Re: [PATCH v3 0/4] Add PLIC support for Renesas RZ/Five SoC / Fix T-HEAD PLIC edge flow

From: Palmer Dabbelt
Date: Tue Jul 12 2022 - 23:19:15 EST


On Fri, 01 Jul 2022 07:28:48 PDT (-0700), Marc Zyngier wrote:
On 2022-06-30 11:02, Samuel Holland wrote:
This patch series adds PLIC support for Renesas RZ/Five SoC.

Since the T-HEAD C900 PLIC has the same behavior, it also applies the
fix for that variant.

This series is an update of v2 of the RZ/Five series[0], and replaces
the separate T-HEAD series[1].

[0]:
https://lore.kernel.org/linux-riscv/20220626004326.8548-1-prabhakar.mahadev-lad.rj@xxxxxxxxxxxxxx/
[1]:
https://lore.kernel.org/linux-riscv/20220627051257.38543-1-samuel@xxxxxxxxxxxx/

Changes in v3:
- Add a more detailed explanation for why #interrupt-cells differs
- Add andestech,nceplic100 as a fallback compatible
- Separate the conditional part of the binding into two blocks (one
for
the PLIC implementation and the other for the SoC integration)
- Use a quirk bit for selecting the flow instead of a variant ID
- Use the andestech,nceplic100 compatible to select the new behavior
- Use handle_edge_irq instead of handle_fasteoi_ack_irq so .irq_ack
always gets called
- Do not set the handler name, as RISC-V selects
GENERIC_IRQ_SHOW_LEVEL
- Use the same name for plic_edge_chip as plic_chip

Changes in v2:
- Fixed review comments pointed by Marc and Krzysztof.

Changes in v1:
- Fixed review comments pointed by Rob and Geert.
- Changed implementation for EDGE interrupt handling on Renesas
RZ/Five
SoC.

Lad Prabhakar (2):
dt-bindings: interrupt-controller: sifive,plic: Document Renesas
RZ/Five SoC
irqchip/sifive-plic: Add support for Renesas RZ/Five SoC

Samuel Holland (2):
dt-bindings: interrupt-controller: Require trigger type for T-HEAD
PLIC
irqchip/sifive-plic: Fix T-HEAD PLIC edge trigger handling

.../sifive,plic-1.0.0.yaml | 65 +++++++++++++--
drivers/irqchip/irq-sifive-plic.c | 80 +++++++++++++++++--
2 files changed, 135 insertions(+), 10 deletions(-)

I'm going to provisionally queue this into -next so that it
can get some testing. I'd still want the DT changes to be
Ack'ed before the next merge window though.

+David, as IIRC he still tests on SiFive hardware.

Acked-by: Palmer Dabbelt <palmer@xxxxxxxxxxxx>

Though I also wait for Rob on DT stuff (I saw the other thread), so not sure that helps any.

Thanks!