the usb-storage bug is fixed in kernel 349 (see http://people.redhat.com). Actually it's worked around. cdrecord has a bug where it doesn't check a return value and interprets an error value as 255 sectors-io-size. Which is/was invalid for usb storage. SG just silently ignores the fact that that is an invalid value for usb storage and continues, at the risk of breaking the burn. As I said fixed in 349 by making usb-storage also grok 255 sector sized IO's.
I don't know about cdrecord, but grip and cdparanoia are still broken with 349 and my USB CD-RW drive. I still get:
Error trying to open /dev/sg0 exclusively (No such device or address). retrying in 1 second
...and failure to rip any tracks.
Also, 349 seems to ignore "selinux=0" argument?
:j
