(me again, have I used up my questions quota for the month yet?)
Can anybody shed any light on how hardware (block) devices are mapped to /dev/ entries in Linux. In the bad old days of System V if you had a device on say
scsi0 Channel: 00 Id: 00 Lun: 00
Then that would be found on something like
/dev/dsk/s0u0p0d0
scsi0 Channel: 00 Id: 01 Lun: 00
would be on
/dev/dsk/s0u0p1d0
But on Linux I'm buggered if I can understand what logic is used to map those scsi numbers (output from "# cat /proc/scsi/scsi") to the block device files in /dev. It seems to me when I add a device to the system the /dev/sdc0 or whatever block file that was associated with a device changes!. Is this so and if so why on earth did Linux adopt this approach instead of the seemingly more logical approach of system V?
Cheers
P. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
