Hans, On Fri, Mar 22, 2013 at 12:45 PM, Hans Verkuil <hverkuil@xxxxxxxxx> wrote: > On Fri March 22 2013 17:26:57 edubezval@xxxxxxxxx wrote: >> Hello Hans, >> >> On Fri, Mar 22, 2013 at 10:04 AM, Hans Verkuil <hverkuil@xxxxxxxxx> wrote: >> <snip> >> >> >> >> >> # on your branch on the other hand I get a NULL pointer: >> > >> > I've fixed that (v4l2_dev was never initialized), and I've also rebased my tree >> > to the latest code. Can you try again? >> > >> >> This time I get a kernel crash at _power. Unfortunately I cannot fetch >> the crash log as I am not having access to a serial line (using vga >> console) and in my setup mtdoops is not working somehow. >> >> >> Sequence is v4l2_ioctl->video_usercopy->__video_do_ioctl->v4l_s_ctrl->v4l2_s_ctrl->set_ctrl_lock->try_or_set_cluster->si4713_s_ctrl->si4713_set_power_state->mutex_lock_nested->lock_acquire. >> >> >> I 'd need to spend some time on it to understand better your patches >> and help you to get this working. And for that I'd prob need to spend >> some time to either hack a serial line or get mtdoops to work :-) > > You're doing fine: that was all the info I needed. Can you do a git pull and > try again? > Sure. Your patch removes the locks but I believe the set_power is used in other places as well. And has a misspell: s/clocks/locks. But now I still get the nested lock issue. But this time around si4713_setup, that is called from _s_ctrl and which in turns calls v4l2_ctrl_handler_setup, which call s_ctrl again. > Regards, > > Hans -- Eduardo Bezerra Valentin -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
