Closed bwilcutt closed 8 years ago
After further investigation, it appears this lock up issue goes back several years with the USB BBB stack.
Anyone have any further developments on this issue?
I've been running more experiments. It would seem that the USB crashing issue all but disappears if you run the beaglebone without the mezzanine board plugged in. I've been able to get one of my tests to run nearly twice as long [so far] than ever before without a lock up.
The plot thickens...
Resolution:
I've given up sending AT commands to the USB Modem. Now, I'm sending cqdm commands to ttyUSB1 instead of AT commands to ttyUSB0. No issues found, everything working great... case closed ... for me... but the ttyUSB0 issue is still there, I think.
Having an issue-- found it with both 3.14.58 and 4.1.15 so it must be systemic?
Hardware:
Beaglebone black with USB wireless (Novatel USB760).
I'm running a simple script that sends an "AT" to the modem, the modem responds with "OK". I print OK to the screen and repeat in a while() loop. That's it. Simple.
This script works just fine by after 300+ queries the USB driver crashes with the dump listed below. Note that I've run this on two versions and get the same results. It's intermittent, meaning unpredictable when it occurs but given enough time it will occur. I think there may be some kind of bug in the USB host driver? I've included my config file dump below the kernel dump below.
My .config file contains (this is 4.1.15):