spicywolf that will most likely not produce *some specific* scsi errors. If you want to create specific errors, then you would need to create method to inject such errors.

true.

but even accounting for that I think I'd need some specific hardware or really good software to test luxadm. Do we have a way to test luxadm? I definitely don't have a SunFire 880 or any other FibreChannel Arbitrated Loop devices. :p

you're going to be unable to test luxadm

we should probably delete it, but I have lost track of what it supports

definitely A5000, no idea after that

it might also have been SPARC only, which would make deleting it easier :)

I think we can accept that the luxadm change in code.illumos.org/c/illumos-gate/+/4345 is much more likely than not to improve things.

→ CODE REVIEW 4345: 17536 overflow checks ruined by int promotion 17540 kmdb should be built as c99 (NEW) | illumos.org/issues/17536

oops, wrong link

→ CODE REVIEW 4351: 17516 Some readable output transposes newline and full stop (NEW) | illumos.org/issues/17516

sommerfeld: right, I just mean in general I don't know of any way to really use it anymore