[8939] in linux-scsi channel archive
Re: LINUX Jobs for 2.4 update [scsi]
daemon@ATHENA.MIT.EDU (Eric Youngdale)
Sat Jun 3 17:47:58 2000
Message-ID: <00b401bfcda5$16cc5110$0f17a8c0@eric.home>
From: "Eric Youngdale" <eric@andante.org>
To: =?iso-8859-1?Q?G=E9rard_Roudier?= <groudier@club-internet.fr>,
"Jens Axboe" <axboe@suse.de>
Cc: "Paul Gortmaker" <p_gortmaker@yahoo.com>,
"Alan Cox" <alan@lxorguk.ukuu.org.uk>,
<linux-kernel@vger.rutgers.edu>, <linux-scsi@vger.rutgers.edu>
Date: Sat, 3 Jun 2000 17:45:51 -0400
MIME-Version: 1.0
Content-Type: multipart/alternative;
boundary="----=_NextPart_000_00B1_01BFCD83.8F484030"
This is a multi-part message in MIME format.
------=_NextPart_000_00B1_01BFCD83.8F484030
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
----- Original Message -----=20
From: "G=E9rard Roudier" <groudier@club-internet.fr>
To: "Jens Axboe" <axboe@suse.de>
Cc: "Paul Gortmaker" <p_gortmaker@yahoo.com>; "Alan Cox" =
<alan@lxorguk.ukuu.org.uk>; <linux-kernel@vger.rutgers.edu>; =
<linux-scsi@vger.rutgers.edu>
Sent: Saturday, May 27, 2000 5:44 PM
Subject: Re: LINUX Jobs for 2.4 update [scsi]
>=20
>=20
> On Sat, 27 May 2000, Jens Axboe wrote:
>=20
> > On Sat, May 27 2000, Paul Gortmaker wrote:
> > > > To Do
> > > > -----
> > > > Linux uses TEST_UNIT_READY to chck for device presence on a =
PUN/LUN. The
> > > > INQUIRY is the only valid test allowed by the spec.
> > >=20
> > > The draft of SCSI-2 spec I have here hints that INQUIRY should be =
used
> > > to probe system configuration and that TEST_UNIT_READY is more for
> > > polling on devices with removable media.
>=20
> Why only removable ?
> It is intended to check if media can be accessed, regardless the media
> type.
>=20
> > > I tossed the TEST_UNIT_READY
> > > part out and INQUIRY alone works fine (one disk, 2 CDs and a tape =
on
> > > a BusLogic clone - all found as per usual).
> >=20
> > Looks good to me. Although I can't possibly see what harm a =
TEST_UNIT_READY
> > command could do...
>=20
> TEST UNIT READY seems not appropriate for the discovery of devices. It
> also may return CHECK CONDITION when INQUIRY will succeed (and =
preserve a
> UNIT ATTENTION CONDITION if any).
>=20
> On the other hand, it may be risky to try a TUR on an not existing =
LUN,
> given that SCSI devices usually expect INQUIRY. Broken SCSI device
> firmwares may arm on a TUR to a non existing LUN but not harm on an
> INQUIRY.
>=20
> INQUIRY will not fail is a device is here. If the device is not ready,
> INQUIRY may well be the the only command the device wants to accept.
The main problem with not using TUR is that I believe that INQUIRY =
is not required to be sensitive to the LUN number. In other words, an =
INQUIRY in order to detect whether a lun is actually present at a given =
lun number, you also need to do a TUR. =20
I tried digging around in the standards to find something which =
describes this situation. Section 7.5.3 (Selection of an invalid =
logical unit) seems to cover this a little bit:
7.5.3 Selection of an invalid logical unit
The target's response to selection of a logical unit that is not valid =
is described in the following paragraphs.
The logical unit may not be valid because:
a) the target does not support the logical unit (e.g. some targets =
support only one peripheral device). In response
to an INQUIRY command, the target shall return the INQUIRY data with =
the peripheral qualifier set to the value
required in 8.2.5.1. In response to any other command except REQUEST =
SENSE, the target shall terminate
the command with CHECK CONDITION status. In response to a REQUEST =
SENSE command, the target shall
return sense data. The sense key shall be set to ILLEGAL REQUEST and =
the additional sense code shall be
set to LOGICAL UNIT NOT SUPPORTED.
b) the target supports the logical unit, but the peripheral device is =
not currently attached to the target. In response
to an INQUIRY command, the target shall return the INQUIRY data with =
the peripheral qualifier set to the value
required in 8.2.5.1. In response to any other command except REQUEST =
SENSE, the target shall terminate
the command with CHECK CONDITION status. In response to a REQUEST =
SENSE command, the target shall
return sense data. The sense key shall be set to ILLEGAL REQUEST and =
the additional sense code shall be
set to LOGICAL UNIT NOT SUPPORTED.
c) the target supports the logical unit and the peripheral device is =
attached, but not operational. In response to
an INQUIRY command, the target shall return the INQUIRY data with the =
peripheral qualifier set to the value
required in 8.2.5.1. The target's response to any command other than =
INQUIRY and REQUEST SENSE is
vendor-specific.
d) the target supports the logical unit but is incapable of =
determining if the peripheral device is attached or is not
operational when it is not ready. In response to an INQUIRY =
command,the target shall return the INQUIRY data
with the peripheral qualifier set to the value required in 8.2.5.1. In =
response to a REQUEST SENSE command,
the target shall return the REQUEST SENSE data with a sense key of NO =
SENSE unless a contingent
allegiance exists. The target's response to any other command is =
vendor-specific.
Reversing the order of the INQUIRY/TUR wouldn't hurt anything I =
guess. Eliminating it altogether seems dangerous.
In SCSI-3, there is now a command that can be used to ask a device =
what logical units are present.
-Eric
------=_NextPart_000_00B1_01BFCD83.8F484030
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META content=3D"text/html; charset=3Diso-8859-1" =
http-equiv=3DContent-Type>
<META content=3D"MSHTML 5.00.2919.6307" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>----- Original Message ----- </FONT>
<DIV><FONT face=3DArial size=3D2>From: "G=E9rard Roudier" <</FONT><A=20
href=3D"mailto:groudier@club-internet.fr"><FONT face=3DArial=20
size=3D2>groudier@club-internet.fr</FONT></A><FONT face=3DArial=20
size=3D2>></FONT></DIV>
<DIV><FONT face=3DArial size=3D2>To: "Jens Axboe" <</FONT><A=20
href=3D"mailto:axboe@suse.de"><FONT face=3DArial=20
size=3D2>axboe@suse.de</FONT></A><FONT face=3DArial =
size=3D2>></FONT></DIV>
<DIV><FONT face=3DArial size=3D2>Cc: "Paul Gortmaker" <</FONT><A=20
href=3D"mailto:p_gortmaker@yahoo.com"><FONT face=3DArial=20
size=3D2>p_gortmaker@yahoo.com</FONT></A><FONT face=3DArial =
size=3D2>>; "Alan Cox"=20
<</FONT><A href=3D"mailto:alan@lxorguk.ukuu.org.uk"><FONT =
face=3DArial=20
size=3D2>alan@lxorguk.ukuu.org.uk</FONT></A><FONT face=3DArial =
size=3D2>>;=20
<</FONT><A href=3D"mailto:linux-kernel@vger.rutgers.edu"><FONT =
face=3DArial=20
size=3D2>linux-kernel@vger.rutgers.edu</FONT></A><FONT face=3DArial =
size=3D2>>;=20
<</FONT><A href=3D"mailto:linux-scsi@vger.rutgers.edu"><FONT =
face=3DArial=20
size=3D2>linux-scsi@vger.rutgers.edu</FONT></A><FONT face=3DArial=20
size=3D2>></FONT></DIV>
<DIV><FONT face=3DArial size=3D2>Sent: Saturday, May 27, 2000 5:44 =
PM</FONT></DIV>
<DIV><FONT face=3DArial size=3D2>Subject: Re: LINUX Jobs for 2.4 update=20
[scsi]</FONT></DIV></DIV>
<DIV><BR></DIV>
<DIV><FONT face=3DArial size=3D2>> <BR>> <BR>> On Sat, 27 May =
2000, Jens=20
Axboe wrote:<BR>> <BR>> > On Sat, May 27 2000, Paul Gortmaker=20
wrote:<BR>> > > > To Do<BR>> > > > -----<BR>> =
>=20
> > Linux uses TEST_UNIT_READY to chck for device presence on a =
PUN/LUN.=20
The<BR>> > > =
> =20
INQUIRY is the only valid test allowed by the spec.<BR>> > > =
<BR>>=20
> > The draft of SCSI-2 spec I have here hints that INQUIRY should =
be=20
used<BR>> > > to probe system configuration and that =
TEST_UNIT_READY is=20
more for<BR>> > > polling on devices with removable =
media.<BR>>=20
<BR>> Why only removable ?<BR>> It is intended to check if media =
can be=20
accessed, regardless the media<BR>> type.<BR>> <BR>> > > =
I tossed=20
the TEST_UNIT_READY<BR>> > > part out and INQUIRY alone works =
fine (one=20
disk, 2 CDs and a tape on<BR>> > > a BusLogic clone - all found =
as per=20
usual).<BR>> > <BR>> > Looks good to me. Although I can't =
possibly=20
see what harm a TEST_UNIT_READY<BR>> > command could do...<BR>> =
<BR>> TEST UNIT READY seems not appropriate for the discovery of =
devices.=20
It<BR>> also may return CHECK CONDITION when INQUIRY will succeed =
(and=20
preserve a<BR>> UNIT ATTENTION CONDITION if any).<BR>> <BR>> On =
the=20
other hand, it may be risky to try a TUR on an not existing LUN,<BR>> =
given=20
that SCSI devices usually expect INQUIRY. Broken SCSI device<BR>> =
firmwares=20
may arm on a TUR to a non existing LUN but not harm on an<BR>>=20
INQUIRY.<BR>> <BR>> INQUIRY will not fail is a device is here. If =
the=20
device is not ready,<BR>> INQUIRY may well be the the only command =
the device=20
wants to accept.<BR></FONT></DIV>
<DIV><FONT face=3DArial size=3D2> The main problem =
with not using=20
TUR is that I believe that INQUIRY is not required to be sensitive to =
the LUN=20
number. In other words, an INQUIRY in order to detect whether a =
lun is=20
actually present at a given lun number, you also need to do a TUR. =
</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2> I tried digging =
around in the=20
standards to find something which describes this situation. =
Section 7.5.3=20
(Selection of an invalid logical unit) seems to cover this a little=20
bit:</FONT></DIV>
<BLOCKQUOTE style=3D"MARGIN-RIGHT: 0px">
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>7.5.3 Selection of an invalid logical =
unit</FONT></DIV>
<DIV><FONT face=3DArial size=3D2><BR>The target’s response to =
selection of a=20
logical unit that is not valid is described in the following=20
paragraphs.<BR>The logical unit may not be valid because:</FONT></DIV>
<DIV><FONT face=3DArial size=3D2><BR>a) the target does not support =
the logical=20
unit (e.g. some targets support only one peripheral device). In =
response<BR>to=20
an INQUIRY command, the target shall return the INQUIRY data with the=20
peripheral qualifier set to the value<BR>required in 8.2.5.1. In =
response to=20
any other command except REQUEST SENSE, the target shall =
terminate<BR>the=20
command with CHECK CONDITION status. In response to a REQUEST SENSE =
command,=20
the target shall<BR>return sense data. The sense key shall be set to =
ILLEGAL=20
REQUEST and the additional sense code shall be<BR>set to LOGICAL UNIT =
NOT=20
SUPPORTED.</FONT></DIV>
<DIV><FONT face=3DArial size=3D2><BR>b) the target supports the =
logical unit, but=20
the peripheral device is not currently attached to the target. In=20
response<BR>to an INQUIRY command, the target shall return the INQUIRY =
data=20
with the peripheral qualifier set to the value<BR>required in 8.2.5.1. =
In=20
response to any other command except REQUEST SENSE, the target shall=20
terminate<BR>the command with CHECK CONDITION status. In response to a =
REQUEST=20
SENSE command, the target shall<BR>return sense data. The sense key =
shall be=20
set to ILLEGAL REQUEST and the additional sense code shall be<BR>set =
to=20
LOGICAL UNIT NOT SUPPORTED.</FONT></DIV>
<DIV><FONT face=3DArial size=3D2><BR>c) the target supports the =
logical unit and=20
the peripheral device is attached, but not operational. In response =
to<BR>an=20
INQUIRY command, the target shall return the INQUIRY data with the =
peripheral=20
qualifier set to the value<BR>required in 8.2.5.1. The target’s =
response to=20
any command other than INQUIRY and REQUEST SENSE=20
is<BR>vendor-specific.</FONT></DIV>
<DIV><FONT face=3DArial size=3D2><BR>d) the target supports the =
logical unit but=20
is incapable of determining if the peripheral device is attached or is =
not<BR>operational when it is not ready. In response to an INQUIRY =
command,the=20
target shall return the INQUIRY data<BR>with the peripheral qualifier =
set to=20
the value required in 8.2.5.1. In response to a REQUEST SENSE =
command,<BR>the=20
target shall return the REQUEST SENSE data with a sense key of NO =
SENSE unless=20
a contingent<BR>allegiance exists. The target’s response to any =
other command=20
is vendor-specific.</FONT></DIV></BLOCKQUOTE>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2> Reversing the order =
of the=20
INQUIRY/TUR wouldn't hurt anything I guess. Eliminating it =
altogether=20
seems dangerous.</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2> In SCSI-3, there is =
now a=20
command that can be used to ask a device what logical units are=20
present.</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>-Eric</FONT></DIV>
<DIV><BR><FONT face=3DArial =
size=3D2><BR></FONT> </DIV></BODY></HTML>
------=_NextPart_000_00B1_01BFCD83.8F484030--
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.rutgers.edu
