Re: [PATCH v6 0/3] Allow accessing CSR using CSR number

[Palmer Dabbelt]

On Thu, 25 Apr 2019 01:38:24 PDT (-0700), Anup Patel wrote:
> This patch series adds support to access CSR using both CSR name and
> CSR numbers.
>
> Also, we should prefer accessing CSRs using their CSR numbers because:
> 1. It compiles fine with older toolchains.
> 2. We can use latest CSR names in #define macro names of CSR numbers
>    as-per RISC-V spec. (e.g. sptbr => CSR_SATP, sbadaddr => CSR_STVAL, etc.)
> 3. We can access newly added CSRs even if toolchain does not recognize
>    newly addes CSRs by name. (e.g. BSSTATUS, BSIE, SSIP, etc.)
>
> The patchset can be found in riscv_csr_number_v6 branch of
> https//github.com/avpatel/linux.git
>
> Changes since v5:
>  - Drop redundant INTERRUPT_CAUSE_FLAG from kernel/irq.c
>
> Changes since v4:
>  - Express SCAUSE_IRQ_FLAG in-terms of __riscv_xlen
>
> Changes since v3:
>  - Keep old INTERRUPT_xyz defines in kernel/irq.c for PATCH2
>
> Changes since v2:
>  - Dropped PATCH1 which added asm/encoding.h
>  - Added new PATCH1 which beautifies asm/csr.h by using tabs to
>    align macro values
>
> Changes since v1:
>  - Squash PATCH2 into cpatch3
>  - Added new PATCH2 to add interrupt related SCAUSE defines
>    in asm/encoding.h
>
> Anup Patel (3):
>   RISC-V: Use tabs to align macro values in asm/csr.h
>   RISC-V: Add interrupt related SCAUSE defines in asm/csr.h
>   RISC-V: Access CSRs using CSR numbers
>
>  arch/riscv/include/asm/csr.h         | 123 +++++++++++++++++----------
>  arch/riscv/include/asm/irqflags.h    |  10 +--
>  arch/riscv/include/asm/mmu_context.h |   7 +-
>  arch/riscv/kernel/entry.S            |  22 ++---
>  arch/riscv/kernel/head.S             |  12 +--
>  arch/riscv/kernel/irq.c              |  16 +---
>  arch/riscv/kernel/perf_event.c       |   4 +-
>  arch/riscv/kernel/smp.c              |   2 +-
>  arch/riscv/kernel/traps.c            |   6 +-
>  arch/riscv/mm/fault.c                |   6 +-
>  10 files changed, 111 insertions(+), 97 deletions(-)
>
> --
> 2.17.1

Thanks.  I've added these to for-next with a minor merge conflict, but I think
I managed to avoid screwing it up!