Re: MUSB patch (don't start next rx urb if current one failed)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Bin,

thanks. I have reverted commit https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/drivers/usb/musb/musb_host.c?id=dbac5d07d13e330e6706813c9fde477140fb5d80 and communication with both RT2800 and SMSC9500 was much more stable, but we still had less frequent issues with SMSC9500, so I looked at usbnet and smsc95xx drivers and found out that execution of *usb_clear_halt* is not enough if stalling endpoint is detected (bit MUSB_TXCSR_H_RXSTALL is set) and the chip must be reset, but none of these two drivers do it. The reset is mentioned in Microship's documentation and also used in their driver, so maybe someone can adapt smsc95xx driver to reset the chip properly. I also found out that execution of *unlink_urbs* from *usbnet_deferred_kevent * often leads to various errors ("queue 0 timed out", "Could not flush host TX2 fifo") or even system freeze, so it seems that MUSB driver still has some issue with fifo flushing.

Best regards

Tomas


Dne 25.1.2018 v 15:59 Bin Liu napsal(a):
+ cc:linux-usb

Hi Tomas,

It is always a good idea to cc linux-usb list when sending emails to an
individual. (otherwise, for example in my setup the emails go to the
Junk folder if a mailing list is not in to/cc.)

On Thu, Jan 18, 2018 at 03:47:12PM +0100, Tomas Paukrt wrote:
Hi Bin,

I have found out that your patch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/drivers/usb/musb/musb_host.c?id=dbac5d07d13e330e6706813c9fde477140fb5d80
is causing at least issues with SMSC9500 Ethernet chips and RT2800
WiFi chips. Error EPROTO sometimes happens during communication with
these chips and it leads to stopping communication.

We have kernel 3.12.10 with backported selected patches including
yours. Do we miss some patch that should restart sending URBs? In
our situation sending URBs just stop after first failure and is not
resumed unless we execute for example "ifconfig eth2 down" and
"ifconfig eth2 up".
Sorry for causing the regression. I will loop you in another email
thread to try to find a solution.

Regards,
-Bin.
.


--
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



[Index of Archives]     [Linux Media]     [Linux Input]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Old Linux USB Devel Archive]

  Powered by Linux