Comment # 13
on bug 10469
from Anish Bhatt
(In reply to Bill Meier from comment #12)
> >
> > I went through the RFC and a couple of implementations, no else seems to be
> > imposing this restriction. Might be legacy code since the commit is from
> > 2006.
>
> Looking at the commit message (for the patch which added the test), I now
> realize that the test was added to improve the "heuristic test" to verify
> that the packet is actually an iSCSI packet.
>
> Question: would it be correct to say that the "expected data length" field
> for a SCSI_COMMAND must always be non-zero ?
>
> If so, I'll change the code to test for a non-zero field.
Unfortunately no, it's a little more complicated than that.
>
> > I'll post a fix for this
>
> No need to post a patch. It's simple enough for me to make a change.
Sure, I was planning to add some expert_info stuff to iSCSI, so I can roll it
in there and save you the effort.
You are receiving this mail because:
- You are watching all bug changes.