On Fri, May 18, 2001 at 07:30:21AM +0200, Michel Lanners wrote: > > Sure, since there's no FCODE in there (i386 board), it's unbootable > anyway.... Just wanted to see what it spits out ;-) > > >> What doesn't look so good is that it gives a path for /dev/sdb (but > >> there's no such device). > > > > the scsi detection stuff is a huge kludge. can you send me the > > contents of /proc/scsi/scsi and /proc/partitions ? > > [mlan@piglet ~]$ cat /proc/scsi/scsi > Attached devices: > Host: scsi0 Channel: 00 Id: 01 Lun: 00 > Vendor: IBM Model: DCAS-34330 Rev: S65A > Type: Direct-Access ANSI SCSI revision: 02 > Host: scsi0 Channel: 00 Id: 03 Lun: 00 > Vendor: MATSHITA Model: CD-ROM CR-8008 Rev: 8.0e > Type: CD-ROM ANSI SCSI revision: 02 ok, i see. the problem is the test whether a given device actually has a disk associated with it is somewhat naive. ofpath takes the contents of /proc/scsi/scsi and counts how many devices are listed there, then it simply assumes that all of them are disks, so if there are two as in your case it will consider /dev/sda and /dev/sdb to be configured. i am not sure of the best way to fix that... counting how many `Direct-Access' lines there are directly should work... but that negates finding CDROM paths, but ofpath doesn't support that anyway now that i look at it. fixing that is somewhat non-trivial (this is not an easy script to modify). > Well I suppose it mistakes the cdrom for sdb; it probably shouldn't, > since there is no second SCSI disk.... i agree, it also gave a bogus path for it (no scsi id). i will see what i can come up with but the fix will need alot of testing... -- Ethan Benson who wishes there was an easier way to get this info. http://www.alaska.net/~erbenson/
Attachment:
pgpcGch8hS3bE.pgp
Description: PGP signature