os/doc
3
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

brought up-to-date with regards to some aspects of the scsi system

Browse source 
doubtlessly imported new and interesting spelling errors..
This commit is contained in:
Julian Elischer 1995-11-20 05:46:00 +00:00
parent 7825ec568c
commit 7392d17c9e
Notes: svn2git 2020-12-08 03:00:23 +00:00 
svn path=/head/; revision=163
1 changed files with 90 additions and 56 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

146
handbook/scsi.sgml
View file

@ -1,4 +1,4 @@
<!-- $Id: scsi.sgml,v 1.7 1995-11-20 01:10:30 jfieber Exp $ --> <!-- $Id: scsi.sgml,v 1.8 1995-11-20 05:46:00 julian Exp $ -->
<!-- The FreeBSD Documentation Project --> <!-- The FreeBSD Documentation Project -->
<!-- <!--
@ -67,25 +67,25 @@
Elaborate caching schemes, automatic bad block replacement etc Elaborate caching schemes, automatic bad block replacement etc
are all made possible by this 'intelligent device' approach. are all made possible by this 'intelligent device' approach.
On a SCSI bus, each possible pair of devices can communicate. If On a SCSI bus, each possible pair of devices can communicate. Whether
their function allows this is another matter, but the standard does their function allows this is another matter, but the standard does
not restrict it. To avoid signal contention, the 2 devices have to not restrict it. To avoid signal contention, the 2 devices have to
arbitrate for the bus before using it. arbitrate for the bus before using it.
The philosophy of SCSI is to have a standard that allows The philosophy of SCSI is to have a standard that allows
older-standard devices to work with newer-standard ones. So, an older-standard devices to work with newer-standard ones. So, an
old SCSI-1 device should normally work on a SCSI-2 bus. Normally, old SCSI-1 device should normally work on a SCSI-2 bus. I say
because it is not absolutely sure that the implementation of an old Normally, because it is not absolutely sure that the implementation
device follows the (old) standard closely enough to be acceptable of an old device follows the (old) standard closely enough to be
on a new bus. Modern devices are usually more well-behaved, acceptable on a new bus. Modern devices are usually more
because the standardization has become more strict and is better well-behaved, because the standardization has become more strict
adhered to by the device manufacturers. Generally speaking, the and is better adhered to by the device manufacturers.
chances of getting a working set of devices on a single bus is Generally speaking, the chances of getting a working set of
better when all the devices are SCSI-2 or newer. This does not devices on a single bus is better when all the devices are SCSI-2
imply that you have to dump all your old stuff when you get that or newer. This does not imply that you have to dump all your old
shiny 2Gb disk: I own a system on which a pre-SCSI-1 disk, a SCSI-2 stuff when you get that shiny 2Gb disk: I own a system on which a
QIC tape unit, a SCSI-1 helical scan tape unit and 2 SCSI-1 disks pre-SCSI-1 disk, a SCSI-2 QIC tape unit, a SCSI-1 helical scan
work together quite happily. tape unit and 2 SCSI-1 disks work together quite happily.
<sect1><heading>Components of SCSI</heading> <sect1><heading>Components of SCSI</heading>
<p> <p>
@ -114,7 +114,7 @@
hoods with strain reliefs etc are the way to go. Second golden hoods with strain reliefs etc are the way to go. Second golden
rule: don't use cables longer than necessary. I once spent 3 days rule: don't use cables longer than necessary. I once spent 3 days
hunting down a problem with a flaky machine only to discover that hunting down a problem with a flaky machine only to discover that
shortening the SCSI bus with 1 meter solved the problem. And the shortening the SCSI bus by 1 meter solved the problem. And the
original bus length was well within the SCSI specification. original bus length was well within the SCSI specification.
<sect2><heading>SCSI bus types</heading> <sect2><heading>SCSI bus types</heading>
@ -232,19 +232,19 @@
external variants. Almost every SCSI device comes with a external variants. Almost every SCSI device comes with a
number of sockets in which a number of resistor networks can number of sockets in which a number of resistor networks can
(must be!) installed. If you remove terminators from a device, (must be!) installed. If you remove terminators from a device,
carefully stock 'm. You will need them when you ever decide to carefully store 'm. You will need them when you ever decide to
reconfigure your SCSI bus. There is enough variation in even reconfigure your SCSI bus. There is enough variation in even
these simple tiny things to make finding the exact replacement these simple tiny things to make finding the exact replacement
a frustrating business. There are also SCSI devices that have a frustrating business. There are also SCSI devices that have
a single jumper to enable or disable a built-in terminator. a single jumper to enable or disable a built-in terminator.
There are special terminators you can stick onto a flat cable There are special terminators you can stick onto a flat cable
bus. Others look like external connectors, so a connector hood bus. Others look like external connectors, or a connector hood
without a cable. So, lots of choice as you can see. without a cable. So, lots of choice as you can see.
There is much debate going on if and when you should switch There is much debate going on if and when you should switch
from simple resistor (passive) terminators to active from simple resistor (passive) terminators to active
terminators. Active terminators contain more or less elaborate terminators. Active terminators contain slightly more elaborate
circuits to give more clean bus signals. The general consensus circuit to give cleaner bus signals. The general consensus
seems to be that the usefulness of active termination increases seems to be that the usefulness of active termination increases
when you have long buses and/or fast devices. If you ever have when you have long buses and/or fast devices. If you ever have
problems with your SCSI buses you might consider trying an problems with your SCSI buses you might consider trying an
@ -307,8 +307,8 @@
certainly will. Clever external terminators sometimes have a certainly will. Clever external terminators sometimes have a
LED indication that shows whether terminator power is present. LED indication that shows whether terminator power is present.
In newer designs auto-restoring fuses are used who 'reset' In newer designs auto-restoring fuses that 'reset'
themselves after some time. themselves after some time are sometimes used.
On modern devices, sometimes integrated terminators are On modern devices, sometimes integrated terminators are
used. These things are special purpose integrated circuits that used. These things are special purpose integrated circuits that
@ -384,7 +384,8 @@
One might be inclined to think that since SCSI disks are smart One might be inclined to think that since SCSI disks are smart
you can forget about this. Alas, the arcane setup issue is still you can forget about this. Alas, the arcane setup issue is still
present today. The system BIOS needs to know how to access your present today. The system BIOS needs to know how to access your
SCSI disk with the head/cyl/sector method. SCSI disk with the head/cyl/sector method in order to load the
FreeBSD kernel during boot.
The SCSI host adapter or SCSI controller you have put in your The SCSI host adapter or SCSI controller you have put in your
AT/EISA/PCI/whatever bus to connect your disk therefore has it's AT/EISA/PCI/whatever bus to connect your disk therefore has it's
@ -397,7 +398,7 @@
<bf>translated</bf> drive. This means that a fake drive table is <bf>translated</bf> drive. This means that a fake drive table is
constructed that allows the PC to boot the drive. This constructed that allows the PC to boot the drive. This
translation is often (but not always) done using a pseudo drive translation is often (but not always) done using a pseudo drive
with 32 heads and 64 sectors per track. By varying the number of with 64 heads and 32 sectors per track. By varying the number of
cylinders, the SCSI BIOS adapts to the actual drive size. It is cylinders, the SCSI BIOS adapts to the actual drive size. It is
useful to note that 32 * 64 / 2 = the size of your drive in useful to note that 32 * 64 / 2 = the size of your drive in
megabytes. The division by 2 is to get from disk blocks that are megabytes. The division by 2 is to get from disk blocks that are
@ -417,25 +418,32 @@
jumper or software setup selection, to switch the translation the jumper or software setup selection, to switch the translation the
SCSI BIOS uses. SCSI BIOS uses.
It is very important that <bf>all</bf> operating systems on the disk use It is very important that <bf>all</bf> operating systems on the
the <bf>same translation</bf> to get the right idea about where to find disk use the <bf>same translation</bf> to get the right idea about
the relevant partitions. So, when installing FreeBSD you must where to find the relevant partitions. So, when installing
answer any questions about heads/cylinders etc using the FreeBSD you must answer any questions about heads/cylinders
translated values your host adapter uses. etc using the translated values your host adapter uses.
Failing to observe the translation issue might be un-bootable systems or Failing to observe the translation issue might lead to
operating systems overwriting each others partitions. Using fdisk un-bootable systems or operating systems overwriting each
you should be able to see all partitions. others partitions. Using fdisk you should be able to see
all partitions.
As promised earlier: what is this talk about 'lying' devices? As You might have heard some talk of 'lying' devices?
you might already know, the FreeBSD kernel reports the geometry Older FreeBSD kernels used to report the geometry
of SCSI disks when booting. An example from one of my systems: of SCSI disks when booting. An example from one of my systems:
<verb> <verb>
Feb 9 19:33:46 yedi /386bsd: aha0 targ 0 lun 0: <MICROP 1588-15MB1057404HSP4> aha0 targ 0 lun 0: <MICROP 1588-15MB1057404HSP4>
Feb 9 19:33:46 yedi /386bsd: sd0: 636MB (1303250 total sec), 1632 cyl, 15 head, sd0: 636MB (1303250 total sec), 1632 cyl, 15 head, 53 sec, bytes/sec 512
53 sec, bytes/sec 512
</verb> </verb>
Newer kernels usually don't report this information.. e.g.
<verb>
(bt0:0:0): "SEAGATE ST41651 7574" type 0 fixed SCSI 2
sd0(bt0:0:0): Direct-Access 1350MB (2766300 512 byte sectors)
</verb>
Why has this changed?
This info is retrieved from the SCSI disk itself. Newer disks This info is retrieved from the SCSI disk itself. Newer disks
often use a technique called zone bit recording. The idea is that often use a technique called zone bit recording. The idea is that
@ -444,7 +452,11 @@ Feb 9 19:33:46 yedi /386bsd: sd0: 636MB (1303250 total sec), 1632 cyl, 15 head,
have more tracks on outer cylinders than on the inner cylinders have more tracks on outer cylinders than on the inner cylinders
and, last but not least, have more capacity. You can imagine that and, last but not least, have more capacity. You can imagine that
the value reported by the drive when inquiring about the geometry the value reported by the drive when inquiring about the geometry
now becomes fake. now becomes suspect at best, and nearly always misleading. When
asked fro a geometry , it is nearly always better to supply the
geometry used by the BIOS, or <em>if the BIOS is never going to know
about this disk</em>, (e.g. it is not a booting disk) to supply a
ficticious geometry that is convenient.
<sect2><heading>SCSI subsystem design</heading> <sect2><heading>SCSI subsystem design</heading>
<p> <p>
@ -533,8 +545,20 @@ device cd0 at scbus? [the first ever CDROM found, no wiring]
<em>only</em> attach them when they match the target ID and <em>only</em> attach them when they match the target ID and
LUN specified on the corresponding bus. LUN specified on the corresponding bus.
So, if you had a SCSI tape at target ID 2 it would not be Wired down devices get 'first shot' at the unit numbers
configured, but it will attach when it is at target ID 6. so the first non 'wired down' device, is allocated the unit number
one greater than the highest 'wired down' unit number
for that kind of device.
So, if you had a SCSI tape at target ID 2 it would be
configured as st2, as the tape at target ID 6 is wired down
to unit number 1. Note that <em>wired down devices need not
be found</em>
to get their unit number. The unit number for a wired down device
is reserved for thet device, even if it is turned off at boot
time. This allows the device to be turned on and brought
on-line at a later time, without rebooting. Notice that a device's
unit number has <em>no</em> relationship with it's target ID on
the SCSI bus.
Below is another example of a kernel config file as used by Below is another example of a kernel config file as used by
FreeBSD version < 2.0.5. The difference with the first example is FreeBSD version < 2.0.5. The difference with the first example is
@ -562,23 +586,19 @@ ST01/02&rsqb;
controller scbus0 controller scbus0
device sd0 &lsqb;support for 4 SCSI harddisks, sd0 up sd3&rsqb; device sd0 &lsqb;support for 4 SCSI harddisks, sd0 up sd3&rsqb;
device sd1
device sd2
device sd3
device st0 &lsqb;support for 2 SCSI tapes&rsqb; device st0 &lsqb;support for 2 SCSI tapes&rsqb;
device st1
device cd0 #Only need one of these, the code dynamically grows &lsqb;for the cdrom&rsqb; device cd0 #Only need one of these, the code dynamically grows &lsqb;for the cdrom&rsqb;
</verb> </verb>
Both examples support 4 SCSI disks. If during boot more
Both examples support SCSI disks. If during boot more
devices of a specific type (e.g. sd disks) are found than are devices of a specific type (e.g. sd disks) are found than are
configured in the booting kernel, the system will complain. You configured in the booting kernel, the system will simply allocate
will have to build and boot a new kernel (after adapting the kernel more devices, incrementing the unit number starting at the last
configuration file) before you can use all of the devices. It number 'wired down'. If there are no 'wired down' devices
does not hurt to have 'extra' devices in the kernel, the example then counting starts at unit 0.
above will work fine when you have only 2 SCSI disks.
Use <tt>man 4 scsi</tt> to check for the latest info on the SCSI Use <tt>man 4 scsi</tt> to check for the latest info on the SCSI
subsystem. For more detailed info on host adapter drivers use eg subsystem. For more detailed info on host adapter drivers use eg
@ -587,13 +607,15 @@ device cd0 #Only need one of these, the code dynamically grows &lsq
<sect2><heading>Tuning your SCSI kernel setup</heading> <sect2><heading>Tuning your SCSI kernel setup</heading>
<p> <p>
Experience has shown that some devices are slow to respond to INQUIRY Experience has shown that some devices are slow to respond to INQUIRY
commands after a SCSI bus reset. An INQUIRY command is sent by the kernel commands after a SCSI bus reset (which happens at Boot time).
on boot to see what kind of device (disk, tape, cdrom etc) is connected An INQUIRY command is sent by the kernel on boot to see what
to a specific target ID. This process is called device probing by the way. kind of device (disk, tape, cdrom etc) is connected to a
specific target ID. This process is called device probing by the way.
To work around this problem, FreeBSD allows a tunable delay time before To work around this problem, FreeBSD allows a tunable delay time
the SCSI devices are probed following a SCSI bus reset. You can set this before the SCSI devices are probed following a SCSI bus reset.
delay time in your kernel configuration file using a line like: You can set this delay time in your kernel configuration file
using a line like:
<verb> <verb>
options "SCSI_DELAY=15" #Be pessimistic about Joe SCSI device options "SCSI_DELAY=15" #Be pessimistic about Joe SCSI device
@ -627,7 +649,7 @@ Mar 29 21:16:37 yedi /386bsd: st1: Archive Viper 150 is a known rogue
all LUNs on a certain target ID, even if they are actually only one all LUNs on a certain target ID, even if they are actually only one
device. It is easy to see that the kernel might be fooled into device. It is easy to see that the kernel might be fooled into
believing that there are 8 LUNs at that particular target ID. The believing that there are 8 LUNs at that particular target ID. The
confusion this causes is left as an exercise to the user. confusion this causes is left as an exercise to the reader.
The SCSI subsystem of FreeBSD recognizes devices with bad habits by The SCSI subsystem of FreeBSD recognizes devices with bad habits by
looking at the INQUIRY response they send when probed. Because the looking at the INQUIRY response they send when probed. Because the
@ -695,6 +717,18 @@ options "TUNE_1542" #dynamic tune of bus DMA speed
Make a minimal bus config with as little devices as possible. Make a minimal bus config with as little devices as possible.
<item> <item>
If possible, configure your host adapter to use slow bus speeds. If possible, configure your host adapter to use slow bus speeds.
<item>
If you can compile a kernel, make one with the SCSIDEBUG option,
and try accessing the device with debugging turned on for
that device. If your device doesn't even probe at startup,
you may have to define the address of the device that
is failing, and the desired debug level in /sys/scsi/scsidebug.h.
If it probes but just doesn't work, you can use the
<tt>scsi(8)</tt> command to dynamically set a debug level to
it in a running kernel (if SCSIDEBUG is defined).
This will give you COPIOUS debugging output with which to confuse
the gurus. see <tt>man 4 scsi</tt> for more exact information.
Also look at <tt>man 8 scsi</tt>.
</itemize> </itemize>
<sect1><heading>Further reading<label id="scsi:further-reading"></heading> <sect1><heading>Further reading<label id="scsi:further-reading"></heading>