Changes in sysfs-bus-pci

2311ab18 - PCI/DOE: Expose DOE features via sysfs

Thu, 06 Mar 2025 07:52:10 +0000

PCI/DOE: Expose DOE features via sysfsPCIe r6.0 added support for Data Object Exchange (DOE). When DOE issupported, the DOE Discovery Feature must be implemented per PCIe r6.1, sec6.30.1.1. DOE allows a requester to obtain information about the other DOEfeatures supported by the device.The kernel already queries the DOE features supported and caches thevalues. Expose the values in sysfs to allow user space to determine whichDOE features are supported by the PCIe device.By exposing the information to userspace, tools like lspci can relay theinformation to users. By listing all of the supported features we can allowuserspace to parse the list, which might include vendor specific featuresas well as yet to be supported features.As the DOE Discovery feature must always be supported we treat it as aspecial named attribute case. This allows the usual PCI attribute_grouphandling to correctly create the doe_features directory when registeringpci_doe_sysfs_group (otherwise it doesn't and sysfs_add_file_to_group()will seg fault).After this patch is supported you can see something like this whenattaching a DOE device: $ ls /sys/devices/pci0000:00/0000:00:02.0//doe* 0001:01 0001:02 doe_discoveryLink: https://lore.kernel.org/r/20250306075211.1855177-3-alistair@alistair23.meSigned-off-by: Alistair Francis [bhelgaas: drop pci_doe_sysfs_init() stub return, makeDEVICE_ATTR_RO(doe_discovery) static]Signed-off-by: Bjorn Helgaas Reviewed-by: Jonathan Cameron List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

2fa04644 - PCI: Add 'reset_subordinate' to reset hierarchy below bridge

Fri, 25 Oct 2024 22:27:54 +0000

PCI: Add 'reset_subordinate' to reset hierarchy below bridgeThe "bus" and "cxl_bus" reset methods reset a device by asserting SecondaryBus Reset on the bridge leading to the device. These only work if thedevice is the only device below the bridge.Add a sysfs 'reset_subordinate' attribute on bridges that can assertSecondary Bus Reset regardless of how many devices are below the bridge.This resets all the devices below a bridge in a single command, includingthe locking and config space save/restore that reset methods normally do.This may be the only way to reset devices that don't support other resetmethods (ACPI, FLR, PM reset, etc).Link: https://lore.kernel.org/r/20241025222755.3756162-1-kbusch@meta.comSigned-off-by: Keith Busch [bhelgaas: commit log, add capable(CAP_SYS_ADMIN) check]Signed-off-by: Bjorn Helgaas Reviewed-by: Alex Williamson Reviewed-by: Amey Narkhede List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

759ec282 - PCI/NPEM: Add _DSM PCIe SSD status LED management

Wed, 04 Sep 2024 10:48:48 +0000

PCI/NPEM: Add _DSM PCIe SSD status LED managementThe PCIe SSD Status LED Management _DSM defined in PCI Firmware Spec r3.3sec 4.7 provides a way to manage LEDs via ACPI.The design is similar to NPEM defined in PCIe Base Specification r6.1 sec6.28: - Both standards are indication oriented, - _DSM supported bits correspond to NPEM capability register bits, - _DSM control bits correspond to NPEM control register bits._DSM does not support enclosure-specific indications or the special NPEMcommands NPEM_ENABLE and NPEM_RESET._DSM is implemented as a second backend in NPEM driver. The backend used islogged with info priority. The same sysfs interface is used for both NPEMand _DSM.According to spec, _DSM has higher priority, and availability of _DSM innot limited to devices with NPEM support.The Dell implementation of DSM uses acpi ipmi, which may not be availableimmediately (in fact it may take up to 10s for this interface to beavailable). It can determine if DSM is supported (GET_SUPPORTED_STATES_DSMis working) but it cannot serve GET_STATE_DSM or SET_STATE_DSM commands inthis time.From userspace application perspective (primarily configured by systemdservice) it is better to have not working but configured interface ratherthan have it available after few seconds.For that reason, npem->active_indications cache is now loaded lazily, i.e.any GET or SET request want cache to be updated if it is not done yet.Link: https://lore.kernel.org/r/20240904104848.23480-4-mariusz.tkaczyk@linux.intel.comSuggested-by: Lukas Wunner Signed-off-by: Stuart Hayes Signed-off-by: Mariusz Tkaczyk Signed-off-by: Bjorn Helgaas Tested-by: Stuart Hayes Reviewed-by: Christoph Hellwig Reviewed-by: Ilpo Järvinen List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

4e893545 - PCI/NPEM: Add Native PCIe Enclosure Management support

Wed, 04 Sep 2024 10:48:47 +0000

PCI/NPEM: Add Native PCIe Enclosure Management supportNative PCIe Enclosure Management (NPEM, PCIe r6.1 sec 6.28) allows managingLEDs in storage enclosures. NPEM is indication oriented and it does notgive direct access to LEDs. Although each indication *could* represent anindividual LED, multiple indications could also be represented as a single,multi-color LED or a single LED blinking in a specific interval. Thespecification leaves that open.Each enabled indication (capability register bit on) is represented as aledclass_dev which can be controlled through sysfs. For every ledclassdevice only 2 brightness states are allowed: LED_ON (1) or LED_OFF (0).This corresponds to the NPEM control register (Indication bit on/off).Ledclass devices appear in sysfs as child devices (subdirectory) of PCIdevice which has an NPEM Extended Capability and indication is enabled inNPEM capability register. For example, these are LEDs created for pcieport"10000:02:05.0" on my setup: leds/ ├── 10000:02:05.0:enclosure:fail ├── 10000:02:05.0:enclosure:locate ├── 10000:02:05.0:enclosure:ok └── 10000:02:05.0:enclosure:rebuildThey can be also found in "/sys/class/leds" directory. The parent PCIedevice domain/bus/device/function address is used to guarantee uniquenessacross leds subsystem.To enable/disable a "fail" indication, the "brightness" file can be edited: echo 1 > ./leds/10000:02:05.0:enclosure:fail/brightness echo 0 > ./leds/10000:02:05.0:enclosure:fail/brightnessPCIe r6.1, sec 7.9.19.2 defines the possible indications.Multiple indications for same parent PCIe device can conflict and hardwaremay update them when processing new request. To avoid issues, driverrefresh all indications by reading back control register.This driver expects to be the exclusive NPEM extended capability manager.It waits up to 1 second after imposing new request, it doesn't verify ifcontroller is busy before write, and it assumes the mutex lock givesprotection from concurrent updates.If _DSM LED management is available, we assume the platform may be usingNPEM for its own purposes (see PCI Firmware Spec r3.3 sec 4.7), so thedriver does not use NPEM. A future patch will add _DSM support; an infomessage notes whether NPEM or _DSM is being used.NPEM is a PCIe extended capability so it should be registered inpcie_init_capabilities() but it is not possible due to LED dependency. Theparent pci_device must be added earlier for led_classdev_register() to besuccessful. NPEM does not require configuration on kernel side, so it issafe to register LED devices later.Link: https://lore.kernel.org/r/20240904104848.23480-3-mariusz.tkaczyk@linux.intel.comSuggested-by: Lukas Wunner Signed-off-by: Mariusz Tkaczyk Signed-off-by: Bjorn Helgaas Tested-by: Stuart Hayes Reviewed-by: Christoph Hellwig Reviewed-by: Ilpo Järvinen List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

6d4338cb - ABI: sysfs-bus-pci: add documentation for p2pmem allocate

Fri, 21 Oct 2022 17:41:16 +0000

ABI: sysfs-bus-pci: add documentation for p2pmem allocateAdd documentation for the p2pmem/allocate binary file which allowsfor allocating p2pmem buffers in userspace for passing to driversthat support them. (Currently only O_DIRECT to NVMe devices.)Signed-off-by: Logan Gunthorpe Reviewed-by: John Hubbard Reviewed-by: Greg Kroah-Hartman Reviewed-by: Chaitanya Kulkarni Link: https://lore.kernel.org/r/20221021174116.7200-10-logang@deltatee.comSigned-off-by: Jens Axboe List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

91fa1277 - PCI: Expose PCIe Resizable BAR support via sysfs

Fri, 16 Sep 2022 20:44:48 +0000

PCI: Expose PCIe Resizable BAR support via sysfsAdd a simple sysfs interface to Resizable BAR support, largely for thepurposes of assigning such devices to a VM through VFIO. Resizable BARspresent a difficult feature to expose to a VM through emulation, asresizing a BAR is done on the host. It can fail, and often does, but wehave no means via emulation of a PCIe REBAR capability to handle the errorcases.A vfio-pci specific ioctl interface is also cumbersome as there are oftenmultiple devices within the same bridge aperture and handling them is achallenge. In the interface proposed here, expanding a BAR potentiallyrequires such devices to be soft-removed during the resize operation andrescanned after, in order for all the necessary resources to be released.A pci-sysfs interface is also more universal than a vfio specificinterface.Please see the ABI documentation update for usage.Link: https://lore.kernel.org/r/166336088796.3597940.14973499936692558556.stgit@omenSigned-off-by: Alex Williamson Signed-off-by: Bjorn Helgaas Reviewed-by: Christian König Cc: Krzysztof Wilczyński List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

5e3be666 - PCI: Document /sys/bus/pci/devices/.../irq

Wed, 25 Aug 2021 10:26:34 +0000

PCI: Document /sys/bus/pci/devices/.../irqDocument /sys/bus/pci/devices/.../irq.This file contains the IRQ of the INTx interrupt (or zero if the devicedoesn't support INTx interrupts).If the device has enabled MSI (not MSI-X), it contains the first MSI IRQinstead. This is a historical mistake because devices may support severalMSI or MSI-X vectors, and this file can't contain them all. But wepreserve this behavior to avoid breaking userspace.[bhelgaas: commit log]Link: https://lore.kernel.org/r/20210825102636.52757-2-21cnbao@gmail.comSigned-off-by: Barry Song Signed-off-by: Bjorn Helgaas Acked-by: Greg Kroah-Hartman List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

bab2f3c1 - ABI: sysfs-bus-pci: add a alternative What fields

Mon, 27 Sep 2021 13:59:43 +0000

ABI: sysfs-bus-pci: add a alternative What fieldsThere are some PCI ABI that aren't shown under: /sys/bus/pci/drivers/.../Because they're registered with a different class. That'sthe case of, for instance: /sys/bus/i2c/drivers/CHT Whiskey Cove PMIC/unbindThis one is not present under /sys/bus/pci: $ find /sys/bus/pci -name 'CHT Whiskey Cove PMIC'Although clearly this is provided by a PCI driver: /sys/devices/pci0000:00/0000:00:02.0/i2c-4/subsystem/drivers/CHT Whiskey Cove PMIC/unbindSo, add an altertate What location in order to match bind/unbindto such devices.Signed-off-by: Mauro Carvalho Chehab Link: https://lore.kernel.org/r/15ba8c07f1b0fd7359106920c8e34a7b9af7aea6.1632750608.git.mchehab+huawei@kernel.orgSigned-off-by: Greg Kroah-Hartman List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

9919c339 - ABI: sysfs-bus-pci: add documentation for modalias

Mon, 27 Sep 2021 13:59:39 +0000

ABI: sysfs-bus-pci: add documentation for modaliasEven being available since 2005, there's no documentation formodalias.Acked-by: Bjorn Helgaas Signed-off-by: Mauro Carvalho Chehab Link: https://lore.kernel.org/r/9ceb1fcdbef3c0d2d0368dcc2f19084a3e529ad1.1632750608.git.mchehab+huawei@kernel.orgSigned-off-by: Greg Kroah-Hartman List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

24d732a9 - ABI: sysfs-bus-pci: use wildcards on What definitions

Thu, 16 Sep 2021 08:59:41 +0000

ABI: sysfs-bus-pci: use wildcards on What definitionsAn "N" upper letter is not a wildcard, nor can easily be identifiedby script, specially since the USB sysfs define things likebNumInterfaces. Use, instead, , in order to let script/get_abi.plto convert it into a Regex.Acked-by: Bjorn Helgaas Signed-off-by: Mauro Carvalho Chehab Link: https://lore.kernel.org/r/4ede4ec98e295f054f3e5a6f3f9393b5e3d5d2a7.1631782432.git.mchehab+huawei@kernel.orgSigned-off-by: Greg Kroah-Hartman List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

d88f521d - PCI: Allow userspace to query and set device reset mechanism

Tue, 17 Aug 2021 18:04:56 +0000

PCI: Allow userspace to query and set device reset mechanismAdd "reset_method" sysfs attribute to enable user to query and setpreferred device reset methods and their ordering.[bhelgaas: on invalid sysfs input, return error and preserve previousconfig, as in earlier patch versions]Co-developed-by: Alex Williamson Link: https://lore.kernel.org/r/20210817180500.1253-6-ameynarkhede03@gmail.comSigned-off-by: Alex Williamson Signed-off-by: Amey Narkhede Signed-off-by: Bjorn Helgaas Reviewed-by: Raphael Norwitz List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

3e42d1de - docs: typo fixes in Documentation/ABI/

Thu, 13 May 2021 13:31:10 +0000

docs: typo fixes in Documentation/ABI/Fix the following typos in the Documentation/ABI/ directory:- In file obsolete/sysfs-cpuidle, change "obselete" for "obsolete".- In file removed/sysfs-kernel-uids, change "propotional" for "proportional".- In directory stable/, fix the following words: "associtated" for "associated", "hexidecimal" for "hexadecimal", "vlue" for "value", "csed" for "caused" and "wrtie" for "write". This updates a total of five files.- In directory testing/, fix the following words: "subystem" for "subsystem", "isochrnous" for "isochronous", "Desctiptors" for "Descriptors", "picutre" for "picture", "capture" for "capture", "occured" for "ocurred", "connnected" for "connected","agressively" for "aggressively","manufacturee" for "manufacturer" and "transaction" for "transaction", "malformatted" for "incorrectly formated" ,"internel" for "internal", "writtento" for "written to", "specificed" for "specified", "beyound" for "beyond", "Symetric" for "Symmetric". This updates a total of eleven files.Signed-off-by: Carlos Bilbao Reviewed-by: Randy Dunlap Reviewed-by: Mauro Carvalho Chehab Link: https://lore.kernel.org/r/5710038.lOV4Wx5bFT@iron-maidenSigned-off-by: Jonathan Corbet List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

81bbf039 - s390/pci: expose a PCI device's UID as its index

Wed, 24 Feb 2021 10:29:36 +0000

s390/pci: expose a PCI device's UID as its indexOn s390 each PCI device has a user-defined ID (UID) exposed under/sys/bus/pci/devices//uid. This ID was designed to serve as the PCIdevice's primary index and to match the device within Linux to thedevice configured in the hypervisor. To serve as a primary identifierthe UID must be unique within the Linux instance, this is guaranteed bythe platform if and only if the UID Uniqueness Checking flag is setwithin the CLP List PCI Functions response.In this sense the UID serves an analogous function as the SMBIOSinstance number or ACPI index exposed as the "index" respectively"acpi_index" device attributes and used by e.g. systemd to set interfacenames. As s390 does not use and will likely never use ACPI nor SMBIOSthere is no conflict and we can just expose the UID under the "index"attribute whenever UID Uniqueness Checking is active and get systemd'sinterface naming support for free.Link: https://lore.kernel.org/lkml/20210412135905.1434249-1-schnelle@linux.ibm.com/Acked-by: Viktor Mihajlovski Acked-by: Bjorn Helgaas Acked-by: Narendra K Signed-off-by: Niklas Schnelle Signed-off-by: Heiko Carstens List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

c3d5c2d9 - PCI/IOV: Add sysfs MSI-X vector assignment interface

Sun, 04 Apr 2021 07:22:18 +0000

PCI/IOV: Add sysfs MSI-X vector assignment interfaceA typical cloud provider SR-IOV use case is to create many VFs for use byguest VMs. The VFs may not be assigned to a VM until a customer requests aVM of a certain size, e.g., number of CPUs. A VF may need MSI-X vectorsproportional to the number of CPUs in the VM, but there is no standard wayto change the number of MSI-X vectors supported by a VF.Some Mellanox ConnectX devices support dynamic assignment of MSI-X vectorsto SR-IOV VFs. This can be done by the PF driver after VFs are enabled,and it can be done without affecting VFs that are already in use. Thehardware supports a limited pool of MSI-X vectors that can be assigned tothe PF or to individual VFs. This is device-specific behavior thatrequires support in the PF driver.Add a read-only "sriov_vf_total_msix" sysfs file for the PF and a writable"sriov_vf_msix_count" file for each VF. Management software may use theseto learn how many MSI-X vectors are available and to dynamically assignthem to VFs before the VFs are passed through to a VM.If the PF driver implements the ->sriov_get_vf_total_msix() callback,"sriov_vf_total_msix" contains the total number of MSI-X vectors availablefor distribution among VFs.If no driver is bound to the VF, writing "N" to "sriov_vf_msix_count" usesthe PF driver ->sriov_set_msix_vec_count() callback to assign "N" MSI-Xvectors to the VF. When a VF driver subsequently reads the MSI-X MessageControl register, it will see the new Table Size "N".Link: https://lore.kernel.org/linux-pci/20210314124256.70253-2-leon@kernel.orgAcked-by: Bjorn Helgaas Signed-off-by: Leon Romanovsky List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

80a129af - PCI: Add sysfs attribute for device power state

Mon, 02 Nov 2020 14:15:20 +0000

PCI: Add sysfs attribute for device power stateWhile PCI power states D0-D3hot can be queried from user-space via lspci,D3cold cannot. lspci cannot provide an accurate value when the device isin D3cold as it has to restore the device to D0 before it can access itspower state via the configuration space, leading to it reporting D0 oranother on-state. Thus lspci cannot be used to diagnose power consumptionissues for devices that can enter D3cold or to ensure that devices properlyenter D3cold at all.Add a new sysfs device attribute for the PCI power state, showing thecurrent power state as seen by the kernel.[bhelgaas: drop READ_ONCE(), see discussion at the link]Link: https://lore.kernel.org/r/20201102141520.831630-1-luzmaximilian@gmail.comSigned-off-by: Maximilian Luz Signed-off-by: Bjorn Helgaas List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

54a19b4d - docs: ABI: cleanup several ABI documents

Fri, 30 Oct 2020 07:40:50 +0000

docs: ABI: cleanup several ABI documentsThere are some ABI documents that, while they don't generateany warnings, they have issues when parsed by get_abi.pl scripton its output result.Address them, in order to provide a clean output.Reviewed-by: Tom Rix # for fpga-managerReviewed-By: Kajol Jain # for sysfs-bus-event_source-devices-hv_gpci and sysfs-bus-event_source-devices-hv_24x7Acked-by: Jonathan Cameron #for IIOAcked-by: Oded Gabbay # for HabanalabsAcked-by: Vaibhav Jain # for sysfs-bus-papr-pmemAcked-by: Cezary Rojewski # for catptAcked-by: Suzuki K Poulose Acked-by: Ilya Dryomov # for rbdAcked-by: Jonathan Corbet Signed-off-by: Mauro Carvalho Chehab Link: https://lore.kernel.org/r/5bc78e5b68ed1e9e39135173857cb2e753be868f.1604042072.git.mchehab+huawei@kernel.orgSigned-off-by: Greg Kroah-Hartman List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

72ea91af - PCI/ASPM: Add sysfs attributes for controlling ASPM link states

Sat, 05 Oct 2019 12:07:56 +0000

PCI/ASPM: Add sysfs attributes for controlling ASPM link statesAdd sysfs attributes to Endpoints and other Upstream Ports to control ASPM,Clock PM, and L1 PM Substates. The new attributes are: /sys/devices/pci*/.../link/clkpm /sys/devices/pci*/.../link/l0s_aspm /sys/devices/pci*/.../link/l1_aspm /sys/devices/pci*/.../link/l1_1_aspm /sys/devices/pci*/.../link/l1_2_aspm /sys/devices/pci*/.../link/l1_1_pcipm /sys/devices/pci*/.../link/l1_2_pcipmAn attribute is only visible if both ends of the Link leading to the devicesupport the state. Writing y/1/on to the file enables the state; n/0/offdisables it.These attributes can be used to tune the power/performance tradeoff forindividual devices.[bhelgaas: commit log, rename directory to "link"]Link: https://lore.kernel.org/r/b1c83f8a-9bf6-eac5-82d0-cf5b90128fbf@gmail.comSigned-off-by: Heiner Kallweit Signed-off-by: Bjorn Helgaas List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

cbb8ca69 - PCI/P2PDMA: Add sysfs group to display p2pmem stats

Thu, 04 Oct 2018 21:27:36 +0000

PCI/P2PDMA: Add sysfs group to display p2pmem statsAdd a sysfs group to display statistics about P2P memory that is registeredin each PCI device.Attributes in the group display the total amount of P2P memory, the amountavailable and whether it is published or not.Signed-off-by: Logan Gunthorpe Signed-off-by: Bjorn Helgaas Reviewed-by: Christoph Hellwig List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

0e7df224 - PCI: Add sysfs sriov_drivers_autoprobe to control VF driver binding

Wed, 12 Apr 2017 22:51:40 +0000

PCI: Add sysfs sriov_drivers_autoprobe to control VF driver bindingSometimes it is not desirable to bind SR-IOV VFs to drivers. This can savehost side resource usage by VF instances that will be assigned to VMs.Add a new PCI sysfs interface "sriov_drivers_autoprobe" to control thatfrom the PF. To modify it, echo 0/n/N (disable probe) or 1/y/Y (enableprobe) to: /sys/bus/pci/devices//sriov_drivers_autoprobeNote that this must be done before enabling VFs. The change will not takeeffect if VFs are already enabled. Simply, one can disable VFs by settingsriov_numvfs to 0, choose whether to probe or not, and then re-enable theVFs by restoring sriov_numvfs.[bhelgaas: changelog, ABI doc]Signed-off-by: Bodong Wang Signed-off-by: Eli Cohen Signed-off-by: Bjorn Helgaas Reviewed-by: Gavin Shan Reviewed-by: Alex Williamson List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci

88486bee - Documentation: ABI: testing: sysfs-bus-pci: Fix repetition of word 'the'

Thu, 09 Mar 2017 06:17:07 +0000

Documentation: ABI: testing: sysfs-bus-pci: Fix repetition of word 'the'The patch replaces 'the the' with 'the' in the documentation.Signed-off-by: sayli karnik Signed-off-by: Jonathan Corbet List of files: /linux-6.15/Documentation/ABI/testing/sysfs-bus-pci