Patchwork ARM: imx: pwm: Add support for MX53

login
register
mail settings
Submitter Fabio Estevam
Date Aug. 23, 2011, 4:26 p.m.
Message ID <1314116798-12755-1-git-send-email-fabio.estevam@freescale.com>
Download mbox | patch
Permalink /patch/111142/
State New
Headers show

Comments

Fabio Estevam - Aug. 23, 2011, 4:26 p.m.
Instead of adding the check for MX53 CPU type, use the negative logic and let 
PWM to be configured for MX53 as well.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
---
 arch/arm/plat-mxc/pwm.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
Sascha Hauer - Aug. 24, 2011, 8:18 a.m.
On Tue, Aug 23, 2011 at 01:26:38PM -0300, Fabio Estevam wrote:
> Instead of adding the check for MX53 CPU type, use the negative logic and let 
> PWM to be configured for MX53 as well.

Applied for next.

Sascha

> 
> Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
> ---
>  arch/arm/plat-mxc/pwm.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/arch/arm/plat-mxc/pwm.c b/arch/arm/plat-mxc/pwm.c
> index 761c3c9..42d74ea 100644
> --- a/arch/arm/plat-mxc/pwm.c
> +++ b/arch/arm/plat-mxc/pwm.c
> @@ -57,7 +57,7 @@ int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
>  	if (pwm == NULL || period_ns == 0 || duty_ns > period_ns)
>  		return -EINVAL;
>  
> -	if (cpu_is_mx27() || cpu_is_mx3() || cpu_is_mx25() || cpu_is_mx51()) {
> +	if (!(cpu_is_mx1() || cpu_is_mx21())) {
>  		unsigned long long c;
>  		unsigned long period_cycles, duty_cycles, prescale;
>  		u32 cr;
> -- 
> 1.7.1
> 
> 
>
Uwe Kleine-König - Aug. 24, 2011, 8:25 a.m.
Hello Fabio,

On Tue, Aug 23, 2011 at 01:26:38PM -0300, Fabio Estevam wrote:
> Instead of adding the check for MX53 CPU type, use the negative logic and let 
> PWM to be configured for MX53 as well.
I wonder why you prefer negative logic here. It might be shorter, but it
is also less robust. For example it seems to me you not only added
support for i.MX53 but also i.MX50.

The IMHO best solution would be to use device ids and get rid of
cpu_is_mx.. completely.

Best regards
Uwe
>  arch/arm/plat-mxc/pwm.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/arch/arm/plat-mxc/pwm.c b/arch/arm/plat-mxc/pwm.c
> index 761c3c9..42d74ea 100644
> --- a/arch/arm/plat-mxc/pwm.c
> +++ b/arch/arm/plat-mxc/pwm.c
> @@ -57,7 +57,7 @@ int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
>  	if (pwm == NULL || period_ns == 0 || duty_ns > period_ns)
>  		return -EINVAL;
>  
> -	if (cpu_is_mx27() || cpu_is_mx3() || cpu_is_mx25() || cpu_is_mx51()) {
> +	if (!(cpu_is_mx1() || cpu_is_mx21())) {
>  		unsigned long long c;
>  		unsigned long period_cycles, duty_cycles, prescale;
>  		u32 cr;
Lothar Waßmann - Aug. 24, 2011, 8:33 a.m.
Uwe Kleine-König writes:
> Hello Fabio,
> 
> On Tue, Aug 23, 2011 at 01:26:38PM -0300, Fabio Estevam wrote:
> > Instead of adding the check for MX53 CPU type, use the negative logic and let 
> > PWM to be configured for MX53 as well.
> I wonder why you prefer negative logic here. It might be shorter, but it
> is also less robust. For example it seems to me you not only added
> support for i.MX53 but also i.MX50.
> 
> The IMHO best solution would be to use device ids and get rid of
> cpu_is_mx.. completely.
> 
+1


Lothar Waßmann

Patch

diff --git a/arch/arm/plat-mxc/pwm.c b/arch/arm/plat-mxc/pwm.c
index 761c3c9..42d74ea 100644
--- a/arch/arm/plat-mxc/pwm.c
+++ b/arch/arm/plat-mxc/pwm.c
@@ -57,7 +57,7 @@  int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
 	if (pwm == NULL || period_ns == 0 || duty_ns > period_ns)
 		return -EINVAL;
 
-	if (cpu_is_mx27() || cpu_is_mx3() || cpu_is_mx25() || cpu_is_mx51()) {
+	if (!(cpu_is_mx1() || cpu_is_mx21())) {
 		unsigned long long c;
 		unsigned long period_cycles, duty_cycles, prescale;
 		u32 cr;