|CAM(4)||Device Drivers Manual||CAM(4)|
CAMsubsystem provides a uniform and modular system for the implementation of drivers to control various SCSI, ATA, NMVe, and MMC / SD devices, and to utilize different SCSI, ATA, NVMe, and MMC / SD host adapters through host adapter drivers. When the system probes buses, it attaches any devices it finds to the appropriate drivers. The pass(4) driver, if it is configured in the kernel, will attach to all devices.
KERNEL CONFIGURATION¶There are a number of generic kernel configuration options for the
- This option compiles in all the
CAMdebugging printf code. This will not actually cause any debugging information to be printed out when included by itself. See below for details.
- This sets the maximum allowable number of concurrent "high power" commands. A "high power" command is a command that takes more electrical power than most to complete. An example of this is the SCSI START UNIT command. Starting a disk often takes significantly more electrical power than normal operation. This option allows the user to specify how many concurrent high power commands may be outstanding without overloading the power supply on his computer.
- This eliminates text descriptions of each SCSI Additional Sense Code and Additional Sense Code Qualifier pair. Since this is a fairly large text database, eliminating it reduces the size of the kernel somewhat. This is primarily necessary for boot floppies and other low disk space or low memory space environments. In most cases, though, this should be enabled, since it speeds the interpretation of SCSI error messages. Do not let the "kernel bloat" zealots get to you -- leave the sense descriptions in your kernel!
- This disables text descriptions of each SCSI opcode. This option, like the sense string option above, is primarily useful for environments like a boot floppy where kernel size is critical. Enabling this option for normal use is not recommended, since it slows debugging of SCSI problems.
- This is the SCSI "bus settle delay." In
CAM, it is specified in milliseconds, not seconds like the old SCSI layer used to do. When the kernel boots, it sends a bus reset to each SCSI bus to tell each device to reset itself to a default set of transfer negotiations and other settings. Most SCSI devices need some amount of time to recover from a bus reset. Newer disks may need as little as 100ms, while old, slow devices may need much longer. If the
SCSI_DELAYis not specified, it defaults to 2 seconds. The minimum allowable value for
SCSI_DELAYis "100", or 100ms. One special case is that if the
SCSI_DELAYis set to 0, that will be taken to mean the "lowest possible value." In that case, the
SCSI_DELAYwill be reset to 100ms.
All devices and buses support dynamic allocation so that an upper
number of devices and controllers does not need to be configured;
device da will suffice for any number of disk
The devices are either wired so they appear as a particular device unit or counted so that they appear as the next available unused unit.
Units are wired down by setting kernel environment hints. This is usually done either interactively from the loader(8), or automatically via the /boot/device.hints file. The basic syntax is:
CAM bus numbers can be wired
down to specific controllers with a config line similar to the
CAM bus number 0 to the
ahd1 driver instance. For controllers supporting more than
one bus, a particular bus can be assigned as follows:
CAM bus 0 to the bus 1
instance on ahc1. Peripheral drivers can be wired to a
specific bus, target, and lun as so:
hint.da.0.at="scbus0" hint.da.0.target="0" hint.da.0.unit="0"
This assigns da0 to target 0, unit (lun) 0 of
scbus 0. Omitting the target or unit hints will instruct
CAM to treat them as wildcards and use the first
respective counted instances. These examples can be combined together to
allow a peripheral device to be wired to any particular controller, bus,
target, and/or unit instance.
This also works with nvme(4) drives as well.
hint.nvme.4.at="pci7:0:0" hint.scbus.10.at="nvme4" hint.nda.10.at="scbus10" hint.nda.10.target="1" hint.nda.10.unit="12" hint.nda.11.at="scbus10" hint.nda.11.target="1" hint.nda.11.unit="2"
This assigns the NVMe card living at PCI bus 7 to scbus 10 (in PCIe, slot and function are rarely used and usually 0). The target for nda(4) devices is always 1. The unit is the namespace identifier from the drive. The namespace id 1 is exported as nda10 and namespace id 2 is exported as nda11.
When you have a mixture of wired down and counted devices then the counting begins with the first non-wired down unit for a particular type. That is, if you have a disk wired down as device da1, then the first non-wired disk shall come on line as da2.
ADAPTERS¶The system allows common device drivers to work through many different types of adapters. The adapters take requests from the upper layers and do all IO between the SCSI, ATA, NVMe, or MMC / SD bus and the system. The maximum size of a transfer is governed by the adapter. Most adapters can transfer 64KB in a single operation, however many can transfer larger amounts.
TARGET MODE¶Some adapters support target mode in which the system is capable of operating as a device, responding to operations initiated by another system. Target mode is supported for some adapters, but is not yet complete for this version of the
DIAGNOSTICS¶An XPT_DEBUG CCB can be used to enable various amounts of tracing information on any specific bus/device from the list of options compiled into the kernel. There are currently seven debugging flags that may be compiled in and used:
- This flag enables general informational printfs for the device or devices in question.
- This flag enables function-level command flow tracing i.e., kernel printfs will happen at the entrance and exit of various functions.
- This flag enables debugging output internal to various functions.
- This flag will cause the kernel to print out all ATA and SCSI commands sent to a particular device or devices.
- This flag will enable command scheduler tracing.
- This flag will enable peripheral drivers messages.
- This flag will enable devices probe process tracing.
Some of these flags, most notably
CAM_DEBUG_SUBTRACE, will produce kernel printfs in
Users can enable debugging from their kernel config file, by using the following kernel config options:
- This builds into the kernel all possible
- This allows to specify support for which debugging flags described above should be built into the kernel. Flags may be ORed together if the user wishes to see printfs for multiple debugging levels.
- This allows to set the various debugging flags from a kernel config file.
- Specify a bus to debug. To debug all buses, set this to -1.
- Specify a target to debug. To debug all targets, set this to -1.
- Specify a lun to debug. To debug all luns, set this to -1.
SEE ALSO¶ada(4), aha(4), ahc(4), ahci(4), ahd(4), ata(4), bt(4), cd(4), ch(4), da(4), nda(4), nvme(4), pass(4), pt(4), sa(4), xpt(4), camcontrol(8)
CAMSCSI subsystem first appeared in FreeBSD 3.0. The
CAMATA support was added in FreeBSD 8.0.
CAMSCSI subsystem was written by Justin Gibbs and Kenneth Merry. The
CAMATA support was added by Alexander Motin <mav@FreeBSD.org>. The
CAMNVMe support was added by Warner Losh <imp@FreeBSD.org>.
|December 20, 2017||Linux 4.19.0-10-amd64|