On 09/26/2014 07:25 AM, Johannes Stezenbach wrote: > On Fri, Sep 26, 2014 at 10:12:22AM -0300, Mauro Carvalho Chehab wrote: >> Try to add a WARN_ON or printk at em28xx_usb_resume(). > > It is called two times, once during hibernate and once during resume: >
> root@debian:~# echo disk >/sys/power/state On the upside this does look similar to what I have seen when I was debugging suspend/resume on pctv stick that uses em28xx and drx39xyj One thing that helped me debug the problem is testing hibernate in platform mode (which is default) and then Hibernate in reboot mode. I enabled usb debug and device debug to see what is happening at the usb-core and ran the following cases: I enable pm trace: echo 1 > /sys/power/pm_trace Hibernate in platform mode (default and recommended hibernation mode) echo platform > /sys/power/disk echo disk > /sys/power/state Hibernate in reboot mode: (usb bus could go through loss of power as platform might not maintain power to the buses). reset_resume should recover from loss of power or have the force disconnect path handle the case. i.e don't install reset_resume echo reboot > /sys/power/disk echo disk > /sys/power/state I also simply selected suspend from the GUI, this seems to take the usb-bus through a different path. These behave differently when reset_resume is installed vs. not installed. In our case, reset_resume simply points to resume which can't handle the power loss case. It would be good to get data on these different scenarios. I wish I have the WinTV 930, but I don't. If we have full debug for the above three scenarios, it would help debug it further. I also do the following to see resume works in a simple case: no disk involved suspend to ram echo mem > /sys/power/state I am looking at drxk to see if I can figure out anything. Also I dumped em28xx eprom to see if looks ok during these tests. thanks, -- Shuah -- Shuah Khan Sr. Linux Kernel Developer Samsung Research America (Silicon Valley) shua...@osg.samsung.com | (970) 217-8978 -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
