I need help! (Youse guys don't like PMs, so here it is).
Question: is it possible to programatically reduce the amount of time that a drive (in my case an NEC 4550A, running 1.X7) spends trying to read sectors with errors (specifically L-EC UEs)?
After digging out the MMC-3 docs, I've tried quite a few things- set the "RRC" (Read Retry Count) to "1" and "0", changed the Linux "sr" driver to send READ_12 (vs. READ_10) commands with the "Streaming" bit set in CDB, and tried the "RC" (Read Continuous) bit in the Error Recovery mode page (which was only settable in my DVR-105 drive).
None of those approaches worked. I have modified the underlying user-space library to ignore the read error on an -EIO returned from the driver (and just skip forward, returning garbage data (which won't be used anyway)), and I was able to ... recover ... the DVD in question reliably, but because of the number of bad sectors (wait- gotta sneeze - aa aa aaa aa AAARCCOOOS! - whew - excuse me!) it took some 10 hrs, and I'd like to avoid that. Nothing seemed to reduce the amount of time the drive spent sawing on those sectors.
Any hints? The MMC-3 docs seems to describe many "real-time read" options (and it's my understanding a popular Windows program is able to do this reliably) but none of 'em seem to work- did I miss a setting somewhere?