[PATCH] mx35: Fix boot ROM hang in internal boot mode

[PATCH] mx35: Fix boot ROM hang in internal boot mode

[Hans J. Koch]

If a watchdog reset occurs after booting in internal boot mode, the i.MX35
won't boot anymore. The boot ROM code seems to assume that some clocks are
turned on (they are after a power-on reset). This patch turns on the
necessary clocks.

Signed-off-by: Hans J. Koch <hjk@linutronix.de>
---
 arch/arm/mach-mx3/clock-imx35.c |   20 +++++++++++++++++---
 1 files changed, 17 insertions(+), 3 deletions(-)

diff --git a/arch/arm/mach-mx3/clock-imx35.c b/arch/arm/mach-mx3/clock-imx35.c
index d3af0fd..671acb6 100644
--- a/arch/arm/mach-mx3/clock-imx35.c
+++ b/arch/arm/mach-mx3/clock-imx35.c
@@ -485,7 +485,7 @@ static struct clk_lookup lookups[] = {
 
 int __init mx35_clocks_init()
 {
-	unsigned int ll = 0;
+	unsigned int ll = 0, cgr2, cgr3;
 
 #if defined(CONFIG_DEBUG_LL) && !defined(CONFIG_DEBUG_ICEDCC)
 	ll = (3 << 16);
@@ -499,8 +499,22 @@ int __init mx35_clocks_init()
 	__raw_writel((3 << 18), CCM_BASE + CCM_CGR0);
 	__raw_writel((3 << 2) | (3 << 4) | (3 << 6) | (3 << 8) | (3 << 16),
 			CCM_BASE + CCM_CGR1);
-	__raw_writel((3 << 26) | ll, CCM_BASE + CCM_CGR2);
-	__raw_writel(0, CCM_BASE + CCM_CGR3);
+
+	/* Check if we came up in internal boot mode. If yes, we need some
+	 * extra clocks turned on, otherwise the MX35 boot ROM code will
+	 * hang after a watchdog reset.
+	 */
+	if (__raw_readl(CCM_BASE + CCM_RCSR) & (3 << 10)) {
+		cgr2 = (3 << 26);
+		cgr3 = 0;
+	} else {
+		/* Additionally turn on UART1, SCC, and IIM clocks */
+		cgr2 = (3 << 26) | (3 << 16) | (3 << 4);
+		cgr3 = (3 << 2);
+	}
+
+	__raw_writel(cgr2 | ll, CCM_BASE + CCM_CGR2);
+	__raw_writel(cgr3, CCM_BASE + CCM_CGR3);
 
 	mxc_timer_init(&gpt_clk,
 			MX35_IO_ADDRESS(MX35_GPT1_BASE_ADDR), MX35_INT_GPT);
-- 
1.7.1

[PATCH] mx35: Fix boot ROM hang in internal boot mode

[Uwe Kleine-König]

Hello Hans,

On Wed, Aug 11, 2010 at 02:19:33AM +0200, Hans J. Koch wrote:
> If a watchdog reset occurs after booting in internal boot mode, the i.MX35
> won't boot anymore. The boot ROM code seems to assume that some clocks are
> turned on (they are after a power-on reset). This patch turns on the
> necessary clocks.
> 
> Signed-off-by: Hans J. Koch <hjk@linutronix.de>
> ---
>  arch/arm/mach-mx3/clock-imx35.c |   20 +++++++++++++++++---
>  1 files changed, 17 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/arm/mach-mx3/clock-imx35.c b/arch/arm/mach-mx3/clock-imx35.c
> index d3af0fd..671acb6 100644
> --- a/arch/arm/mach-mx3/clock-imx35.c
> +++ b/arch/arm/mach-mx3/clock-imx35.c
> @@ -485,7 +485,7 @@ static struct clk_lookup lookups[] = {
>  
>  int __init mx35_clocks_init()
>  {
> -	unsigned int ll = 0;
> +	unsigned int ll = 0, cgr2, cgr3;
>  
>  #if defined(CONFIG_DEBUG_LL) && !defined(CONFIG_DEBUG_ICEDCC)
>  	ll = (3 << 16);
> @@ -499,8 +499,22 @@ int __init mx35_clocks_init()
>  	__raw_writel((3 << 18), CCM_BASE + CCM_CGR0);
>  	__raw_writel((3 << 2) | (3 << 4) | (3 << 6) | (3 << 8) | (3 << 16),
>  			CCM_BASE + CCM_CGR1);
> -	__raw_writel((3 << 26) | ll, CCM_BASE + CCM_CGR2);
> -	__raw_writel(0, CCM_BASE + CCM_CGR3);
> +
> +	/* Check if we came up in internal boot mode. If yes, we need some
<nitpick> /* on an extra line please. </nitpick>

> +	 * extra clocks turned on, otherwise the MX35 boot ROM code will
> +	 * hang after a watchdog reset.
> +	 */
Better move the comment into the else branch below.

> +	if (__raw_readl(CCM_BASE + CCM_RCSR) & (3 << 10)) {
Can you please introduce a symbol for (3 << 10)?.

> +		cgr2 = (3 << 26);
> +		cgr3 = 0;
> +	} else {
> +		/* Additionally turn on UART1, SCC, and IIM clocks */
really UART1?

> +		cgr2 = (3 << 26) | (3 << 16) | (3 << 4);
> +		cgr3 = (3 << 2);
> +	}
I would prefer having symbolic names here, too, but this affects not
only your patch.  Will put it on my todo list.

Is it sensible to assume that the machine will boot in internal boot
mode again iff it came up this time in internal boot mode?  (I assume
yes, still I would reword the comment.  Something like the following?:

	In internal boot mode the MX35 boot ROM code seems to assume
	that UART1, SCC, and IIM clocks are on and hangs if they are
	not.  These are enabled by hardware after a power-on reset, but
	not after a watchdog reset.  So enable them here.
)

> +
> +	__raw_writel(cgr2 | ll, CCM_BASE + CCM_CGR2);
hmm, can you make this:

-	unsigned int ll = 0;
+	unsigned int cgr2 = 3 << 26, cgr3 = 0;
 
 #if defined(CONFIG_DEBUG_LL) && !defined(CONFIG_DEBUG_ICEDCC)
 	cgr2 |= 3 << 16;
 #endif
 ...
-...
+	if (!__raw_readl(CCM_BASE + CCM_RCSR) & (3 << 10)) {
+		/*
+		 * ...
+		 */
+		cgr2 |= 3 << 16 | 3 << 4;
+		cgr3 |= 3 << 2;
+	}
+
+	__raw_writel(cgr2, CCM_BASE + CCM_CGR2);
+	__raw_writel(cgr3, CCM_BASE + CCM_CGR3);


> +	__raw_writel(cgr3, CCM_BASE + CCM_CGR3);
>  
>  	mxc_timer_init(&gpt_clk,
>  			MX35_IO_ADDRESS(MX35_GPT1_BASE_ADDR), MX35_INT_GPT);

Thanks
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-K?nig            |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |

[PATCH] mx35: Fix boot ROM hang in internal boot mode

[Hans J. Koch]

On Wed, Aug 11, 2010 at 07:27:50AM +0200, Uwe Kleine-K?nig wrote:
> Hello Hans,

Hello Uwe,
thanks for your review. A -v2 patch will follow soon. Some comments below.

Thanks,
Hans

> 
> On Wed, Aug 11, 2010 at 02:19:33AM +0200, Hans J. Koch wrote:
> > If a watchdog reset occurs after booting in internal boot mode, the i.MX35
> > won't boot anymore. The boot ROM code seems to assume that some clocks are
> > turned on (they are after a power-on reset). This patch turns on the
> > necessary clocks.
> > 
> > Signed-off-by: Hans J. Koch <hjk@linutronix.de>
> > ---
> >  arch/arm/mach-mx3/clock-imx35.c |   20 +++++++++++++++++---
> >  1 files changed, 17 insertions(+), 3 deletions(-)
> > 
> > diff --git a/arch/arm/mach-mx3/clock-imx35.c b/arch/arm/mach-mx3/clock-imx35.c
> > index d3af0fd..671acb6 100644
> > --- a/arch/arm/mach-mx3/clock-imx35.c
> > +++ b/arch/arm/mach-mx3/clock-imx35.c
> > @@ -485,7 +485,7 @@ static struct clk_lookup lookups[] = {
> >  
> >  int __init mx35_clocks_init()
> >  {
> > -	unsigned int ll = 0;
> > +	unsigned int ll = 0, cgr2, cgr3;
> >  
> >  #if defined(CONFIG_DEBUG_LL) && !defined(CONFIG_DEBUG_ICEDCC)
> >  	ll = (3 << 16);
> > @@ -499,8 +499,22 @@ int __init mx35_clocks_init()
> >  	__raw_writel((3 << 18), CCM_BASE + CCM_CGR0);
> >  	__raw_writel((3 << 2) | (3 << 4) | (3 << 6) | (3 << 8) | (3 << 16),
> >  			CCM_BASE + CCM_CGR1);
> > -	__raw_writel((3 << 26) | ll, CCM_BASE + CCM_CGR2);
> > -	__raw_writel(0, CCM_BASE + CCM_CGR3);
> > +
> > +	/* Check if we came up in internal boot mode. If yes, we need some
> <nitpick> /* on an extra line please. </nitpick>

alright, if you like it...

> 
> > +	 * extra clocks turned on, otherwise the MX35 boot ROM code will
> > +	 * hang after a watchdog reset.
> > +	 */
> Better move the comment into the else branch below.

There's no else branch anymore. Before the "if" the comment says what's
being checked, and in the if branch the comment says what's being done.

> 
> > +	if (__raw_readl(CCM_BASE + CCM_RCSR) & (3 << 10)) {
> Can you please introduce a symbol for (3 << 10)?.

I just wanted to follow the style of the rest of the file ;-)
Since you say you've got those #defines on your todo list, I better
leave that to you.

> 
> > +		cgr2 = (3 << 26);
> > +		cgr3 = 0;
> > +	} else {
> > +		/* Additionally turn on UART1, SCC, and IIM clocks */
> really UART1?

Well, stupid reference manual. They name the UARTs 1,2,3 instead of 0,1,2.
That also leads to lines like:

_REGISTER_CLOCK("imx-uart.0", NULL, uart1_clk)

I renamed it UART0 in the comment to make it more clear.

> 
> > +		cgr2 = (3 << 26) | (3 << 16) | (3 << 4);
> > +		cgr3 = (3 << 2);
> > +	}
> I would prefer having symbolic names here, too, but this affects not
> only your patch.  Will put it on my todo list.

Good idea, thank you. It's a PITA to need that >3000 pages reference manual
to read the code...

> 
> Is it sensible to assume that the machine will boot in internal boot
> mode again iff it came up this time in internal boot mode?  (I assume
> yes, still I would reword the comment.  Something like the following?:
> 
> 	In internal boot mode the MX35 boot ROM code seems to assume
> 	that UART1, SCC, and IIM clocks are on and hangs if they are
> 	not.  These are enabled by hardware after a power-on reset, but
> 	not after a watchdog reset.  So enable them here.
> )
> 
> > +
> > +	__raw_writel(cgr2 | ll, CCM_BASE + CCM_CGR2);
> hmm, can you make this:
> 
> -	unsigned int ll = 0;
> +	unsigned int cgr2 = 3 << 26, cgr3 = 0;
>  
>  #if defined(CONFIG_DEBUG_LL) && !defined(CONFIG_DEBUG_ICEDCC)
>  	cgr2 |= 3 << 16;
>  #endif
>  ...
> -...
> +	if (!__raw_readl(CCM_BASE + CCM_RCSR) & (3 << 10)) {
> +		/*
> +		 * ...
> +		 */
> +		cgr2 |= 3 << 16 | 3 << 4;
> +		cgr3 |= 3 << 2;
> +	}

Well, OK, I like that, too.

> +
> +	__raw_writel(cgr2, CCM_BASE + CCM_CGR2);
> +	__raw_writel(cgr3, CCM_BASE + CCM_CGR3);
> 
> 
> > +	__raw_writel(cgr3, CCM_BASE + CCM_CGR3);
> >  
> >  	mxc_timer_init(&gpt_clk,
> >  			MX35_IO_ADDRESS(MX35_GPT1_BASE_ADDR), MX35_INT_GPT);
> 
> Thanks
> Uwe
> 
> -- 
> Pengutronix e.K.                           | Uwe Kleine-K?nig            |
> Industrial Linux Solutions                 | http://www.pengutronix.de/  |

[PATCH] mx35: Fix boot ROM hang in internal boot mode

[Uwe Kleine-König]

Hello Hans,

On Thu, Aug 12, 2010 at 01:20:14AM +0200, Hans J. Koch wrote:
> > > diff --git a/arch/arm/mach-mx3/clock-imx35.c b/arch/arm/mach-mx3/clock-imx35.c
> > > index d3af0fd..671acb6 100644
> > > --- a/arch/arm/mach-mx3/clock-imx35.c
> > > +++ b/arch/arm/mach-mx3/clock-imx35.c
> > > @@ -485,7 +485,7 @@ static struct clk_lookup lookups[] = {
> > >  
> > >  int __init mx35_clocks_init()
> > >  {
> > > -	unsigned int ll = 0;
> > > +	unsigned int ll = 0, cgr2, cgr3;
> > >  
> > >  #if defined(CONFIG_DEBUG_LL) && !defined(CONFIG_DEBUG_ICEDCC)
> > >  	ll = (3 << 16);
> > > @@ -499,8 +499,22 @@ int __init mx35_clocks_init()
> > >  	__raw_writel((3 << 18), CCM_BASE + CCM_CGR0);
> > >  	__raw_writel((3 << 2) | (3 << 4) | (3 << 6) | (3 << 8) | (3 << 16),
> > >  			CCM_BASE + CCM_CGR1);
> > > -	__raw_writel((3 << 26) | ll, CCM_BASE + CCM_CGR2);
> > > -	__raw_writel(0, CCM_BASE + CCM_CGR3);
> > > +
> > > +	/* Check if we came up in internal boot mode. If yes, we need some
> > <nitpick> /* on an extra line please. </nitpick>
> 
> alright, if you like it...
> 
> > 
> > > +	 * extra clocks turned on, otherwise the MX35 boot ROM code will
> > > +	 * hang after a watchdog reset.
> > > +	 */
> > Better move the comment into the else branch below.
> 
> There's no else branch anymore. Before the "if" the comment says what's
> being checked, and in the if branch the comment says what's being done.
> 
> > 
> > > +	if (__raw_readl(CCM_BASE + CCM_RCSR) & (3 << 10)) {
> > Can you please introduce a symbol for (3 << 10)?.
> 
> I just wanted to follow the style of the rest of the file ;-)
:-)

> Since you say you've got those #defines on your todo list, I better
> leave that to you.
> 
> > 
> > > +		cgr2 = (3 << 26);
> > > +		cgr3 = 0;
> > > +	} else {
> > > +		/* Additionally turn on UART1, SCC, and IIM clocks */
> > really UART1?
> 
> Well, stupid reference manual. They name the UARTs 1,2,3 instead of 0,1,2.
> That also leads to lines like:
> 
> _REGISTER_CLOCK("imx-uart.0", NULL, uart1_clk)
> 
> I renamed it UART0 in the comment to make it more clear.
Actually I already thought about fixing that.  (On the software side
obviously.)  There are only little problems and one bigger to make them
match.  The bigger one is that it breaks people's kernel commandlines
that usually say:

	console=ttymxc0...

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-K?nig            |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |