On Thu, 28 Jun 2007, Alex Chiang wrote: > Hello, > > I have a Teac DV-28E-N cdrom, and I've been seeing some buffer > I/O errors on a heavily loaded system (uptime indicates system > load of 390 or so). > > Turning on USB debugging for usbcore and mass storage indicates > that something is timing out (hanging?), which makes the SCSI > mid-layer issue an abort, and then a bus reset. > > This problem occurs on 2.6.16 but seems to go away in 2.6.18. > Upgrading to 2.6.18 isn't an option. > > I've been poking through the git logs to try and figure out > what's changed between 2.6.16 and 2.6.18, and found some patches > that looked promising: > > usb-storage-get-rid-of-timer.patch > http://marc.info/?l=linux-usb-devel&m=115092809714473&w=2 > > usb-core-get-rid-of-timer.patch > http://marc.info/?l=linux-usb-devel&m=115938791628537&w=2 > > But applying those didn't seem to help. > > If someone could point me in the right direction, either with an > explanation of why my device is timing out, further > instrumentation I could add to my kernel to figure out what's > going on, or a patch that magically solves this problem ;), I'd > really appreciate it.
The errors shown in your logs are not problems in the usb-storage driver, so whatever fixed them in 2.6.18 was a change made somewhere else. Probably in the USB host controller driver, but maybe in the USB scatter-gather library or somewhere else entirely. The log showed that all the requested data had been transferred but nevertheless the request didn't complete. Alan Stern ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
