Hi,

On Fri, Oct 26, 2012 at 10:03:11PM +0000, Paul Walmsley wrote:

Hi Felipe,

On Fri, 26 Oct 2012, Felipe Balbi wrote:

quoted

On Thu, Oct 25, 2012 at 01:12:12PM +0300, Felipe Balbi wrote:

quoted

On Tue, Oct 23, 2012 at 09:23:03PM +0200, Jean Pihet wrote:

quoted

On Tue, Oct 23, 2012 at 9:19 PM, Paul Walmsley [off-list ref] wrote:

quoted

On Mon, 22 Oct 2012, Jean Pihet wrote:

quoted

On Mon, Oct 22, 2012 at 6:12 PM, Jean Pihet [off-list ref] wrote:

quoted

Do you have CPU_IDLE enabled?

FYI the issue is not present with CPU_IDLE enabled.

Hmm, how can you tell?  I thought you weren't able to reproduce it with
CPU_IDLE disabled either?

I could not reproduce the issue, with and without CPU_IDLE enabled.
What puzzles me is that the PM QoS code only has influence on the
states chosen by cpuidle, so the change should not have any impact
with CPU_IDLE enabled. I reallt need to reproduce the issue.
Let me try with the same setup as yours (bootloader images,
omap2pus_defconfig, angstrom roots).

I just sent a patch to fix a bug I found on OMAP4 panda but while
reading this thread again, I think it could be that it's the same bug
which is just easier to reproduce on Paul's setup.

Paul, Aaro, can you see if [1] makes the problem go away ? that would be
another reason to push [1] during this -rc cycle.

[1] http://marc.info/?l=linux-omap&m=135115602407925&w=2

Thanks for mentioning it, but this patch doesn't fix the I2C timeout 
problem here.  Log fragment below from the 3530ES3 Beagle.

that's too bad :-(

Can you compile with:

CONFIG_I2C_DEBUG_CORE=y
CONFIG_I2C_DEBUG_ALGO=y
CONFIG_I2C_DEBUG_BUS=y
CONFIG_DEFAULT_MESSAGE_LOGLEVEL=9

so that I can see all transfers ?

-- 
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20121029/c02377b7/attachment-0001.sig>