CuBox-i kernel 4a10065b77 HDMI hotplug

[Rudi]

Hi folks,

I have a problem with the Apr 22 kernel (4a10065b77) from Jon's repo. When unplugging/replugging the HDMI, the video output is lost.
This happens only on the I4, not on the C1. It does not happen with the Feb 28 kernel (a305f2d8f6). Since my I4 is somewhat flaky,
I'm wondering if others see this issue as well or if it's just my hardware...


--

Ruediger "Rudi" Ihle

[Rudi]

On 27.04.2014 09:50, Rudi wrote:

I just noticed the following, which seems to happen only on a real replug (not just an input change on the AVR):



INFO: rcu_sched self-detected stall on CPU
0: (2099 ticks this GP) idle=797/140000000000001/0 softirq=15762/15762
(t=2100 jiffies g=1478 c=1477 q=704)
CPU: 0 PID: 25 Comm: kworker/0:1 Not tainted 3.10.30 #1
Workqueue: events hotplug_worker
[<80014220>] (unwind_backtrace) from [<8001124c>] (show_stack+0x10/0x14)
[<8001124c>] (show_stack) from [<80088608>] (rcu_check_callbacks+0x3d0/0x7f8)
[<80088608>] (rcu_check_callbacks) from [<80034158>] (update_process_times+0x40/0x6c)
[<80034158>] (update_process_times) from [<80069de8>] (tick_sched_timer+0x4c/0x78)
[<80069de8>] (tick_sched_timer) from [<800482b8>] (__run_hrtimer.isra.32+0x54/0xe4)
[<800482b8>] (__run_hrtimer.isra.32) from [<80048bc8>] (hrtimer_interrupt+0x11c/0x2d0)
[<80048bc8>] (hrtimer_interrupt) from [<80013ba0>] (twd_handler+0x34/0x44)
[<80013ba0>] (twd_handler) from [<800829f0>] (handle_percpu_devid_irq+0x6c/0x84)
[<800829f0>] (handle_percpu_devid_irq) from [<8007f064>] (generic_handle_irq+0x2c/0x3c)
[<8007f064>] (generic_handle_irq) from [<8000e600>] (handle_IRQ+0x40/0x90)
[<8000e600>] (handle_IRQ) from [<800085b8>] (gic_handle_irq+0x2c/0x5c)
[<800085b8>] (gic_handle_irq) from [<8000d980>] (__irq_svc+0x40/0x50)
Exception stack(0xdc0e9bb0 to 0xdc0e9bf8)
9ba0: 000d7960 ffffffff 000025ed 80252718
9bc0: 00000010 8084b8dc dc077a34 dc077810 dc0778d0 00000000 dc0e8038 81700600
9be0: 00000000 dc0e9bf8 8029b8d8 80252744 20000013 ffffffff
[<8000d980>] (__irq_svc) from [<80252744>] (__loop_delay+0x0/0xc)

[Thomas Genty]

Le 27/04/2014 09:59, Rudi a écrit :
> On 27.04.2014 09:50, Rudi wrote:
>
> I just noticed the following, which seems to happen only on a real
> replug (not just an input change on the AVR):

Didn't notice that
I can replug without this
but I will test with your latest changes

Tom

[Rudi]

Hi Tom,

> Didn't notice that
> I can replug without this
> but I will test with your latest changes

Now it works here as well. Maybe part of the problem was HDMI plug not
correctly inserted... But the oops should not happen in that case either...


Cheers !


--

Ruediger "Rudi" Ihle