Ticket #163 (closed defect: fixed)

Opened 11 months ago

Last modified 8 months ago

does not work

Reported by: mdrons Owned by: somebody
Priority: normal Component: USB IR Transceiver
Version: Severity: critical
Keywords: Cc:

Description

It works sometimes after reboot, but after a few days it stops working. Log file from today attached. I have tried the .99 and the pre1.0 release, both eventually fail.

I am running kernel version: myth ~ # uname -a Linux myth 2.6.31.1-smp #1 SMP Tue Sep 29 17:03:09 EDT 2009 x86_64 AMD Phenom(tm) 9850 Quad-Core Processor AuthenticAMD GNU/Linux

Oct 27 14:32:50 2009 ERROR: failed to write data packet: Failed to write (interrupt end point): No error Oct 27 14:32:50 2009 ERROR: Failed to send IR data. Oct 27 14:32:50 2009 ERROR: Device transaction (0x15) failed Oct 27 14:32:50 2009 ERROR: handleClientRequest(0x00) failed with: 22 (Invalid argument) Oct 27 14:32:50 2009 INFO: FAILED to write packet back to client: 0x0 Oct 27 14:34:20 2009 ERROR: failed to write control packet: Failed to write (interrupt end point): No error Oct 27 14:34:20 2009 ERROR: Device transaction (0x12) failed Oct 27 14:34:20 2009 ERROR: handleClientRequest(0x00) failed with: 4 (Interrupted system call) Oct 27 14:34:20 2009 INFO: FAILED to write packet back to client: 0x0 Oct 27 14:34:20 2009 INFO: Found client using protocol version 1 Oct 27 14:34:20 2009 INFO: Request handled within daemon: 0xfe Oct 27 14:34:20 2009 INFO: FAILED to write packet back to client: 0xfe Oct 27 14:34:20 2009 INFO: Found client using protocol version 1 Oct 27 14:34:20 2009 INFO: Request handled within daemon: 0xfe Oct 27 14:34:20 2009 INFO: FAILED to write packet back to client: 0xfe Oct 27 14:34:20 2009 INFO: Found client using protocol version 1 Oct 27 14:34:20 2009 INFO: Request handled within daemon: 0xfe Oct 27 14:34:20 2009 INFO: FAILED to write packet back to client: 0xfe Oct 27 14:34:20 2009 INFO: Found client using protocol version 1 Oct 27 14:34:20 2009 INFO: Request handled within daemon: 0xfe Oct 27 14:34:20 2009 INFO: FAILED to write packet back to client: 0xfe Oct 27 14:34:20 2009 INFO: Found client using protocol version 1 Oct 27 14:34:20 2009 INFO: Request handled within daemon: 0xfe Oct 27 14:34:20 2009 INFO: FAILED to write packet back to client: 0xfe Oct 27 14:34:20 2009 INFO: Found client using protocol version 1 Oct 27 14:34:20 2009 INFO: Request handled within daemon: 0xfe Oct 27 14:34:20 2009 INFO: FAILED to write packet back to client: 0xfe Oct 27 14:34:20 2009 INFO: Found client using protocol version 1 Oct 27 14:34:20 2009 INFO: Request handled within daemon: 0xfe Oct 27 14:34:20 2009 INFO: Found client using protocol version 1 Oct 27 14:34:20 2009 INFO: Request handled within daemon: 0xfe Oct 27 14:34:20 2009 ERROR: failed to write control packet Oct 27 14:34:20 2009 ERROR: Device transaction (0x12) failed Oct 27 14:34:20 2009 ERROR: handleClientRequest(0x00) failed with: 6 (No such de vice or address) Oct 27 14:34:20 2009 INFO: FAILED to write packet back to client: 0x0 Oct 27 14:34:21 2009 INFO: Device 0 released Oct 27 14:34:21 2009 INFO: Worker 0 exiting Oct 27 14:34:21 2009 DEBUG: Reaped child: 0x7f794da0b910 Oct 27 18:13:35 2009 DEBUG: Forking into the background. Oct 27 18:13:35 2009 DEBUG: Parameters: Oct 27 18:13:35 2009 DEBUG: recvTimeout: 1000 Oct 27 18:13:35 2009 DEBUG: sendTimeout: 1000 Oct 27 18:13:35 2009 DEBUG: Handling 1 device(s): Oct 27 18:13:35 2009 INFO: Worker 0 starting Oct 27 18:13:35 2009 DEBUG: 0) usb:3.2 id=0 addr=0x60acf0 Oct 27 18:13:36 2009 ERROR: failed to write control packet: Failed to write (interrupt end point): No error Oct 27 18:13:36 2009 ERROR: Failed to get version.

Change History

Changed 10 months ago by jdunn

This problem should be fixed in the next release. The problem is that libusb does not handle threads correctly. Some report that changing the command line options used for the daemon makes a difference. Using the --no-threads option and increasing the receive timeout should work around the problem, but we've received mixed reports on that. Check the /etc/default/iguanaIR for details.

Changed 10 months ago by bluey

Please upgrade to iguanair 1.0pre2 (and use new default config if asked). Install libusb-1.0-0. Run

igdaemon -nvvv

in terminal and check that the 4th line should say libusb.so and not not libusbpre1.so

If so, hopefully that will fix your problems. Let us know.

Changed 10 months ago by bluey

  • priority changed from highest to normal

Changed 9 months ago by sqrammi

I'm seeing the same exact error message:

Dec 23 22:32:32 2009 ERROR: failed to write control packet: Failed to write (interrupt end point): Input/output error

I've tried all USB ports on the system (just in case there's a hub issue), on versions 0.98 and 1.0pre2. I need this working ASAP. I'm willing to try new firmwares, etc.

It works perfectly on my 2-year-old Core 2 Duo box with 0.98 (but only after using the new libusb 1.0 -- the old libusb stops working after a random period of time), but I am getting this error on this new Core i7 box. I think this should be higher priority... perhaps it only happens on Core i7 usb host controllers?

Changed 9 months ago by sqrammi

btw my igdaemon *does* say libusb.so and not libusbpre1.so.

Changed 8 months ago by sqrammi

Any feedback on this issue? I'm about to give up and use my HD-PVR's IR emitter instead.

Changed 8 months ago by sqrammi

I was able to get it to work. I had to disable CONFIG_USB_DEBUG in my Linux kernel config, because I was getting:

usb8-2: BOGUS urb flags, 0x1 --> 0x0 usb8-2: usbfs: usb_submit_urb returned -22 uhci_hcd 0000:00:1d.2: release dev 2 ep81-INT, period 8, phase 4, 118 us

in my kernel log. This is triggered by usb_ir/igdaemon sending bad USB urbs, so I'm not sure why/how igdaemon is sending bad urbs, but when I disable debugging, it at least doesn't choke.

I also had to use the "--no-labels" argument to igdaemon, and now I can transmit!!

Changed 8 months ago by jdunn

  • status changed from new to closed
  • resolution set to fixed

Glad to hear you got it working, and since we have not heard back from the original poster I'm going to close this ticket. Thank you for the note about CONFIG_USB_DEBUG. I'll keep that in mind if other people report the same error code.

Note: See TracTickets for help on using tickets.