Sujith Manoharan <c_manoha@...> writes:

David Madore wrote:

quoted

In other words: loading the module seems to place the dongle in a bad
state that goes away after ifconfig up (but not by removing the
modules), and if a reboot happens while the dongle is in that bad
state, it will fail permanently.

Right, there's a difference here, a 'reboot' of the card is done only when
if the interface is brought up. The device status from the USB subsystem
is used to handle this - it works on x86.

On your platform, I think that a full reboot is required even when doing just
a simple load/unload cycle. I'll come up with a patch to test this.

Thanks for the detailed log.

Sujith
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo@...
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Hi Sujith,
I seem to running into this same issue, mine is that the driver emits the
"Target unresponsive" error if the unit is powered-up with the TL-WN721N plugged
in already. I think my issue is that the USB subsytem event is not recieved by
the atk9k_htc driver, and therefore the hardware is not reset.

Has this ben fixed already? Can you give me the commit sha1 in wireless-testing?
I am using compat-wireless, so I am lagging master a bit.

If not, can you point me at where in the driver code the reset happens? Maybe I
can come up with a patch.