reading SysV tape from 4.2

[sm at cadre.UUCP]

< take this >

In article <1133 at sunybcs.UUCP> loverso at sunybcs.UUCP writes:
>We run 4.2 on a few vaxes, and have recently gotten a distribution
>tape for SysV.2.  Well, I can read the tape with dd and see file
>headers, but I cant extract the stuff with tar.  The tape has a blocksize
>of 5120, but even so:
>	dd -ibs </dev/rmt12 | tar tvf -
>doesnt work.  It gives checksum mismatch.

We had the same problem, only worse. We wanted to boot SysV on a backup
disk (RM05) and couldn't because we have our TU77 and RM05 on the same
massbus adapter and an RP07 on the second. The standard DEC midpriced
11/780 configuration (of four years ago), was one tape drive on one
massbus adapter and the disk on the other. Apparently Bell doesn't
add peripherals to existing Vaxen, they simply buy new ones. The result
is that SysV will only boot (from tape), on a machine that has a disk
on mba0 and a tape on mba1.

What we had to do (I hate to admit it), in order to get the SysV sources
was to extract cpio(1) from the distribution tape (all but the boot
section are in cpio NOT tar format), and then use cpio and mt to get
the stuff we wanted. I got cpio by using mt to get to the right file
(system sources) and then cp'ing the whole thing onto a large partition.
Then I edited out the cpio sources and compiled it. It is no fun.

If you send a copy of your SysV license I'll send you cpio and you
can do it that way.

Sean McLinden
Decision Systems Laboratory
1360 Scaife Hall
Pittsburgh, PA. 15261