Re: Issues with ondemand governor

2010-11-23 Thread Vishwanath Sripathy 
Thanks David for the inputs.
I tried your patch. In addition to that I reduced transition_latency.
With these 2 changes, I do see much better results (worst case
performance of ondemand is 88%).

Vishwa


On Mon, Nov 22, 2010 at 9:39 PM, David C Niemi  wrote:
> The general problem here is that the ondemand governor is aimed more at
> power savings than performance.  In cases where the ondemand governor
> performs worse than the performance governor, the "sampling_down_factor"
> tunable is often useful.  I submitted the patch to add this tunable a
> few weeks ago and it was acked by Venki, but I don't know what happened
> to it after that.  It helps in two ways:
>
> 1) the governor does not spend as much overhead on the governor when the
> CPU is truly busy
>
> 2) the governor is a lot less eager to downshift when the CPU is busy --
> without this patch, even on a busy system ondemand will blip down in
> clock speed surprisingly often, hurting performance.
>
> This patch is all about improving peak load performance.  On quite a few
> loads I've tried this patch with a sampling_down_factor of 100 matches
> the performance governor quite well while the original ondemand
> performance was poor.  On the other hand, it is not much help if you are
> trying to minimize power consumption on light to medium loads.  If you
> set sampling_down_factor to "1" it preserves default behavior.
>
> David C Niemi
>
> Vishwanath Sripathy wrote:
>> Hi,
>>
>> I was trying to investigate performance issues that we were seeing
>> with some usecases like Video playback on OMAP Platforms with ondemand
>> governor.
>> As part of this, I found a tool called cpufreq-bench
>> (http://lwn.net/Articles/339862) which can be used determine the
>> performance impact of ondemand governor compared to performacne
>> governor.
>> When I ran this tool on OMAP3 (ZOOM3) platform using 2.6.36 kernel
>> with below command, the worstcase ondemand performance is 35% compared
>> to performance governor.
>> cpufreq-bench -l 5 -s 10 -x 5 -y 10 -g ondemand -r 5 -n 5 -v
>>
>> I tried the same on x86 platforms and there the worstcase performance
>> is around 88%.
>> Attached are the cpufreq-bench logs for x86 and omap3.
>>
>> Questions:
>> 1. Is this is known limitaiton of ondemand governor?
>> 2. How do we support system usecases (like video playback etc) with
>> ondemand governor if governor is not able to scale the frequencies in
>> realtime? Are applications expected to play with scaling_min_freq to
>> increase mpu frequency?
>>
>> Regards
>> Vishwa
>
>

___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Re: New problem with BeagleBoard

2010-11-23 Thread Ira Rosen 
On 22 November 2010 21:07, Michael Hudson  wrote:
> On Mon, 22 Nov 2010 15:09:29 +0200, Ira Rosen  wrote:
>> On 22 November 2010 15:02, Anand Gadiyar  wrote:
>> >
>> > If it's the XM (as your bootloader claims), then there is no on-board NAND.
>> > Your SD card is the only storage, and you should be able to recover by
>> > reformatting and updating the images on the SD card.
>>
>> I tried exactly that, but it didn't work.
>
> Do you have another SD card you can try?  They rot and die,
> unfortunately :/

After redoing everything once again with the newer packages, I tend to
think that the problem is indeed in the card. I don't get "Reading
from mmc" message, and if I press the "user" button, it doesn't
respond at all. I'll try to get another SD card.

Thanks,
Ira

>
> Cheers,
> mwh
>

___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Re: Issues with ondemand governor

2010-11-23 Thread Amit Kucheria 
Vishwa,

Have you had a chance to do some usetime tests with these changes?

It would be interesting to measure the power consumption with and
without these changes.

/Amit

On Tue, Nov 23, 2010 at 5:59 PM, Vishwanath Sripathy
 wrote:
> Thanks David for the inputs.
> I tried your patch. In addition to that I reduced transition_latency.
> With these 2 changes, I do see much better results (worst case
> performance of ondemand is 88%).
>
> Vishwa
>
>
> On Mon, Nov 22, 2010 at 9:39 PM, David C Niemi  wrote:
>> The general problem here is that the ondemand governor is aimed more at
>> power savings than performance.  In cases where the ondemand governor
>> performs worse than the performance governor, the "sampling_down_factor"
>> tunable is often useful.  I submitted the patch to add this tunable a
>> few weeks ago and it was acked by Venki, but I don't know what happened
>> to it after that.  It helps in two ways:
>>
>> 1) the governor does not spend as much overhead on the governor when the
>> CPU is truly busy
>>
>> 2) the governor is a lot less eager to downshift when the CPU is busy --
>> without this patch, even on a busy system ondemand will blip down in
>> clock speed surprisingly often, hurting performance.
>>
>> This patch is all about improving peak load performance.  On quite a few
>> loads I've tried this patch with a sampling_down_factor of 100 matches
>> the performance governor quite well while the original ondemand
>> performance was poor.  On the other hand, it is not much help if you are
>> trying to minimize power consumption on light to medium loads.  If you
>> set sampling_down_factor to "1" it preserves default behavior.
>>
>> David C Niemi
>>
>> Vishwanath Sripathy wrote:
>>> Hi,
>>>
>>> I was trying to investigate performance issues that we were seeing
>>> with some usecases like Video playback on OMAP Platforms with ondemand
>>> governor.
>>> As part of this, I found a tool called cpufreq-bench
>>> (http://lwn.net/Articles/339862) which can be used determine the
>>> performance impact of ondemand governor compared to performacne
>>> governor.
>>> When I ran this tool on OMAP3 (ZOOM3) platform using 2.6.36 kernel
>>> with below command, the worstcase ondemand performance is 35% compared
>>> to performance governor.
>>> cpufreq-bench -l 5 -s 10 -x 5 -y 10 -g ondemand -r 5 -n 5 -v
>>>
>>> I tried the same on x86 platforms and there the worstcase performance
>>> is around 88%.
>>> Attached are the cpufreq-bench logs for x86 and omap3.
>>>
>>> Questions:
>>> 1. Is this is known limitaiton of ondemand governor?
>>> 2. How do we support system usecases (like video playback etc) with
>>> ondemand governor if governor is not able to scale the frequencies in
>>> realtime? Are applications expected to play with scaling_min_freq to
>>> increase mpu frequency?
>>>
>>> Regards
>>> Vishwa
>>
>>
>
> ___
> linaro-dev mailing list
> linaro-dev@lists.linaro.org
> http://lists.linaro.org/mailman/listinfo/linaro-dev
>

___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Re: omap4 panda ehci failure

2010-11-23 Thread Anand Gadiyar 
On 23 November 2010 11:38, John Rigby  wrote:

> All omap4 experts,
>
> I have a packaged kernel for omap3/omap4 that boots to a shell prompt
> on a Panda and a Beagle XM.  One problem on Panda is that I have not
> network.  When I looked at the dmesg I noticed that the ehci driver
> failed to initialize:
>
> [0.761352] ehci_hcd: USB 2.0 'Enhanced' Host Controller (EHCI) Driver
> [0.768280] ehci-omap: probe of ehci-omap.0 failed with error -2
>
> My kernel is linux-linaro-2.6.36 from Nicolas plus some Ubuntu sauce
> and  packaging.
>
> Here is my packaged kernel:
> http://git.linaro.org/gitweb?p=ubuntu/linux-linaro-natty.git;a=summary
>
> Here is linux-linaro-2.6.36
> http://git.linaro.org/gitweb?p=kernel/linux-linaro-2.6.36.git;a=summary
>
> and attached is my config.
>
> If anyone knows what is wrong with ehci it would be greatly appreciated.
>
>
Well, the EHCI driver isn't upstream for one. I only posted it out this
weekend. [1]

I'll see if I can rebase against linux-linaro-2.6.36 and send a pull
request.
There are a couple of minor issues with the original series (including
me forgetting to add a patch description for one patch). I'll fix those
up while I'm at it.

Do I need to post the patches here for review first? Or is it sufficient
to point to a git branch for people to take a look?




> One more thing, my pandaboard is an old black 6 layer version.
>
>
Is it an ES1.0 board or an ES2.0 board? (It's ES1.0 if there's a black
USB cable running along the bottom of the board).

If it's ES2.0, the 6 layer board should work.

- Anand

[1] http://marc.info/?l=linux-usb&m=129036484509328&w=2
___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Re: omap4 panda ehci failure

2010-11-23 Thread John Rigby 
Anand,

Thanks for the info.

On Tue, Nov 23, 2010 at 9:12 AM, Anand Gadiyar  wrote:
> On 23 November 2010 11:38, John Rigby  wrote:
>>
>> All omap4 experts,
>>
>> I have a packaged kernel for omap3/omap4 that boots to a shell prompt
>> on a Panda and a Beagle XM.  One problem on Panda is that I have not
>> network.  When I looked at the dmesg I noticed that the ehci driver
>> failed to initialize:
>>
>> [    0.761352] ehci_hcd: USB 2.0 'Enhanced' Host Controller (EHCI) Driver
>> [    0.768280] ehci-omap: probe of ehci-omap.0 failed with error -2
>>
>> My kernel is linux-linaro-2.6.36 from Nicolas plus some Ubuntu sauce
>> and  packaging.
>>
>> Here is my packaged kernel:
>> http://git.linaro.org/gitweb?p=ubuntu/linux-linaro-natty.git;a=summary
>>
>> Here is linux-linaro-2.6.36
>> http://git.linaro.org/gitweb?p=kernel/linux-linaro-2.6.36.git;a=summary
>>
>> and attached is my config.
>>
>> If anyone knows what is wrong with ehci it would be greatly appreciated.
>>
>
> Well, the EHCI driver isn't upstream for one. I only posted it out this
> weekend. [1]
Well that makes sense.  I assumed that since the panda board file has
calls to initialize ehci that it was supposed to work.
>
> I'll see if I can rebase against linux-linaro-2.6.36 and send a pull
> request.
> There are a couple of minor issues with the original series (including
> me forgetting to add a patch description for one patch). I'll fix those
> up while I'm at it.
>
> Do I need to post the patches here for review first? Or is it sufficient
> to point to a git branch for people to take a look?
That is a question of Nicolas.  I think the basic rule is that
linux-linaro-2.6.xx is supposed to have things that are headed
upstream.
>
>
>
>>
>> One more thing, my pandaboard is an old black 6 layer version.
>>
>
> Is it an ES1.0 board or an ES2.0 board? (It's ES1.0 if there's a black
> USB cable running along the bottom of the board).
>
> If it's ES2.0, the 6 layer board should work.
No USB cable so I guess it should work.
>
> - Anand
>
> [1] http://marc.info/?l=linux-usb&m=129036484509328&w=2
>
>
>

___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Re: omap4 panda ehci failure

2010-11-23 Thread Anand Gadiyar 
On 23 November 2010 21:42, Anand Gadiyar  wrote:

> On 23 November 2010 11:38, John Rigby  wrote:
>
>> All omap4 experts,
>>
>> I have a packaged kernel for omap3/omap4 that boots to a shell prompt
>> on a Panda and a Beagle XM.  One problem on Panda is that I have not
>> network.  When I looked at the dmesg I noticed that the ehci driver
>> failed to initialize:
>>
>> [0.761352] ehci_hcd: USB 2.0 'Enhanced' Host Controller (EHCI) Driver
>> [0.768280] ehci-omap: probe of ehci-omap.0 failed with error -2
>>
>> My kernel is linux-linaro-2.6.36 from Nicolas plus some Ubuntu sauce
>> and  packaging.
>>
>> Here is my packaged kernel:
>> http://git.linaro.org/gitweb?p=ubuntu/linux-linaro-natty.git;a=summary
>>
>> Here is linux-linaro-2.6.36
>> http://git.linaro.org/gitweb?p=kernel/linux-linaro-2.6.36.git;a=summary
>>
>> and attached is my config.
>>
>> If anyone knows what is wrong with ehci it would be greatly appreciated.
>>
>>
> Well, the EHCI driver isn't upstream for one. I only posted it out this
> weekend. [1]
>
> I'll see if I can rebase against linux-linaro-2.6.36 and send a pull
> request.
> There are a couple of minor issues with the original series (including
> me forgetting to add a patch description for one patch). I'll fix those
> up while I'm at it.
>
> Do I need to post the patches here for review first? Or is it sufficient
> to point to a git branch for people to take a look?
>

...


> [1] http://marc.info/?l=linux-usb&m=129036484509328&w=2
>
>
>
I just checked - the patches I posted apply cleanly against
linux-linaro-2.6.36. Patch 3 has some line offsets, but otherwise
applies cleanly. EHCI works on Panda after applying these
patches, however I didn't get a chance to check if EHCI on
OMAP3 is still working.

The series (against v2.6.37-rc2) is available here:
git://dev.omapzoom.org/pub/scm/anand/linux-omap-usb.git
in the omap4-ehci-upstream-v1 branch.

- Anand
___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Re: omap4 panda ehci failure

2010-11-23 Thread Anand Gadiyar 
On 23 November 2010 22:17, John Rigby  wrote:
>
> On Tue, Nov 23, 2010 at 9:12 AM, Anand Gadiyar  
> wrote:
> > On 23 November 2010 11:38, John Rigby  wrote:
> > I'll see if I can rebase against linux-linaro-2.6.36 and send a pull
> > request.
> > There are a couple of minor issues with the original series (including
> > me forgetting to add a patch description for one patch). I'll fix those
> > up while I'm at it.
> >
> > Do I need to post the patches here for review first? Or is it sufficient
> > to point to a git branch for people to take a look?
> That is a question of Nicolas.  I think the basic rule is that
> linux-linaro-2.6.xx is supposed to have things that are headed
> upstream.

Okay - I haven't received any review comments yet - but that could be
because most people are out due to holidays this week.

I'll wait until they're queued up in the linux-usb tree, and then post that
version here then.

- Anand

___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Re: omap4 panda ehci failure

2010-11-23 Thread Nicolas Pitre 
On Tue, 23 Nov 2010, John Rigby wrote:

> On Tue, Nov 23, 2010 at 9:12 AM, Anand Gadiyar  
> wrote:
> > Do I need to post the patches here for review first? Or is it sufficient
> > to point to a git branch for people to take a look?
> That is a question of Nicolas.  I think the basic rule is that
> linux-linaro-2.6.xx is supposed to have things that are headed
> upstream.

Preferably, if you can post it upstream (e.g. on the USB mailing list) 
and you get positive feedback there then I'll be happy to merge it.

If this is stuff that is intended to be in upstream someday but is not 
in a shape for it yet then we need to discuss.  I would be inclined to 
merge it to give it exposure as long as it is reasonably certain that 
nothing else might break due to that merge.


Nicolas

___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Re: omap4 panda ehci failure

2010-11-23 Thread Nicolas Pitre 
On Tue, 23 Nov 2010, Anand Gadiyar wrote:

> Okay - I haven't received any review comments yet - but that could be
> because most people are out due to holidays this week.

Either that, or there is nothing blatently controvertial about those 
patches.

> I'll wait until they're queued up in the linux-usb tree, and then post that
> version here then.

If you get them accepted in the linux-usb tree then I don't think you'd 
need to repost them again here.


Nicolas

___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Re: omap4 panda ehci failure

2010-11-23 Thread Nicolas Pitre 
On Tue, 23 Nov 2010, Anand Gadiyar wrote:

> I just checked - the patches I posted apply cleanly against
> linux-linaro-2.6.36. Patch 3 has some line offsets, but otherwise
> applies cleanly. EHCI works on Panda after applying these
> patches, however I didn't get a chance to check if EHCI on
> OMAP3 is still working.

If you can do that (check that nothing else is broken) then I'll be 
happy to merge them.


Nicolas

___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Re: omap4 panda ehci failure

2010-11-23 Thread John Rigby 
On Tue, Nov 23, 2010 at 11:31 AM, Nicolas Pitre
 wrote:
> On Tue, 23 Nov 2010, Anand Gadiyar wrote:
>
>> I just checked - the patches I posted apply cleanly against
>> linux-linaro-2.6.36. Patch 3 has some line offsets, but otherwise
>> applies cleanly. EHCI works on Panda after applying these
>> patches, however I didn't get a chance to check if EHCI on
>> OMAP3 is still working.
>
> If you can do that (check that nothing else is broken) then I'll be
> happy to merge them.
>
I checked these on my Beagle xM the usb seems to work (usb keyboard
works, and ifconfig shows usb0 interface).

John

___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

Minutes & actions: Kernel Working Group meeting of Nov 22, 2010

2010-11-23 Thread Mounir Bsaibes 
Greetings,

Enclosed you'll find a link to the agenda, minutes and actions from the
Linaro kernel working group weekly meeting of  November 22nd, 2010.

https://wiki.linaro.org/WorkingGroups/KernelConsolidation/Meetings/2010-11-22

Regards,
Mounir


___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev

[RFC] thermal sensor drivers under drivers/hwmon

2010-11-23 Thread Yong Shen 
Hi there,

In power management group, we have a working item of exposing thermal
information to user space through sysfs. However, so far, the thermal
sensor drivers under 'drivers/hwmon' expose their information in
various nodes under sysfs, which makes information collection
difficult.
Intel's thermal framework is located under 'drivers/thermal', but all
the thermal drivers under hwmon don't make use of it due to some
reason (part of the reason for this situation may be that intel's
thermal framework is comparatively new therefore when those sensor
drivers were implemented they did not have a framework to register
themselves.)
We like to find out a simple and unified way to expose thermal related
information, intel's thermal interface could be a choice. Using
thermal framework is straightforward, below is a sample driver for
this purpose. Thus all the information goes to /sys/class/thermal,
like temperature, mode...

...
#include 
#include 
#include 
#include 
static int thermal_get_temp(struct thermal_zone_device *thermal,
unsigned long *temp)
{
*temp = xxx;
return 0;
}
static struct device *hwmon;
static struct thermal_zone_device *thermal;
static struct thermal_zone_device_ops ops = {
.get_temp = thermal_get_temp,
};
static int __init sensor_init(void)
{
...
thermal = thermal_zone_device_register("sample", 0, 0,
&ops, 0, 0, 0, 0);
}
static void __exit sensor_cleanup(void)
{
...
}
module_init(sensor_init);
module_exit(sensor_cleanup);

Are there any better ideas about this? Your comments are highly appreciated.

Yong

___
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev