From: Kito Cheng <kito.cheng@gmail.com>
To: 钟居哲 <juzhe.zhong@rivai.ai>
Cc: GCC Patches <gcc-patches@gcc.gnu.org>,
Palmer Dabbelt <palmer@dabbelt.com>
Subject: Re: [PATCH] RISC-V: Add testcases for VSETVL PASS 3
Date: Mon, 19 Dec 2022 23:08:54 +0800 [thread overview]
Message-ID: <CA+yXCZDYu0XWUQbRXWwYabWUHzOSqe0ZmGm22AkYi_khKqMSUw@mail.gmail.com> (raw)
In-Reply-To: <20221214081548.253313-1-juzhe.zhong@rivai.ai>
[-- Attachment #1: Type: text/plain, Size: 66984 bytes --]
Commited to trunk
<juzhe.zhong@rivai.ai> 於 2022年12月14日 週三 16:16 寫道：
> From: Ju-Zhe Zhong <juzhe.zhong@rivai.ai>
>
> gcc/testsuite/ChangeLog:
>
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-1.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-10.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-11.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-12.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-13.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-14.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-15.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-16.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-17.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-18.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-19.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-2.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-20.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-21.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-22.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-23.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-24.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-25.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-26.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-27.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-28.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-3.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-4.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-5.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-6.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-7.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-8.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-9.c: New test.
>
> ---
> .../riscv/rvv/vsetvl/vlmax_miss_default-1.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-10.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-11.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-12.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-13.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-14.c | 189 ++++++++++++++
> .../riscv/rvv/vsetvl/vlmax_miss_default-15.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-16.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-17.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-18.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-19.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-2.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-20.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-21.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-22.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-23.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-24.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-25.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-26.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-27.c | 38 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-28.c | 231 ++++++++++++++++++
> .../riscv/rvv/vsetvl/vlmax_miss_default-3.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-4.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-5.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-6.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-7.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-8.c | 32 +++
> .../riscv/rvv/vsetvl/vlmax_miss_default-9.c | 32 +++
> 28 files changed, 1330 insertions(+)
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-1.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-10.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-11.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-12.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-13.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-14.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-15.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-16.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-17.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-18.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-19.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-2.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-20.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-21.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-22.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-23.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-24.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-25.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-26.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-27.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-28.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-3.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-4.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-5.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-6.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-7.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-8.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-9.c
>
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-1.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-1.c
> new file mode 100644
> index 00000000000..d801428004e
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-1.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 100);
> + *(vint8mf8_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 200);
> + *(vint8mf8_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 300);
> + *(vint8mf8_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf8_t v = *(vint8mf8_t*)(in + i);
> + *(vint8mf8_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-10.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-10.c
> new file mode 100644
> index 00000000000..d96e9b2d7ab
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-10.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vuint16mf2_t v = *(vuint16mf2_t*)(in + 100);
> + *(vuint16mf2_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vuint16mf2_t v = *(vuint16mf2_t*)(in + 200);
> + *(vuint16mf2_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vuint16mf2_t v = *(vuint16mf2_t*)(in + 300);
> + *(vuint16mf2_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint16mf2_t v = *(vuint16mf2_t*)(in + i);
> + *(vuint16mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-11.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-11.c
> new file mode 100644
> index 00000000000..de967f78bc6
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-11.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 100);
> + *(vint32mf2_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 300);
> + *(vint32mf2_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint32mf2_t v = *(vint32mf2_t*)(in + i);
> + *(vint32mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-12.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-12.c
> new file mode 100644
> index 00000000000..8342edb353a
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-12.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vuint32mf2_t v = *(vuint32mf2_t*)(in + 100);
> + *(vuint32mf2_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vuint32mf2_t v = *(vuint32mf2_t*)(in + 200);
> + *(vuint32mf2_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vuint32mf2_t v = *(vuint32mf2_t*)(in + 300);
> + *(vuint32mf2_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint32mf2_t v = *(vuint32mf2_t*)(in + i);
> + *(vuint32mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-13.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-13.c
> new file mode 100644
> index 00000000000..82cd0d36913
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-13.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 100);
> + *(vfloat32mf2_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 200);
> + *(vfloat32mf2_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 300);
> + *(vfloat32mf2_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + i);
> + *(vfloat32mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-14.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-14.c
> new file mode 100644
> index 00000000000..5e08b23178b
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-14.c
> @@ -0,0 +1,189 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vbool64_t v = *(vbool64_t*)(in + 100);
> + *(vbool64_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vbool64_t v = *(vbool64_t*)(in + 200);
> + *(vbool64_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vbool64_t v = *(vbool64_t*)(in + 300);
> + *(vbool64_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool64_t v = *(vbool64_t*)(in + i);
> + *(vbool64_t*)(out + i) = v;
> + }
> +}
> +
> +void f2 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vbool32_t v = *(vbool32_t*)(in + 100);
> + *(vbool32_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vbool32_t v = *(vbool32_t*)(in + 200);
> + *(vbool32_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vbool32_t v = *(vbool32_t*)(in + 300);
> + *(vbool32_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool32_t v = *(vbool32_t*)(in + i);
> + *(vbool32_t*)(out + i) = v;
> + }
> +}
> +
> +void f3 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vbool16_t v = *(vbool16_t*)(in + 100);
> + *(vbool16_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vbool16_t v = *(vbool16_t*)(in + 200);
> + *(vbool16_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vbool16_t v = *(vbool16_t*)(in + 300);
> + *(vbool16_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool16_t v = *(vbool16_t*)(in + i);
> + *(vbool16_t*)(out + i) = v;
> + }
> +}
> +
> +void f4 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vbool8_t v = *(vbool8_t*)(in + 100);
> + *(vbool8_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vbool8_t v = *(vbool8_t*)(in + 200);
> + *(vbool8_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vbool8_t v = *(vbool8_t*)(in + 300);
> + *(vbool8_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool8_t v = *(vbool8_t*)(in + i);
> + *(vbool8_t*)(out + i) = v;
> + }
> +}
> +
> +void f5 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vbool4_t v = *(vbool4_t*)(in + 100);
> + *(vbool4_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vbool4_t v = *(vbool4_t*)(in + 200);
> + *(vbool4_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vbool4_t v = *(vbool4_t*)(in + 300);
> + *(vbool4_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool4_t v = *(vbool4_t*)(in + i);
> + *(vbool4_t*)(out + i) = v;
> + }
> +}
> +
> +void f6 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vbool2_t v = *(vbool2_t*)(in + 100);
> + *(vbool2_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vbool2_t v = *(vbool2_t*)(in + 200);
> + *(vbool2_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vbool2_t v = *(vbool2_t*)(in + 300);
> + *(vbool2_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool2_t v = *(vbool2_t*)(in + i);
> + *(vbool2_t*)(out + i) = v;
> + }
> +}
> +
> +void f7 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vbool1_t v = *(vbool1_t*)(in + 100);
> + *(vbool1_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vbool1_t v = *(vbool1_t*)(in + 200);
> + *(vbool1_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vbool1_t v = *(vbool1_t*)(in + 300);
> + *(vbool1_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool1_t v = *(vbool1_t*)(in + i);
> + *(vbool1_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m1,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m8,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m1,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m2,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m4,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m8,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-15.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-15.c
> new file mode 100644
> index 00000000000..267d1bd5cf6
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-15.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vint8mf8_t v = *(vint8mf8_t*)(in + 100);
> + *(vint8mf8_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vint8mf8_t v = *(vint8mf8_t*)(in + 200);
> + *(vint8mf8_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vint8mf8_t v = *(vint8mf8_t*)(in + 300);
> + *(vint8mf8_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf8_t v = *(vint8mf8_t*)(in + i);
> + *(vint8mf8_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-16.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-16.c
> new file mode 100644
> index 00000000000..3e6c51ad9af
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-16.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 100);
> + *(vuint8mf8_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 200);
> + *(vuint8mf8_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 300);
> + *(vuint8mf8_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + i);
> + *(vuint8mf8_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-17.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-17.c
> new file mode 100644
> index 00000000000..ae2caedff27
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-17.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vint8mf4_t v = *(vint8mf4_t*)(in + 100);
> + *(vint8mf4_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vint8mf4_t v = *(vint8mf4_t*)(in + 200);
> + *(vint8mf4_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vint8mf4_t v = *(vint8mf4_t*)(in + 300);
> + *(vint8mf4_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf4_t v = *(vint8mf4_t*)(in + i);
> + *(vint8mf4_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-18.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-18.c
> new file mode 100644
> index 00000000000..2378b4006c0
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-18.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vuint8mf4_t v = *(vuint8mf4_t*)(in + 100);
> + *(vuint8mf4_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vuint8mf4_t v = *(vuint8mf4_t*)(in + 200);
> + *(vuint8mf4_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vuint8mf4_t v = *(vuint8mf4_t*)(in + 300);
> + *(vuint8mf4_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint8mf4_t v = *(vuint8mf4_t*)(in + i);
> + *(vuint8mf4_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-19.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-19.c
> new file mode 100644
> index 00000000000..1de8fc461dc
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-19.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vint8mf2_t v = *(vint8mf2_t*)(in + 100);
> + *(vint8mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vint8mf2_t v = *(vint8mf2_t*)(in + 200);
> + *(vint8mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vint8mf2_t v = *(vint8mf2_t*)(in + 300);
> + *(vint8mf2_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf2_t v = *(vint8mf2_t*)(in + i);
> + *(vint8mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-2.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-2.c
> new file mode 100644
> index 00000000000..6c77a94ae63
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-2.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 100);
> + *(vuint8mf8_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 200);
> + *(vuint8mf8_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 300);
> + *(vuint8mf8_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + i);
> + *(vuint8mf8_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-20.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-20.c
> new file mode 100644
> index 00000000000..41100cb04a3
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-20.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vuint8mf2_t v = *(vuint8mf2_t*)(in + 100);
> + *(vuint8mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vuint8mf2_t v = *(vuint8mf2_t*)(in + 200);
> + *(vuint8mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vuint8mf2_t v = *(vuint8mf2_t*)(in + 300);
> + *(vuint8mf2_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint8mf2_t v = *(vuint8mf2_t*)(in + i);
> + *(vuint8mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-21.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-21.c
> new file mode 100644
> index 00000000000..070899e6e49
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-21.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vint16mf4_t v = *(vint16mf4_t*)(in + 100);
> + *(vint16mf4_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vint16mf4_t v = *(vint16mf4_t*)(in + 200);
> + *(vint16mf4_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vint16mf4_t v = *(vint16mf4_t*)(in + 300);
> + *(vint16mf4_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint16mf4_t v = *(vint16mf4_t*)(in + i);
> + *(vint16mf4_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-22.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-22.c
> new file mode 100644
> index 00000000000..1e96191a9fc
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-22.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vuint16mf4_t v = *(vuint16mf4_t*)(in + 100);
> + *(vuint16mf4_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vuint16mf4_t v = *(vuint16mf4_t*)(in + 200);
> + *(vuint16mf4_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vuint16mf4_t v = *(vuint16mf4_t*)(in + 300);
> + *(vuint16mf4_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint16mf4_t v = *(vuint16mf4_t*)(in + i);
> + *(vuint16mf4_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-23.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-23.c
> new file mode 100644
> index 00000000000..fa6420822f2
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-23.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vint16mf2_t v = *(vint16mf2_t*)(in + 100);
> + *(vint16mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vint16mf2_t v = *(vint16mf2_t*)(in + 200);
> + *(vint16mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vint16mf2_t v = *(vint16mf2_t*)(in + 300);
> + *(vint16mf2_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint16mf2_t v = *(vint16mf2_t*)(in + i);
> + *(vint16mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-24.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-24.c
> new file mode 100644
> index 00000000000..d0352fccacf
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-24.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vuint16mf2_t v = *(vuint16mf2_t*)(in + 100);
> + *(vuint16mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vuint16mf2_t v = *(vuint16mf2_t*)(in + 200);
> + *(vuint16mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vuint16mf2_t v = *(vuint16mf2_t*)(in + 300);
> + *(vuint16mf2_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint16mf2_t v = *(vuint16mf2_t*)(in + i);
> + *(vuint16mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-25.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-25.c
> new file mode 100644
> index 00000000000..ad76c24d294
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-25.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vint32mf2_t v = *(vint32mf2_t*)(in + 100);
> + *(vint32mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vint32mf2_t v = *(vint32mf2_t*)(in + 300);
> + *(vint32mf2_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint32mf2_t v = *(vint32mf2_t*)(in + i);
> + *(vint32mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-26.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-26.c
> new file mode 100644
> index 00000000000..71a3f4a06f6
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-26.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vuint32mf2_t v = *(vuint32mf2_t*)(in + 100);
> + *(vuint32mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vuint32mf2_t v = *(vuint32mf2_t*)(in + 200);
> + *(vuint32mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vuint32mf2_t v = *(vuint32mf2_t*)(in + 300);
> + *(vuint32mf2_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint32mf2_t v = *(vuint32mf2_t*)(in + i);
> + *(vuint32mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-27.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-27.c
> new file mode 100644
> index 00000000000..613a028089b
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-27.c
> @@ -0,0 +1,38 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 100);
> + *(vfloat32mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 200);
> + *(vfloat32mf2_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 300);
> + *(vfloat32mf2_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + i);
> + *(vfloat32mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-28.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-28.c
> new file mode 100644
> index 00000000000..6a2011b0290
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-28.c
> @@ -0,0 +1,231 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vbool64_t v = *(vbool64_t*)(in + 100);
> + *(vbool64_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vbool64_t v = *(vbool64_t*)(in + 200);
> + *(vbool64_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vbool64_t v = *(vbool64_t*)(in + 300);
> + *(vbool64_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool64_t v = *(vbool64_t*)(in + i);
> + *(vbool64_t*)(out + i) = v;
> + }
> +}
> +
> +void f2 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vbool32_t v = *(vbool32_t*)(in + 100);
> + *(vbool32_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vbool32_t v = *(vbool32_t*)(in + 200);
> + *(vbool32_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vbool32_t v = *(vbool32_t*)(in + 300);
> + *(vbool32_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool32_t v = *(vbool32_t*)(in + i);
> + *(vbool32_t*)(out + i) = v;
> + }
> +}
> +
> +void f3 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vbool16_t v = *(vbool16_t*)(in + 100);
> + *(vbool16_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vbool16_t v = *(vbool16_t*)(in + 200);
> + *(vbool16_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vbool16_t v = *(vbool16_t*)(in + 300);
> + *(vbool16_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool16_t v = *(vbool16_t*)(in + i);
> + *(vbool16_t*)(out + i) = v;
> + }
> +}
> +
> +void f4 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vbool8_t v = *(vbool8_t*)(in + 100);
> + *(vbool8_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vbool8_t v = *(vbool8_t*)(in + 200);
> + *(vbool8_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vbool8_t v = *(vbool8_t*)(in + 300);
> + *(vbool8_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool8_t v = *(vbool8_t*)(in + i);
> + *(vbool8_t*)(out + i) = v;
> + }
> +}
> +
> +void f5 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vbool4_t v = *(vbool4_t*)(in + 100);
> + *(vbool4_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vbool4_t v = *(vbool4_t*)(in + 200);
> + *(vbool4_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vbool4_t v = *(vbool4_t*)(in + 300);
> + *(vbool4_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool4_t v = *(vbool4_t*)(in + i);
> + *(vbool4_t*)(out + i) = v;
> + }
> +}
> +
> +void f6 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vbool2_t v = *(vbool2_t*)(in + 100);
> + *(vbool2_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vbool2_t v = *(vbool2_t*)(in + 200);
> + *(vbool2_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vbool2_t v = *(vbool2_t*)(in + 300);
> + *(vbool2_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool2_t v = *(vbool2_t*)(in + i);
> + *(vbool2_t*)(out + i) = v;
> + }
> +}
> +
> +void f7 (void * restrict in, void * restrict out, int n, int cond)
> +{
> + switch (cond)
> + {
> + case 1:{
> + vbool1_t v = *(vbool1_t*)(in + 100);
> + *(vbool1_t*)(out + 100) = v;
> + break;
> + }
> + case 2:{
> + vbool1_t v = *(vbool1_t*)(in + 200);
> + *(vbool1_t*)(out + 100) = v;
> + break;
> + }
> + case 3:{
> + vbool1_t v = *(vbool1_t*)(in + 300);
> + *(vbool1_t*)(out + 100) = v;
> + break;
> + }
> + default:{
> + break;
> + }
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vbool1_t v = *(vbool1_t*)(in + i);
> + *(vbool1_t*)(out + i) = v;
> + }
> +}
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m1,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m8,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m1,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m2,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m4,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*m8,\s*t[au],\s*m[au]\s+\.L[0-9]+\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-3.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-3.c
> new file mode 100644
> index 00000000000..a03f36ab1c5
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-3.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vint8mf4_t v = *(vint8mf4_t*)(in + 100);
> + *(vint8mf4_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vint8mf4_t v = *(vint8mf4_t*)(in + 200);
> + *(vint8mf4_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vint8mf4_t v = *(vint8mf4_t*)(in + 300);
> + *(vint8mf4_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf4_t v = *(vint8mf4_t*)(in + i);
> + *(vint8mf4_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-4.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-4.c
> new file mode 100644
> index 00000000000..ab16ef2e148
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-4.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vuint8mf4_t v = *(vuint8mf4_t*)(in + 100);
> + *(vuint8mf4_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vuint8mf4_t v = *(vuint8mf4_t*)(in + 200);
> + *(vuint8mf4_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vuint8mf4_t v = *(vuint8mf4_t*)(in + 300);
> + *(vuint8mf4_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint8mf4_t v = *(vuint8mf4_t*)(in + i);
> + *(vuint8mf4_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf4,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-5.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-5.c
> new file mode 100644
> index 00000000000..7722b9eef1f
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-5.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vint8mf2_t v = *(vint8mf2_t*)(in + 100);
> + *(vint8mf2_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vint8mf2_t v = *(vint8mf2_t*)(in + 200);
> + *(vint8mf2_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vint8mf2_t v = *(vint8mf2_t*)(in + 300);
> + *(vint8mf2_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf2_t v = *(vint8mf2_t*)(in + i);
> + *(vint8mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-6.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-6.c
> new file mode 100644
> index 00000000000..7ec497129b1
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-6.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vuint8mf2_t v = *(vuint8mf2_t*)(in + 100);
> + *(vuint8mf2_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vuint8mf2_t v = *(vuint8mf2_t*)(in + 200);
> + *(vuint8mf2_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vuint8mf2_t v = *(vuint8mf2_t*)(in + 300);
> + *(vuint8mf2_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint8mf2_t v = *(vuint8mf2_t*)(in + i);
> + *(vuint8mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e8,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-7.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-7.c
> new file mode 100644
> index 00000000000..2772de06151
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-7.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vint16mf4_t v = *(vint16mf4_t*)(in + 100);
> + *(vint16mf4_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vint16mf4_t v = *(vint16mf4_t*)(in + 200);
> + *(vint16mf4_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vint16mf4_t v = *(vint16mf4_t*)(in + 300);
> + *(vint16mf4_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint16mf4_t v = *(vint16mf4_t*)(in + i);
> + *(vint16mf4_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-8.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-8.c
> new file mode 100644
> index 00000000000..124c09b485a
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-8.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vuint16mf4_t v = *(vuint16mf4_t*)(in + 100);
> + *(vuint16mf4_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vuint16mf4_t v = *(vuint16mf4_t*)(in + 200);
> + *(vuint16mf4_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vuint16mf4_t v = *(vuint16mf4_t*)(in + 300);
> + *(vuint16mf4_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vuint16mf4_t v = *(vuint16mf4_t*)(in + i);
> + *(vuint16mf4_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-9.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-9.c
> new file mode 100644
> index 00000000000..7699a4408ad
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_miss_default-9.c
> @@ -0,0 +1,32 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +/* The for loop body should not have vsetvl instruction. */
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 1)
> + {
> + vint16mf2_t v = *(vint16mf2_t*)(in + 100);
> + *(vint16mf2_t*)(out + 100) = v;
> + }
> + else if (cond == 2)
> + {
> + vint16mf2_t v = *(vint16mf2_t*)(in + 200);
> + *(vint16mf2_t*)(out + 200) = v;
> + }
> + else if (cond == 3)
> + {
> + vint16mf2_t v = *(vint16mf2_t*)(in + 300);
> + *(vint16mf2_t*)(out + 300) = v;
> + }
> + for (int i = 0; i < n; i++)
> + {
> + vint16mf2_t v = *(vint16mf2_t*)(in + i);
> + *(vint16mf2_t*)(out + i) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf2,\s*t[au],\s*m[au]}
> 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+(?:ra|[sgtf]p|t[0-6]|s[0-9]|s10|s11|a[0-7]),\s*zero,\s*e16,\s*mf2,\s*t[au],\s*m[au]\s+\.L[0-9]\:}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops"
> no-opts "-g" } } } } */
> --
> 2.36.3
>
>
prev parent reply other threads:[~2022-12-19 15:09 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-12-14 8:15 juzhe.zhong
2022-12-19 15:08 ` Kito Cheng [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CA+yXCZDYu0XWUQbRXWwYabWUHzOSqe0ZmGm22AkYi_khKqMSUw@mail.gmail.com \
--to=kito.cheng@gmail.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=juzhe.zhong@rivai.ai \
--cc=palmer@dabbelt.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).