1# Device Tree GPIO Naming in OpenBMC 2 3Author: Andrew Geissler (geissonator) 4 5Other contributors: 6 < None > 7 8Created: April 3, 2020 9 10## Problem Description 11The Linux kernel has deprecated the use of sysfs to interact with the GPIO 12subsystem. The replacement is a "descriptor-based" character device interface. 13 14[libgpiod][1] is a suite of tools and library implemented in C and C++ which 15provides an abstraction to this new character device gpio interface. 16 17libgpiod provides a feature where you can access gpios by a name given to 18them in the kernel device tree files. The problem is there are no naming 19conventions for these GPIO names and if you want userspace code to be able 20to be consistent across different machines, these names would need to be 21consistent. 22 23## Background and References 24The kernel [documentation][2] has a good summary of the GPIO subsystem. The 25specific field used to name the GPIOs in the DTS is `gpio-line-names`. 26This [patch][3] shows an example of naming the GPIOs for a system. 27 28GPIOs are used for arbitrary things. It's pretty hard to have a coherent naming 29scheme in the face of a universe of potential use-cases. 30 31Scoping the problem down to just the vastness of OpenBMC narrows the 32possibilities quite a bit and allows the possibility of a naming scheme to 33emerge. 34 35## Requirements 36- Ensure common function GPIOs within OpenBMC use the same naming convention 37 38## Proposed Design 39Below are the standard categories. The "Pattern" in each section describes the 40naming convention and then the sub bullets list the common GPIO names to 41be used (when available on an OpenBMC system). This naming convention must be 42followed for all common GPIOs. 43 44This list below includes all common GPIOs within OpenBMC. Any OpenBMC 45system which provides one of the below GPIOs must name it as listed in 46this document. This document must be updated as new common GPIOs are added. 47 48### Buttons 49Pattern: `*-button` 50 51Buttons should always be considered 'input' to the BMC. There may be cases 52where BMC-less machines use a button to trigger system behavior and in a 53BMC-managed system this signal is emulated by the BMC (as output). These 54should not be considered a button in this document but enumerated as another 55signal type. 56 57#### power-button 58 59### Host Ready 60Below are input GPIO names specific to Host ready. The name of Host ready GPIO 61depends on the index of Host and the active state is high or low. 62 63Pattern: 64- `host*-ready`: Host ready, active high 65- `host*-ready-n`: Host ready, active low 66 67Defined: 68- host0-ready 69- host1-ready-n 70- ... 71 72### LEDs 73Pattern: `led-*` 74 75#### led-fault 76#### led-identify 77#### led-power 78#### led-sys-boot-status 79#### led-attention 80#### led-hdd-fault 81#### led-rear-fault 82#### led-rear-power 83#### led-rear-id 84 85### Power and Regulators 86Pattern: `power-*`, `regulator-*` 87 88#### power-chassis-control 89Set to initiate power-on or power-off of the chassis. 90 91#### power-chassis-good 92Indicates the power good state of the chassis. 93 94#### power-config-full-load 95Output GPIO set by the power managing application that indicates to the hardware 96the threshold of power supplies that are expected to be present and working for 97this type of system for the case where a system has a model that supports two 98different number of power supplies (example 2 or 4). Default is the lowest 99number of power supplies expected of the two models. In the case when the number 100of power supplies that are present are less than the ones indicated by this 101GPIO, the hardware can then take actions such as reducing the system's 102performance so that it can maintain its powered on state. 103 104#### power-ffs-sync-history 105Output GPIO set by the power managing applications. The IBM Common Form Factor 106power supplies use this line as an input. When set low and left low the power 107supply fans run at full speed (Fans Full Speed). When toggled low, then high, 108it resets the power supply input history (average and maximum power values over 109a time range). 110 111#### regulator-standby-faulted 112This GPIO value represents the status of standby power regulator fault detection 113logic. This GPIO is an input only. The status will reflect a regulator 114non-faulted condition after AC power cycle when no standby power regulator fault 115condition is present. The status will reflect a standby regulator power faulted 116condition when an unexpected drop in standby power is detected. 117 118#### rtc-battery-voltage-read-enable 119Output pin that enables the ADC to read the board RTC battery voltage. 120 121### Presence 122Pattern: `presence-*` 123 124#### presence-ps0, presence-ps1, ..., presence-ps\<N> 125 126### Reset Cause 127These are GPIOs that provide more detail on the reason for a BMC reset. BMC 128hardware generally provides some information on a BMC reboot, like a EXTRST 129(i.e. a BMC reset was reset by some external source). At times though, 130firmware needs more details on the cause of a reset. Hardware can be configured 131to latch an event into a GPIO for firmware to then utilize for different 132software logic. 133 134Pattern: `reset-cause-*` 135 136#### reset-cause-pinhole 137The pinhole reset cause will be utilized by BMC firmware to know when it 138has been reset due to a user initiated pinhole reset. This is commonly done in 139error scenarios where the BMC is hanging or otherwise unresponsive. Note that 140this GPIO is not utilized to cause the actual reset, it is a GPIO that can be 141read after the BMC reset to know the reason for the reboot was a pinhole reset. 142 143### Secure Boot 144 145#### bmc-secure-boot 146Input pin that indicates that the BMC is configured to boot with security 147protections enforced. 148 149Pulled up by default (secure). Placing the jumper will pull the pin down 150(bypass security). 151 152### Special 153These are special case and/or grandfathered in pin names. 154 155#### air-water 156Indicates whether system is air or water cooled 157 158#### factory-reset-toggle 159The software records the state of this GPIO and checks upon reboot if the state 160has changed since the last reboot. If it has, it indicates that a factory reset 161should be performed. 162 163### POWER Specific GPIOs 164Below are GPIO names specific to the POWER processor based servers. 165 166#### Special 167These are special case and/or grandfathered in pin names. 168 169##### cfam-reset 170Utilized to issue a processor logic reset to a IBM POWER processor. 171 172##### checkstop 173Utilized to indicate a IBM POWER processor has entered an unrecoverable error 174state. 175 176## Alternatives Considered 177- Continue to hard code a config file per system type that has the 178gpio bank and pin number. This removes a dependency on the device tree to 179have consistent names but adds overhead in supporting each new system. 180 181- Have the device tree GPIO names match the hardware schematics and then 182have another userspace config file that maps between the schematic names 183and logical pin names. This makes the GPIO to schematic mapping easy but 184adds an additional layer of work with the userspace config. 185 186## Impacts 187Need to ensure OpenBMC device trees conform to the above naming conventions. 188 189## Testing 190Userspace utilization of the GPIO names will provide some testing coverage 191during CI. 192 193[1]: https://git.kernel.org/pub/scm/libs/libgpiod/libgpiod.git/about/ 194[2]: https://www.kernel.org/doc/html/latest/driver-api/gpio/index.html 195[3]: https://lore.kernel.org/linux-arm-kernel/20200306170218.79698-1-geissonator@yahoo.com/ 196