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 5
Date: Mon, 19 Dec 2022 23:09:27 +0800 [thread overview]
Message-ID: <CA+yXCZDBn773ehgmi5wS8zvdKswRcR9D_HHj5XuAJ2KrB+omaw@mail.gmail.com> (raw)
In-Reply-To: <20221214082558.261570-1-juzhe.zhong@rivai.ai>
[-- Attachment #1: Type: text/plain, Size: 91436 bytes --]
Commited to trunk
<juzhe.zhong@rivai.ai> 於 2022年12月14日 週三 16:27 寫道：
> From: Ju-Zhe Zhong <juzhe.zhong@rivai.ai>
>
> gcc/testsuite/ChangeLog:
>
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-1.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-10.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-11.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-12.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-13.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-14.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-15.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-16.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-17.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-18.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-19.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-2.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-20.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-21.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-22.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-23.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-24.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-25.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-26.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-27.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-28.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-29.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-3.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-30.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-31.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-32.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-33.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-34.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-35.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-36.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-37.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-38.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-39.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-4.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-40.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-41.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-42.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-43.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-44.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-45.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-46.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-5.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-6.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-7.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-8.c: New test.
> * gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-9.c: New test.
>
> ---
> .../riscv/rvv/vsetvl/vlmax_back_prop-1.c | 36 ++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-10.c | 59 +++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-11.c | 63 +++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-12.c | 64 ++++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-13.c | 64 ++++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-14.c | 58 +++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-15.c | 143 ++++++++++++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-16.c | 54 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-17.c | 59 +++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-18.c | 58 +++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-19.c | 48 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-2.c | 50 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-20.c | 59 +++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-21.c | 50 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-22.c | 58 +++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-23.c | 41 +++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-24.c | 41 +++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-25.c | 96 +++++++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-26.c | 89 ++++++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-27.c | 51 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-28.c | 54 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-29.c | 54 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-3.c | 47 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-30.c | 44 +++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-31.c | 46 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-32.c | 46 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-33.c | 45 +++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-34.c | 45 +++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-35.c | 38 +++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-36.c | 47 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-37.c | 41 +++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-38.c | 73 +++++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-39.c | 20 +++
> .../riscv/rvv/vsetvl/vlmax_back_prop-4.c | 104 ++++++++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-40.c | 22 +++
> .../riscv/rvv/vsetvl/vlmax_back_prop-41.c | 26 +++
> .../riscv/rvv/vsetvl/vlmax_back_prop-42.c | 26 +++
> .../riscv/rvv/vsetvl/vlmax_back_prop-43.c | 27 +++
> .../riscv/rvv/vsetvl/vlmax_back_prop-44.c | 28 ++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-45.c | 34 ++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-46.c | 32 ++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-5.c | 48 ++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-6.c | 155 ++++++++++++++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-7.c | 43 +++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-8.c | 149 +++++++++++++++++
> .../riscv/rvv/vsetvl/vlmax_back_prop-9.c | 44 +++++
> 46 files changed, 2579 insertions(+)
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-1.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-10.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-11.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-12.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-13.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-14.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-15.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-16.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-17.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-18.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-19.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-2.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-20.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-21.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-22.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-23.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-24.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-25.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-26.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-27.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-28.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-29.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-3.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-30.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-31.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-32.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-33.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-34.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-35.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-36.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-37.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-38.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-39.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-4.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-40.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-41.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-42.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-43.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-44.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-45.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-46.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-5.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-6.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-7.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-8.c
> create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-9.c
>
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-1.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-1.c
> new file mode 100644
> index 00000000000..47645ee7110
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-1.c
> @@ -0,0 +1,36 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" } } } }
> */
> +/* { dg-final { scan-assembler-times {vsetvli} 2 { 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_back_prop-10.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-10.c
> new file mode 100644
> index 00000000000..d36df955a43
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-10.c
> @@ -0,0 +1,59 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond,
> int cond2, int cond3)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0)
> + {
> + if (cond2 == 11)
> + {
> + for (int i = 0; i < n; i++)
> + {
> + out[i] = out[i] + 2;
> + }
> + }
> + else if (cond2 == 111)
> + {
> + if (cond3 == 300)
> + {
> + for (int i = 0; i < n; i++)
> + {
> + out[i] = out[i] + out[i];
> + }
> + }
> + }
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 2 { target { no-opts "-O0"
> no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-11.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-11.c
> new file mode 100644
> index 00000000000..fa818aa3b1c
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-11.c
> @@ -0,0 +1,63 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond,
> int cond2, int cond3)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + i);
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200 + i);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300 + i);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400 + i);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500 + i);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0)
> + {
> + if (cond2 == 11)
> + {
> + for (int i = 0; i < n; i++)
> + {
> + out[i] = out[i] + 2;
> + }
> + }
> + else if (cond2 == 111)
> + {
> + if (cond3 == 300)
> + {
> + for (int i = 0; i < n; i++)
> + {
> + out[i] = out[i] + out[i];
> + }
> + } else {
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 2000 + i);
> + *(vfloat32mf2_t*)(out + i + 4000) = v;
> + }
> + }
> + }
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 2 { target { no-opts "-O0"
> no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-12.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-12.c
> new file mode 100644
> index 00000000000..324e38d3fc6
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-12.c
> @@ -0,0 +1,64 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond,
> int cond2, int cond3)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + i);
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200 + i);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300 + i);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400 + i);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500 + i);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0)
> + {
> + if (cond2 == 11)
> + {
> + for (int i = 0; i < n; i++)
> + {
> + out[i] = out[i] + 2;
> + }
> + }
> + else if (cond2 == 111)
> + {
> + if (cond3 == 300)
> + {
> + for (int i = 0; i < n; i++)
> + {
> + out[i] = out[i] + out[i];
> + }
> + } else {
> + for (int i = 0; i < n; i++) {
> + vint8mf2_t v = *(vint8mf2_t*)(in + 2000 + i);
> + *(vint8mf2_t*)(out + i + 4000) = v;
> + }
> + }
> + }
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" no-opts "-flto" } } } } */
> +/* { 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" no-opts "-flto" } } } } */
> +/* { 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" no-opts "-flto" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 3 { target { no-opts "-O0"
> no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" no-opts
> "-flto" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-13.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-13.c
> new file mode 100644
> index 00000000000..23d21557d03
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-13.c
> @@ -0,0 +1,64 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond,
> int cond2, int cond3)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + i);
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200 + i);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300 + i);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400 + i);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500 + i);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0)
> + {
> + if (cond2 == 11)
> + {
> + for (int i = 0; i < n; i++)
> + {
> + out[i] = out[i] + 2;
> + }
> + }
> + else if (cond2 == 111)
> + {
> + if (cond3 == 300)
> + {
> + for (int i = 0; i < n; i++)
> + {
> + out[i] = out[i] + out[i];
> + }
> + } else {
> + for (int i = 0; i < n; i++) {
> + vint8mf2_t v = *(vint8mf2_t*)(in + 2000 + i);
> + *(vint8mf2_t*)(out + i + 4000) = v;
> + }
> + }
> + }
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16m1_t v;
> + *(vint16m1_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" no-opts "-flto" } } } } */
> +/* { 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*m1,\s*t[au],\s*m[au]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" no-opts "-flto" } } } } */
> +/* { 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" no-opts "-flto" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 3 { target { no-opts "-O0"
> no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" no-opts
> "-flto" } } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-14.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-14.c
> new file mode 100644
> index 00000000000..da48ce2f1f3
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-14.c
> @@ -0,0 +1,58 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0) {
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333);
> + *(vint8mf8_t*)(out + i + 333) = v;
> + }
> + } else if (cond == 1){
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 444);
> + *(vint32mf2_t*)(out + i + 444) = v;
> + }
> + } else if (cond == 2) {
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 555);
> + *(vint64m1_t*)(out + i + 555) = v;
> + }
> + } else {
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 666);
> + *(vfloat64m1_t*)(out + i + 666) = v;
> + }
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 2 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-15.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-15.c
> new file mode 100644
> index 00000000000..7dd931c9df8
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-15.c
> @@ -0,0 +1,143 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0) {
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333);
> + *(vint8mf8_t*)(out + i + 333) = v;
> + }
> + } else if (cond == 1){
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 444);
> + *(vint32mf2_t*)(out + i + 444) = v;
> + }
> + } else if (cond == 2) {
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 555);
> + *(vint64m1_t*)(out + i + 555) = v;
> + }
> + } else {
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 666);
> + *(vfloat64m1_t*)(out + i + 666) = v;
> + }
> + }
> +
> + if (cond == 0) {
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333);
> + *(vint8mf8_t*)(out + i + 333) = v;
> + }
> + } else if (cond == 1){
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 444);
> + *(vint32mf2_t*)(out + i + 444) = v;
> + }
> + } else if (cond == 2) {
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 555);
> + *(vint64m1_t*)(out + i + 555) = v;
> + }
> + } else {
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 666);
> + *(vfloat64m1_t*)(out + i + 666) = v;
> + }
> + }
> +
> + if (cond == 0) {
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333);
> + *(vint8mf8_t*)(out + i + 333) = v;
> + }
> + } else if (cond == 1){
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 444);
> + *(vint32mf2_t*)(out + i + 444) = v;
> + }
> + } else if (cond == 2) {
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 555);
> + *(vint64m1_t*)(out + i + 555) = v;
> + }
> + } else {
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 666);
> + *(vfloat64m1_t*)(out + i + 666) = v;
> + }
> + }
> + if (cond == 0) {
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333);
> + *(vint8mf8_t*)(out + i + 333) = v;
> + }
> + } else if (cond == 1){
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 444);
> + *(vint32mf2_t*)(out + i + 444) = v;
> + }
> + } else if (cond == 2) {
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 555);
> + *(vint64m1_t*)(out + i + 555) = v;
> + }
> + } else {
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 666);
> + *(vfloat64m1_t*)(out + i + 666) = v;
> + }
> + }
> + if (cond == 0) {
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333);
> + *(vint8mf8_t*)(out + i + 333) = v;
> + }
> + } else if (cond == 1){
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 444);
> + *(vint32mf2_t*)(out + i + 444) = v;
> + }
> + } else if (cond == 2) {
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 555);
> + *(vint64m1_t*)(out + i + 555) = v;
> + }
> + } else {
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 666);
> + *(vfloat64m1_t*)(out + i + 666) = v;
> + }
> + }
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 2 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-16.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-16.c
> new file mode 100644
> index 00000000000..84abe55a2b0
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-16.c
> @@ -0,0 +1,54 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 771);
> + *(vint8mf8_t*)(out + 771) = v;
> + vint32mf2_t v2 = *(vint32mf2_t*)(in + 71);
> + *(vint32mf2_t*)(out + 71) = v2;
> + vfloat32mf2_t v3 = *(vfloat32mf2_t*)(in + 17);
> + *(vfloat32mf2_t*)(out + 17) = v3;
> + vuint32mf2_t v4 = *(vuint32mf2_t*)(in + 117);
> + *(vuint32mf2_t*)(out + 117) = v4;
> + } else {
> + vfloat32mf2_t v0 = *(vfloat32mf2_t*)(in + 1123);
> + *(vfloat32mf2_t*)(out + 1123) = v0;
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333);
> + *(vint8mf8_t*)(out + 333) = v;
> + vbool64_t v2 = *(vbool64_t*)(in + 91);
> + *(vbool64_t*)(out + 91) = v2;
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops"
> no-opts "-g" no-opts "-flto" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 3 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-17.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-17.c
> new file mode 100644
> index 00000000000..dce21cc8dbc
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-17.c
> @@ -0,0 +1,59 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0) {
> + for (int i = 0; i < n; i++){
> + vint8mf8_t v = *(vint8mf8_t*)(in + 771 + i);
> + *(vint8mf8_t*)(out + 771 + i) = v;
> + vint32mf2_t v2 = *(vint32mf2_t*)(in + 71 + i);
> + *(vint32mf2_t*)(out + 71 + i) = v2;
> + vfloat32mf2_t v3 = *(vfloat32mf2_t*)(in + 17 + i);
> + *(vfloat32mf2_t*)(out + 17 + i) = v3;
> + vuint32mf2_t v4 = *(vuint32mf2_t*)(in + 117 + i);
> + *(vuint32mf2_t*)(out + 117 + i) = v4;
> + }
> + } else {
> + for (int i = 0; i < n; i++){
> + vfloat32mf2_t v0 = *(vfloat32mf2_t*)(in + 1123 + i);
> + *(vfloat32mf2_t*)(out + 1123 + i) = v0;
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333 + i);
> + *(vint8mf8_t*)(out + 333 + i) = v;
> + vbool64_t v2 = *(vbool64_t*)(in + 91 + i);
> + *(vbool64_t*)(out + 91 + i) = v2;
> + }
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 2 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-18.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-18.c
> new file mode 100644
> index 00000000000..18c44d6479d
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-18.c
> @@ -0,0 +1,58 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0) {
> + for (int i = 0; i < n; i++){
> + vint8mf8_t v = *(vint8mf8_t*)(in + 771 + i);
> + *(vint8mf8_t*)(out + 771 + i) = v;
> + vint32mf2_t v2 = *(vint32mf2_t*)(in + 71 + i);
> + *(vint32mf2_t*)(out + 71 + i) = v2;
> + vfloat32mf2_t v3 = *(vfloat32mf2_t*)(in + 17 + i);
> + *(vfloat32mf2_t*)(out + 17 + i) = v3;
> + vuint32mf2_t v4 = *(vuint32mf2_t*)(in + 117 + i);
> + *(vuint32mf2_t*)(out + 117 + i) = v4;
> + }
> + } else {
> + vfloat32mf2_t v0 = *(vfloat32mf2_t*)(in + 1123);
> + *(vfloat32mf2_t*)(out + 1123) = v0;
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333);
> + *(vint8mf8_t*)(out + 333) = v;
> + vbool64_t v2 = *(vbool64_t*)(in + 91);
> + *(vbool64_t*)(out + 91) = v2;
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 3 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-19.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-19.c
> new file mode 100644
> index 00000000000..0c6a572671a
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-19.c
> @@ -0,0 +1,48 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 771);
> + *(vint8mf8_t*)(out + 771) = v;
> + vint32mf2_t v2 = *(vint32mf2_t*)(in + 71);
> + *(vint32mf2_t*)(out + 71) = v2;
> + vfloat32mf2_t v3 = *(vfloat32mf2_t*)(in + 17);
> + *(vfloat32mf2_t*)(out + 17) = v3;
> + vuint32mf2_t v4 = *(vuint32mf2_t*)(in + 117);
> + *(vuint32mf2_t*)(out + 117) = v4;
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = v;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 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*mf8,\s*t[au],\s*m[au]}
> 2 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 3 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-2.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-2.c
> new file mode 100644
> index 00000000000..3e7d8f4030f
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-2.c
> @@ -0,0 +1,50 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + out[i] = out[i] + 2;
> + }
> + for (int i = 0; i < n; i++) {
> + out[i] = out[i] + out[i];
> + }
> + for (int i = 0; i < n; i++) {
> + out[i] = out[i] * 2;
> + }
> + for (int i = 0; i < n; i++) {
> + out[i] = out[i] * out[i];
> + }
> + for (int i = 0; i < n; i++) {
> + out[i] = out[i] * out[i] + 100;
> + }
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" } } } }
> */
> +/* { dg-final { scan-assembler-times {vsetvli} 2 { 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_back_prop-20.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-20.c
> new file mode 100644
> index 00000000000..dce21cc8dbc
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-20.c
> @@ -0,0 +1,59 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0) {
> + for (int i = 0; i < n; i++){
> + vint8mf8_t v = *(vint8mf8_t*)(in + 771 + i);
> + *(vint8mf8_t*)(out + 771 + i) = v;
> + vint32mf2_t v2 = *(vint32mf2_t*)(in + 71 + i);
> + *(vint32mf2_t*)(out + 71 + i) = v2;
> + vfloat32mf2_t v3 = *(vfloat32mf2_t*)(in + 17 + i);
> + *(vfloat32mf2_t*)(out + 17 + i) = v3;
> + vuint32mf2_t v4 = *(vuint32mf2_t*)(in + 117 + i);
> + *(vuint32mf2_t*)(out + 117 + i) = v4;
> + }
> + } else {
> + for (int i = 0; i < n; i++){
> + vfloat32mf2_t v0 = *(vfloat32mf2_t*)(in + 1123 + i);
> + *(vfloat32mf2_t*)(out + 1123 + i) = v0;
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333 + i);
> + *(vint8mf8_t*)(out + 333 + i) = v;
> + vbool64_t v2 = *(vbool64_t*)(in + 91 + i);
> + *(vbool64_t*)(out + 91 + i) = v2;
> + }
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 2 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-21.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-21.c
> new file mode 100644
> index 00000000000..7c2435ab726
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-21.c
> @@ -0,0 +1,50 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0) {
> + for (int i = 0; i < n; i++){
> + vint8mf8_t v = *(vint8mf8_t*)(in + 771 + i);
> + *(vint8mf8_t*)(out + 771 + i) = v;
> + vint32mf2_t v2 = *(vint32mf2_t*)(in + 71 + i);
> + *(vint32mf2_t*)(out + 71 + i) = v2;
> + vfloat32mf2_t v3 = *(vfloat32mf2_t*)(in + 17 + i);
> + *(vfloat32mf2_t*)(out + 17 + i) = v3;
> + vuint32mf2_t v4 = *(vuint32mf2_t*)(in + 117 + i);
> + *(vuint32mf2_t*)(out + 117 + i) = v4;
> + }
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 2 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-22.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-22.c
> new file mode 100644
> index 00000000000..222e0c6cbee
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-22.c
> @@ -0,0 +1,58 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0) {
> + for (int i = 0; i < n; i++){
> + vint8mf8_t v = *(vint8mf8_t*)(in + 771 + i);
> + *(vint8mf8_t*)(out + 771 + i) = v;
> + vint32mf2_t v2 = *(vint32mf2_t*)(in + 71 + i);
> + *(vint32mf2_t*)(out + 71 + i) = v2;
> + vfloat32mf2_t v3 = *(vfloat32mf2_t*)(in + 17 + i);
> + *(vfloat32mf2_t*)(out + 17 + i) = v3;
> + vuint32mf2_t v4 = *(vuint32mf2_t*)(in + 117 + i);
> + *(vuint32mf2_t*)(out + 117 + i) = v4;
> + }
> + } else {
> + vfloat32mf2_t v0 = *(vfloat32mf2_t*)(in + 1123);
> + *(vfloat32mf2_t*)(out + 1123) = v0;
> + vint8mf8_t v = *(vint8mf8_t*)(in + 333);
> + *(vint8mf8_t*)(out + 333) = v;
> + vbool64_t v2 = *(vbool64_t*)(in + 91);
> + *(vbool64_t*)(out + 91) = v2;
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 3 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-23.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-23.c
> new file mode 100644
> index 00000000000..1dd55cdbb0d
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-23.c
> @@ -0,0 +1,41 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + vint32mf2_t v;
> + *(vint32mf2_t*)(out + 7000) = v;
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 4 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-24.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-24.c
> new file mode 100644
> index 00000000000..931bba5389d
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-24.c
> @@ -0,0 +1,41 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + vint32mf2_t v;
> + *(vint32mf2_t*)(out + 7000) = v;
> +
> + for (int i = 0; i < n; i++) {
> + vbool64_t v;
> + *(vbool64_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 3 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-25.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-25.c
> new file mode 100644
> index 00000000000..93015e0c5f5
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-25.c
> @@ -0,0 +1,96 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 0) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 100);
> + *(vint8mf8_t*)(out + 100) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint16mf4_t v2;
> + *(vint16mf4_t*)(out + i + 100) = v2;
> + }
> + } else if (cond == 1) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 200);
> + *(vint8mf8_t*)(out + 200) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint32mf2_t v2;
> + *(vint32mf2_t*)(out + i + 200) = v2;
> + }
> + } else if (cond == 2) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 300);
> + *(vint8mf8_t*)(out + 300) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf8_t v2;
> + *(vint8mf8_t*)(out + i + 300) = v2;
> + }
> + } else if (cond == 3) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 400);
> + *(vint8mf8_t*)(out + 400) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint64m1_t v2;
> + *(vint64m1_t*)(out + i + 400) = v2;
> + }
> + } else if (cond == 4) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 500);
> + *(vint8mf8_t*)(out + 500) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vfloat32mf2_t v2;
> + *(vfloat32mf2_t*)(out + i + 500) = v2;
> + }
> + } else if (cond == 5) {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 600);
> + *(vuint8mf8_t*)(out + 600) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vuint16mf4_t v2;
> + *(vuint16mf4_t*)(out + i + 600) = v2;
> + }
> + } else if (cond == 6) {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 700);
> + *(vuint8mf8_t*)(out + 700) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vuint32mf2_t v2;
> + *(vuint32mf2_t*)(out + i + 700) = v2;
> + }
> + } else if (cond == 7) {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 800);
> + *(vuint8mf8_t*)(out + 800) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vuint8mf8_t v2;
> + *(vuint8mf8_t*)(out + i + 800) = v2;
> + }
> + } else if (cond == 8) {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 900);
> + *(vuint8mf8_t*)(out + 900) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vuint64m1_t v2;
> + *(vuint64m1_t*)(out + i + 900) = v2;
> + }
> + } else {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 1000);
> + *(vuint8mf8_t*)(out + 1000) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vfloat32mf2_t v2;
> + *(vfloat32mf2_t*)(out + i + 1000) = v2;
> + }
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]} 4 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 2 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e64,\s*m1,\s*t[au],\s*m[au]} 2 { target {
> no-opts "-O0" no-opts "-O1" 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]}
> 10 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]} 2 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 20 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-26.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-26.c
> new file mode 100644
> index 00000000000..adb831d3ee1
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-26.c
> @@ -0,0 +1,89 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (void * restrict in, void * restrict out, int n, int cond)
> +{
> + if (cond == 0) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 100);
> + *(vint8mf8_t*)(out + 100) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint16mf4_t v2;
> + *(vint16mf4_t*)(out + i + 100) = v2;
> + }
> + } else if (cond == 1) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 200);
> + *(vint8mf8_t*)(out + 200) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint32mf2_t v2;
> + *(vint32mf2_t*)(out + i + 200) = v2;
> + }
> + } else if (cond == 2) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 300);
> + *(vint8mf8_t*)(out + 300) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf8_t v2;
> + *(vint8mf8_t*)(out + i + 300) = v2;
> + }
> + } else if (cond == 3) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 400);
> + *(vint8mf8_t*)(out + 400) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint64m1_t v2;
> + *(vint64m1_t*)(out + i + 400) = v2;
> + }
> + } else if (cond == 4) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 500);
> + *(vint8mf8_t*)(out + 500) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vfloat32mf2_t v2;
> + *(vfloat32mf2_t*)(out + i + 500) = v2;
> + }
> + } else if (cond == 5) {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 600);
> + *(vuint8mf8_t*)(out + 600) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vuint16mf4_t v2;
> + *(vuint16mf4_t*)(out + i + 600) = v2;
> + }
> + } else if (cond == 6) {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 700);
> + *(vuint8mf8_t*)(out + 700) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vuint32mf2_t v2;
> + *(vuint32mf2_t*)(out + i + 700) = v2;
> + }
> + } else if (cond == 7) {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 800);
> + *(vuint8mf8_t*)(out + 800) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vuint8mf8_t v2;
> + *(vuint8mf8_t*)(out + i + 800) = v2;
> + }
> + } else if (cond == 8) {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 900);
> + *(vuint8mf8_t*)(out + 900) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vuint64m1_t v2;
> + *(vuint64m1_t*)(out + i + 900) = v2;
> + }
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]} 3 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e64,\s*m1,\s*t[au],\s*m[au]} 2 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]} 2 { target {
> no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 8 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 17 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-27.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-27.c
> new file mode 100644
> index 00000000000..16a5f2d98fb
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-27.c
> @@ -0,0 +1,51 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (void * restrict in, void * restrict out, void * restrict in2,
> void * restrict out2, int n, int cond)
> +{
> + if (cond == 0) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 100);
> + *(vint8mf8_t*)(out + 100) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint16mf4_t v2;
> + *(vint16mf4_t*)(out + i + 100) = v2;
> + }
> + } else {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 1000);
> + *(vuint8mf8_t*)(out + 1000) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vfloat32mf2_t v2;
> + *(vfloat32mf2_t*)(out + i + 1000) = v2;
> + vbool64_t v3;
> + *(vbool64_t*)(out + i + 2000) = v3;
> + vbool32_t v4;
> + *(vbool32_t*)(out + i + 3000) = v4;
> + vbool16_t v5;
> + *(vbool16_t*)(out + i + 4000) = v5;
> + vbool8_t v6;
> + *(vbool8_t*)(out + i + 5000) = v6;
> + vbool4_t v7;
> + *(vbool4_t*)(out + i + 6000) = v7;
> + vbool2_t v8;
> + *(vbool2_t*)(out + i + 7000) = v8;
> + vbool1_t v9;
> + *(vbool1_t*)(out + i + 8000) = v9;
> + }
> + }
> +}
> +
> +/* { 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 9 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-28.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-28.c
> new file mode 100644
> index 00000000000..366b5cf4925
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-28.c
> @@ -0,0 +1,54 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (void * restrict in, void * restrict out, void * restrict in2,
> void * restrict out2, int n, int cond)
> +{
> + if (cond == 0) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 100);
> + *(vint8mf8_t*)(out + 100) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint16mf4_t v2;
> + *(vint16mf4_t*)(out + i + 100) = v2;
> + }
> + } else {
> + vuint8mf8_t v = *(vuint8mf8_t*)(in + 1000);
> + *(vuint8mf8_t*)(out + 1000) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vfloat32mf2_t v2;
> + *(vfloat32mf2_t*)(out + i + 1000) = v2;
> + vbool64_t v3;
> + *(vbool64_t*)(out + i + 2000) = v3;
> + vbool32_t v4;
> + *(vbool32_t*)(out + i + 3000) = v4;
> + vbool16_t v5;
> + *(vbool16_t*)(out + i + 4000) = v5;
> + vbool8_t v6;
> + *(vbool8_t*)(out + i + 5000) = v6;
> + vbool4_t v7;
> + *(vbool4_t*)(out + i + 6000) = v7;
> + vbool2_t v8;
> + *(vbool2_t*)(out + i + 7000) = v8;
> + vbool1_t v9;
> + *(vbool1_t*)(out + i + 8000) = v9;
> + vuint32mf2_t v10;
> + *(vuint32mf2_t*)(out + i + 100000) = v10;
> + }
> + }
> +}
> +
> +/* { 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 10 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-29.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-29.c
> new file mode 100644
> index 00000000000..cbd7a8b1a6b
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-29.c
> @@ -0,0 +1,54 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (void * restrict in, void * restrict out, void * restrict in2,
> void * restrict out2, int n, int cond)
> +{
> + if (cond == 0) {
> + vint8mf8_t v = *(vint8mf8_t*)(in + 100);
> + *(vint8mf8_t*)(out + 100) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vint16mf4_t v2;
> + *(vint16mf4_t*)(out + i + 100) = v2;
> + }
> + } else {
> + vuint8mf8_t v;
> + *(vuint8mf8_t*)(out + 1000) = v;
> + for (int i = 0; i < n; i++)
> + {
> + vfloat32mf2_t v2;
> + *(vfloat32mf2_t*)(out + i + 1000) = v2;
> + vbool64_t v3;
> + *(vbool64_t*)(out + i + 2000) = v3;
> + vbool32_t v4;
> + *(vbool32_t*)(out + i + 3000) = v4;
> + vbool16_t v5;
> + *(vbool16_t*)(out + i + 4000) = v5;
> + vbool8_t v6;
> + *(vbool8_t*)(out + i + 5000) = v6;
> + vbool4_t v7;
> + *(vbool4_t*)(out + i + 6000) = v7;
> + vbool2_t v8;
> + *(vbool2_t*)(out + i + 7000) = v8;
> + vbool1_t v9;
> + *(vbool1_t*)(out + i + 8000) = v9;
> + vuint32mf2_t v10;
> + *(vuint32mf2_t*)(out + i + 100000) = v10;
> + }
> + }
> +}
> +
> +/* { 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]}
> 2 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e32,\s*mf2,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 11 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-3.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-3.c
> new file mode 100644
> index 00000000000..f6ac96fe900
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-3.c
> @@ -0,0 +1,47 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (int32_t * restrict in, int32_t * restrict out, int n, int cond)
> +{
> + for (int i = 0; i < n; i++) {
> + vint8mf8_t v = *(vint8mf8_t*)in;
> + *(vint8mf8_t*)(out + i + 200) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint32mf2_t v = *(vint32mf2_t*)(in + 200);
> + *(vint32mf2_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vint64m1_t v = *(vint64m1_t*)(in + 300);
> + *(vint64m1_t*)(out + i + 400) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat32mf2_t v = *(vfloat32mf2_t*)(in + 400);
> + *(vfloat32mf2_t*)(out + i + 500) = v;
> + }
> + for (int i = 0; i < n; i++) {
> + vfloat64m1_t v = *(vfloat64m1_t*)(in + 500);
> + *(vfloat64m1_t*)(out + i + 600) = v;
> + }
> +
> + if (cond == 0) {
> + for (int i = 0; i < n; i++) {
> + out[i] = out[i] + 2;
> + }
> + } else {
> + for (int i = 0; i < n; i++) {
> + out[i] = out[i] + 3;
> + }
> + }
> +
> + for (int i = 0; i < n; i++) {
> + vint16mf4_t v;
> + *(vint16mf4_t*)(out + i + 700) = 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e16,\s*mf4,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts
> "-g" } } } } */
> +/* { dg-final { scan-assembler-times {vsetvli} 2 { 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_back_prop-30.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-30.c
> new file mode 100644
> index 00000000000..3d800a1daef
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-30.c
> @@ -0,0 +1,44 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (void * restrict in, void * restrict out, void * restrict in2,
> void * restrict out2, int n, int cond)
> +{
> + for (int i = 0; i < n; i++)
> + {
> + vfloat32mf2_t v2 = *(vfloat32mf2_t*)(in + i + 1000);
> + *(vfloat32mf2_t*)(out + i + 1000) = v2;
> + vbool32_t v4;
> + *(vbool32_t*)(out + i + 3000) = v4;
> + vbool16_t v5;
> + *(vbool16_t*)(out + i + 4000) = v5;
> + vbool8_t v6;
> + *(vbool8_t*)(out + i + 5000) = v6;
> + vbool4_t v7;
> + *(vbool4_t*)(out + i + 6000) = v7;
> + vbool2_t v8;
> + *(vbool2_t*)(out + i + 7000) = v8;
> + vbool1_t v9;
> + *(vbool1_t*)(out + i + 8000) = v9;
> + vuint32mf2_t v10 = *(vuint32mf2_t*)(in + i + 100000);
> + *(vuint32mf2_t*)(out + i + 100000) = v10;
> + }
> +
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf8_t v1;
> + *(vint8mf8_t*)(out + i + 10) = v1;
> + }
> +}
> +
> +/* { 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]}
> 2 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +
> +/* { dg-final { scan-assembler-times {vsetvli} 9 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-31.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-31.c
> new file mode 100644
> index 00000000000..ad954f7c51c
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-31.c
> @@ -0,0 +1,46 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (void * restrict in, void * restrict out, void * restrict in2,
> void * restrict out2, int n, int cond)
> +{
> + for (int i = 0; i < n; i++)
> + {
> + vuint16mf4_t v2 = *(vuint16mf4_t*)(in + i + 1000);
> + *(vuint16mf4_t*)(out + i + 1000) = v2;
> + vbool32_t v4;
> + *(vbool32_t*)(out + i + 3000) = v4;
> + vbool16_t v5;
> + *(vbool16_t*)(out + i + 4000) = v5;
> + vbool8_t v6;
> + *(vbool8_t*)(out + i + 5000) = v6;
> + vbool4_t v7;
> + *(vbool4_t*)(out + i + 6000) = v7;
> + vbool2_t v8;
> + *(vbool2_t*)(out + i + 7000) = v8;
> + vbool1_t v9;
> + *(vbool1_t*)(out + i + 8000) = v9;
> + vuint32mf2_t v10 = *(vuint32mf2_t*)(in + i + 100000);
> + *(vuint32mf2_t*)(out + i + 100000) = v10;
> + }
> +
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf8_t v1;
> + *(vint8mf8_t*)(out + i + 10) = v1;
> + }
> +}
> +
> +/* { dg-final { scan-assembler-times
> {vsetvli\s+zero,\s*zero,\s*e8,\s*mf8,\s*t[au],\s*m[au]} 1 { target {
> no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts
> "-funroll-loops" no-opts "-g" } } } } */
> +
> +/* { dg-final { scan-assembler-times {vsetvli} 9 { target { no-opts
> "-O0" no-opts "-O1" no-opts "-Os" no-opts "-funroll-loops" no-opts "-g" }
> } } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-32.c
> b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-32.c
> new file mode 100644
> index 00000000000..aa4dac4e000
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/vsetvl/vlmax_back_prop-32.c
> @@ -0,0 +1,46 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv32gcv -mabi=ilp32 -fno-tree-vectorize" } */
> +
> +#include "riscv_vector.h"
> +
> +void f (void * restrict in, void * restrict out, void * restrict in2,
> void * restrict out2, int n, int cond)
> +{
> + for (int i = 0; i < n; i++)
> + {
> + vuint16mf4_t v2 = *(vuint16mf4_t*)(in + i + 1000);
> + *(vuint16mf4_t*)(out + i + 1000) = v2;
> + vbool32_t v4;
> + *(vbool32_t*)(out + i + 3000) = v4;
> + vbool16_t v5;
> + *(vbool16_t*)(out + i + 4000) = v5;
> + vbool8_t v6;
> + *(vbool8_t*)(out + i + 5000) = v6;
> + vbool4_t v7;
> + *(vbool4_t*)(out + i + 6000) = v7;
> + vbool2_t v8;
> + *(vbool2_t*)(out + i + 7000) = v8;
> + vbool1_t v9;
> + *(vbool1_t*)(out + i + 8000) = v9;
> + vuint32mf2_t v10;
> + *(vuint32mf2_t*)(out + i + 100000) = v10;
> + }
> +
> + for (int i = 0; i < n; i++)
> + {
> + vint8mf8_t v1 = *(vint8mf8_t*)(in + i + 100000);
> + *(vint8mf8_t*)(out + i + 10) = v1;
> + }
> +}
> +
> +/* { 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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]}
> 1 { target { no-opts "-O0" no-opts "-O1" 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 "-O1" 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]}
> 1 { target {
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 8:25 juzhe.zhong
2022-12-19 15:09 ` 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+yXCZDBn773ehgmi5wS8zvdKswRcR9D_HHj5XuAJ2KrB+omaw@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).