Re: PATCH: exclude certain commands from emulated SCSI hosts

[Matthew Dharm <mdharm-scsi@one-eyed-alien.net>]

[-- Attachment #1: Type: text/plain, Size: 2764 bytes --]

On Mon, Apr 21, 2003 at 02:23:36PM -0500, James Bottomley wrote:
> On Mon, 2003-04-21 at 12:01, Matthew Dharm wrote:
> > My biggest complaints are:
> > (1) Your patch doesn't compile on my system.  RH7.3 -- it chokes on an
> > array in a struct with no length.  I doubt I'm the only person with this
> > problem.  Considering that in other places we bother to support this
> > compiler (-fomit-frame-pointer anyone?), it seems reasonable that we want
> > to support it here.
> >
> > (2) You're creating a struct to associate two things (white/black selection
> > and the list) which don't really need to be associated.  A list is a list
> > -- if we treat it as whitelist or blacklist really is a separate question.
> 
> To be honest, I don't care much.  I think I had a merge problem where
> the function template mismatched.

Then can we have it the way I proposed?   I have tested patches that use
this code as I proposed it waiting to be merged as soon as this is
accepted.

> The blacklist only exists because you apparently have a device that lies
> about the inquiry length it supports and then goes out to lunch when an
> inquiry of that length is sent to it.
> 
> Therefore, the only current use of the inquiry filter is to black/white
> list 36 and 58 byte inquiries.
> 
> Can you send us the inquiry strings of such a problem device?  We've had
> absolutely no takers for the inquiry limitations in the current
> blacklist, which does seem to imply that the code changes made to probe
> all luns fixed the problem for almost everyone else.

I think takers may be waiting in the wings to see this code merged.
Honestly, tho, I don't think I can send you such strings.  I work pretty
closely with some vendors, so most of the devices in my possession have
'fixed' firmware (to report meaningful INQUIRY length).  However, I get
report after report of devices that blow this in new and creative ways from
end-users.

You may recall that my first approach was to set the BLIST flag for 36-byte
INQUIRY -- Linus shot that down in favor of a command-filter approach.

In the end, the limitation of what INQUIRY can be filtered is a minor
complaint compared to the others.  If I have to live with it, I will --
however I will suggest that instead of allowing a length < 127 to be
specified, why not just use a couple of bits to mean 36 and 58 -- i.e.
treat this like the EVPD logic.  If it's not going to be fully flexible,
then let's make it mean exactly what we want.

Matt

-- 
Matthew Dharm                              Home: mdharm-usb@one-eyed-alien.net 
Maintainer, Linux USB Mass Storage Driver

I see you've been reading alt.sex.chubby.sheep voraciously.
					-- Tanya
User Friendly, 11/24/97

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]