[ath9k-devel] ath9k_htc, 2.6.36 and "Target is unresponsive" error

["Petr Štetiar" <ynezz@true.cz>]

Mohammed Shafi <shafi.wireless@gmail.com> [2010-11-22 11:15:17]:

> On Fri, Nov 19, 2010 at 4:35 PM, Petr ?tetiar <ynezz@true.cz> wrote:
> >
> > 1. line 128 in htc_drv_init.c time_left =
> > wait_for_completion_timeout(&priv->htc->target_wait, HZ);
> Can you please try out ..  htc_drv_init.c time_left =
> wait_for_completion_timeout(&priv->htc->target_wait, HZ/4); this should
> obviously fail

I'll try that later today, but it I think, that it will fail also.

> htc_drv_init.c time_left =
> wait_for_completion_timeout(&priv->htc->target_wait, 30*HZ);

I've even tried 100*HZ, but it fails anyway and I think, that the root of the
problem is somewhere else. This point is just the consequence of the failure
in htc_process_target_ready(), so even adding the 1000*HZ wouldn't help it
much :-)

As I understand it, host downloads the firmware to the MCU with
FIRMWARE_DOWNLOAD, then if it's sending of the firmware complete the host
sends FIRMWARE_DOWNLOAD_COMP and then MCU should respond with
HTC_MSG_READY_ID, but it didn't in my case for some reason. It's responding
with msg 0x18 on USB endpoint 0x1, instead of msg 0x1(HTC_MSG_READY_ID) on USB
endpoint 0.

I don't have the sources of the firmware, so I don't know how it works
internally, but maybe it's just wrong timing assumption somewhere or something
similar. That ARM board I'm using is kind of slow, it's like 200MHz(99
Bogomips), so everything takes quite longer time to complete, than on one is
used on desktop PC.

-- ynezz