USB errors and kernel panics with USB modem

[matt...@gmail.com]

Hey, I've found some similar posts around but no solutions, I'm wondering if anyone knows how to fix this...

# uname -a
Linux arm 3.8.13-bone64 #1 SMP PREEMPT Fri Aug 22 12:11:18 ADT 2014 armv7l GNU/Linux

I've got a cellular/GPS modem that gives me 4 ttyUSB devices, one for cell data, one for modem control and one for GPS data. Everything seems to be working except for the GPS. I can read data from /dev/ttyUSB1 (the GPS device), but when I am done reading and try to close the file, I get varying errors and occasionally a kernel panic. Using cat on /dev/ttyUSB1 generates the problem, but I've written a small C program to verify; the failure always happens when the program tries to call close().

It looks like an issue with the USB driver, or maybe I did something when I compiled the kernel (RT-PREEMPT maybe)?

Here is a section from dmesg:

[  100.271254] WARNING: at drivers/usb/musb/musb_host.c:125 musb_h_tx_flush_fifo+0x5f/0x6c()
[  100.271281] Could not flush host TX2 fifo: csr: 2003
[  100.271303] Modules linked in: option usb_wwan usbserial omap_rng
[  100.271425] [<c00111d5>] (unwind_backtrace+0x1/0x98) from [<c002efdf>] (warn_slowpath_common+0x33/0x48)
[  100.271475] [<c002efdf>] (warn_slowpath_common+0x33/0x48) from [<c002f04d>] (warn_slowpath_fmt+0x1d/0x28)
[  100.271522] [<c002f04d>] (warn_slowpath_fmt+0x1d/0x28) from [<c03253b7>] (musb_h_tx_flush_fifo+0x5f/0x6c)
[  100.271571] [<c03253b7>] (musb_h_tx_flush_fifo+0x5f/0x6c) from [<c0326143>] (musb_cleanup_urb+0x1b/0x5c)
[  100.271617] [<c0326143>] (musb_cleanup_urb+0x1b/0x5c) from [<c0326221>] (musb_urb_dequeue+0x9d/0xc8)
[  100.271674] [<c0326221>] (musb_urb_dequeue+0x9d/0xc8) from [<c03114d5>] (unlink1+0x19/0xbc)
[  100.271724] [<c03114d5>] (unlink1+0x19/0xbc) from [<c0312319>] (usb_hcd_unlink_urb+0x31/0x7c)
[  100.271770] [<c0312319>] (usb_hcd_unlink_urb+0x31/0x7c) from [<c0312d71>] (usb_kill_urb+0x35/0x90)
[  100.271841] [<c0312d71>] (usb_kill_urb+0x35/0x90) from [<bf83446d>] (usb_wwan_close+0x44/0x5c [usb_wwan])
[  100.271932] [<bf83446d>] (usb_wwan_close+0x44/0x5c [usb_wwan]) from [<bf80401b>] (serial_down+0x1a/0x1c [usbserial])
[  100.272012] [<bf80401b>] (serial_down+0x1a/0x1c [usbserial]) from [<c027bb31>] (tty_port_shutdown+0x3d/0x40)
[  100.272070] [<c027bb31>] (tty_port_shutdown+0x3d/0x40) from [<c027c239>] (tty_port_close+0x15/0x34)
[  100.272135] [<c027c239>] (tty_port_close+0x15/0x34) from [<bf804409>] (serial_close+0x10/0x14 [usbserial])
[  100.272201] [<bf804409>] (serial_close+0x10/0x14 [usbserial]) from [<c0276cef>] (tty_release+0x9f/0x344)
[  100.272261] [<c0276cef>] (tty_release+0x9f/0x344) from [<c00bd5df>] (__fput+0x5b/0x15c)
[  100.272320] [<c00bd5df>] (__fput+0x5b/0x15c) from [<c004315d>] (task_work_run+0x6d/0xa4)
[  100.272384] [<c004315d>] (task_work_run+0x6d/0xa4) from [<c000f1bd>] (do_work_pending+0x6d/0x70)
[  100.272434] [<c000f1bd>] (do_work_pending+0x6d/0x70) from [<c000c833>] (work_pending+0x9/0x1a)
[  100.272463] ---[ end trace 4c88d925e8d40d87 ]---

Here is more dmesg:
http://pastebin.com/dsFKuzha

And here is the output from a kernel panic:
http://pastebin.com/JPbVPpWB

Anyone have any ideas on how to prevent this?

Thanks,

Matt

[gary.s...@jamexvending.com]

Looks like it's been quite a while, but were you able to find a solution to this issue?