1 This driver is for Compaq's SMART Array Controllers.
6 This driver is known to work with the following cards:
15 * SA 6400 U320 Expansion Module
24 If nodes are not already created in the /dev/cciss directory
28 Where ctlrs is the number of controllers you have (defaults to 1 if not
34 You need some entries in /dev for the cciss device. The mkdev.cciss script
35 can make device nodes for you automatically. Currently the device setup
45 b7 b6 b5 b4 b3 b2 b1 b0
46 |----+----| |----+----|
48 | +-------- Partition ID (0=wholedev, 1-15 partition)
50 +-------------------- Logical Volume number
52 The suggested device naming scheme is:
53 /dev/cciss/c0d0 Controller 0, disk 0, whole device
54 /dev/cciss/c0d0p1 Controller 0, disk 0, partition 1
55 /dev/cciss/c0d0p2 Controller 0, disk 0, partition 2
56 /dev/cciss/c0d0p3 Controller 0, disk 0, partition 3
58 /dev/cciss/c1d1 Controller 1, disk 1, whole device
59 /dev/cciss/c1d1p1 Controller 1, disk 1, partition 1
60 /dev/cciss/c1d1p2 Controller 1, disk 1, partition 2
61 /dev/cciss/c1d1p3 Controller 1, disk 1, partition 3
63 Support for more than 8 controllers
64 -----------------------------------
65 Originally the driver only supports 8 controllers in the system,
66 and this is due to the major numbers assigned to the driver
69 The driver can now support up to 32 controllers in the system.
71 For the ninth controller and beyond, the major numbers will be
72 assigned dynamically by the kernel when it is discovered,
73 and there is no guarantee what the major number you will get,
74 but most likely it will start from 254 and goes down from there.
76 You can check the assigned major numbers by typing
78 And look for cciss controllers
80 Once you have this, you need to create device nodes in
81 /dev/cciss directory. The nodes for the first 8 controllers
82 should already be created by mkdev.cciss script or
83 /etc/makedev.d/cciss script. You can add the major number(s)
84 in those scripts, or create the nodes manually by using
87 You can also use mknod_dyn.cciss and rmnod_dyn.cciss scripts
88 to create or remove nodes easily. These scripts can be found
89 in the Documentation directory.
91 Then you can mount the devices and create partitions
92 (You also need to make nodes for these partitions).
94 As for the minor number, the disk device will have a minor
95 number divisible by 16 (e.g: 0, 16, 32 etc), and the
96 partitions on those disk devices will have the minor number
97 of the disk device plus the partition number (1-15).
98 For example, disk d2 will have minor number 32, and its
99 partitions 1 and 2 will have minor numbers 33 and 34.
101 Look at the mkdev.cciss script for example.
104 In 2.4 kernel, partition names are hard coded in
105 /usr/src/linux/fs/partitions/check.c
106 and only for the first 8 cciss controllers. The rest
107 will be reported as ccissXX. This should not affect
108 I/O operation or performance. Please apply the
109 cciss_2.4_partition_name.patch to address this. This
110 will not be an issue under 2.5 kernel, since partition
111 names will be handled by the driver.
113 SCSI tape drive and medium changer support
114 ------------------------------------------
116 SCSI sequential access devices and medium changer devices are supported and
117 appropriate device nodes are automatically created. (e.g.
118 /dev/st0, /dev/st1, etc. See the "st" man page for more details.)
119 You must enable "SCSI tape drive support for Smart Array 5xxx" and
120 "SCSI support" in your kernel configuration to be able to use SCSI
121 tape drives with your Smart Array 5xxx controller.
123 Additionally, note that the driver will not engage the SCSI core at init
124 time. The driver must be directed to dynamically engage the SCSI core via
125 the /proc filesystem entry which the "block" side of the driver creates as
126 /proc/driver/cciss/cciss* at runtime. This is because at driver init time,
127 the SCSI core may not yet be initialized (because the driver is a block
128 driver) and attempting to register it with the SCSI core in such a case
129 would cause a hang. This is best done via an initialization script
130 (typically in /etc/init.d, but could vary depending on distibution).
133 for x in /proc/driver/cciss/cciss[0-9]*
135 echo "engage scsi" > $x
138 Once the SCSI core is engaged by the driver, it cannot be disengaged
139 (except by unloading the driver, if it happens to be linked as a module.)
141 Note also that if no sequential access devices or medium changers are
142 detected, the SCSI core will not be engaged by the action of the above
145 Hot plug support for SCSI tape drives
146 -------------------------------------
148 Hot plugging of SCSI tape drives is supported, with some caveats.
149 The cciss driver must be informed that changes to the SCSI bus
150 have been made, in addition to and prior to informing the SCSI
151 mid layer. This may be done via the /proc filesystem. For example:
153 echo "rescan" > /proc/scsi/cciss0/1
155 This causes the adapter to query the adapter about changes to the
156 physical SCSI buses and/or fibre channel arbitrated loop and the
157 driver to make note of any new or removed sequential access devices
158 or medium changers. The driver will output messages indicating what
159 devices have been added or removed and the controller, bus, target and
160 lun used to address the device. Once this is done, the SCSI mid layer
161 can be informed of changes to the virtual SCSI bus which the driver
162 presents to it in the usual way. For example:
164 echo scsi add-single-device 3 2 1 0 > /proc/scsi/scsi
166 to add a device on controller 3, bus 2, target 1, lun 0. Note that
167 the driver makes an effort to preserve the devices positions
168 in the virtual SCSI bus, so if you are only moving tape drives
169 around on the same adapter and not adding or removing tape drives
170 from the adapter, informing the SCSI mid layer may not be necessary.
172 Note that the naming convention of the /proc filesystem entries
173 contains a number in addition to the driver name. (E.g. "cciss0"
174 instead of just "cciss" which you might expect.) This is because
175 of changes to the 2.4 kernel PCI interface related to PCI hot plug
176 that imply the driver must register with the SCSI mid layer once per
177 adapter instance rather than once per driver.
179 Note: ONLY sequential access devices and medium changers are presented
180 as SCSI devices to the SCSI mid layer by the cciss driver. Specifically,
181 physical SCSI disk drives are NOT presented to the SCSI mid layer. The
182 physical SCSI disk drives are controlled directly by the array controller
183 hardware and it is important to prevent the OS from attempting to directly
184 access these devices too, as if the array controller were merely a SCSI
185 controller in the same way that we are allowing it to access SCSI tape drives.
190 For multipath configurations (acheived via a higher level driver, such
191 as the "md" driver) it is important that failure of a controller is detected.
192 Ordinarily, the driver is entirely interrupt driven. If a failure occurs
193 in such a way that the processor cannot receive interrupts from an adapter,
194 the driver will wait forever for i/o's to complete. In a multipath
195 configuration this is undesirable, as the md driver relies on i/o's being
196 reported as failed by the low level driver to trigger failing over to an
197 alternate controller. The monitor threads allow the driver to detect such
198 situations and report outstanding i/o's as having failed so that recovery
199 actions such switching to an alternate controller can occur. The monitor
200 threads periodically sends a trivial "no-operation" command down to
201 the controllers and expect them to complete within a a reasonable (short)
202 time period. The firmware on the adapter is designed such that no matter
203 how busy the adapter is serving i/o, it can respond quickly to a
204 "no-operation" command. In the event that a deadline elapses before a no
205 operation command completes, all outstanding commands on that controller
206 are reported back to the upper layers as having failed, and any new commands
207 sent to the controller are immediately reported back as failed.
209 To enable the monitor threads, the compile time option must be enabled
210 (via the usual linux kernel configuration) and the monitor thread must
211 be enabled at runtime as well. A system may have many adapters, but
212 perhaps only a single pair operating in a multipath configuration.
213 In this way, it is possible to run monitoring threads only for those
214 adapters which require it.
216 To start a monitoring thread on the first cciss adapter, "cciss0" with
217 a polling interval of 30 seconds, execute the following command:
219 echo "monitor 30" > /proc/driver/cciss/cciss0
221 To change the polling interval, to say, 60 seconds:
223 echo "monitor 60" > /proc/driver/cciss/cciss0
225 (Note, the change will not take effect until the previous polling
228 To disable the monitoring thread, set the polling interval to 0 seconds:
230 echo "monitor 0" > /proc/driver/cciss/cciss0
232 (Again, the monitoring thread will not exit until the previous polling
235 The minimum monitoring period is 10 seconds, and the maximum monitoring
236 period is 3600 seconds (1 hour). The no-operation command must complete
237 with 5 seconds of submission in all cases or the controller will be presumed