.\" .\" Copyright (c) 1999 Kenneth D. Merry. .\" All rights reserved. .\" .\" Redistribution and use in source and binary forms, with or without .\" modification, are permitted provided that the following conditions .\" are met: .\" 1. Redistributions of source code must retain the above copyright .\" notice, this list of conditions and the following disclaimer. .\" 2. The name of the author may not be used to endorse or promote products .\" derived from this software without specific prior written permission. .\" .\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE .\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF .\" SUCH DAMAGE. .\" .\" $FreeBSD: releng/12.2/share/man/man4/pci.4 349769 2019-07-05 22:21:02Z zeising $ .\" .Dd June 17, 2019 .Dt PCI 4 .Os .Sh NAME .Nm pci .Nd generic PCI bus driver .Sh SYNOPSIS To compile the PCI bus driver into the kernel, place the following line in your kernel configuration file: .Bd -ragged -offset indent .Cd device pci .Ed .Pp To compile in support for Single Root I/O Virtualization .Pq SR-IOV : .Bd -ragged -offset indent .Cd options PCI_IOV .Ed .Pp To compile in support for native PCI-express HotPlug: .Bd -ragged -offset indent .Cd options PCI_HP .Ed .Sh DESCRIPTION The .Nm driver provides support for .Tn PCI devices in the kernel and limited access to .Tn PCI devices for userland. .Pp The .Nm driver provides a .Pa /dev/pci character device that can be used by userland programs to read and write .Tn PCI configuration registers. Programs can also use this device to get a list of all .Tn PCI devices, or all .Tn PCI devices that match various patterns. .Pp Since the .Nm driver provides a write interface for .Tn PCI configuration registers, system administrators should exercise caution when granting access to the .Nm device. If used improperly, this driver can allow userland applications to crash a machine or cause data loss. .Pp The .Nm driver implements the .Tn PCI bus in the kernel. It enumerates any devices on the .Tn PCI bus and gives .Tn PCI client drivers the chance to attach to them. It assigns resources to children, when the BIOS does not. It takes care of routing interrupts when necessary. It reprobes the unattached .Tn PCI children when .Tn PCI client drivers are dynamically loaded at runtime. The .Nm driver also includes support for PCI-PCI bridges, various platform-specific Host-PCI bridges, and basic support for .Tn PCI VGA adapters. .Sh IOCTLS The following .Xr ioctl 2 calls are supported by the .Nm driver. They are defined in the header file .In sys/pciio.h . .Bl -tag -width 012345678901234 .It PCIOCGETCONF This .Xr ioctl 2 takes a .Va pci_conf_io structure. It allows the user to retrieve information on all .Tn PCI devices in the system, or on .Tn PCI devices matching patterns supplied by the user. The call may set .Va errno to any value specified in either .Xr copyin 9 or .Xr copyout 9 . The .Va pci_conf_io structure consists of a number of fields: .Bl -tag -width match_buf_len .It pat_buf_len The length, in bytes, of the buffer filled with user-supplied patterns. .It num_patterns The number of user-supplied patterns. .It patterns Pointer to a buffer filled with user-supplied patterns. .Va patterns is a pointer to .Va num_patterns .Va pci_match_conf structures. The .Va pci_match_conf structure consists of the following elements: .Bl -tag -width pd_vendor .It pc_sel .Tn PCI domain, bus, slot and function. .It pd_name .Tn PCI device driver name. .It pd_unit .Tn PCI device driver unit number. .It pc_vendor .Tn PCI vendor ID. .It pc_device .Tn PCI device ID. .It pc_class .Tn PCI device class. .It flags The flags describe which of the fields the kernel should match against. A device must match all specified fields in order to be returned. The match flags are enumerated in the .Va pci_getconf_flags structure. Hopefully the flag values are obvious enough that they do not need to described in detail. .El .It match_buf_len Length of the .Va matches buffer allocated by the user to hold the results of the .Dv PCIOCGETCONF query. .It num_matches Number of matches returned by the kernel. .It matches Buffer containing matching devices returned by the kernel. The items in this buffer are of type .Va pci_conf , which consists of the following items: .Bl -tag -width pc_subvendor .It pc_sel .Tn PCI domain, bus, slot and function. .It pc_hdr .Tn PCI header type. .It pc_subvendor .Tn PCI subvendor ID. .It pc_subdevice .Tn PCI subdevice ID. .It pc_vendor .Tn PCI vendor ID. .It pc_device .Tn PCI device ID. .It pc_class .Tn PCI device class. .It pc_subclass .Tn PCI device subclass. .It pc_progif .Tn PCI device programming interface. .It pc_revid .Tn PCI revision ID. .It pd_name Driver name. .It pd_unit Driver unit number. .El .It offset The offset is passed in by the user to tell the kernel where it should start traversing the device list. The value passed out by the kernel points to the record immediately after the last one returned. The user may pass the value returned by the kernel in subsequent calls to the .Dv PCIOCGETCONF ioctl. If the user does not intend to use the offset, it must be set to zero. .It generation .Tn PCI configuration generation. This value only needs to be set if the offset is set. The kernel will compare the current generation number of its internal device list to the generation passed in by the user to determine whether its device list has changed since the user last called the .Dv PCIOCGETCONF ioctl. If the device list has changed, a status of .Va PCI_GETCONF_LIST_CHANGED will be passed back. .It status The status tells the user the disposition of his request for a device list. The possible status values are: .Bl -ohang .It PCI_GETCONF_LAST_DEVICE This means that there are no more devices in the PCI device list matching the specified criteria after the ones returned in the .Va matches buffer. .It PCI_GETCONF_LIST_CHANGED This status tells the user that the .Tn PCI device list has changed since his last call to the .Dv PCIOCGETCONF ioctl and he must reset the .Va offset and .Va generation to zero to start over at the beginning of the list. .It PCI_GETCONF_MORE_DEVS This tells the user that his buffer was not large enough to hold all of the remaining devices in the device list that match his criteria. .It PCI_GETCONF_ERROR This indicates a general error while servicing the user's request. If the .Va pat_buf_len is not equal to .Va num_patterns times .Fn sizeof "struct pci_match_conf" , .Va errno will be set to .Er EINVAL . .El .El .It PCIOCREAD This .Xr ioctl 2 reads the .Tn PCI configuration registers specified by the passed-in .Va pci_io structure. The .Va pci_io structure consists of the following fields: .Bl -tag -width pi_width .It pi_sel A .Va pcisel structure which specifies the domain, bus, slot and function the user would like to query. If the specific bus is not found, errno will be set to ENODEV and -1 returned from the ioctl. .It pi_reg The .Tn PCI configuration registers the user would like to access. .It pi_width The width, in bytes, of the data the user would like to read. This value may be either 1, 2, or 4. 3-byte reads and reads larger than 4 bytes are not supported. If an invalid width is passed, errno will be set to EINVAL. .It pi_data The data returned by the kernel. .El .It PCIOCWRITE This .Xr ioctl 2 allows users to write to the .Tn PCI configuration registers specified in the passed-in .Va pci_io structure. The .Va pci_io structure is described above. The limitations on data width described for reading registers, above, also apply to writing .Tn PCI configuration registers. .It PCIOCATTACHED This .Xr ioctl 2 allows users to query if a driver is attached to the .Tn PCI device specified in the passed-in .Va pci_io structure. The .Va pci_io structure is described above, however, the .Va pi_reg and .Va pi_width fields are not used. The status of the device is stored in the .Va pi_data field. A value of 0 indicates no driver is attached, while a value larger than 0 indicates that a driver is attached. .It PCIOCBARMMAP This .Xr ioctl 2 command allows userspace processes to .Xr mmap 2 the memory-mapped PCI BAR into its address space. The input parameters and results are passed in the .Va pci_bar_mmap structure, which has the following fields: .Bl -tag -width Vt struct pcise pbm_sel .It Vt uint64_t pbm_map_base Reports the established mapping base to the caller. If .Va PCIIO_BAR_MMAP_FIXED flag was specified, then this field must be filled before the call with the desired address for the mapping. .It Vt uint64_t pbm_map_length Reports the mapped length of the BAR, in bytes. Its .Vt uint64_t value is always multiple of machine pages. .It Vt int64_t pbm_bar_length Reports length of the bar as exposed by the device. .It Vt int pbm_bar_off Reports offset from the mapped base to the start of the first register in the bar. .It Vt struct pcisel pbm_sel Should be filled before the call. Describes the device to operate on. .It Vt int pbm_reg The BAR index to mmap. .It Vt int pbm_flags Flags which augments the operation. See below. .It Vt int pbm_memattr The caching attribute for the mapping. Typical values are .Dv VM_MEMATTR_UNCACHEABLE for control registers BARs, and .Dv VM_MEMATTR_WRITE_COMBINING for frame buffers. Regular memory-like BAR should be mapped with .Dv VM_MEMATTR_DEFAULT attribute. .El .Pp Currently defined flags are: .Bl -tag -width PCIIO_BAR_MMAP_ACTIVATE .It PCIIO_BAR_MMAP_FIXED The resulted mappings should be established at the address specified by the .Va pbm_map_base member, otherwise fail. .It PCIIO_BAR_MMAP_EXCL Must be used together with .Vd PCIIO_BAR_MMAP_FIXED If the specified base contains already established mappings, the operation fails instead of implicitly unmapping them. .It PCIIO_BAR_MMAP_RW The requested mapping allows both reading and writing. Without the flag, read-only mapping is established. Note that it is common for the device registers to have side-effects even on reads. .It PCIIO_BAR_MMAP_ACTIVATE (Unimplemented) If the BAR is not activated, activate it in the course of mapping. Currently attempt to mmap an inactive BAR results in error. .El .El .Sh LOADER TUNABLES Tunables can be set at the .Xr loader 8 prompt before booting the kernel, or stored in .Xr loader.conf 5 . The current value of these tunables can be examined at runtime via .Xr sysctl 8 nodes of the same name. Unless otherwise specified, each of these tunables is a boolean that can be enabled by setting the tunable to a non-zero value. .Bl -tag -width indent .It Va hw.pci.clear_bars Pq Defaults to 0 Ignore any firmware-assigned memory and I/O port resources. This forces the .Tn PCI bus driver to allocate resource ranges for memory and I/O port resources from scratch. .It Va hw.pci.clear_buses Pq Defaults to 0 Ignore any firmware-assigned bus number registers in PCI-PCI bridges. This forces the .Tn PCI bus driver and PCI-PCI bridge driver to allocate bus numbers for secondary buses behind PCI-PCI bridges. .It Va hw.pci.clear_pcib Pq Defaults to 0 Ignore any firmware-assigned memory and I/O port resource windows in PCI-PCI bridges. This forces the PCI-PCI bridge driver to allocate memory and I/O port resources for resource windows from scratch. .Pp By default the PCI-PCI bridge driver will allocate windows that contain the firmware-assigned resources devices behind the bridge. In addition, the PCI-PCI bridge driver will suballocate from existing window regions when possible to satisfy a resource request. As a result, both .Va hw.pci.clear_bars and .Va hw.pci.clear_pcib must be enabled to fully ignore firmware-supplied resource assignments. .It Va hw.pci.default_vgapci_unit Pq Defaults to -1 By default, the first .Tn PCI VGA adapter encountered by the system is assumed to be the boot display device. This tunable can be set to choose a specific VGA adapter by specifying the unit number of the associated .Va vgapci Ns Ar X device. .It Va hw.pci.do_power_nodriver Pq Defaults to 0 Place devices into a low power state .Pq D3 when a suitable device driver is not found. Can be set to one of the following values: .Bl -tag -width indent .It 3 Powers down all .Tn PCI devices without a device driver. .It 2 Powers down most devices without a device driver. PCI devices with the display, memory, and base peripheral device classes are not powered down. .It 1 Similar to a setting of 2 except that storage controllers are also not powered down. .It 0 All devices are left fully powered. .El .Pp A .Tn PCI device must support power management to be powered down. Placing a device into a low power state may not reduce power consumption. .It Va hw.pci.do_power_resume Pq Defaults to 1 Place .Tn PCI devices into the fully powered state when resuming either the system or an individual device. Setting this to zero is discouraged as the system will not attempt to power up non-powered PCI devices after a suspend. .It Va hw.pci.do_power_suspend Pq Defaults to 1 Place .Tn PCI devices into a low power state when suspending either the system or individual devices. Normally the D3 state is used as the low power state, but firmware may override the desired power state during a system suspend. .It Va hw.pci.enable_ari Pq Defaults to 1 Enable support for PCI-express Alternative RID Interpretation. This is often used in conjunction with SR-IOV. .It Va hw.pci.enable_io_modes Pq Defaults to 1 Enable memory or I/O port decoding in a PCI device's command register if it has firmware-assigned memory or I/O port resources. The firmware .Pq BIOS in some systems does not enable memory or I/O port decoding for some devices even when it has assigned resources to the device. This enables decoding for such resources during bus probe. .It Va hw.pci.enable_msi Pq Defaults to 1 Enable support for Message Signalled Interrupts .Pq MSI . MSI interrupts can be disabled by setting this tunable to 0. .It Va hw.pci.enable_msix Pq Defaults to 1 Enable support for extended Message Signalled Interrupts .Pq MSI-X . MSI-X interrupts can be disabled by setting this tunable to 0. .It Va hw.pci.enable_pcie_hp Pq Defaults to 1 Enable support for native PCI-express HotPlug. .It Va hw.pci.honor_msi_blacklist Pq Defaults to 1 MSI and MSI-X interrupts are disabled for certain chipsets known to have broken MSI and MSI-X implementations when this tunable is set. It can be set to zero to permit use of MSI and MSI-X interrupts if the chipset match is a false positive. .It Va hw.pci.iov_max_config Pq Defaults to 1MB The maximum amount of memory permitted for the configuration parameters used when creating Virtual Functions via SR-IOV. This tunable can also be changed at runtime via .Xr sysctl 8 . .It Va hw.pci.realloc_bars Pq Defaults to 0 Attempt to allocate a new resource range during the initial device scan for any memory or I/O port resources with firmware-assigned ranges that conflict with another active resource. .It Va hw.pci.usb_early_takeover Pq Defaults to 1 on Tn amd64 and Tn i386 Disable legacy device emulation of USB devices during the initial device scan. Set this tunable to zero to use USB devices via legacy emulation when using a custom kernel without USB controller drivers. .It Va hw.pci...INT

.irq These tunables can be used to override the interrupt routing for legacy PCI INTx interrupts. Unlike other tunables in this list, these do not have corresponding sysctl nodes. The tunable name includes the address of the PCI device as well as the pin of the desired INTx IRQ to override: .Bl -tag -width indent .It The domain .Pq or segment of the PCI device in decimal. .It The bus address of the PCI device in decimal. .It The slot of the PCI device in decimal. .It

The interrupt pin of the PCI slot to override. One of .Ql A , .Ql B , .Ql C , or .Ql D . .El .Pp The value of the tunable is the raw IRQ value to use for the INTx interrupt pin identified by the tunable name. Mapping of IRQ values to platform interrupt sources is machine dependent. .El .Sh DEVICE WIRING You can wire the device unit at a given location with device.hints. Entries of the form .Va hints...at="pci::" or .Va hints...at="pci:::" will force the driver .Va name to probe and attach at unit .Va unit for any PCI device found to match the specification, where: .Bl -tag -width -indent .It The domain .Pq or segment of the PCI device in decimal. Defaults to 0 if unspecified .It The bus address of the PCI device in decimal. .It The slot of the PCI device in decimal. .It The function of the PCI device in decimal. .El .Pp The code to do the matching requires an exact string match. Do not specify the angle brackets .Pq < > in the hints file. Wiring multiple devices to the same .Va name and .Va unit produces undefined results. .Ss Examples Given the following lines in .Pa /boot/device.hints : .Cd hint.nvme.3.at="pci6:0:0" .Cd hint.igb.8.at="pci14:0:0" If there is a device that supports .Xr igb 4 at PCI bus 14 slot 0 function 0, then it will be assigned igb8 for probe and attach. Likewise, if there is an .Xr nvme 4 card at PCI bus 6 slot 0 function 0, then it will be assigned nvme3 for probe and attach. If another type of card is in either of these locations, the name and unit of that card will be the default names and will be unaffected by these hints. If other igb or nvme cards are located elsewhere, they will be assigned their unit numbers sequentially, skipping the unit numbers that have 'at' hints. .Sh FILES .Bl -tag -width /dev/pci -compact .It Pa /dev/pci Character device for the .Nm driver. .El .Sh SEE ALSO .Xr pciconf 8 .Sh HISTORY The .Nm driver (not the kernel's .Tn PCI support code) first appeared in .Fx 2.2 , and was written by Stefan Esser and Garrett Wollman. Support for device listing and matching was re-implemented by Kenneth Merry, and first appeared in .Fx 3.0 . .Sh AUTHORS .An Kenneth Merry Aq Mt ken@FreeBSD.org .Sh BUGS It is not possible for users to specify an accurate offset into the device list without calling the .Dv PCIOCGETCONF at least once, since they have no way of knowing the current generation number otherwise. This probably is not a serious problem, though, since users can easily narrow their search by specifying a pattern or patterns for the kernel to match against.