Bacula fails on FreeBSD 10.x / "mt fsf" infinitely proceeds

[Joerg Wunsch]

As Kenneth D. Merry wrote:

> Ahh.  Now I know who to send sa(4) driver patches to test.  :)

The only thing is that I don't love to reboot my main server machine
gratuitously.  But everything I could do on old scratch hardware is
possible.  I've also got a Tandberg SLR5 still around, I simply can't
throw it away ...

(Replacing the L9 by some SAS-attached LTO-x library is already
planned.)

> {black-pearl:/usr/home/kenm:!:0} dtrace -s tapetest.d
> dtrace: script 'tapetest.d' matched 1 probe
> CPU     ID                    FUNCTION:NAME
>   0    401                    saerror:entry Opcode 08, Status 0xc Data: Len 524290, Resid 2, Sense: Len 252, Resid 0
> 
>   0    401                    saerror:entry Opcode 08, Status 0xc Data: Len 524290, Resid 2, Sense: Len 252, Resid 0

Here's the DTrace output, first a dd with a blocksize of 64 KiB (on a
standard tar tape with 10 KiB records), then I did a rewind, followed
by "mt fsf 32767".

root at uriah:~ # dtrace -s /tmp/tapetest.d
dtrace: script '/tmp/tapetest.d' matched 1 probe
CPU     ID                    FUNCTION:NAME
  3    406                    saerror:entry Opcode 08, Status 0xc Data: Len 65536, Resid 55296, Sense: Len 252, Resid 223

  0    406                    saerror:entry Opcode 11, Status 0xc Data: Len 0, Resid 0, Sense: Len 252, Resid 223

It seems we are getting no residual for the SPACE command?!
-- 
cheers, Joerg               .-.-.   --... ...--   -.. .  DL8DTL

http://www.sax.de/~joerg/
Never trust an operating system you don't have sources for. ;-)