FEATURES prints

FEATURES prints

[Nishanth Menon]

Folks,

With the addition of FEATURES in l-o, the following prints:
  - l2cache : Y
  - iva : Y
  - sgx : Y
  - neon : Y
  - isp : Y

comes up on SDP3430 -> now that we will introduce half a dozen features 
here and there, we will soon clutter this up. we should introduce a 
sysfs entry + remove the above noise..

any comments?

-- 
Regards,
Nishanth Menon

RE: FEATURES prints

[Premi, Sanjeev]

> -----Original Message-----
> From: Menon, Nishanth 
> Sent: Friday, October 09, 2009 12:43 AM
> To: Premi, Sanjeev; linux-omap@vger.kernel.org
> Subject: FEATURES prints
> 
> Folks,
> 
> With the addition of FEATURES in l-o, the following prints:
>   - l2cache : Y
>   - iva : Y
>   - sgx : Y
>   - neon : Y
>   - isp : Y
> 
> comes up on SDP3430 -> now that we will introduce half a 
> dozen features 

[sp] The intention here is to print the features that help in
     identifying the processor. Not really the board/EVM.
     Even if there are more features, we decide which ones to
     show.

> here and there, we will soon clutter this up. we should introduce a 
> sysfs entry + remove the above noise..

[sp] I believe the sysfs would be good for board level features.

> 
> any comments?
> 
> -- 
> Regards,
> Nishanth Menon
> 

Re: FEATURES prints

[Nishanth Menon]

Premi, Sanjeev had written, on 10/09/2009 05:46 AM, the following:
>> -----Original Message-----
>> From: Menon, Nishanth 
>> Sent: Friday, October 09, 2009 12:43 AM
>> To: Premi, Sanjeev; linux-omap@vger.kernel.org
>> Subject: FEATURES prints
>>
>> Folks,
>>
>> With the addition of FEATURES in l-o, the following prints:
>>   - l2cache : Y
>>   - iva : Y
>>   - sgx : Y
>>   - neon : Y
>>   - isp : Y
>>
>> comes up on SDP3430 -> now that we will introduce half a 
>> dozen features 
> 
> [sp] The intention here is to print the features that help in
>      identifying the processor. Not really the board/EVM.
>      Even if there are more features, we decide which ones to
>      show.
this still a bunch of noise which is unusable for applications -> e.g. 
if an app wanted to use ISP, how would it do it?


-- 
Regards,
Nishanth Menon

RE: FEATURES prints

[Premi, Sanjeev]

> -----Original Message-----
> From: Menon, Nishanth 
> Sent: Friday, October 09, 2009 7:45 PM
> To: Premi, Sanjeev
> Cc: linux-omap@vger.kernel.org
> Subject: Re: FEATURES prints
> 
> Premi, Sanjeev had written, on 10/09/2009 05:46 AM, the following:
> >> -----Original Message-----
> >> From: Menon, Nishanth 
> >> Sent: Friday, October 09, 2009 12:43 AM
> >> To: Premi, Sanjeev; linux-omap@vger.kernel.org
> >> Subject: FEATURES prints
> >>
> >> Folks,
> >>
> >> With the addition of FEATURES in l-o, the following prints:
> >>   - l2cache : Y
> >>   - iva : Y
> >>   - sgx : Y
> >>   - neon : Y
> >>   - isp : Y
> >>
> >> comes up on SDP3430 -> now that we will introduce half a 
> >> dozen features 
> > 
> > [sp] The intention here is to print the features that help in
> >      identifying the processor. Not really the board/EVM.
> >      Even if there are more features, we decide which ones to
> >      show.
> this still a bunch of noise which is unusable for 
> applications -> e.g. 
> if an app wanted to use ISP, how would it do it?

[sp] At this stage, they help in identifying the variants in the OMAP35x family.

~sanjeev

> 
> 
> -- 
> Regards,
> Nishanth Menon
> 

Re: FEATURES prints

[Kevin Hilman]

Nishanth Menon <nm@ti.com> writes:

> Folks,
>
> With the addition of FEATURES in l-o, the following prints:
>  - l2cache : Y
>  - iva : Y
>  - sgx : Y
>  - neon : Y
>  - isp : Y
>
> comes up on SDP3430 -> now that we will introduce half a dozen
> features here and there, we will soon clutter this up. we should
> introduce a sysfs entry + remove the above noise..
>

Like Nishanth, I don't like the multi-line noise here.  The patch
below results in a single line output like this instead

OMAP3430/3530 ES3.0 (l2cache iva sgx neon isp )

Not sure why we need to dump features that are not there, but if that
s considered important, maybe prefixing each feature with a '+' or '-'
would still allow this to be collapsed into a single line.

Even with this, I think adding the display of these features into an
OMAP-specific section of /proc/cpuinfo would be even better.

Comments?

Kevin


commit 24f7422bad970cea2ed71d71e3994eeed70f175f
Author: Kevin Hilman <khilman@deeprootsystems.com>
Date:   Tue Oct 13 14:42:00 2009 -0700

    OMAP3: collapse chip feature prints to single line
    
    Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com>

diff --git a/arch/arm/mach-omap2/id.c b/arch/arm/mach-omap2/id.c
index 71d5568..b90fcf1 100644
--- a/arch/arm/mach-omap2/id.c
+++ b/arch/arm/mach-omap2/id.c
@@ -249,11 +249,8 @@ void __init omap3_check_revision(void)
 }
 
 #define OMAP3_SHOW_FEATURE(feat)		\
-	if (omap3_has_ ##feat()) {		\
-		pr_info (" - "#feat" : Y");	\
-	} else {				\
-		pr_info (" - "#feat" : N");	\
-	}
+	if (omap3_has_ ##feat())		\
+		printk (#feat" ");      	\
 
 void __init omap3_cpuinfo(void)
 {
@@ -307,13 +304,14 @@ void __init omap3_cpuinfo(void)
 	/*
 	 * Print verbose information
 	 */
-	pr_info("OMAP%s ES%s\n", cpu_name, cpu_rev);
+	pr_info("OMAP%s ES%s (", cpu_name, cpu_rev);
 
 	OMAP3_SHOW_FEATURE(l2cache);
 	OMAP3_SHOW_FEATURE(iva);
 	OMAP3_SHOW_FEATURE(sgx);
 	OMAP3_SHOW_FEATURE(neon);
 	OMAP3_SHOW_FEATURE(isp);
+	printk(")\n");
 }
 
 /*

RE: FEATURES prints

[Premi, Sanjeev]

> -----Original Message-----
> From: Kevin Hilman [mailto:khilman@deeprootsystems.com] 
> Sent: Wednesday, October 14, 2009 3:17 AM
> To: Menon, Nishanth
> Cc: Premi, Sanjeev; linux-omap@vger.kernel.org
> Subject: Re: FEATURES prints
> 
> Nishanth Menon <nm@ti.com> writes:
> 
> > Folks,
> >
> > With the addition of FEATURES in l-o, the following prints:
> >  - l2cache : Y
> >  - iva : Y
> >  - sgx : Y
> >  - neon : Y
> >  - isp : Y
> >
> > comes up on SDP3430 -> now that we will introduce half a dozen
> > features here and there, we will soon clutter this up. we should
> > introduce a sysfs entry + remove the above noise..
> >
> 
> Like Nishanth, I don't like the multi-line noise here.  The patch
> below results in a single line output like this instead
> 
> OMAP3430/3530 ES3.0 (l2cache iva sgx neon isp )
> 
> Not sure why we need to dump features that are not there, but if that
> s considered important, maybe prefixing each feature with a '+' or '-'
> would still allow this to be collapsed into a single line.
> 
> Even with this, I think adding the display of these features into an
> OMAP-specific section of /proc/cpuinfo would be even better.
> 

[sp] Single line prints look good. We can also add details in cpuinfo.

~sanjeev

> Comments?
> 
> Kevin
> 
> 
> commit 24f7422bad970cea2ed71d71e3994eeed70f175f
> Author: Kevin Hilman <khilman@deeprootsystems.com>
> Date:   Tue Oct 13 14:42:00 2009 -0700
> 
>     OMAP3: collapse chip feature prints to single line
>     
>     Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com>
> 
> diff --git a/arch/arm/mach-omap2/id.c b/arch/arm/mach-omap2/id.c
> index 71d5568..b90fcf1 100644
> --- a/arch/arm/mach-omap2/id.c
> +++ b/arch/arm/mach-omap2/id.c
> @@ -249,11 +249,8 @@ void __init omap3_check_revision(void)
>  }
>  
>  #define OMAP3_SHOW_FEATURE(feat)		\
> -	if (omap3_has_ ##feat()) {		\
> -		pr_info (" - "#feat" : Y");	\
> -	} else {				\
> -		pr_info (" - "#feat" : N");	\
> -	}
> +	if (omap3_has_ ##feat())		\
> +		printk (#feat" ");      	\
>  
>  void __init omap3_cpuinfo(void)
>  {
> @@ -307,13 +304,14 @@ void __init omap3_cpuinfo(void)
>  	/*
>  	 * Print verbose information
>  	 */
> -	pr_info("OMAP%s ES%s\n", cpu_name, cpu_rev);
> +	pr_info("OMAP%s ES%s (", cpu_name, cpu_rev);
>  
>  	OMAP3_SHOW_FEATURE(l2cache);
>  	OMAP3_SHOW_FEATURE(iva);
>  	OMAP3_SHOW_FEATURE(sgx);
>  	OMAP3_SHOW_FEATURE(neon);
>  	OMAP3_SHOW_FEATURE(isp);
> +	printk(")\n");
>  }
>  
>  /*
> 
> 

RE: FEATURES prints

[Shilimkar, Santosh]

Sanjeev,

> -----Original Message-----
> From: linux-omap-owner@vger.kernel.org [mailto:linux-omap-
> owner@vger.kernel.org] On Behalf Of Premi, Sanjeev
> Sent: Wednesday, October 14, 2009 3:26 AM
> To: Kevin Hilman; Menon, Nishanth
> Cc: linux-omap@vger.kernel.org
> Subject: RE: FEATURES prints
> 
> 
> > -----Original Message-----
> > From: Kevin Hilman [mailto:khilman@deeprootsystems.com]
> > Sent: Wednesday, October 14, 2009 3:17 AM
> > To: Menon, Nishanth
> > Cc: Premi, Sanjeev; linux-omap@vger.kernel.org
> > Subject: Re: FEATURES prints
> >
> > Nishanth Menon <nm@ti.com> writes:
> >
> > > Folks,
> > >
> > > With the addition of FEATURES in l-o, the following prints:
> > >  - l2cache : Y
> > >  - iva : Y
> > >  - sgx : Y
> > >  - neon : Y
> > >  - isp : Y
> > >
> > > comes up on SDP3430 -> now that we will introduce half a dozen
> > > features here and there, we will soon clutter this up. we should
> > > introduce a sysfs entry + remove the above noise..
> > >
> >
> > Like Nishanth, I don't like the multi-line noise here.  The patch
> > below results in a single line output like this instead
> >
> > OMAP3430/3530 ES3.0 (l2cache iva sgx neon isp )
> >
> > Not sure why we need to dump features that are not there, but if that
> > s considered important, maybe prefixing each feature with a '+' or '-'
> > would still allow this to be collapsed into a single line.
> >
> > Even with this, I think adding the display of these features into an
> > OMAP-specific section of /proc/cpuinfo would be even better.
> >
> 
> [sp] Single line prints look good. We can also add details in cpuinfo.

If you are ok with proc entry then we don't even need this noise at all in the boot. It's just that adding proc entries is discouraged to some extent.

Regards,
Santosh

Re: FEATURES prints

[Kevin Hilman]

Shilimkar, Santosh wrote:
> Sanjeev,
> 
>> -----Original Message-----
>> From: linux-omap-owner@vger.kernel.org [mailto:linux-omap-
>> owner@vger.kernel.org] On Behalf Of Premi, Sanjeev
>> Sent: Wednesday, October 14, 2009 3:26 AM
>> To: Kevin Hilman; Menon, Nishanth
>> Cc: linux-omap@vger.kernel.org
>> Subject: RE: FEATURES prints
>>
>>
>>> -----Original Message-----
>>> From: Kevin Hilman [mailto:khilman@deeprootsystems.com]
>>> Sent: Wednesday, October 14, 2009 3:17 AM
>>> To: Menon, Nishanth
>>> Cc: Premi, Sanjeev; linux-omap@vger.kernel.org
>>> Subject: Re: FEATURES prints
>>>
>>> Nishanth Menon <nm@ti.com> writes:
>>>
>>>> Folks,
>>>>
>>>> With the addition of FEATURES in l-o, the following prints:
>>>>  - l2cache : Y
>>>>  - iva : Y
>>>>  - sgx : Y
>>>>  - neon : Y
>>>>  - isp : Y
>>>>
>>>> comes up on SDP3430 -> now that we will introduce half a dozen
>>>> features here and there, we will soon clutter this up. we should
>>>> introduce a sysfs entry + remove the above noise..
>>>>
>>> Like Nishanth, I don't like the multi-line noise here.  The patch
>>> below results in a single line output like this instead
>>>
>>> OMAP3430/3530 ES3.0 (l2cache iva sgx neon isp )
>>>
>>> Not sure why we need to dump features that are not there, but if that
>>> s considered important, maybe prefixing each feature with a '+' or '-'
>>> would still allow this to be collapsed into a single line.
>>>
>>> Even with this, I think adding the display of these features into an
>>> OMAP-specific section of /proc/cpuinfo would be even better.
>>>
>> [sp] Single line prints look good. We can also add details in cpuinfo.
> 
> If you are ok with proc entry then we don't even need this noise at all in the boot. It's just that adding proc entries is discouraged to some extent.

FWIW, I was not thinking adding a new proc entry.  I was thinking of 
extending  /proc/cpuinfo with some platform specific entries.

Kevin

Re: FEATURES prints

[Tony Lindgren]

* Kevin Hilman <khilman@deeprootsystems.com> [091013 15:01]:
> Nishanth Menon <nm@ti.com> writes:
> 
> > Folks,
> >
> > With the addition of FEATURES in l-o, the following prints:
> >  - l2cache : Y
> >  - iva : Y
> >  - sgx : Y
> >  - neon : Y
> >  - isp : Y
> >
> > comes up on SDP3430 -> now that we will introduce half a dozen
> > features here and there, we will soon clutter this up. we should
> > introduce a sysfs entry + remove the above noise..
> >
> 
> Like Nishanth, I don't like the multi-line noise here.  The patch
> below results in a single line output like this instead
> 
> OMAP3430/3530 ES3.0 (l2cache iva sgx neon isp )
> 
> Not sure why we need to dump features that are not there, but if that
> s considered important, maybe prefixing each feature with a '+' or '-'
> would still allow this to be collapsed into a single line.
> 
> Even with this, I think adding the display of these features into an
> OMAP-specific section of /proc/cpuinfo would be even better.
> 
> Comments?

I like. Looks like this patch needs to be refreshed.

Tony
 
> Kevin
> 
> 
> commit 24f7422bad970cea2ed71d71e3994eeed70f175f
> Author: Kevin Hilman <khilman@deeprootsystems.com>
> Date:   Tue Oct 13 14:42:00 2009 -0700
> 
>     OMAP3: collapse chip feature prints to single line
>     
>     Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com>
> 
> diff --git a/arch/arm/mach-omap2/id.c b/arch/arm/mach-omap2/id.c
> index 71d5568..b90fcf1 100644
> --- a/arch/arm/mach-omap2/id.c
> +++ b/arch/arm/mach-omap2/id.c
> @@ -249,11 +249,8 @@ void __init omap3_check_revision(void)
>  }
>  
>  #define OMAP3_SHOW_FEATURE(feat)		\
> -	if (omap3_has_ ##feat()) {		\
> -		pr_info (" - "#feat" : Y");	\
> -	} else {				\
> -		pr_info (" - "#feat" : N");	\
> -	}
> +	if (omap3_has_ ##feat())		\
> +		printk (#feat" ");      	\
>  
>  void __init omap3_cpuinfo(void)
>  {
> @@ -307,13 +304,14 @@ void __init omap3_cpuinfo(void)
>  	/*
>  	 * Print verbose information
>  	 */
> -	pr_info("OMAP%s ES%s\n", cpu_name, cpu_rev);
> +	pr_info("OMAP%s ES%s (", cpu_name, cpu_rev);
>  
>  	OMAP3_SHOW_FEATURE(l2cache);
>  	OMAP3_SHOW_FEATURE(iva);
>  	OMAP3_SHOW_FEATURE(sgx);
>  	OMAP3_SHOW_FEATURE(neon);
>  	OMAP3_SHOW_FEATURE(isp);
> +	printk(")\n");
>  }
>  
>  /*
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: FEATURES prints

[Kevin Hilman]

Tony Lindgren <tony@atomide.com> writes:

> * Kevin Hilman <khilman@deeprootsystems.com> [091013 15:01]:
>> Nishanth Menon <nm@ti.com> writes:
>> 
>> > Folks,
>> >
>> > With the addition of FEATURES in l-o, the following prints:
>> >  - l2cache : Y
>> >  - iva : Y
>> >  - sgx : Y
>> >  - neon : Y
>> >  - isp : Y
>> >
>> > comes up on SDP3430 -> now that we will introduce half a dozen
>> > features here and there, we will soon clutter this up. we should
>> > introduce a sysfs entry + remove the above noise..
>> >
>> 
>> Like Nishanth, I don't like the multi-line noise here.  The patch
>> below results in a single line output like this instead
>> 
>> OMAP3430/3530 ES3.0 (l2cache iva sgx neon isp )
>> 
>> Not sure why we need to dump features that are not there, but if that
>> s considered important, maybe prefixing each feature with a '+' or '-'
>> would still allow this to be collapsed into a single line.
>> 
>> Even with this, I think adding the display of these features into an
>> OMAP-specific section of /proc/cpuinfo would be even better.
>> 
>> Comments?
>
> I like. Looks like this patch needs to be refreshed.
>

OK, will send refreshed version.

Kevin