Description of problem: USB Pentax Optio cameras (at least the S and the 430RS), which appear as mass storage, cause 'mount' to go to D state. Also, looking at /proc/scsi/scsi shows '(null)'s to appear in several places (sorry, don't have it handy). Version-Release number of selected component (if applicable): 2.4.20-18.9 How reproducible: 100%
I haven't looked at the details tim reported, but when i plug in my Pentax Optio S camera all USB devices silently hang, including my mouse...
I need a dmesg and <alt><sysrq>"t", taken after the lockup. First, grab dmesg > /tmp/dmesg.out, because sysrq-t will overflow it. Then, sysrq-t, and dmesg > /tmp/dmesg2.out. This may have enough space to capture sysrq-t, or may not, depending. Best of all do it from init runlevel 3, without X. Fewer processes, less overlow. Do not drop dmesgs into comments box, please. Attach them to the bug instead.
See also bug 89491. This reminds me, I need /proc/bus/usb/devices for both cameras. Just make the OS limp somehow far enough to capture that, so I can get vendor strings.
Created attachment 92382 [details] dmesg
Created attachment 92383 [details] /proc/bus/usb/devices
With the camera, there are two mass storage devices on this machine; one is a flash card reader on the printer.
The 330 is known to work with vanilla kernels, dunno about the 430RS. 430 is also reported working with SuSE and idents added http://www.chip.de/forum/thread.html?bwthreadid=382836
The 0x0a17/0x0004 part is fixed upstream, I'm going to poke Arjan to take a patch.
CVS 2.4.20-19+ It is in upstream, indeed.
Nicely fixed in kernel-2.4.20-20.9! Thanks to all involved.
Tim, don't forget to close 'modified' bugs, please.
This is still broken for the Pentax Optio S. Upgrading to 2.4.25 fixed it.