1This driver is for Compaq's SMART Array Controllers. 2 3Supported Cards: 4---------------- 5 6This driver is known to work with the following cards: 7 8 * SA 5300 9 * SA 5i 10 * SA 532 11 * SA 5312 12 * SA 641 13 * SA 642 14 * SA 6400 15 * SA 6400 U320 Expansion Module 16 * SA 6i 17 * SA P600 18 * SA P800 19 * SA P400 20 * SA E200 21 * SA E200i 22 * SA 344 23 * SA 544 24 * SA 380 25 * SA 580 26 * SA 380i 27 * SA 580i 28 * SA 308 29 * SA 508 30 * SA 388 31 * SA 588 32 33If nodes are not already created in the /dev/cciss directory 34 35# mkdev.cciss [ctlrs] 36 37Where ctlrs is the number of controllers you have (defaults to 1 if not 38specified). 39 40Device Naming: 41-------------- 42 43You need some entries in /dev for the cciss device. The mkdev.cciss script 44can make device nodes for you automatically. Currently the device setup 45is as follows: 46 47Major numbers: 48 104 cciss0 49 105 cciss1 50 106 cciss2 51 etc... 52 53Minor numbers: 54 b7 b6 b5 b4 b3 b2 b1 b0 55 |----+----| |----+----| 56 | | 57 | +-------- Partition ID (0=wholedev, 1-15 partition) 58 | 59 +-------------------- Logical Volume number 60 61The suggested device naming scheme is: 62/dev/cciss/c0d0 Controller 0, disk 0, whole device 63/dev/cciss/c0d0p1 Controller 0, disk 0, partition 1 64/dev/cciss/c0d0p2 Controller 0, disk 0, partition 2 65/dev/cciss/c0d0p3 Controller 0, disk 0, partition 3 66 67/dev/cciss/c1d1 Controller 1, disk 1, whole device 68/dev/cciss/c1d1p1 Controller 1, disk 1, partition 1 69/dev/cciss/c1d1p2 Controller 1, disk 1, partition 2 70/dev/cciss/c1d1p3 Controller 1, disk 1, partition 3 71 72SCSI tape drive and medium changer support 73------------------------------------------ 74 75SCSI sequential access devices and medium changer devices are supported and 76appropriate device nodes are automatically created. (e.g. 77/dev/st0, /dev/st1, etc. See the "st" man page for more details.) 78You must enable "SCSI tape drive support for Smart Array 5xxx" and 79"SCSI support" in your kernel configuration to be able to use SCSI 80tape drives with your Smart Array 5xxx controller. 81 82Additionally, note that the driver will not engage the SCSI core at init 83time. The driver must be directed to dynamically engage the SCSI core via 84the /proc filesystem entry which the "block" side of the driver creates as 85/proc/driver/cciss/cciss* at runtime. This is because at driver init time, 86the SCSI core may not yet be initialized (because the driver is a block 87driver) and attempting to register it with the SCSI core in such a case 88would cause a hang. This is best done via an initialization script 89(typically in /etc/init.d, but could vary depending on distibution). 90For example: 91 92 for x in /proc/driver/cciss/cciss[0-9]* 93 do 94 echo "engage scsi" > $x 95 done 96 97Once the SCSI core is engaged by the driver, it cannot be disengaged 98(except by unloading the driver, if it happens to be linked as a module.) 99 100Note also that if no sequential access devices or medium changers are 101detected, the SCSI core will not be engaged by the action of the above 102script. 103 104Hot plug support for SCSI tape drives 105------------------------------------- 106 107Hot plugging of SCSI tape drives is supported, with some caveats. 108The cciss driver must be informed that changes to the SCSI bus 109have been made, in addition to and prior to informing the SCSI 110mid layer. This may be done via the /proc filesystem. For example: 111 112 echo "rescan" > /proc/scsi/cciss0/1 113 114This causes the adapter to query the adapter about changes to the 115physical SCSI buses and/or fibre channel arbitrated loop and the 116driver to make note of any new or removed sequential access devices 117or medium changers. The driver will output messages indicating what 118devices have been added or removed and the controller, bus, target and 119lun used to address the device. Once this is done, the SCSI mid layer 120can be informed of changes to the virtual SCSI bus which the driver 121presents to it in the usual way. For example: 122 123 echo scsi add-single-device 3 2 1 0 > /proc/scsi/scsi 124 125to add a device on controller 3, bus 2, target 1, lun 0. Note that 126the driver makes an effort to preserve the devices positions 127in the virtual SCSI bus, so if you are only moving tape drives 128around on the same adapter and not adding or removing tape drives 129from the adapter, informing the SCSI mid layer may not be necessary. 130 131Note that the naming convention of the /proc filesystem entries 132contains a number in addition to the driver name. (E.g. "cciss0" 133instead of just "cciss" which you might expect.) This is because 134of changes to the 2.4 kernel PCI interface related to PCI hot plug 135that imply the driver must register with the SCSI mid layer once per 136adapter instance rather than once per driver. 137 138Note: ONLY sequential access devices and medium changers are presented 139as SCSI devices to the SCSI mid layer by the cciss driver. Specifically, 140physical SCSI disk drives are NOT presented to the SCSI mid layer. The 141physical SCSI disk drives are controlled directly by the array controller 142hardware and it is important to prevent the OS from attempting to directly 143access these devices too, as if the array controller were merely a SCSI 144controller in the same way that we are allowing it to access SCSI tape drives. 145 146SCSI error handling for tape drives and medium changers 147------------------------------------------------------- 148 149The linux SCSI mid layer provides an error handling protocol which 150kicks into gear whenever a SCSI command fails to complete within a 151certain amount of time (which can vary depending on the command). 152The cciss driver participates in this protocol to some extent. The 153normal protocol is a four step process. First the device is told 154to abort the command. If that doesn't work, the device is reset. 155If that doesn't work, the SCSI bus is reset. If that doesn't work 156the host bus adapter is reset. Because the cciss driver is a block 157driver as well as a SCSI driver and only the tape drives and medium 158changers are presented to the SCSI mid layer, and unlike more 159straightforward SCSI drivers, disk i/o continues through the block 160side during the SCSI error recovery process, the cciss driver only 161implements the first two of these actions, aborting the command, and 162resetting the device. Additionally, most tape drives will not oblige 163in aborting commands, and sometimes it appears they will not even 164obey a reset coommand, though in most circumstances they will. In 165the case that the command cannot be aborted and the device cannot be 166reset, the device will be set offline. 167 168In the event the error handling code is triggered and a tape drive is 169successfully reset or the tardy command is successfully aborted, the 170tape drive may still not allow i/o to continue until some command 171is issued which positions the tape to a known position. Typically you 172must rewind the tape (by issuing "mt -f /dev/st0 rewind" for example) 173before i/o can proceed again to a tape drive which was reset. 174 175