This document is named "Linux and Devicetree" so I think we should write a bit more about how the device tree fits into the Linux kernel frameworks these days after we have created the fwnode and swnode. Cc: Grant Likely <grant.likely@xxxxxxxxxx> Signed-off-by: Linus Walleij <linus.walleij@xxxxxxxxxx> --- I bet you get the idea but Rob and Krzysztof probably have some ideas on where this goes and what goes into it so it it at least a starting point for discussions. --- Documentation/devicetree/usage-model.rst | 45 +++++++++++++++++++++--- 1 file changed, 41 insertions(+), 4 deletions(-) diff --git a/Documentation/devicetree/usage-model.rst b/Documentation/devicetree/usage-model.rst index b6a287955ee5..789744b45ab2 100644 --- a/Documentation/devicetree/usage-model.rst +++ b/Documentation/devicetree/usage-model.rst @@ -75,7 +75,44 @@ out of mainline (nios) have some level of DT support. If you haven't already read the Device Tree Usage\ [1]_ page, then go read it now. It's okay, I'll wait.... -2.1 High Level View +2.1 Linux Kernel Firmware Abstractions +-------------------------------------- + +The Linux kernel supports sevaral different hardware description +frameworks and DT is just one of them. The closest sibling is the +:ref:`Documentation/firmware-guide/acpi/index.rst ACPI` +DSDT (Differentiated System Description Table). + +To make it possible to write a device driver that will adapt to DT +or other hardware description models, the kernel has grown some +abstractions, first and foremost the firmware node API, exposing +device properties. The firmware node "fwnode" internals can be found +in ``<linux/fwnode.h>`` while the device driver-facing API can be +found in ``<linux/property.h>``. The idea is that if a driver is using +the firmware node API, it should be trivial to support DT and ACPI +DSDT alike in the same driver. + +The fwnode framework also makes it possible to modify and extend the +Linux in-kernel model with software-managed nodes "swnodes" to apply +quirks or manage registration of devices that cannot be handled any +other way. This API can also be found in ``<linux/property.h>``. + +Further, when the DT core register devices these need to fold into the +Linux device driver model, which essentially means that some kind of +``struct device`` has to be created to match a corresponding +``struct device_driver``. This API can be explored in +detail in :ref:`Documentation/driver-api/driver-model/index.rst the driver API documentation` +but what you need to know is that the Linux DT parser code will on its +own mostly spawn platform devices and AMBA devices on the platform +and AMBA bus respectively, and apart from that it will augment devices +spawn on other buses where applicable. + +Every Linux kernel subsystem that want to supply additional data to +detected devices using the device tree, or that want to provide +resources to other devices in the DT, will need to implement calls into +the DT abstractions. + +2.2 High Level View ------------------- The most important thing to understand is that the DT is simply a data structure that describes the hardware. There is nothing magical about @@ -97,7 +134,7 @@ Linux uses DT data for three major purposes: 2) runtime configuration, and 3) device population. -2.2 Platform Identification +2.3 Platform Identification --------------------------- First and foremost, the kernel will use data in the DT to identify the specific machine. In a perfect world, the specific platform shouldn't @@ -180,7 +217,7 @@ However, this approach does not take into account the priority of the compatible list, and probably should be avoided for new architecture support. -2.3 Runtime configuration +2.4 Runtime configuration ------------------------- In most cases, a DT will be the sole method of communicating data from firmware to the kernel, so also gets used to pass in runtime and @@ -217,7 +254,7 @@ On ARM, the function setup_machine_fdt() is responsible for early scanning of the device tree after selecting the correct machine_desc that supports the board. -2.4 Device population +2.5 Device population --------------------- After the board has been identified, and after the early configuration data has been parsed, then kernel initialization can proceed in the normal -- 2.39.0
