[V2,0/4] Add versal-pci driver

Message

Yidong Zhang Dec. 10, 2024, 6:37 p.m. UTC

This patchset introduces a new Linux Kernel Driver, versal-pci for AMD
Alevo Versal based PCIe Card. The driver is based on Linux fpga driver
framework.

The AMD Alevo Versal based PCIe Card, including V70, is the first Alevo
production card leveraging AMD XDNA architecture with AI Engines. It is
designed for AI inference efficiency and is tuned for video analytics and
natural language processing applications [1].

This versal-pci driver provides services, including:
  - leveraging linux firmware and FPGA framework to download management
    firmware
  - program additional bit-streams for AMD Xilinx specific hardware
  - communicate with PCIe user function
  - communicate with firmware running on the PCIe Card
  - monitor device health

The driver is licensed under GPL-2.0.

The firmware and bit-streams are distributed as a closed binary, delivered
by AMD. Please see [1] for more information.

[1] https://www.amd.com/en/products/accelerators/alveo/v70.html

Refactor driver to address all comments from v1.
Changes since v1:
- Add driver architecture description.
- Change the driver name to versal-pci 
- Remove unnecessary memcpy in versal-pci-comm-chan.c
- Keep mod_timer because we need single_threaded_queue with delayed_work.
- Change the "comms" to "comm_chan".
- Remove regmap, use base+offset directly.
- Add skeleton ops with on implementation (TODO) for fpga_manager in early
  patch (0001) and add implementation in later patch(0004).
- Remove br_ops and FPGA region, no need.
- Add load_xclbin request from user PF driver via comm_chan service, and
  read firmware from local cached data. 
- Remove pcim_iomap_table and pcim_iomap_regions. Use pcim_iomap_region.
- Redo patches, remove no-used definitions for each patch itself. Use
  /* TODO */ to mark dependencies that will be addressed in later patch.
- Fix memory leak in rm_check_health.
- Fix cmd with error status in rm_check_health.
- Change to vzalloc in all places.
- Add rm_queue_withdraw_cmd in fw_cancel.

Yidong Zhang (4):
  drivers/fpga/amd: Add new driver amd versal-pci
  drivers/fpga/amd: Add communication channel
  drivers/fpga/amd: Add remote queue
  drivers/fpga/amd: Add load xclbin and load firmware

 MAINTAINERS                              |   6 +
 drivers/fpga/Kconfig                     |   3 +
 drivers/fpga/Makefile                    |   3 +
 drivers/fpga/amd/Kconfig                 |  15 +
 drivers/fpga/amd/Makefile                |   8 +
 drivers/fpga/amd/versal-pci-comm-chan.c  | 271 ++++++++++++
 drivers/fpga/amd/versal-pci-comm-chan.h  |  14 +
 drivers/fpga/amd/versal-pci-main.c       | 445 ++++++++++++++++++++
 drivers/fpga/amd/versal-pci-rm-queue.c   | 324 +++++++++++++++
 drivers/fpga/amd/versal-pci-rm-queue.h   |  21 +
 drivers/fpga/amd/versal-pci-rm-service.c | 497 +++++++++++++++++++++++
 drivers/fpga/amd/versal-pci-rm-service.h | 229 +++++++++++
 drivers/fpga/amd/versal-pci.h            |  90 ++++
 13 files changed, 1926 insertions(+)
 create mode 100644 drivers/fpga/amd/Kconfig
 create mode 100644 drivers/fpga/amd/Makefile
 create mode 100644 drivers/fpga/amd/versal-pci-comm-chan.c
 create mode 100644 drivers/fpga/amd/versal-pci-comm-chan.h
 create mode 100644 drivers/fpga/amd/versal-pci-main.c
 create mode 100644 drivers/fpga/amd/versal-pci-rm-queue.c
 create mode 100644 drivers/fpga/amd/versal-pci-rm-queue.h
 create mode 100644 drivers/fpga/amd/versal-pci-rm-service.c
 create mode 100644 drivers/fpga/amd/versal-pci-rm-service.h
 create mode 100644 drivers/fpga/amd/versal-pci.h