omap3evm: touchscreen delays on pm branch

omap3evm: touchscreen delays on pm branch

[Premi, Sanjeev]

Hi,

I am finding the response of touchscreen on the omap3evm very slow.

Here is my test:
On console, I run : watch -n2 "cat /proc/interrupts"
Then, I tap the touchscreen approximately once per second. However,
(usually) no interrupts are registered. As I increase the frequency
of 'taps' more and more interrupts are registered. But still not
matching exact taps.

However, when I keep the cpu busy with "cat /dev/zero > /dev/null &"
each tap is recognized.

EVM uses GPIO175 for touchscreen. I notice that Zoom2 uses GPIO102.
Both are not on BANK0.

Is the behavior same on ZOOM2 as well?

Best regards,
Sanjeev

Re: omap3evm: touchscreen delays on pm branch

[Hemanth V]

----- Original Message ----- 
From: "Premi, Sanjeev" <premi@ti.com>
To: <linux-omap@vger.kernel.org>
Sent: Thursday, November 26, 2009 10:32 PM
Subject: omap3evm: touchscreen delays on pm branch


> Hi,
> 
> I am finding the response of touchscreen on the omap3evm very slow.
> 
> Here is my test:
> On console, I run : watch -n2 "cat /proc/interrupts"
> Then, I tap the touchscreen approximately once per second. However,
> (usually) no interrupts are registered. As I increase the frequency
> of 'taps' more and more interrupts are registered. But still not
> matching exact taps.
> 
> However, when I keep the cpu busy with "cat /dev/zero > /dev/null &"
> each tap is recognized.
> 
> EVM uses GPIO175 for touchscreen. I notice that Zoom2 uses GPIO102.
> Both are not on BANK0.
> 
> Is the behavior same on ZOOM2 as well?

I dont see this problem on Zoom2/Zoom3, I am able to see interrupts being
incremented for every touch.

> 
> Best regards,
> Sanjeev
> --
> 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: omap3evm: touchscreen delays on pm branch

[Romit Dasgupta]

Premi, Sanjeev wrote:
> Hi,
> 
> I am finding the response of touchscreen on the omap3evm very slow.
> 
> Here is my test:
> On console, I run : watch -n2 "cat /proc/interrupts"
> Then, I tap the touchscreen approximately once per second. However,
> (usually) no interrupts are registered. As I increase the frequency
> of 'taps' more and more interrupts are registered. But still not
> matching exact taps.
> 
> However, when I keep the cpu busy with "cat /dev/zero > /dev/null &"
> each tap is recognized.
> 
Do you see this even if we don't enable OFF?

RE: omap3evm: touchscreen delays on pm branch

[Premi, Sanjeev]

> -----Original Message-----
> From: Dasgupta, Romit 
> Sent: Friday, November 27, 2009 2:10 PM
> To: Premi, Sanjeev
> Cc: linux-omap@vger.kernel.org
> Subject: Re: omap3evm: touchscreen delays on pm branch
> 
> Premi, Sanjeev wrote:
> > Hi,
> > 
> > I am finding the response of touchscreen on the omap3evm very slow.
> > 
> > Here is my test:
> > On console, I run : watch -n2 "cat /proc/interrupts"
> > Then, I tap the touchscreen approximately once per second. However,
> > (usually) no interrupts are registered. As I increase the frequency
> > of 'taps' more and more interrupts are registered. But still not
> > matching exact taps.
> > 
> > However, when I keep the cpu busy with "cat /dev/zero > /dev/null &"
> > each tap is recognized.
> > 
> Do you see this even if we don't enable OFF?
> 
Yes. Sleep_while_idle=0; enable_off_mode=0
~sanjeev

RE: omap3evm: touchscreen delays on pm branch

[Romit Dasgupta]

On Fri, 2009-11-27 at 14:11 +0530, Premi, Sanjeev wrote:
> > -----Original Message-----
> > From: Dasgupta, Romit 
> > Sent: Friday, November 27, 2009 2:10 PM
> > To: Premi, Sanjeev
> > Cc: linux-omap@vger.kernel.org
> > Subject: Re: omap3evm: touchscreen delays on pm branch
> > 
> > Premi, Sanjeev wrote:
> > > Hi,
> > > 
> > > I am finding the response of touchscreen on the omap3evm very slow.
> > > 
> > > Here is my test:
> > > On console, I run : watch -n2 "cat /proc/interrupts"
> > > Then, I tap the touchscreen approximately once per second. However,
> > > (usually) no interrupts are registered. As I increase the frequency
> > > of 'taps' more and more interrupts are registered. But still not
> > > matching exact taps.
> > > 
> > > However, when I keep the cpu busy with "cat /dev/zero > /dev/null &"
> > > each tap is recognized.
> > > 
> > Do you see this even if we don't enable OFF?
> > 
> Yes. Sleep_while_idle=0; enable_off_mode=0
> ~sanjeev

Hopefully you have the same TSC driver. Nevertheless, can you please try
this (just to see if clock domain idling is causing any problem or not):

diff --git a/arch/arm/mach-omap2/cpuidle34xx.c b/arch/arm/mach-omap2/cpuidle34xx.c
index 1cfa5a6..79710a1 100644
--- a/arch/arm/mach-omap2/cpuidle34xx.c
+++ b/arch/arm/mach-omap2/cpuidle34xx.c
@@ -248,7 +248,8 @@ void omap_init_power_states(void)
 			cpuidle_params_table[OMAP3_STATE_C2].threshold;
 	omap3_power_states[OMAP3_STATE_C2].mpu_state = PWRDM_POWER_ON;
 	omap3_power_states[OMAP3_STATE_C2].core_state = PWRDM_POWER_ON;
-	omap3_power_states[OMAP3_STATE_C2].flags = CPUIDLE_FLAG_TIME_VALID;
+	omap3_power_states[OMAP3_STATE_C2].flags = CPUIDLE_FLAG_TIME_VALID |
+				CPUIDLE_FLAG_CHECK_BM;
 
 	/* C3 . MPU CSWR + Core inactive */
 	omap3_power_states[OMAP3_STATE_C3].valid =

RE: omap3evm: touchscreen delays on pm branch

[Premi, Sanjeev]

> -----Original Message-----
> From: Dasgupta, Romit 
> Sent: Friday, November 27, 2009 2:38 PM
> To: Premi, Sanjeev
> Cc: linux-omap@vger.kernel.org
> Subject: RE: omap3evm: touchscreen delays on pm branch
> 
> On Fri, 2009-11-27 at 14:11 +0530, Premi, Sanjeev wrote:
> > > -----Original Message-----
> > > From: Dasgupta, Romit 
> > > Sent: Friday, November 27, 2009 2:10 PM
> > > To: Premi, Sanjeev
> > > Cc: linux-omap@vger.kernel.org
> > > Subject: Re: omap3evm: touchscreen delays on pm branch
> > > 
> > > Premi, Sanjeev wrote:
> > > > Hi,
> > > > 
> > > > I am finding the response of touchscreen on the 
> omap3evm very slow.
> > > > 
> > > > Here is my test:
> > > > On console, I run : watch -n2 "cat /proc/interrupts"
> > > > Then, I tap the touchscreen approximately once per 
> second. However,
> > > > (usually) no interrupts are registered. As I increase 
> the frequency
> > > > of 'taps' more and more interrupts are registered. But still not
> > > > matching exact taps.
> > > > 
> > > > However, when I keep the cpu busy with "cat /dev/zero > 
> /dev/null &"
> > > > each tap is recognized.
> > > > 
> > > Do you see this even if we don't enable OFF?
> > > 
> > Yes. Sleep_while_idle=0; enable_off_mode=0
> > ~sanjeev
> 
> Hopefully you have the same TSC driver. Nevertheless, can you 
> please try
> this (just to see if clock domain idling is causing any 
> problem or not):

It is the same driver at SDP3430. I had earlier tried removing
cpuidle altogether and did not see this issue. I too believe that
issue is caused by clocks being going to (auto)idle.

But then, Hemanth should be seeing the same behavior.

~sanjeev

> 
> diff --git a/arch/arm/mach-omap2/cpuidle34xx.c 
> b/arch/arm/mach-omap2/cpuidle34xx.c
> index 1cfa5a6..79710a1 100644
> --- a/arch/arm/mach-omap2/cpuidle34xx.c
> +++ b/arch/arm/mach-omap2/cpuidle34xx.c
> @@ -248,7 +248,8 @@ void omap_init_power_states(void)
>  			cpuidle_params_table[OMAP3_STATE_C2].threshold;
>  	omap3_power_states[OMAP3_STATE_C2].mpu_state = PWRDM_POWER_ON;
>  	omap3_power_states[OMAP3_STATE_C2].core_state = PWRDM_POWER_ON;
> -	omap3_power_states[OMAP3_STATE_C2].flags = 
> CPUIDLE_FLAG_TIME_VALID;
> +	omap3_power_states[OMAP3_STATE_C2].flags = 
> CPUIDLE_FLAG_TIME_VALID |
> +				CPUIDLE_FLAG_CHECK_BM;
>  
>  	/* C3 . MPU CSWR + Core inactive */
>  	omap3_power_states[OMAP3_STATE_C3].valid =
> 
> 

Re: omap3evm: touchscreen delays on pm branch

[Romit Dasgupta]

> 
> It is the same driver at SDP3430. I had earlier tried removing
> cpuidle altogether and did not see this issue. I too believe that
> issue is caused by clocks being going to (auto)idle.
> 
> But then, Hemanth should be seeing the same behavior.
Do you see PER powerdomain entering retention?

Re: omap3evm: touchscreen delays on pm branch

[Hemanth V]

----- Original Message ----- 
From: "Premi, Sanjeev" <premi@ti.com>
To: "Dasgupta, Romit" <romit@ti.com>
Cc: <linux-omap@vger.kernel.org>
Sent: Friday, November 27, 2009 3:20 PM
Subject: RE: omap3evm: touchscreen delays on pm branch


>> -----Original Message-----
>> From: Dasgupta, Romit 
>> Sent: Friday, November 27, 2009 2:38 PM
>> To: Premi, Sanjeev
>> Cc: linux-omap@vger.kernel.org
>> Subject: RE: omap3evm: touchscreen delays on pm branch
>> 
>> On Fri, 2009-11-27 at 14:11 +0530, Premi, Sanjeev wrote:
>> > > -----Original Message-----
>> > > From: Dasgupta, Romit 
>> > > Sent: Friday, November 27, 2009 2:10 PM
>> > > To: Premi, Sanjeev
>> > > Cc: linux-omap@vger.kernel.org
>> > > Subject: Re: omap3evm: touchscreen delays on pm branch
>> > > 
>> > > Premi, Sanjeev wrote:
>> > > > Hi,
>> > > > 
>> > > > I am finding the response of touchscreen on the 
>> omap3evm very slow.
>> > > > 
>> > > > Here is my test:
>> > > > On console, I run : watch -n2 "cat /proc/interrupts"
>> > > > Then, I tap the touchscreen approximately once per 
>> second. However,
>> > > > (usually) no interrupts are registered. As I increase 
>> the frequency
>> > > > of 'taps' more and more interrupts are registered. But still not
>> > > > matching exact taps.
>> > > > 
>> > > > However, when I keep the cpu busy with "cat /dev/zero > 
>> /dev/null &"
>> > > > each tap is recognized.
>> > > > 
>> > > Do you see this even if we don't enable OFF?
>> > > 
>> > Yes. Sleep_while_idle=0; enable_off_mode=0
>> > ~sanjeev
>> 
>> Hopefully you have the same TSC driver. Nevertheless, can you 
>> please try
>> this (just to see if clock domain idling is causing any 
>> problem or not):
> 
> It is the same driver at SDP3430. I had earlier tried removing
> cpuidle altogether and did not see this issue. I too believe that
> issue is caused by clocks being going to (auto)idle.
> 
> But then, Hemanth should be seeing the same behavior.

Zoom2/Zoom3 use a different touchscreen driver compared to SDP.
Its uses Synaptic Touchscreen over I2C.

Re: omap3evm: touchscreen delays on pm branch

[Sriram V]

Hi,
  Have you enabled CPUFREQ? We faced a similar issue and was due to
ondemand governor.
 Selecting performance governors solved the issue.


Regards,
sriram





On Fri, Nov 27, 2009 at 4:04 PM, Hemanth V <hemanthv@ti.com> wrote:
> ----- Original Message ----- From: "Premi, Sanjeev" <premi@ti.com>
> To: "Dasgupta, Romit" <romit@ti.com>
> Cc: <linux-omap@vger.kernel.org>
> Sent: Friday, November 27, 2009 3:20 PM
> Subject: RE: omap3evm: touchscreen delays on pm branch
>
>
>>> -----Original Message-----
>>> From: Dasgupta, Romit Sent: Friday, November 27, 2009 2:38 PM
>>> To: Premi, Sanjeev
>>> Cc: linux-omap@vger.kernel.org
>>> Subject: RE: omap3evm: touchscreen delays on pm branch
>>>
>>> On Fri, 2009-11-27 at 14:11 +0530, Premi, Sanjeev wrote:
>>> > > -----Original Message-----
>>> > > From: Dasgupta, Romit > > Sent: Friday, November 27, 2009 2:10 PM
>>> > > To: Premi, Sanjeev
>>> > > Cc: linux-omap@vger.kernel.org
>>> > > Subject: Re: omap3evm: touchscreen delays on pm branch
>>> > > > > Premi, Sanjeev wrote:
>>> > > > Hi,
>>> > > > > > > I am finding the response of touchscreen on the omap3evm very
>>> > > > > > > slow.
>>> > > > > > > Here is my test:
>>> > > > On console, I run : watch -n2 "cat /proc/interrupts"
>>> > > > Then, I tap the touchscreen approximately once per second. However,
>>> > > > (usually) no interrupts are registered. As I increase the frequency
>>> > > > of 'taps' more and more interrupts are registered. But still not
>>> > > > matching exact taps.
>>> > > > > > > However, when I keep the cpu busy with "cat /dev/zero >
>>> > > > > > > /dev/null &"
>>> > > > each tap is recognized.
>>> > > > > > Do you see this even if we don't enable OFF?
>>> > > > Yes. Sleep_while_idle=0; enable_off_mode=0
>>> > ~sanjeev
>>>
>>> Hopefully you have the same TSC driver. Nevertheless, can you please try
>>> this (just to see if clock domain idling is causing any problem or not):
>>
>> It is the same driver at SDP3430. I had earlier tried removing
>> cpuidle altogether and did not see this issue. I too believe that
>> issue is caused by clocks being going to (auto)idle.
>>
>> But then, Hemanth should be seeing the same behavior.
>
> Zoom2/Zoom3 use a different touchscreen driver compared to SDP.
> Its uses Synaptic Touchscreen over I2C.
> --
> 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
>
--
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: omap3evm: touchscreen delays on pm branch

[Premi, Sanjeev]

> -----Original Message-----
> From: Sriram V [mailto:vshrirama@gmail.com] 
> Sent: Tuesday, December 01, 2009 3:36 PM
> To: V, Hemanth
> Cc: Premi, Sanjeev; Dasgupta, Romit; linux-omap@vger.kernel.org
> Subject: Re: omap3evm: touchscreen delays on pm branch
> 
> Hi,
>   Have you enabled CPUFREQ? We faced a similar issue and was due to
> ondemand governor.
>  Selecting performance governors solved the issue.

Nope. No cpufreq. I did notice that CPIO175 was not being set
properly; have made local changes. It is a definite improvement;
but there are still very noticeable delays.

I will post the GPIO related patches in a few days; currently
drowned in few debug issues :(

Best regards,
Sanjeev

> 
> 
> Regards,
> sriram
> 
> 
> 
> 
> 
> On Fri, Nov 27, 2009 at 4:04 PM, Hemanth V <hemanthv@ti.com> wrote:
> > ----- Original Message ----- From: "Premi, Sanjeev" <premi@ti.com>
> > To: "Dasgupta, Romit" <romit@ti.com>
> > Cc: <linux-omap@vger.kernel.org>
> > Sent: Friday, November 27, 2009 3:20 PM
> > Subject: RE: omap3evm: touchscreen delays on pm branch
> >
> >
> >>> -----Original Message-----
> >>> From: Dasgupta, Romit Sent: Friday, November 27, 2009 2:38 PM
> >>> To: Premi, Sanjeev
> >>> Cc: linux-omap@vger.kernel.org
> >>> Subject: RE: omap3evm: touchscreen delays on pm branch
> >>>
> >>> On Fri, 2009-11-27 at 14:11 +0530, Premi, Sanjeev wrote:
> >>> > > -----Original Message-----
> >>> > > From: Dasgupta, Romit > > Sent: Friday, November 27, 
> 2009 2:10 PM
> >>> > > To: Premi, Sanjeev
> >>> > > Cc: linux-omap@vger.kernel.org
> >>> > > Subject: Re: omap3evm: touchscreen delays on pm branch
> >>> > > > > Premi, Sanjeev wrote:
> >>> > > > Hi,
> >>> > > > > > > I am finding the response of touchscreen on 
> the omap3evm very
> >>> > > > > > > slow.
> >>> > > > > > > Here is my test:
> >>> > > > On console, I run : watch -n2 "cat /proc/interrupts"
> >>> > > > Then, I tap the touchscreen approximately once per 
> second. However,
> >>> > > > (usually) no interrupts are registered. As I 
> increase the frequency
> >>> > > > of 'taps' more and more interrupts are registered. 
> But still not
> >>> > > > matching exact taps.
> >>> > > > > > > However, when I keep the cpu busy with "cat 
> /dev/zero >
> >>> > > > > > > /dev/null &"
> >>> > > > each tap is recognized.
> >>> > > > > > Do you see this even if we don't enable OFF?
> >>> > > > Yes. Sleep_while_idle=0; enable_off_mode=0
> >>> > ~sanjeev
> >>>
> >>> Hopefully you have the same TSC driver. Nevertheless, can 
> you please try
> >>> this (just to see if clock domain idling is causing any 
> problem or not):
> >>
> >> It is the same driver at SDP3430. I had earlier tried removing
> >> cpuidle altogether and did not see this issue. I too believe that
> >> issue is caused by clocks being going to (auto)idle.
> >>
> >> But then, Hemanth should be seeing the same behavior.
> >
> > Zoom2/Zoom3 use a different touchscreen driver compared to SDP.
> > Its uses Synaptic Touchscreen over I2C.
> > --
> > 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
> >
> --
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: omap3evm: touchscreen delays on pm branch

[Romit Dasgupta]

>>  Selecting performance governors solved the issue.
> 
> Nope. No cpufreq. I did notice that CPIO175 was not being set
> properly; have made local changes. It is a definite improvement;
> but there are still very noticeable delays.
> 

If PER Power domain is entering RET then module wakeups wont work! If at C2 you
are seeing this then it might explain.