Re: [PATCH] OMAP: USB : Fix the EHCI enumeration and core retention issue

[Felipe Balbi <balbi-l0cyMroinI0@public.gmane.org>]

[-- Attachment #1: Type: text/plain, Size: 3333 bytes --]

Hi,

On Fri, Jun 22, 2012 at 07:14:32AM -0700, Kevin Hilman wrote:
> Felipe Balbi <balbi-l0cyMroinI0@public.gmane.org> writes:
> 
> > Hi,
> >
> > On Fri, Jun 22, 2012 at 01:00:39PM +0530, Munegowda, Keshava wrote:
> >> On Fri, Jun 22, 2012 at 12:32 AM, Kevin Hilman <khilman-l0cyMroinI0@public.gmane.org> wrote:
> >> > "Munegowda, Keshava" <keshava_mgowda-l0cyMroinI0@public.gmane.org> writes:
> >> >
> >> >> On Thu, Jun 21, 2012 at 7:12 PM, Keshava Munegowda
> >> >> <keshava_mgowda-l0cyMroinI0@public.gmane.org> wrote:
> >> >>> This commit 354ab8567ae3107a8cbe7228c3181990ba598aac titled
> >> >>> "Fix OMAP EHCI suspend/resume failure (i693)" is causing
> >> >>> the usb hub and device detection fails in beagle XM
> >> >>> causeing NFS not functional. This affects the core retention too.
> >> >>> The same commit logic needs to be revisted adhering to hwmod and
> >> >>> device tree framework.
> >> >>> for now, this commit id 354ab8567ae3107a8cbe7228c3181990ba598aac
> >> >>> titled "Fix OMAP EHCI suspend/resume failure (i693)" reverted.
> >> >>>
> >> >>> This patch is validated on BeagleXM with NFS support over
> >> >>> usb ethernet and USB mass storage and other device detection.
> >> >>>
> >> >>> Signed-off-by: Keshava Munegowda <keshava_mgowda-l0cyMroinI0@public.gmane.org>
> >> >
> >> > [...]
> >> >
> >> >>
> >> >> hi kevin
> >> >>
> >> >> here is pm count log on beagle XM with the above patch:
> >> >
> >> > What are you meaning to show by this log?
> >> >
> >> > This dump shows that neither PER or CORE are hitting retention in idle.
> >> > Which sounds to me like you have not enabled UART runtime suspend:
> >> >
> >> > echo 3000 > /sys/devices/platform/omap_uart.0/power/autosuspend_delay_ms
> >> > echo 3000 > /sys/devices/platform/omap_uart.1/power/autosuspend_delay_ms
> >> > echo 3000 > /sys/devices/platform/omap_uart.2/power/autosuspend_delay_ms
> >> > echo 3000 > /sys/devices/platform/omap_uart.3/power/autosuspend_delay_ms
> >> >
> >> > My test with your patch shows that it fixes the oops during boot, and
> >> > doesn't hang during suspend, but that USB host is still preventing CORE
> >> > retention during idle (after UART runtime suspend is enabled.)
> >> >
> >> > This happens on 3530/Overo, 3630/Beagle-xM and 3730/Overo
> >> >
> >> > Setting CONFIG_MFD_OMAP_USB_HOST=n allows CORE to hit retention again.
> >> >
> >> > Kevin
> >> 
> >> 
> >> 
> >> Hi kevin
> >>        It woks. only the log was wrong. I was using no_console_suspend
> >> in boot args.
> >> i removed it. now I can see the core retention hits with USB host in Beagle XM.
> >> below is the log:
> >
> > the fact is that we can't really survive without that workaround. Kevin,
> 
> I don't know what workaround you're talking about.    Are you talking
> about the revert proposed in $SUBJECT patch?
> 
> I don't have a problem with that revert.  The problem I have is that it
> does not fix the problem I initially reported: USB host prevents CORE
> retention in *idle*.

Keshava is reverting a fix for a HW errata. I can't accept it as it will
cause regressions. Granted, regression by regression, there's no change,
but I simply can't knowingly cause a regression to the driver just to
have PM working. We need a real fix for this issue.

-- 
balbi

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]