This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: [PATCH 1/2] Correct the requirement of compressed floating point instructions.
- From: Jim Wilson <jimw at sifive dot com>
- To: Kito Cheng <kito dot cheng at gmail dot com>
- Cc: Binutils <binutils at sourceware dot org>, RISC-V Patches <patches at groups dot riscv dot org>, Ana Pazos <apazos at codeaurora dot org>, Alex Bradbury <asb at lowrisc dot org>
- Date: Thu, 30 Aug 2018 13:31:32 -0700
- Subject: Re: [PATCH 1/2] Correct the requirement of compressed floating point instructions.
- References: <CA+yXCZDEeVhM_y=--CifVkYDuxHa4St1SdqREgXZmqhAnZ+y1w@mail.gmail.com>
On Wed, Aug 29, 2018 at 1:07 AM, Kito Cheng <kito.cheng@gmail.com> wrote:
> ChangeLog
>
> 2018-08-29 Kito Cheng <kito@andestech.com>
>
> gas/
> * config/tc-riscv.c (riscv_subset_supports): New argument:
> xlen_required.
> (riscv_multi_subset_supports): New function, able to check more
> than one extension.
> (riscv_ip): Use riscv_multi_subset_supports instead of
> riscv_subset_supports.
> (riscv_set_arch): Update call-site for riscv_subset_supports.
> (riscv_after_parse_args): Likewise.
>
> include/
> *opcode/riscv.h (MAX_SUBSET_NUM): New.
> (riscv_opcode): Add xlen_requirement field and change type of
> subset.
>
> opcodes/
> * riscv-dis.c (riscv_disassemble_insn): Check XLEN by
> riscv_opcode.xlen_requirement.
> * riscv-opc.c (riscv_opcodes): Update for struct change.
I committed the first part.
Jim