[v2,0/6] Add SiFive FU740 PCIe host controller driver support
mbox series

Message ID cover.1615954045.git.greentime.hu@sifive.com
Headers show
Series
  • Add SiFive FU740 PCIe host controller driver support
Related show

Message

Greentime Hu March 18, 2021, 6:08 a.m. UTC
This patchset includes SiFive FU740 PCIe host controller driver. We also
add pcie_aux clock and pcie_power_on_reset controller to prci driver for
PCIe driver to use it.

This is tested with e1000e: Intel(R) PRO/1000 Network Card, AMD Radeon R5
230 graphics card and SP M.2 PCIe Gen 3 SSD in SiFive Unmatched based on
v5.11 Linux kernel.

Changes in v2:
 - Refine codes based on reviewers' feedback
 - Remove define and use the common one
 - Replace __raw_writel with writel_relaxed
 - Split fu740_phyregreadwrite to write function
 - Use readl_poll_timeout in stead of while loop checking
 - Use dwc common codes
 - Use gpio descriptors and the gpiod_* api.
 - Replace devm_ioremap_resource with devm_platform_ioremap_resource_byname
 - Replace devm_reset_control_get with devm_reset_control_get_exclusive
 - Add more comments for delay and sleep
 - Remove "phy ? x : y" expressions
 - Refine code logic to remove possible infinite loop
 - Replace magic number with meaningful define
 - Remove fu740_pcie_pm_ops
 - Use builtin_platform_driver

Greentime Hu (5):
  clk: sifive: Add pcie_aux clock in prci driver for PCIe driver
  clk: sifive: Use reset-simple in prci driver for PCIe driver
  MAINTAINERS: Add maintainers for SiFive FU740 PCIe driver
  dt-bindings: PCI: Add SiFive FU740 PCIe host controller
  riscv: dts: Add PCIe support for the SiFive FU740-C000 SoC

Paul Walmsley (1):
  PCI: fu740: Add SiFive FU740 PCIe host controller driver

 .../bindings/pci/sifive,fu740-pcie.yaml       | 119 +++++++
 MAINTAINERS                                   |   8 +
 arch/riscv/boot/dts/sifive/fu740-c000.dtsi    |  34 ++
 drivers/clk/sifive/Kconfig                    |   2 +
 drivers/clk/sifive/fu740-prci.c               |  11 +
 drivers/clk/sifive/fu740-prci.h               |   2 +-
 drivers/clk/sifive/sifive-prci.c              |  54 +++
 drivers/clk/sifive/sifive-prci.h              |  13 +
 drivers/pci/controller/dwc/Kconfig            |   9 +
 drivers/pci/controller/dwc/Makefile           |   1 +
 drivers/pci/controller/dwc/pcie-fu740.c       | 324 ++++++++++++++++++
 drivers/reset/Kconfig                         |   3 +-
 include/dt-bindings/clock/sifive-fu740-prci.h |   1 +
 13 files changed, 579 insertions(+), 2 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/pci/sifive,fu740-pcie.yaml
 create mode 100644 drivers/pci/controller/dwc/pcie-fu740.c

Comments

Stephen Boyd March 29, 2021, 7:12 p.m. UTC | #1
Quoting Greentime Hu (2021-03-17 23:08:07)
> This patchset includes SiFive FU740 PCIe host controller driver. We also
> add pcie_aux clock and pcie_power_on_reset controller to prci driver for
> PCIe driver to use it.
> 
> This is tested with e1000e: Intel(R) PRO/1000 Network Card, AMD Radeon R5
> 230 graphics card and SP M.2 PCIe Gen 3 SSD in SiFive Unmatched based on
> v5.11 Linux kernel.

Can I merge the clk patches to clk-next? Or is the dts patch going to be
sent in for the merge window? I'd like to merge the clk patches if the
other patches are going to miss the next merge window.

> 
> Changes in v2:
>  - Refine codes based on reviewers' feedback
>  - Remove define and use the common one
Greentime Hu April 1, 2021, 6:16 a.m. UTC | #2
Stephen Boyd <sboyd@kernel.org> 於 2021年3月30日 週二 上午3:12寫道:
>
> Quoting Greentime Hu (2021-03-17 23:08:07)
> > This patchset includes SiFive FU740 PCIe host controller driver. We also
> > add pcie_aux clock and pcie_power_on_reset controller to prci driver for
> > PCIe driver to use it.
> >
> > This is tested with e1000e: Intel(R) PRO/1000 Network Card, AMD Radeon R5
> > 230 graphics card and SP M.2 PCIe Gen 3 SSD in SiFive Unmatched based on
> > v5.11 Linux kernel.
>
> Can I merge the clk patches to clk-next? Or is the dts patch going to be
> sent in for the merge window? I'd like to merge the clk patches if the
> other patches are going to miss the next merge window.

Hi Stephen,

Thank you for reviewing. I am ok with either way. :)