp, is there a good reason for doing that?

> +
> +	if (!IS_ERR(nfc->reset)) {
> +		ret = reset_control_deassert(nfc->reset);
> +		if (ret) {
> +			dev_err(dev, "reset err %d\n", ret);
> +			goto out_mod_clk_unprepare;
> +		}
> +	}
> +

Re: [PATCH 2/2] mtd: nand: sunxi: add reset line support

[Boris Brezillon]

On Sun, 19 Jun 2016 20:41:09 +0800
icenowy@aosc.xyz wrote:

> To be honest, I copied them from sunxi-mmc.c.
> 
> What function should be chosen better?

You did the right thing (except for the error detection part). My
question was addressed to Philipp (the reset subsystem maintainer).

> 
> 
> 19.06.2016, 20:06, "Boris Brezillon" <boris.brezillon@free-electrons.com>:
> > +Philipp
> >
> > On Sun, 19 Jun 2016 19:37:39 +0800
> > Icenowy Zheng <icenowy@aosc.xyz> wrote:
> >  
> >>  The NAND controller on some sun8i chips needs its reset line to be deasserted
> >>  before they can enter working state. This commit added the reset line process
> >>  to the driver.
> >>
> >>  Signed-off-by: Icenowy Zheng <icenowy@aosc.xyz>
> >>  ---
> >>   drivers/mtd/nand/sunxi_nand.c | 14 ++++++++++++++
> >>   1 file changed, 14 insertions(+)
> >>
> >>  diff --git a/drivers/mtd/nand/sunxi_nand.c b/drivers/mtd/nand/sunxi_nand.c
> >>  index a83a690..1502748 100644
> >>  --- a/drivers/mtd/nand/sunxi_nand.c
> >>  +++ b/drivers/mtd/nand/sunxi_nand.c
> >>  @@ -39,6 +39,7 @@
> >>   #include <linux/gpio.h>
> >>   #include <linux/interrupt.h>
> >>   #include <linux/iopoll.h>
> >>  +#include <linux/reset.h>
> >>
> >>   #define NFC_REG_CTL 0x0000
> >>   #define NFC_REG_ST 0x0004
> >>  @@ -269,6 +270,7 @@ struct sunxi_nfc {
> >>           void __iomem *regs;
> >>           struct clk *ahb_clk;
> >>           struct clk *mod_clk;
> >>  + struct reset_control *reset;
> >>           unsigned long assigned_cs;
> >>           unsigned long clk_rate;
> >>           struct list_head chips;
> >>  @@ -1871,6 +1873,18 @@ static int sunxi_nfc_probe(struct platform_device *pdev)
> >>           if (ret)
> >>                   goto out_ahb_clk_unprepare;
> >>
> >>  + nfc->reset = devm_reset_control_get_optional(dev, "ahb");
> >>  + if (PTR_ERR(nfc->reset) == -EPROBE_DEFER)
> >>  + return PTR_ERR(nfc->reset);  
> >
> > Actually you should test for != -ENOENT, because all error codes except
> > this one should stop the ->probe().
> >
> > BTW, this devm_reset_control_get_optional() is really weird. While most
> > _optional() methods return NULL when the element is not defined in the
> > DT, this one returns -ENOTENT, which makes it impossible to
> > differentiate a real error from a undefined reset line (which is a
> > valid case for _optional()).
> >
> > Philipp, is there a good reason for doing that?
> >  
> >>  +
> >>  + if (!IS_ERR(nfc->reset)) {
> >>  + ret = reset_control_deassert(nfc->reset);
> >>  + if (ret) {
> >>  + dev_err(dev, "reset err %d\n", ret);
> >>  + goto out_mod_clk_unprepare;
> >>  + }
> >>  + }
> >>  +  

Re: [PATCH 2/2] mtd: nand: sunxi: add reset line support

[Boris Brezillon]

On Sun, 19 Jun 2016 21:11:00 +0800
icenowy@aosc.xyz wrote:

> Then I will soon make the v2 patch set, with the error detection part fixed.
> 
> But why does sunxi-mmc check only EPROBE_DEFER?

I guess someone had a probe-dependency problem and fixed this case but
ignored all the possible errors. But there may be other reasons for
devm_reset_control_get_optional() to fail. The only one that is really
reflecting that the reset line is not defined in the DT is -ENOENT.

> 
> 19.06.2016, 20:53, "Boris Brezillon" <boris.brezillon@free-electrons.com>:
> > On Sun, 19 Jun 2016 20:41:09 +0800
> > icenowy@aosc.xyz wrote:
> >  
> >>  To be honest, I copied them from sunxi-mmc.c.
> >>
> >>  What function should be chosen better?  
> >
> > You did the right thing (except for the error detection part). My
> > question was addressed to Philipp (the reset subsystem maintainer).
> >  
> >>  19.06.2016, 20:06, "Boris Brezillon" <boris.brezillon@free-electrons.com>:  
> >>  > +Philipp
> >>  >
> >>  > On Sun, 19 Jun 2016 19:37:39 +0800
> >>  > Icenowy Zheng <icenowy@aosc.xyz> wrote:
> >>  >  
> >>  >>  The NAND controller on some sun8i chips needs its reset line to be deasserted
> >>  >>  before they can enter working state. This commit added the reset line process
> >>  >>  to the driver.
> >>  >>
> >>  >>  Signed-off-by: Icenowy Zheng <icenowy@aosc.xyz>
> >>  >>  ---
> >>  >>   drivers/mtd/nand/sunxi_nand.c | 14 ++++++++++++++
> >>  >>   1 file changed, 14 insertions(+)
> >>  >>
> >>  >>  diff --git a/drivers/mtd/nand/sunxi_nand.c b/drivers/mtd/nand/sunxi_nand.c
> >>  >>  index a83a690..1502748 100644
> >>  >>  --- a/drivers/mtd/nand/sunxi_nand.c
> >>  >>  +++ b/drivers/mtd/nand/sunxi_nand.c
> >>  >>  @@ -39,6 +39,7 @@
> >>  >>   #include <linux/gpio.h>
> >>  >>   #include <linux/interrupt.h>
> >>  >>   #include <linux/iopoll.h>
> >>  >>  +#include <linux/reset.h>
> >>  >>
> >>  >>   #define NFC_REG_CTL 0x0000
> >>  >>   #define NFC_REG_ST 0x0004
> >>  >>  @@ -269,6 +270,7 @@ struct sunxi_nfc {
> >>  >>           void __iomem *regs;
> >>  >>           struct clk *ahb_clk;
> >>  >>           struct clk *mod_clk;
> >>  >>  + struct reset_control *reset;
> >>  >>           unsigned long assigned_cs;
> >>  >>           unsigned long clk_rate;
> >>  >>           struct list_head chips;
> >>  >>  @@ -1871,6 +1873,18 @@ static int sunxi_nfc_probe(struct platform_device *pdev)
> >>  >>           if (ret)
> >>  >>                   goto out_ahb_clk_unprepare;
> >>  >>
> >>  >>  + nfc->reset = devm_reset_control_get_optional(dev, "ahb");
> >>  >>  + if (PTR_ERR(nfc->reset) == -EPROBE_DEFER)
> >>  >>  + return PTR_ERR(nfc->reset);  
> >>  >
> >>  > Actually you should test for != -ENOENT, because all error codes except
> >>  > this one should stop the ->probe().
> >>  >
> >>  > BTW, this devm_reset_control_get_optional() is really weird. While most
> >>  > _optional() methods return NULL when the element is not defined in the
> >>  > DT, this one returns -ENOTENT, which makes it impossible to
> >>  > differentiate a real error from a undefined reset line (which is a
> >>  > valid case for _optional()).
> >>  >
> >>  > Philipp, is there a good reason for doing that?
> >>  >  
> >>  >>  +
> >>  >>  + if (!IS_ERR(nfc->reset)) {
> >>  >>  + ret = reset_control_deassert(nfc->reset);
> >>  >>  + if (ret) {
> >>  >>  + dev_err(dev, "reset err %d\n", ret);
> >>  >>  + goto out_mod_clk_unprepare;
> >>  >>  + }
> >>  >>  + }
> >>  >>  +  

Re: [PATCH 2/2] mtd: nand: sunxi: add reset line support

[Philipp Zabel]

Am Sonntag, den 19.06.2016, 14:06 +0200 schrieb Boris Brezillon:
> +Philipp
> 
> On Sun, 19 Jun 2016 19:37:39 +0800
> Icenowy Zheng <icenowy@aosc.xyz> wrote:
> 
> > The NAND controller on some sun8i chips needs its reset line to be deasserted
> > before they can enter working state. This commit added the reset line process
> > to the driver.
> > 
> > Signed-off-by: Icenowy Zheng <icenowy@aosc.xyz>
> > ---
> >  drivers/mtd/nand/sunxi_nand.c | 14 ++++++++++++++
> >  1 file changed, 14 insertions(+)
> > 
> > diff --git a/drivers/mtd/nand/sunxi_nand.c b/drivers/mtd/nand/sunxi_nand.c
> > index a83a690..1502748 100644
> > --- a/drivers/mtd/nand/sunxi_nand.c
> > +++ b/drivers/mtd/nand/sunxi_nand.c
> > @@ -39,6 +39,7 @@
> >  #include <linux/gpio.h>
> >  #include <linux/interrupt.h>
> >  #include <linux/iopoll.h>
> > +#include <linux/reset.h>
> >  
> >  #define NFC_REG_CTL		0x0000
> >  #define NFC_REG_ST		0x0004
> > @@ -269,6 +270,7 @@ struct sunxi_nfc {
> >  	void __iomem *regs;
> >  	struct clk *ahb_clk;
> >  	struct clk *mod_clk;
> > +	struct reset_control *reset;
> >  	unsigned long assigned_cs;
> >  	unsigned long clk_rate;
> >  	struct list_head chips;
> > @@ -1871,6 +1873,18 @@ static int sunxi_nfc_probe(struct platform_device *pdev)
> >  	if (ret)
> >  		goto out_ahb_clk_unprepare;
> >  
> > +	nfc->reset = devm_reset_control_get_optional(dev, "ahb");
> > +	if (PTR_ERR(nfc->reset) == -EPROBE_DEFER)
> > +		return PTR_ERR(nfc->reset);
> 
> Actually you should test for != -ENOENT, because all error codes except
> this one should stop the ->probe().
> 
> BTW, this devm_reset_control_get_optional() is really weird. While most
> _optional() methods return NULL when the element is not defined in the
> DT, this one returns -ENOTENT, which makes it impossible to
> differentiate a real error from a undefined reset line (which is a
> valid case for _optional()).

Of course it's possible, -ENOENT is only returned if the reset line is
not defined in the device tree.

Note that gpiod_get_(index_)optional do nothing more that replacing
-ENOENT with NULL. And phydev_optional_get replaces -ENODEV with NULL.
And regulator_get_optional, if I understand it correctly, never returns
NULL.

> Philipp, is there a good reason for doing that?

Historically, NULL has not been a valid value for rstc. I suppose we
could add NULL checks to the reset_control_assert/deassert/reset/status
functions and align the reset API a bit with gpiod. I just wouldn't want
to see any IS_ERR_OR_NULL error handling in the drivers.

regards
Philipp

Re: [PATCH 2/2] mtd: nand: sunxi: add reset line support

[Boris Brezillon]

Hi Philipp,

On Mon, 20 Jun 2016 14:05:54 +0200
Philipp Zabel <p.zabel@pengutronix.de> wrote:

> Am Sonntag, den 19.06.2016, 14:06 +0200 schrieb Boris Brezillon:
> > +Philipp
> > 
> > On Sun, 19 Jun 2016 19:37:39 +0800
> > Icenowy Zheng <icenowy@aosc.xyz> wrote:
> >   
> > > The NAND controller on some sun8i chips needs its reset line to be deasserted
> > > before they can enter working state. This commit added the reset line process
> > > to the driver.
> > > 
> > > Signed-off-by: Icenowy Zheng <icenowy@aosc.xyz>
> > > ---
> > >  drivers/mtd/nand/sunxi_nand.c | 14 ++++++++++++++
> > >  1 file changed, 14 insertions(+)
> > > 
> > > diff --git a/drivers/mtd/nand/sunxi_nand.c b/drivers/mtd/nand/sunxi_nand.c
> > > index a83a690..1502748 100644
> > > --- a/drivers/mtd/nand/sunxi_nand.c
> > > +++ b/drivers/mtd/nand/sunxi_nand.c
> > > @@ -39,6 +39,7 @@
> > >  #include <linux/gpio.h>
> > >  #include <linux/interrupt.h>
> > >  #include <linux/iopoll.h>
> > > +#include <linux/reset.h>
> > >  
> > >  #define NFC_REG_CTL		0x0000
> > >  #define NFC_REG_ST		0x0004
> > > @@ -269,6 +270,7 @@ struct sunxi_nfc {
> > >  	void __iomem *regs;
> > >  	struct clk *ahb_clk;
> > >  	struct clk *mod_clk;
> > > +	struct reset_control *reset;
> > >  	unsigned long assigned_cs;
> > >  	unsigned long clk_rate;
> > >  	struct list_head chips;
> > > @@ -1871,6 +1873,18 @@ static int sunxi_nfc_probe(struct platform_device *pdev)
> > >  	if (ret)
> > >  		goto out_ahb_clk_unprepare;
> > >  
> > > +	nfc->reset = devm_reset_control_get_optional(dev, "ahb");
> > > +	if (PTR_ERR(nfc->reset) == -EPROBE_DEFER)
> > > +		return PTR_ERR(nfc->reset);  
> > 
> > Actually you should test for != -ENOENT, because all error codes except
> > this one should stop the ->probe().
> > 
> > BTW, this devm_reset_control_get_optional() is really weird. While most
> > _optional() methods return NULL when the element is not defined in the
> > DT, this one returns -ENOTENT, which makes it impossible to
> > differentiate a real error from a undefined reset line (which is a
> > valid case for _optional()).  
> 
> Of course it's possible, -ENOENT is only returned if the reset line is
> not defined in the device tree.

Okay, testing for err != -ENOENT is the right thing to do here. Maybe
this could be documented.

> 
> Note that gpiod_get_(index_)optional do nothing more that replacing
> -ENOENT with NULL. And phydev_optional_get replaces -ENODEV with NULL.
> And regulator_get_optional, if I understand it correctly, never returns
> NULL.
> 
> > Philipp, is there a good reason for doing that?  
> 
> Historically, NULL has not been a valid value for rstc. I suppose we
> could add NULL checks to the reset_control_assert/deassert/reset/status
> functions and align the reset API a bit with gpiod. I just wouldn't want
> to see any IS_ERR_OR_NULL error handling in the drivers.

Well, returning NULL is mainly a convenient way to differentiate real
errors from missing optional reset lines (which are not errors). Now,
if you say checking for -ENOENT is the right thing to do here, I'm
fine with that.

Regards,

Boris