We would like users to explicitly set that, so that implication rule
won't screw anything up or unexpect -mabi, that's kind of the
conclusion of most RISC-V GCC maintainers (Palmer/Jim Willsom/me).
Also the behavior is there for years, we don't want to make surprise
to user for the behavior change
> ok, thanks :)
This has likely been discussed at length before, but why need to
specify the additional -mabi with -march (instead of -march implying
a matching abi)?
-patches
> CC: kito.cheng; palmer; jeffreyalaw; Juzhe-Zhong
> Subject: [PATCH] RISC-V: Fix RVV binary auto-vectorizaiton test fails
> From: Juzhe-Zhong
>
> In rv32:
> FAIL: gcc.target/riscv/rvv/autovec/vmax-rv64gcv.c -O3 -ftree-vectorize (test
> for excess errors)
> FAIL: gcc
This patch has tested on both RV32/RV64, and all fails in RVV are cleaned up.
Ok for trunk?
juzhe.zh...@rivai.ai
From: juzhe.zhong
Date: 2023-05-12 07:29
To: gcc-patches
CC: kito.cheng; palmer; jeffreyalaw; Juzhe-Zhong
Subject: [PATCH] RISC-V: Fix RVV binary auto-vectorizaiton test fails
From
From: Juzhe-Zhong
In rv32:
FAIL: gcc.target/riscv/rvv/autovec/vmax-rv64gcv.c -O3 -ftree-vectorize (test
for excess errors)
FAIL: gcc.target/riscv/rvv/autovec/vmin-run.c -O3 -ftree-vectorize (test for
excess errors)
FAIL: gcc.target/riscv/rvv/autovec/vadd-rv64gcv.c -O3 -ftree-vectorize (test
fo
From: Juzhe-Zhong
In rv32:
FAIL: gcc.target/riscv/rvv/autovec/vmax-rv64gcv.c -O3 -ftree-vectorize (test
for excess errors)
FAIL: gcc.target/riscv/rvv/autovec/vmin-run.c -O3 -ftree-vectorize (test for
excess errors)
FAIL: gcc.target/riscv/rvv/autovec/vadd-rv64gcv.c -O3 -ftree-vectorize (test
fo
