time@oxtrap.uucp (Tim Endres) (10/19/89)
I recently acquired a DataFrame XP330, and through some finesse managed to partition it with HD Setup. It has a MacOS partition that works flawlessly. I created two A/UX partitions (along with the eschatology and other misc) and went to mkfs them. When I do this I get the following: /dev/dsk/c4d0s0 device write error: ... sector info ... Timeout. This happens in only about 5% of the sectors (very rough estimate). It seems also to happen more frequently with higher numbered sectors. And on some occasions the thing reports that "retry #2 succeeded" and it thinks things are ok. OK, so I try to fsck the file system when mkfs is done, and voila, no problems. Reads have no problems. then mount, no problem, then sync ... device write error. So, it looks like writes are a problem, and the problem happens on both partitions. A quick reboot, and some MacOS software verifies that the sectors complained about are indeed writable and show no problems. Is this a timing problem with the A/UX SCSI driver? Is it possible the thing was formatted with a poor interleave? Does a long cable possibly cause problems for the A/UX driver & not the MacOS? DataFrame is on a SCSI daisy chain, but has a terminator and uses Apple cables.
kaufman@Neon.Stanford.EDU (Marc T. Kaufman) (10/20/89)
In article <1989Oct19.161340.27893@oxtrap.uucp> time@oxtrap.uucp (Tim Endres) writes:
-I recently acquired a DataFrame XP330, and through some finesse managed to
-partition it with HD Setup. It has a MacOS partition that works flawlessly.
-I created two A/UX partitions (along with the eschatology and other misc)
-and went to mkfs them. When I do this I get the following:
-/dev/dsk/c4d0s0 device write error: ... sector info ... Timeout.
Call SuperMac and get the upgraded ROM for the XP330 controller board. The
original ROM didn't handle disconnect/reselect correctly.
Marc Kaufman (kaufman@Neon.stanford.edu)