On 10/05/2013 10:01 AM, Alan Stern wrote:
On Sat, 5 Oct 2013, Arokux X wrote:
Hi all,
first of all thank you all for your help. I now have some news to
report. Using your hint about timing I've inserted a bunch of udelays
around the read/write functions that get called from _rtl92c_write_fw
and got rid of the "detected XactErr len 0/0 retry" errors. Then I
just tried to use the WLAN adapter and it worked. Afterwards, I've
removed udelays and tried it again and guess what... everything still
worked.
I do not really know if I should consider these errors harmless but it
looks as if they are. What do you think?
It's hard to tell. But at any rate, it seems clear that your glue
driver is working okay.
I do however have something fishy to report, but this is probably an
off-topic for linux-usb mailing list so I'll just mention it here
briefly for completeness. Fist, I've compared the speed of the
rtl8192cu vs 8192cu in 3.4 kernel. The latter one is the Realtek's
driver. I've did many download tests of the same file and it turns out
8192cu is 1.5 times faster. Things haven't improved since then. The
3.12-rc1 rtl8192cu shows the same performance as in 3.4. In addition
with 3.12-rc1 I get some strange messages in the dmesg [1].
You may want to bring this to the attention of the wireless
maintainers.
Actually he has. From the logged messages, he must have loaded the driver with
extra debugging enabled. (debug=2)
As for the performance, he appears to be complaining that the Linux version
performs better than the vendor one. As I do not support the latter, there is
nothing for me to do.
Larry
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
