On Thu, Jan 12, 2012 at 5:40 PM, Felipe Contreras
<[email protected]> wrote:
> On Thu, Jan 12, 2012 at 1:56 PM, Grazvydas Ignotas <[email protected]> wrote:
>> On Mon, Dec 19, 2011 at 10:01 PM, Felipe Contreras
>> <[email protected]> wrote:
>>> From: Felipe Contreras <[email protected]>
>>>
>>> In musb_init_controller() there's a pm_runtime_put(), but there's no
>>> pm_runtime_get(), which creates a mismatch that causes the driver to
>>> sleep when it shouldn't.
>>
>> This patch is causing problems here, the device never suspends, I've
>> verified musb_runtime_suspend and omap2430_runtime_suspend are never
>> called in my setup, unless I revert this patch. I'm using musb with
>> ethernet gadget; cable removal also doesn't suspend musb.
>>
>> It seems pm_runtime_put() that was removed was balancing out
>> pm_runtime_get in musb_platform_init(), which is called at the
>> beginning of musb_init_controller(). I think this pm_runtime_get()
>> from musb_platform_init() should be moved to musb_init_controller() to
>> stop confusing people and pm_runtime_put() needs to be brought back. I
>> can do a patch if nobody objects.
>
> You mean have both get and put in musb_init_controller()? That's OK,
> it would be balanced then, but you would have to remove the
> pm_runtime_put() from omap2430_remove() that presumably was balancing
> the pm_runtime_get_sync() in omap2430_musb_init().
>
>>> This was introduced in 7acc619[1], but it wasn't triggered in my setup
>>> until 18a2689[2] was merged to Linus' branch at point df0914[3]. IOW;
>>> when PM is working as it was supposed to.
>>>
>>> However, it seems most of the time this is used in a way that keeps the
>>> counter above 0, so nobody noticed. Also, it seems to depend on the
>>> configuration used in versions before 3.1, but not later (or in it).
>>>
>>> I found the problem by loading isp1704_charger before any usb gadgets:
>>> http://article.gmane.org/gmane.linux.kernel/1226122
>>
>> isp1704_charger seems to be doing otg_io_read(), and it's perfectly
>> normal that musb can be suspended at the time isp1704 decides to do
>> this. I think proper fix is to wake up must on musb_ulpi_read() and
>> musb_ulpi_write() instead.
>
> I guess so. If you send these patches I can give them a try.
Sent them a week ago, would be nice if you tried them.
--
GraÅžvydas
On Mon, Feb 13, 2012 at 2:54 PM, Grazvydas Ignotas <[email protected]> wrote:
> On Thu, Jan 12, 2012 at 5:40 PM, Felipe Contreras
> <[email protected]> wrote:
>> On Thu, Jan 12, 2012 at 1:56 PM, Grazvydas Ignotas <[email protected]> wrote:
>>> On Mon, Dec 19, 2011 at 10:01 PM, Felipe Contreras
>>> <[email protected]> wrote:
>>>> From: Felipe Contreras <[email protected]>
>>>>
>>>> In musb_init_controller() there's a pm_runtime_put(), but there's no
>>>> pm_runtime_get(), which creates a mismatch that causes the driver to
>>>> sleep when it shouldn't.
>>>
>>> This patch is causing problems here, the device never suspends, I've
>>> verified musb_runtime_suspend and omap2430_runtime_suspend are never
>>> called in my setup, unless I revert this patch. I'm using musb with
>>> ethernet gadget; cable removal also doesn't suspend musb.
>>>
>>> It seems pm_runtime_put() that was removed was balancing out
>>> pm_runtime_get in musb_platform_init(), which is called at the
>>> beginning of musb_init_controller(). I think this pm_runtime_get()
>>> from musb_platform_init() should be moved to musb_init_controller() to
>>> stop confusing people and pm_runtime_put() needs to be brought back. I
>>> can do a patch if nobody objects.
>>
>> You mean have both get and put in musb_init_controller()? That's OK,
>> it would be balanced then, but you would have to remove the
>> pm_runtime_put() from omap2430_remove() that presumably was balancing
>> the pm_runtime_get_sync() in omap2430_musb_init().
>>
>>>> This was introduced in 7acc619[1], but it wasn't triggered in my setup
>>>> until 18a2689[2] was merged to Linus' branch at point df0914[3]. IOW;
>>>> when PM is working as it was supposed to.
>>>>
>>>> However, it seems most of the time this is used in a way that keeps the
>>>> counter above 0, so nobody noticed. Also, it seems to depend on the
>>>> configuration used in versions before 3.1, but not later (or in it).
>>>>
>>>> I found the problem by loading isp1704_charger before any usb gadgets:
>>>> http://article.gmane.org/gmane.linux.kernel/1226122
>>>
>>> isp1704_charger seems to be doing otg_io_read(), and it's perfectly
>>> normal that musb can be suspended at the time isp1704 decides to do
>>> this. I think proper fix is to wake up must on musb_ulpi_read() and
>>> musb_ulpi_write() instead.
>>
>> I guess so. If you send these patches I can give them a try.
>
> Sent them a week ago, would be nice if you tried them.
Sorry, I just started hacking with my N900 again. I will probably test
them later today.
Cheers.
--
Felipe Contreras