On Wed, Feb 14, 2001 at 08:16:30AM +0100, Oliver Neukum
<[EMAIL PROTECTED]> wrote:
>
> > However, in light of other email in this thread, it looks like maybe the
> > table can go altogether, so hopefully this question is now moot.
>
> Oh yes, I saw the mail you are refering to.
> Let me put it this way: If we get rid of 'direction unknown'
> then I'd be comfortable with killing it.
> I'll put a printk into the drivers I use the test this.
I don't know if you looked at my patch, but all the drivers in question
assume that if they don't know a command is a write, then it's a read
command. There are only eight drivers that do things this way, and they
don't seem to be particularly new ones, in general, though I might be
wrong about that with a few of them. So I think that trusting what the
kernel thinks is the direction might be as good as trusting what the
driver thinks is the direction currently. On the other hand, given that
MESH at least locks up hard when the driver guesses wrong, it's
important to verify that the kernel is right at least as often as the
drivers currently are about the direction.
-Daniel
--
Daniel E. Eisenbud
[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to [EMAIL PROTECTED]
