Skip to contentSkip to conversation

USB device not accepting new address

DISCUSSION Pete Zaitcev 3 participants 4 messages Sep 5, 2001 – Sep 10, 2001 1 message not in the archive

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

#1 lore raw
Quoted 5 lines

I have just come across another USB address problem, which happens sporadically and is not easy to reproduce.

1: [cfefa240] link (00000001) e0 IOC Stalled CRC/Timeo Length=7ff MaxLen=7ff DT1 EndPt=0 Dev=0, PID=69(IN) (buf=00000000)

If usb_set_address() ends in timeouts, something is bad with the hadrware, most likely. Microcode crash in the device, perhaps. Someone, I think it was Oliver, posted a patch that retries usb_set_address(). It may help you, look in linux-usb-devel archives.

-- Pete

#2 lore raw
In reply to a message not in the archive
Quoted 4 lines · Pete Zaitcev
Quoted 3 lines

I have just come across another USB address problem, which happens sporadically and is not easy to reproduce.

Quoted 8 lines
Quoted 2 lines

1: [cfefa240] link (00000001) e0 IOC Stalled CRC/Timeo Length=7ff MaxLen=7ff DT1 EndPt=0 Dev=0, PID=69(IN) (buf=00000000)

If usb_set_address() ends in timeouts, something is bad with the hadrware, most likely. Microcode crash in the device, perhaps. Someone, I think it was Oliver, posted a patch that retries usb_set_address(). It may help you, look in linux-usb-devel archives.

Maybe it's a hardware problem, but this problem has never occured before Alan started merging bits of 2.4.9 into his tree.

-Udo.

#3 lore raw
In reply to a message not in the archive
Quoted 8 lines · Pete Zaitcev

Interesting. I just dealt with a bug which is most definitely a hardware problem, and the owner swore that "it worked with 2.4.3-12 perfectly" (it was the last errata that we shipped). The thing (a Kaweth compatible Ethernet) was found NOT to work on earlier kernels reliably either. Such things do happen:
something stops working just in time.
What regressions did you do? Did you run 2.4.7-ac _after_ it broke?

I've just tried all -ac1 kernels from 2.4.5-ac1 to 2.4.9-ac1 and the problem exists with all of them. So it would appear that it didn't work reliably with earlier kernels either. I find it interesting however, that it is much harder to trigger in earlier kernels and that it doesn't happen every time.

-Udo.

#4 lore raw
In reply to a message not in the archive
Quoted 5 lines · Udo A. Steinberg

I've just tried all -ac1 kernels from 2.4.5-ac1 to 2.4.9-ac1 and the problem exists with all of them. So it would appear that it didn't work reliably with earlier kernels either. I find it interesting however, that it is much harder to trigger in earlier kernels and that it doesn't happen every time.

I have seen a similar problem when I plugged in a creative nomad jukebox (which was misbehaving with windows as well). It had been behaving properly in the past, but was of course an unrecognized usb device. However this time it failed to get a device id assigned. The solution:
replace the usb cable. Turns out one of the wires had a break in it, and would loose the connection in one direction if you didn't bend it "just right". if you have usb problems, treat it like scsi: blame the cable first. With a different usb cable all the problems went away instantly and it accepted it's usb device id.

Len Sorensen