diff options
Diffstat (limited to 'static/freebsd/man4/nvme.4 3.html')
| -rw-r--r-- | static/freebsd/man4/nvme.4 3.html | 220 |
1 files changed, 0 insertions, 220 deletions
diff --git a/static/freebsd/man4/nvme.4 3.html b/static/freebsd/man4/nvme.4 3.html deleted file mode 100644 index e189e085..00000000 --- a/static/freebsd/man4/nvme.4 3.html +++ /dev/null @@ -1,220 +0,0 @@ -<table class="head"> - <tr> - <td class="head-ltitle">NVME(4)</td> - <td class="head-vol">Device Drivers Manual</td> - <td class="head-rtitle">NVME(4)</td> - </tr> -</table> -<div class="manual-text"> -<section class="Sh"> -<h1 class="Sh" id="NAME"><a class="permalink" href="#NAME">NAME</a></h1> -<p class="Pp"><code class="Nm">nvme</code> — <span class="Nd">NVM Express - core driver</span></p> -</section> -<section class="Sh"> -<h1 class="Sh" id="SYNOPSIS"><a class="permalink" href="#SYNOPSIS">SYNOPSIS</a></h1> -<p class="Pp">To compile this driver into your kernel, place the following line - in your kernel configuration file:</p> -<div class="Bd Pp Bd-indent"><code class="Cd">device nvme</code></div> -<p class="Pp">Or, to load the driver as a module at boot, place the following - line in <a class="Xr">loader.conf(5)</a>:</p> -<div class="Bd Pp Bd-indent Li"> -<pre>nvme_load="YES"</pre> -</div> -<p class="Pp">Most users will also want to enable <a class="Xr">nvd(4)</a> or - <a class="Xr">nda(4)</a> to expose NVM Express namespaces as disk devices - which can be partitioned. Note that in NVM Express terms, a namespace is - roughly equivalent to a SCSI LUN.</p> -</section> -<section class="Sh"> -<h1 class="Sh" id="DESCRIPTION"><a class="permalink" href="#DESCRIPTION">DESCRIPTION</a></h1> -<p class="Pp">The <code class="Nm">nvme</code> driver provides support for NVM - Express (NVMe) controllers, such as:</p> -<ul class="Bl-bullet"> - <li>Hardware initialization</li> - <li>Per-CPU IO queue pairs</li> - <li>API for registering NVMe namespace consumers such as - <a class="Xr">nvd(4)</a> or <a class="Xr">nda(4)</a></li> - <li>API for submitting NVM commands to namespaces</li> - <li>Ioctls for controller and namespace configuration and management</li> -</ul> -<p class="Pp">The <code class="Nm">nvme</code> driver creates controller device - nodes in the format <span class="Pa">/dev/nvmeX</span> and namespace device - nodes in the format <span class="Pa">/dev/nvmeXnsY</span>. Note that the NVM - Express specification starts numbering namespaces at 1, not 0, and this - driver follows that convention.</p> -</section> -<section class="Sh"> -<h1 class="Sh" id="CONFIGURATION"><a class="permalink" href="#CONFIGURATION">CONFIGURATION</a></h1> -<p class="Pp">By default, <code class="Nm">nvme</code> will create an I/O queue - pair for each CPU, provided enough MSI-X vectors and NVMe queue pairs can be - allocated. If not enough vectors or queue pairs are available, nvme(4) will - use a smaller number of queue pairs and assign multiple CPUs per queue - pair.</p> -<p class="Pp">To force a single I/O queue pair shared by all CPUs, set the - following tunable value in <a class="Xr">loader.conf(5)</a>:</p> -<div class="Bd Pp Bd-indent Li"> -<pre>hw.nvme.per_cpu_io_queues=0</pre> -</div> -<p class="Pp">To assign more than one CPU per I/O queue pair, thereby reducing - the number of MSI-X vectors consumed by the device, set the following - tunable value in <a class="Xr">loader.conf(5)</a>:</p> -<div class="Bd Pp Bd-indent Li"> -<pre>hw.nvme.min_cpus_per_ioq=X</pre> -</div> -<p class="Pp">To force legacy interrupts for all <code class="Nm">nvme</code> - driver instances, set the following tunable value in - <a class="Xr">loader.conf(5)</a>:</p> -<div class="Bd Pp Bd-indent Li"> -<pre>hw.nvme.force_intx=1</pre> -</div> -<p class="Pp">Note that use of INTx implies disabling of per-CPU I/O queue - pairs.</p> -<p class="Pp">To control maximum amount of system RAM in bytes to use as Host - Memory Buffer for capable devices, set the following tunable:</p> -<div class="Bd Pp Bd-indent Li"> -<pre>hw.nvme.hmb_max</pre> -</div> -<p class="Pp">The default value is 5% of physical memory size per device.</p> -<p class="Pp">To enable Autonomous Power State Transition (APST), set the - following tunable value in <a class="Xr">loader.conf(5)</a>:</p> -<div class="Bd Pp Bd-indent Li"> -<pre>hw.nvme.apst_enable=1</pre> -</div> -<p class="Pp">The default vendor-provided settings, if any, will be applied. To - override this, set the following tunable:</p> -<div class="Bd Pp Bd-indent Li"> -<pre>hw.nvme.apst_data</pre> -</div> -<p class="Pp">The string must contain up to 32 encoded integers, e.g. - "0x6418 0 0 0x3e820". Each value corresponds to a specific - available power state starting from the lowest, and defines the target state - (bits 3..7) to transition to, as well as the idle time in milliseconds (bits - 8..31) to wait before that transition. Bits 0..2 must be zero.</p> -<p class="Pp">The <a class="Xr">nvd(4)</a> driver is used to provide a disk - driver to the system by default. The <a class="Xr">nda(4)</a> driver can - also be used instead. The <a class="Xr">nvd(4)</a> driver performs better - with smaller transactions and few TRIM commands. It sends all commands - directly to the drive immediately. The <a class="Xr">nda(4)</a> driver - performs better with larger transactions and also collapses TRIM commands - giving better performance. It can queue commands to the drive; combine - <code class="Dv">BIO_DELETE</code> commands into a single trip; and use the - CAM I/O scheduler to bias one type of operation over another. To select the - <a class="Xr">nda(4)</a> driver, set the following tunable value in - <a class="Xr">loader.conf(5)</a>:</p> -<div class="Bd Pp Bd-indent Li"> -<pre>hw.nvme.use_nvd=0</pre> -</div> -<p class="Pp">This value may also be set in the kernel config file with</p> -<div class="Bd Pp Bd-indent Li"> -<pre><code class="Cd">options NVME_USE_NVD=0</code></pre> -</div> -<p class="Pp">When there is an error, <code class="Nm">nvme</code> prints only - the most relevant information about the command by default. To enable - dumping of all information about the command, set the following tunable - value in <a class="Xr">loader.conf(5)</a>:</p> -<div class="Bd Pp Bd-indent Li"> -<pre>hw.nvme.verbose_cmd_dump=1</pre> -</div> -<p class="Pp">Prior versions of the driver reset the card twice on boot. This - proved to be unnecessary and inefficient, so the driver now resets drive - controller only once. The old behavior may be restored in the kernel config - file with</p> -<div class="Bd Pp Bd-indent Li"> -<pre><code class="Cd">options NVME_2X_RESET</code></pre> -</div> -</section> -<section class="Sh"> -<h1 class="Sh" id="SYSCTL_VARIABLES"><a class="permalink" href="#SYSCTL_VARIABLES">SYSCTL - VARIABLES</a></h1> -<p class="Pp">The following controller-level sysctls are currently - implemented:</p> -<dl class="Bl-tag"> - <dt id="dev.nvme.0.num_cpus_per_ioq"><var class="Va">dev.nvme.0.num_cpus_per_ioq</var></dt> - <dd>(R) Number of CPUs associated with each I/O queue pair.</dd> - <dt id="dev.nvme.0.int_coal_time"><var class="Va">dev.nvme.0.int_coal_time</var></dt> - <dd>(R/W) Interrupt coalescing timer period in microseconds. Set to 0 to - disable.</dd> - <dt id="dev.nvme.0.int_coal_threshold"><var class="Va">dev.nvme.0.int_coal_threshold</var></dt> - <dd>(R/W) Interrupt coalescing threshold in number of command completions. Set - to 0 to disable.</dd> -</dl> -<p class="Pp">The following queue pair-level sysctls are currently implemented. - Admin queue sysctls take the format of dev.nvme.0.adminq and I/O queue - sysctls take the format of dev.nvme.0.ioq0.</p> -<dl class="Bl-tag"> - <dt id="dev.nvme.0.ioq0.num_entries"><var class="Va">dev.nvme.0.ioq0.num_entries</var></dt> - <dd>(R) Number of entries in this queue pair's command and completion - queue.</dd> - <dt id="dev.nvme.0.ioq0.num_tr"><var class="Va">dev.nvme.0.ioq0.num_tr</var></dt> - <dd>(R) Number of nvme_tracker structures currently allocated for this queue - pair.</dd> - <dt id="dev.nvme.0.ioq0.num_prp_list"><var class="Va">dev.nvme.0.ioq0.num_prp_list</var></dt> - <dd>(R) Number of nvme_prp_list structures currently allocated for this queue - pair.</dd> - <dt id="dev.nvme.0.ioq0.sq_head"><var class="Va">dev.nvme.0.ioq0.sq_head</var></dt> - <dd>(R) Current location of the submission queue head pointer as observed by - the driver. The head pointer is incremented by the controller as it takes - commands off of the submission queue.</dd> - <dt id="dev.nvme.0.ioq0.sq_tail"><var class="Va">dev.nvme.0.ioq0.sq_tail</var></dt> - <dd>(R) Current location of the submission queue tail pointer as observed by - the driver. The driver increments the tail pointer after writing a command - into the submission queue to signal that a new command is ready to be - processed.</dd> - <dt id="dev.nvme.0.ioq0.cq_head"><var class="Va">dev.nvme.0.ioq0.cq_head</var></dt> - <dd>(R) Current location of the completion queue head pointer as observed by - the driver. The driver increments the head pointer after finishing with a - completion entry that was posted by the controller.</dd> - <dt id="dev.nvme.0.ioq0.num_cmds"><var class="Va">dev.nvme.0.ioq0.num_cmds</var></dt> - <dd>(R) Number of commands that have been submitted on this queue pair.</dd> - <dt id="dev.nvme.0.ioq0.dump_debug"><var class="Va">dev.nvme.0.ioq0.dump_debug</var></dt> - <dd>(W) Writing 1 to this sysctl will dump the full contents of the submission - and completion queues to the console.</dd> -</dl> -<p class="Pp">In addition to the typical pci attachment, the - <code class="Nm">nvme</code> driver supports attaching to a - <a class="Xr">ahci(4)</a> device. Intel's Rapid Storage Technology (RST) - hides the nvme device behind the AHCI device due to limitations in Windows. - However, this effectively hides it from the <span class="Ux">FreeBSD</span> - kernel. To work around this limitation, <span class="Ux">FreeBSD</span> - detects that the AHCI device supports RST and when it is enabled. See - <a class="Xr">ahci(4)</a> for more details.</p> -</section> -<section class="Sh"> -<h1 class="Sh" id="DIAGNOSTICS"><a class="permalink" href="#DIAGNOSTICS">DIAGNOSTICS</a></h1> -<dl class="Bl-diag"> - <dt>nvme%d: System interrupt issues?</dt> - <dd>The driver found a timed-out transaction had a pending completion record, - indicating an interrupt had not been delivered. The system is either not - configuring interrupts properly, or the system drops them under load. This - message will appear at most once per boot per controller.</dd> -</dl> -</section> -<section class="Sh"> -<h1 class="Sh" id="SEE_ALSO"><a class="permalink" href="#SEE_ALSO">SEE - ALSO</a></h1> -<p class="Pp"><a class="Xr">nda(4)</a>, <a class="Xr">nvd(4)</a>, - <a class="Xr">pci(4)</a>, <a class="Xr">nvmecontrol(8)</a>, - <a class="Xr">disk(9)</a></p> -</section> -<section class="Sh"> -<h1 class="Sh" id="HISTORY"><a class="permalink" href="#HISTORY">HISTORY</a></h1> -<p class="Pp">The <code class="Nm">nvme</code> driver first appeared in - <span class="Ux">FreeBSD 9.2</span>.</p> -</section> -<section class="Sh"> -<h1 class="Sh" id="AUTHORS"><a class="permalink" href="#AUTHORS">AUTHORS</a></h1> -<p class="Pp">The <code class="Nm">nvme</code> driver was developed by Intel and - originally written by <span class="An">Jim Harris</span> - <<a class="Mt" href="mailto:jimharris@FreeBSD.org">jimharris@FreeBSD.org</a>>, - with contributions from <span class="An">Joe Golio</span> at EMC.</p> -<p class="Pp">This man page was written by <span class="An">Jim Harris</span> - <<a class="Mt" href="mailto:jimharris@FreeBSD.org">jimharris@FreeBSD.org</a>>.</p> -</section> -</div> -<table class="foot"> - <tr> - <td class="foot-date">June 6, 2020</td> - <td class="foot-os">FreeBSD 15.0</td> - </tr> -</table> |