From 1a7d3a01967733cace38130322d514c4c1418191 Mon Sep 17 00:00:00 2001
From: Brandon Wu <brandon.wu@sifive.com>
Date: Tue, 12 Dec 2023 13:17:47 +0800
Subject: [PATCH 1/4] Recommit [RISCV] Update the interface of sifive vqmaccqoq
 (#74284)

The
spec(https://sifive.cdn.prismic.io/sifive/60d5a660-3af0-49a3-a904-d2bbb1a21517_int8-matmul-spec.pdf)
is updated.
---
 .../clang/Basic/riscv_sifive_vector.td        | 26 +++++---
 .../non-overloaded/sf_vqmacc_4x8x4.c          | 24 +++----
 .../non-overloaded/sf_vqmaccsu_4x8x4.c        | 24 +++----
 .../non-overloaded/sf_vqmaccu_4x8x4.c         | 24 +++----
 .../non-overloaded/sf_vqmaccus_4x8x4.c        | 24 +++----
 .../non-policy/overloaded/sf_vqmacc_4x8x4.c   | 24 +++----
 .../non-policy/overloaded/sf_vqmaccsu_4x8x4.c | 24 +++----
 .../non-policy/overloaded/sf_vqmaccu_4x8x4.c  | 24 +++----
 .../non-policy/overloaded/sf_vqmaccus_4x8x4.c | 24 +++----
 .../policy/non-overloaded/sf_vqmacc_4x8x4.c   | 24 +++----
 .../policy/non-overloaded/sf_vqmaccsu_4x8x4.c | 24 +++----
 .../policy/non-overloaded/sf_vqmaccu_4x8x4.c  | 24 +++----
 .../policy/non-overloaded/sf_vqmaccus_4x8x4.c | 24 +++----
 .../policy/overloaded/sf_vqmacc_4x8x4.c       | 24 +++----
 .../policy/overloaded/sf_vqmaccsu_4x8x4.c     | 24 +++----
 .../policy/overloaded/sf_vqmaccu_4x8x4.c      | 24 +++----
 .../policy/overloaded/sf_vqmaccus_4x8x4.c     | 24 +++----
 clang/test/Sema/rvv-required-features.c       | 18 +-----
 llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td    | 64 ++++++++++++-------
 .../CodeGen/RISCV/rvv/sf_vfwmacc_4x4x4.ll     | 20 +++---
 .../test/CodeGen/RISCV/rvv/sf_vqmacc_4x8x4.ll | 58 ++++++++---------
 .../CodeGen/RISCV/rvv/sf_vqmaccsu_4x8x4.ll    | 58 ++++++++---------
 .../CodeGen/RISCV/rvv/sf_vqmaccu_4x8x4.ll     | 58 ++++++++---------
 .../CodeGen/RISCV/rvv/sf_vqmaccus_4x8x4.ll    | 58 ++++++++---------
 24 files changed, 373 insertions(+), 371 deletions(-)

diff --git a/clang/include/clang/Basic/riscv_sifive_vector.td b/clang/include/clang/Basic/riscv_sifive_vector.td
index bb54e26641861..e19a34f7632fd 100644
--- a/clang/include/clang/Basic/riscv_sifive_vector.td
+++ b/clang/include/clang/Basic/riscv_sifive_vector.td
@@ -112,7 +112,7 @@ multiclass RVVVFWMACCBuiltinSet<list<list<string>> suffixes_prototypes> {
     defm NAME : RVVOutOp1Op2BuiltinSet<NAME, "b", suffixes_prototypes>;
 }
 
-multiclass RVVVQMACCBuiltinSet<list<list<string>> suffixes_prototypes> {
+multiclass RVVVQMACCDODBuiltinSet<list<list<string>> suffixes_prototypes> {
   let OverloadedName = NAME,
       Name = NAME,
       HasMasked = false,
@@ -120,6 +120,14 @@ multiclass RVVVQMACCBuiltinSet<list<list<string>> suffixes_prototypes> {
     defm NAME : RVVOutOp1Op2BuiltinSet<NAME, "i", suffixes_prototypes>;
 }
 
+multiclass RVVVQMACCQOQBuiltinSet<list<list<string>> suffixes_prototypes> {
+   let OverloadedName = NAME,
+       Name = NAME,
+       HasMasked = false,
+       Log2LMUL = [-1, 0, 1, 2] in
+     defm NAME : RVVOutOp1Op2BuiltinSet<NAME, "s", suffixes_prototypes>;
+}
+
 multiclass RVVVFNRCLIPBuiltinSet<string suffix, string prototype, string type_range> {
   let Log2LMUL = [-3, -2, -1, 0, 1, 2],
       Name = NAME,
@@ -130,18 +138,18 @@ multiclass RVVVFNRCLIPBuiltinSet<string suffix, string prototype, string type_ra
 
 let UnMaskedPolicyScheme = HasPolicyOperand in
   let RequiredFeatures = ["Xsfvqmaccdod"] in {
-    defm sf_vqmaccu_2x8x2 : RVVVQMACCBuiltinSet<[["", "v", "vv(FixedSEW:8)SUv(FixedSEW:8)Uv"]]>;
-    defm sf_vqmacc_2x8x2 : RVVVQMACCBuiltinSet<[["", "v", "vv(FixedSEW:8)Sv(FixedSEW:8)v"]]>;
-    defm sf_vqmaccus_2x8x2 : RVVVQMACCBuiltinSet<[["", "v", "vv(FixedSEW:8)SUv(FixedSEW:8)v"]]>;
-    defm sf_vqmaccsu_2x8x2 : RVVVQMACCBuiltinSet<[["", "v", "vv(FixedSEW:8)Sv(FixedSEW:8)Uv"]]>;
+    defm sf_vqmaccu_2x8x2 : RVVVQMACCDODBuiltinSet<[["", "v", "vv(FixedSEW:8)SUv(FixedSEW:8)Uv"]]>;
+    defm sf_vqmacc_2x8x2 : RVVVQMACCDODBuiltinSet<[["", "v", "vv(FixedSEW:8)Sv(FixedSEW:8)v"]]>;
+    defm sf_vqmaccus_2x8x2 : RVVVQMACCDODBuiltinSet<[["", "v", "vv(FixedSEW:8)SUv(FixedSEW:8)v"]]>;
+    defm sf_vqmaccsu_2x8x2 : RVVVQMACCDODBuiltinSet<[["", "v", "vv(FixedSEW:8)Sv(FixedSEW:8)Uv"]]>;
   }
 
 let UnMaskedPolicyScheme = HasPolicyOperand in
   let RequiredFeatures = ["Xsfvqmaccqoq"] in {
-    defm sf_vqmaccu_4x8x4 : RVVVQMACCBuiltinSet<[["", "v", "vv(FixedSEW:8)SUv(FixedSEW:8)Uv"]]>;
-    defm sf_vqmacc_4x8x4 : RVVVQMACCBuiltinSet<[["", "v", "vv(FixedSEW:8)Sv(FixedSEW:8)v"]]>;
-    defm sf_vqmaccus_4x8x4 : RVVVQMACCBuiltinSet<[["", "v", "vv(FixedSEW:8)SUv(FixedSEW:8)v"]]>;
-    defm sf_vqmaccsu_4x8x4 : RVVVQMACCBuiltinSet<[["", "v", "vv(FixedSEW:8)Sv(FixedSEW:8)Uv"]]>;
+    defm sf_vqmaccu_4x8x4 : RVVVQMACCQOQBuiltinSet<[["", "w", "ww(FixedSEW:8)SUv(FixedSEW:8)Uv"]]>;
+    defm sf_vqmacc_4x8x4 : RVVVQMACCQOQBuiltinSet<[["", "w", "ww(FixedSEW:8)Sv(FixedSEW:8)v"]]>;
+    defm sf_vqmaccus_4x8x4 : RVVVQMACCQOQBuiltinSet<[["", "w", "ww(FixedSEW:8)SUv(FixedSEW:8)v"]]>;
+    defm sf_vqmaccsu_4x8x4 : RVVVQMACCQOQBuiltinSet<[["", "w", "ww(FixedSEW:8)Sv(FixedSEW:8)Uv"]]>;
   }
 
 let UnMaskedPolicyScheme = HasPolicyOperand in
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmacc_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmacc_4x8x4.c
index 935cb2e007d35..80e1c443eb54b 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmacc_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmacc_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmacc_4x8x4_i32m1
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmacc_4x8x4_i32m1(vint32m1_t vd, vint8m1_t vs1, vint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmacc_4x8x4_i32m1(vint32m1_t vd, vint8m1_t vs1, vint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_i32m1(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmacc_4x8x4_i32m2
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmacc_4x8x4_i32m2(vint32m2_t vd, vint8m1_t vs1, vint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmacc_4x8x4_i32m2(vint32m2_t vd, vint8m1_t vs1, vint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_i32m2(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmacc_4x8x4_i32m4
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmacc_4x8x4_i32m4(vint32m4_t vd, vint8m1_t vs1, vint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmacc_4x8x4_i32m4(vint32m4_t vd, vint8m1_t vs1, vint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_i32m4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmacc_4x8x4_i32m8
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmacc_4x8x4_i32m8(vint32m8_t vd, vint8m1_t vs1, vint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmacc_4x8x4_i32m8(vint32m8_t vd, vint8m1_t vs1, vint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_i32m8(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccsu_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccsu_4x8x4.c
index f34517b24bcf2..8c0a6218c1d2f 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccsu_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccsu_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccsu_4x8x4_i32m1
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccsu_4x8x4_i32m1(vint32m1_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccsu_4x8x4_i32m1(vint32m1_t vd, vint8m1_t vs1, vuint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_i32m1(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccsu_4x8x4_i32m2
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccsu_4x8x4_i32m2(vint32m2_t vd, vint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccsu_4x8x4_i32m2(vint32m2_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_i32m2(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccsu_4x8x4_i32m4
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccsu_4x8x4_i32m4(vint32m4_t vd, vint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccsu_4x8x4_i32m4(vint32m4_t vd, vint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_i32m4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccsu_4x8x4_i32m8
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccsu_4x8x4_i32m8(vint32m8_t vd, vint8m1_t vs1, vuint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccsu_4x8x4_i32m8(vint32m8_t vd, vint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_i32m8(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccu_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccu_4x8x4.c
index ab7f6627ad1fb..b40891f417f2c 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccu_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccu_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccu_4x8x4_i32m1
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccu_4x8x4_i32m1(vint32m1_t vd, vuint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccu_4x8x4_i32m1(vint32m1_t vd, vuint8m1_t vs1, vuint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_i32m1(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccu_4x8x4_i32m2
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccu_4x8x4_i32m2(vint32m2_t vd, vuint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccu_4x8x4_i32m2(vint32m2_t vd, vuint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_i32m2(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccu_4x8x4_i32m4
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccu_4x8x4_i32m4(vint32m4_t vd, vuint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccu_4x8x4_i32m4(vint32m4_t vd, vuint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_i32m4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccu_4x8x4_i32m8
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccu_4x8x4_i32m8(vint32m8_t vd, vuint8m1_t vs1, vuint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccu_4x8x4_i32m8(vint32m8_t vd, vuint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_i32m8(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccus_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccus_4x8x4.c
index d0bcdcbf40cc3..d106aab64c514 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccus_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/non-overloaded/sf_vqmaccus_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccus_4x8x4_i32m1
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccus_4x8x4_i32m1(vint32m1_t vd, vuint8m1_t vs1, vint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccus_4x8x4_i32m1(vint32m1_t vd, vuint8m1_t vs1, vint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_i32m1(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccus_4x8x4_i32m2
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccus_4x8x4_i32m2(vint32m2_t vd, vuint8m1_t vs1, vint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccus_4x8x4_i32m2(vint32m2_t vd, vuint8m1_t vs1, vint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_i32m2(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccus_4x8x4_i32m4
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccus_4x8x4_i32m4(vint32m4_t vd, vuint8m1_t vs1, vint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccus_4x8x4_i32m4(vint32m4_t vd, vuint8m1_t vs1, vint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_i32m4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccus_4x8x4_i32m8
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccus_4x8x4_i32m8(vint32m8_t vd, vuint8m1_t vs1, vint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccus_4x8x4_i32m8(vint32m8_t vd, vuint8m1_t vs1, vint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_i32m8(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmacc_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmacc_4x8x4.c
index 839d09c4f9a98..88fae73069440 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmacc_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmacc_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmacc_4x8x4_i32m1
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmacc_4x8x4_i32m1(vint32m1_t vd, vint8m1_t vs1, vint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmacc_4x8x4_i32m1(vint32m1_t vd, vint8m1_t vs1, vint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmacc_4x8x4_i32m2
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmacc_4x8x4_i32m2(vint32m2_t vd, vint8m1_t vs1, vint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmacc_4x8x4_i32m2(vint32m2_t vd, vint8m1_t vs1, vint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmacc_4x8x4_i32m4
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmacc_4x8x4_i32m4(vint32m4_t vd, vint8m1_t vs1, vint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmacc_4x8x4_i32m4(vint32m4_t vd, vint8m1_t vs1, vint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmacc_4x8x4_i32m8
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmacc_4x8x4_i32m8(vint32m8_t vd, vint8m1_t vs1, vint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmacc_4x8x4_i32m8(vint32m8_t vd, vint8m1_t vs1, vint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccsu_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccsu_4x8x4.c
index b18853043e924..0aec4bfd9fe22 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccsu_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccsu_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccsu_4x8x4_i32m1
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccsu_4x8x4_i32m1(vint32m1_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccsu_4x8x4_i32m1(vint32m1_t vd, vint8m1_t vs1, vuint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccsu_4x8x4_i32m2
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccsu_4x8x4_i32m2(vint32m2_t vd, vint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccsu_4x8x4_i32m2(vint32m2_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccsu_4x8x4_i32m4
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccsu_4x8x4_i32m4(vint32m4_t vd, vint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccsu_4x8x4_i32m4(vint32m4_t vd, vint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccsu_4x8x4_i32m8
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccsu_4x8x4_i32m8(vint32m8_t vd, vint8m1_t vs1, vuint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccsu_4x8x4_i32m8(vint32m8_t vd, vint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccu_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccu_4x8x4.c
index 4cb966b08f237..81965e86f77c8 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccu_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccu_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccu_4x8x4_i32m1
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccu_4x8x4_i32m1(vint32m1_t vd, vuint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccu_4x8x4_i32m1(vint32m1_t vd, vuint8m1_t vs1, vuint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccu_4x8x4_i32m2
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccu_4x8x4_i32m2(vint32m2_t vd, vuint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccu_4x8x4_i32m2(vint32m2_t vd, vuint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccu_4x8x4_i32m4
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccu_4x8x4_i32m4(vint32m4_t vd, vuint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccu_4x8x4_i32m4(vint32m4_t vd, vuint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccu_4x8x4_i32m8
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccu_4x8x4_i32m8(vint32m8_t vd, vuint8m1_t vs1, vuint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccu_4x8x4_i32m8(vint32m8_t vd, vuint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccus_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccus_4x8x4.c
index f558151f88a3f..f2544cf3ef2ae 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccus_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/non-policy/overloaded/sf_vqmaccus_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccus_4x8x4_i32m1
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccus_4x8x4_i32m1(vint32m1_t vd, vuint8m1_t vs1, vint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccus_4x8x4_i32m1(vint32m1_t vd, vuint8m1_t vs1, vint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccus_4x8x4_i32m2
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccus_4x8x4_i32m2(vint32m2_t vd, vuint8m1_t vs1, vint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccus_4x8x4_i32m2(vint32m2_t vd, vuint8m1_t vs1, vint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccus_4x8x4_i32m4
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccus_4x8x4_i32m4(vint32m4_t vd, vuint8m1_t vs1, vint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccus_4x8x4_i32m4(vint32m4_t vd, vuint8m1_t vs1, vint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccus_4x8x4_i32m8
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 3)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 3)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccus_4x8x4_i32m8(vint32m8_t vd, vuint8m1_t vs1, vint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccus_4x8x4_i32m8(vint32m8_t vd, vuint8m1_t vs1, vint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmacc_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmacc_4x8x4.c
index 05c10840cabfa..8fdeac62a31f1 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmacc_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmacc_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmacc_4x8x4_i32m1_tu
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmacc_4x8x4_i32m1_tu(vint32m1_t vd, vint8m1_t vs1, vint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmacc_4x8x4_i32m1_tu(vint32m1_t vd, vint8m1_t vs1, vint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_i32m1_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmacc_4x8x4_i32m2_tu
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmacc_4x8x4_i32m2_tu(vint32m2_t vd, vint8m1_t vs1, vint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmacc_4x8x4_i32m2_tu(vint32m2_t vd, vint8m1_t vs1, vint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_i32m2_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmacc_4x8x4_i32m4_tu
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmacc_4x8x4_i32m4_tu(vint32m4_t vd, vint8m1_t vs1, vint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmacc_4x8x4_i32m4_tu(vint32m4_t vd, vint8m1_t vs1, vint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_i32m4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmacc_4x8x4_i32m8_tu
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmacc_4x8x4_i32m8_tu(vint32m8_t vd, vint8m1_t vs1, vint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmacc_4x8x4_i32m8_tu(vint32m8_t vd, vint8m1_t vs1, vint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_i32m8_tu(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccsu_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccsu_4x8x4.c
index bce1a4e9443fd..e02c790dfbeb7 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccsu_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccsu_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccsu_4x8x4_i32m1_tu
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccsu_4x8x4_i32m1_tu(vint32m1_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccsu_4x8x4_i32m1_tu(vint32m1_t vd, vint8m1_t vs1, vuint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_i32m1_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccsu_4x8x4_i32m2_tu
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccsu_4x8x4_i32m2_tu(vint32m2_t vd, vint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccsu_4x8x4_i32m2_tu(vint32m2_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_i32m2_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccsu_4x8x4_i32m4_tu
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccsu_4x8x4_i32m4_tu(vint32m4_t vd, vint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccsu_4x8x4_i32m4_tu(vint32m4_t vd, vint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_i32m4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccsu_4x8x4_i32m8_tu
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccsu_4x8x4_i32m8_tu(vint32m8_t vd, vint8m1_t vs1, vuint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccsu_4x8x4_i32m8_tu(vint32m8_t vd, vint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_i32m8_tu(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccu_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccu_4x8x4.c
index 36aaae9caebf6..ddeb6de007164 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccu_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccu_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccu_4x8x4_i32m1_tu
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccu_4x8x4_i32m1_tu(vint32m1_t vd, vuint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccu_4x8x4_i32m1_tu(vint32m1_t vd, vuint8m1_t vs1, vuint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_i32m1_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccu_4x8x4_i32m2_tu
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccu_4x8x4_i32m2_tu(vint32m2_t vd, vuint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccu_4x8x4_i32m2_tu(vint32m2_t vd, vuint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_i32m2_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccu_4x8x4_i32m4_tu
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccu_4x8x4_i32m4_tu(vint32m4_t vd, vuint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccu_4x8x4_i32m4_tu(vint32m4_t vd, vuint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_i32m4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccu_4x8x4_i32m8_tu
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccu_4x8x4_i32m8_tu(vint32m8_t vd, vuint8m1_t vs1, vuint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccu_4x8x4_i32m8_tu(vint32m8_t vd, vuint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_i32m8_tu(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccus_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccus_4x8x4.c
index f5ac2bf0f1f3a..397e406c2ee58 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccus_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/non-overloaded/sf_vqmaccus_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccus_4x8x4_i32m1_tu
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccus_4x8x4_i32m1_tu(vint32m1_t vd, vuint8m1_t vs1, vint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccus_4x8x4_i32m1_tu(vint32m1_t vd, vuint8m1_t vs1, vint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_i32m1_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccus_4x8x4_i32m2_tu
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccus_4x8x4_i32m2_tu(vint32m2_t vd, vuint8m1_t vs1, vint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccus_4x8x4_i32m2_tu(vint32m2_t vd, vuint8m1_t vs1, vint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_i32m2_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccus_4x8x4_i32m4_tu
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccus_4x8x4_i32m4_tu(vint32m4_t vd, vuint8m1_t vs1, vint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccus_4x8x4_i32m4_tu(vint32m4_t vd, vuint8m1_t vs1, vint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_i32m4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccus_4x8x4_i32m8_tu
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccus_4x8x4_i32m8_tu(vint32m8_t vd, vuint8m1_t vs1, vint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccus_4x8x4_i32m8_tu(vint32m8_t vd, vuint8m1_t vs1, vint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_i32m8_tu(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmacc_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmacc_4x8x4.c
index 531bc2b2b9425..7b3b25a203315 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmacc_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmacc_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmacc_4x8x4_i32m1_tu
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmacc_4x8x4_i32m1_tu(vint32m1_t vd, vint8m1_t vs1, vint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmacc_4x8x4_i32m1_tu(vint32m1_t vd, vint8m1_t vs1, vint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmacc_4x8x4_i32m2_tu
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmacc_4x8x4_i32m2_tu(vint32m2_t vd, vint8m1_t vs1, vint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmacc_4x8x4_i32m2_tu(vint32m2_t vd, vint8m1_t vs1, vint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmacc_4x8x4_i32m4_tu
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmacc_4x8x4_i32m4_tu(vint32m4_t vd, vint8m1_t vs1, vint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmacc_4x8x4_i32m4_tu(vint32m4_t vd, vint8m1_t vs1, vint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmacc_4x8x4_i32m8_tu
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmacc_4x8x4_i32m8_tu(vint32m8_t vd, vint8m1_t vs1, vint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmacc_4x8x4_i32m8_tu(vint32m8_t vd, vint8m1_t vs1, vint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmacc_4x8x4_tu(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccsu_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccsu_4x8x4.c
index 23bba523aaa44..2f3cbeec26fc9 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccsu_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccsu_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccsu_4x8x4_i32m1_tu
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccsu_4x8x4_i32m1_tu(vint32m1_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccsu_4x8x4_i32m1_tu(vint32m1_t vd, vint8m1_t vs1, vuint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccsu_4x8x4_i32m2_tu
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccsu_4x8x4_i32m2_tu(vint32m2_t vd, vint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccsu_4x8x4_i32m2_tu(vint32m2_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccsu_4x8x4_i32m4_tu
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccsu_4x8x4_i32m4_tu(vint32m4_t vd, vint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccsu_4x8x4_i32m4_tu(vint32m4_t vd, vint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccsu_4x8x4_i32m8_tu
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccsu_4x8x4_i32m8_tu(vint32m8_t vd, vint8m1_t vs1, vuint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccsu_4x8x4_i32m8_tu(vint32m8_t vd, vint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccsu_4x8x4_tu(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccu_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccu_4x8x4.c
index 950688c6c7851..1f2b2a1c86451 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccu_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccu_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccu_4x8x4_i32m1_tu
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccu_4x8x4_i32m1_tu(vint32m1_t vd, vuint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccu_4x8x4_i32m1_tu(vint32m1_t vd, vuint8m1_t vs1, vuint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccu_4x8x4_i32m2_tu
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccu_4x8x4_i32m2_tu(vint32m2_t vd, vuint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccu_4x8x4_i32m2_tu(vint32m2_t vd, vuint8m1_t vs1, vuint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccu_4x8x4_i32m4_tu
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccu_4x8x4_i32m4_tu(vint32m4_t vd, vuint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccu_4x8x4_i32m4_tu(vint32m4_t vd, vuint8m1_t vs1, vuint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccu_4x8x4_i32m8_tu
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccu_4x8x4_i32m8_tu(vint32m8_t vd, vuint8m1_t vs1, vuint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccu_4x8x4_i32m8_tu(vint32m8_t vd, vuint8m1_t vs1, vuint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccu_4x8x4_tu(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccus_4x8x4.c b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccus_4x8x4.c
index 7bdce95043ee4..923234fe8e2b6 100644
--- a/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccus_4x8x4.c
+++ b/clang/test/CodeGen/RISCV/rvv-intrinsics-autogenerated/policy/overloaded/sf_vqmaccus_4x8x4.c
@@ -7,41 +7,41 @@
 #include <sifive_vector.h>
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 2 x i32> @test_sf_vqmaccus_4x8x4_i32m1_tu
-// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-SAME: (<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv4i8.i64(<vscale x 2 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 4 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 2 x i32> [[TMP0]]
 //
-vint32m1_t test_sf_vqmaccus_4x8x4_i32m1_tu(vint32m1_t vd, vuint8m1_t vs1, vint8m1_t vs2, size_t vl) {
+vint32m1_t test_sf_vqmaccus_4x8x4_i32m1_tu(vint32m1_t vd, vuint8m1_t vs1, vint8mf2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 4 x i32> @test_sf_vqmaccus_4x8x4_i32m2_tu
-// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 4 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv16i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv8i8.i64(<vscale x 4 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 8 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 4 x i32> [[TMP0]]
 //
-vint32m2_t test_sf_vqmaccus_4x8x4_i32m2_tu(vint32m2_t vd, vuint8m1_t vs1, vint8m2_t vs2, size_t vl) {
+vint32m2_t test_sf_vqmaccus_4x8x4_i32m2_tu(vint32m2_t vd, vuint8m1_t vs1, vint8m1_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 8 x i32> @test_sf_vqmaccus_4x8x4_i32m4_tu
-// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 8 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv32i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv16i8.i64(<vscale x 8 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 16 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x i32> [[TMP0]]
 //
-vint32m4_t test_sf_vqmaccus_4x8x4_i32m4_tu(vint32m4_t vd, vuint8m1_t vs1, vint8m4_t vs2, size_t vl) {
+vint32m4_t test_sf_vqmaccus_4x8x4_i32m4_tu(vint32m4_t vd, vuint8m1_t vs1, vint8m2_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_tu(vd, vs1, vs2, vl);
 }
 
 // CHECK-RV64-LABEL: define dso_local <vscale x 16 x i32> @test_sf_vqmaccus_4x8x4_i32m8_tu
-// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
+// CHECK-RV64-SAME: (<vscale x 16 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0]] {
 // CHECK-RV64-NEXT:  entry:
-// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv64i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 64 x i8> [[VS2]], i64 [[VL]], i64 2)
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv32i8.i64(<vscale x 16 x i32> [[VD]], <vscale x 8 x i8> [[VS1]], <vscale x 32 x i8> [[VS2]], i64 [[VL]], i64 2)
 // CHECK-RV64-NEXT:    ret <vscale x 16 x i32> [[TMP0]]
 //
-vint32m8_t test_sf_vqmaccus_4x8x4_i32m8_tu(vint32m8_t vd, vuint8m1_t vs1, vint8m8_t vs2, size_t vl) {
+vint32m8_t test_sf_vqmaccus_4x8x4_i32m8_tu(vint32m8_t vd, vuint8m1_t vs1, vint8m4_t vs2, size_t vl) {
   return __riscv_sf_vqmaccus_4x8x4_tu(vd, vs1, vs2, vl);
 }
diff --git a/clang/test/Sema/rvv-required-features.c b/clang/test/Sema/rvv-required-features.c
index 2714ef04b9bf2..5846f338aa801 100644
--- a/clang/test/Sema/rvv-required-features.c
+++ b/clang/test/Sema/rvv-required-features.c
@@ -1,8 +1,6 @@
 // REQUIRES: riscv-registered-target
 // RUN: %clang_cc1 -triple riscv64 -target-feature +v -target-feature +xsfvcp \
-// RUN:     -target-feature +xsfvqmaccdod -target-feature +xsfvqmaccqoq \
-// RUN:     -target-feature +experimental-zvfbfmin -target-feature +xsfvfwmaccqqq \
-// RUN:     -target-feature +xsfvfnrclipxfqf %s -fsyntax-only -verify
+// RUN:     -target-feature +xsfvqmaccdod -target-feature +xsfvqmaccqoq %s -fsyntax-only -verify
 
 // expected-no-diagnostics
 
@@ -25,18 +23,6 @@ void test_xsfvqmaccdod(vint32m1_t vd, vint8m1_t vs1, vint8m1_t vs2, size_t vl) {
   __riscv_sf_vqmacc_2x8x2(vd, vs1, vs2, vl);
 }
 
-void test_xsfvqmaccqoq(vint32m1_t vd, vint8m1_t vs1, vint8m1_t vs2, size_t vl) {
+void test_xsfvqmaccqoq(vint32m1_t vd, vint8m1_t vs1, vint8mf2_t vs2, size_t vl) {
   __riscv_sf_vqmacc_4x8x4(vd, vs1, vs2, vl);
 }
-
-void test_xsfvfwmaccqqq(vfloat32m1_t vd, vbfloat16m1_t vs1, vbfloat16mf2_t vs2, size_t vl) {
-  __riscv_sf_vfwmacc_4x4x4(vd, vs1, vs2, vl);
-}
-
-void test_xsfvfnrclipxufqf(vfloat32m1_t vs1, float rs2, size_t vl) {
-  __riscv_sf_vfnrclip_xu_f_qf(vs1, rs2, vl);
-}
-
-void test_xsfvfnrclipxfqf(vfloat32m1_t vs1, float rs2, size_t vl) {
-  __riscv_sf_vfnrclip_x_f_qf(vs1, rs2, vl);
-}
diff --git a/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td b/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
index a16fa7e769929..572eed0de0d29 100644
--- a/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
+++ b/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
@@ -349,16 +349,23 @@ multiclass VPseudoSiFiveVMACC<string mx, VReg vd_type, VReg vs2_type,
       : VPseudoTernaryNoMaskWithPolicy<vd_type, V_M1.vrclass, vs2_type, Constraint>;
 }
 
-multiclass VPseudoSiFiveVQMACC<string Constraint = ""> {
+multiclass VPseudoSiFiveVQMACCDOD<string Constraint = ""> {
   foreach m = MxListVF8 in
     let VLMul = m.value in
     defm NAME : VPseudoSiFiveVMACC<m.MX, m.vrclass, m.vrclass, Constraint>;
 }
 
-multiclass VPseudoSiFiveVFWMACC<string Constraint = ""> {
-  foreach m = MxListFW in
+multiclass VPseudoSiFiveVQMACCQOQ<string Constraint = ""> {
+  foreach m = MxListVF8 in
     let VLMul = m.value in
-    defm NAME : VPseudoSiFiveVMACC<m.MX, m.wvrclass, m.vrclass, Constraint>;
+    defm NAME : VPseudoSiFiveVMACC<m.MX, m.vrclass, m.f2vrclass, Constraint>;
+}
+
+multiclass VPseudoSiFiveVFWMACC<string Constraint = ""> {
+  foreach i = [0, 1, 2, 3, 4] in
+    let VLMul = MxListVF4[i].value in
+    defm NAME : VPseudoSiFiveVMACC<MxListVF4[i].MX, MxListFW[i].wvrclass,
+                                   MxListFW[i].vrclass, Constraint>;
 }
 
 multiclass VPseudoSiFiveVFNRCLIP<string Constraint = "@earlyclobber $rd"> {
@@ -400,17 +407,17 @@ let Predicates = [HasVendorXSfvcp] in {
 }
 
 let Predicates = [HasVendorXSfvqmaccdod] in {
-  defm VQMACCU_2x8x2  : VPseudoSiFiveVQMACC;
-  defm VQMACC_2x8x2   : VPseudoSiFiveVQMACC;
-  defm VQMACCUS_2x8x2 : VPseudoSiFiveVQMACC;
-  defm VQMACCSU_2x8x2 : VPseudoSiFiveVQMACC;
+  defm VQMACCU_2x8x2  : VPseudoSiFiveVQMACCDOD;
+  defm VQMACC_2x8x2   : VPseudoSiFiveVQMACCDOD;
+  defm VQMACCUS_2x8x2 : VPseudoSiFiveVQMACCDOD;
+  defm VQMACCSU_2x8x2 : VPseudoSiFiveVQMACCDOD;
 }
 
 let Predicates = [HasVendorXSfvqmaccqoq] in {
-  defm VQMACCU_4x8x4  : VPseudoSiFiveVQMACC;
-  defm VQMACC_4x8x4   : VPseudoSiFiveVQMACC;
-  defm VQMACCUS_4x8x4 : VPseudoSiFiveVQMACC;
-  defm VQMACCSU_4x8x4 : VPseudoSiFiveVQMACC;
+  defm VQMACCU_4x8x4  : VPseudoSiFiveVQMACCQOQ;
+  defm VQMACC_4x8x4   : VPseudoSiFiveVQMACCQOQ;
+  defm VQMACCUS_4x8x4 : VPseudoSiFiveVQMACCQOQ;
+  defm VQMACCSU_4x8x4 : VPseudoSiFiveVQMACCQOQ;
 }
 
 let Predicates = [HasVendorXSfvfwmaccqqq] in {
@@ -561,21 +568,30 @@ multiclass VPatVMACC<string intrinsic, string instruction, string kind,
     def : VPatTernaryNoMaskWithPolicy<"int_riscv_sf_" # intrinsic,
                                       "Pseudo" # instruction, kind, VdInfo.Vector,
                                       vec_m1, Vs2Info.Vector,
-                                      Vs2Info.Log2SEW, Vs2Info.LMul,
+                                      Vs2Info.Log2SEW, VdInfo.LMul,
                                       VdInfo.RegClass, VR, Vs2Info.RegClass>;
   }
 }
 
-defset list<VTypeInfoToWide> VQMACCInfoPairs = {
+defset list<VTypeInfoToWide> VQMACCDODInfoPairs = {
   def : VTypeInfoToWide<VI8M1, VI32M1>;
   def : VTypeInfoToWide<VI8M2, VI32M2>;
   def : VTypeInfoToWide<VI8M4, VI32M4>;
   def : VTypeInfoToWide<VI8M8, VI32M8>;
 }
 
-multiclass VPatVQMACC<string intrinsic, string instruction, string kind>
-    : VPatVMACC<intrinsic, instruction, kind, VQMACCInfoPairs, vint8m1_t>;
+defset list<VTypeInfoToWide> VQMACCQOQInfoPairs = {
+   def : VTypeInfoToWide<VI8MF2, VI32M1>;
+   def : VTypeInfoToWide<VI8M1, VI32M2>;
+   def : VTypeInfoToWide<VI8M2, VI32M4>;
+   def : VTypeInfoToWide<VI8M4, VI32M8>;
+}
+
+multiclass VPatVQMACCDOD<string intrinsic, string instruction, string kind>
+    : VPatVMACC<intrinsic, instruction, kind, VQMACCDODInfoPairs, vint8m1_t>;
 
+multiclass VPatVQMACCQOQ<string intrinsic, string instruction, string kind>
+    : VPatVMACC<intrinsic, instruction, kind, VQMACCQOQInfoPairs, vint8m1_t>;
 
 multiclass VPatVFWMACC<string intrinsic, string instruction, string kind>
     : VPatVMACC<intrinsic, instruction, kind, AllWidenableBFloatToFloatVectors,
@@ -637,17 +653,17 @@ let Predicates = [HasVendorXSfvcp] in {
 }
 
 let Predicates = [HasVendorXSfvqmaccdod] in {
-  defm : VPatVQMACC<"vqmaccu_2x8x2", "VQMACCU", "2x8x2">;
-  defm : VPatVQMACC<"vqmacc_2x8x2", "VQMACC", "2x8x2">;
-  defm : VPatVQMACC<"vqmaccus_2x8x2", "VQMACCUS", "2x8x2">;
-  defm : VPatVQMACC<"vqmaccsu_2x8x2", "VQMACCSU", "2x8x2">;
+  defm : VPatVQMACCDOD<"vqmaccu_2x8x2", "VQMACCU", "2x8x2">;
+  defm : VPatVQMACCDOD<"vqmacc_2x8x2", "VQMACC", "2x8x2">;
+  defm : VPatVQMACCDOD<"vqmaccus_2x8x2", "VQMACCUS", "2x8x2">;
+  defm : VPatVQMACCDOD<"vqmaccsu_2x8x2", "VQMACCSU", "2x8x2">;
 }
 
 let Predicates = [HasVendorXSfvqmaccqoq] in {
-  defm : VPatVQMACC<"vqmaccu_4x8x4", "VQMACCU", "4x8x4">;
-  defm : VPatVQMACC<"vqmacc_4x8x4", "VQMACC", "4x8x4">;
-  defm : VPatVQMACC<"vqmaccus_4x8x4", "VQMACCUS", "4x8x4">;
-  defm : VPatVQMACC<"vqmaccsu_4x8x4", "VQMACCSU", "4x8x4">;
+  defm : VPatVQMACCQOQ<"vqmaccu_4x8x4", "VQMACCU", "4x8x4">;
+  defm : VPatVQMACCQOQ<"vqmacc_4x8x4", "VQMACC", "4x8x4">;
+  defm : VPatVQMACCQOQ<"vqmaccus_4x8x4", "VQMACCUS", "4x8x4">;
+  defm : VPatVQMACCQOQ<"vqmaccsu_4x8x4", "VQMACCSU", "4x8x4">;
 }
 
 let Predicates = [HasVendorXSfvfwmaccqqq] in {
diff --git a/llvm/test/CodeGen/RISCV/rvv/sf_vfwmacc_4x4x4.ll b/llvm/test/CodeGen/RISCV/rvv/sf_vfwmacc_4x4x4.ll
index e0da3e846759f..3417beee4ef6d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sf_vfwmacc_4x4x4.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sf_vfwmacc_4x4x4.ll
@@ -13,7 +13,7 @@ declare <vscale x 1 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv1f32.nxv4bf16.nxv1b
 define <vscale x 1 x float> @intrinsic_vfwmacc_4x4x4_tu_f32mf2(<vscale x 1 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 1 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32mf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, tu, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -29,7 +29,7 @@ entry:
 define <vscale x 1 x float> @intrinsic_vfwmacc_4x4x4_ta_f32mf2(<vscale x 1 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 1 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32mf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -51,7 +51,7 @@ declare <vscale x 2 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv2f32.nxv4bf16.nxv2b
 define <vscale x 2 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m1(<vscale x 2 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 2 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, tu, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -67,7 +67,7 @@ entry:
 define <vscale x 2 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m1(<vscale x 2 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 2 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -89,7 +89,7 @@ declare <vscale x 4 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv4f32.nxv4bf16.nxv4b
 define <vscale x 4 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m2(<vscale x 4 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 4 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, tu, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -105,7 +105,7 @@ entry:
 define <vscale x 4 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m2(<vscale x 4 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 4 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -127,7 +127,7 @@ declare <vscale x 8 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv8f32.nxv4bf16.nxv8b
 define <vscale x 8 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m4(<vscale x 8 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 8 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, tu, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -143,7 +143,7 @@ entry:
 define <vscale x 8 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m4(<vscale x 8 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 8 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -165,7 +165,7 @@ declare <vscale x 16 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv16f32.nxv4bf16.nxv
 define <vscale x 16 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m8(<vscale x 16 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 16 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m8, tu, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
@@ -181,7 +181,7 @@ entry:
 define <vscale x 16 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m8(<vscale x 16 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 16 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/sf_vqmacc_4x8x4.ll b/llvm/test/CodeGen/RISCV/rvv/sf_vqmacc_4x8x4.ll
index 2d591be2adc21..3c86acc7e233c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sf_vqmacc_4x8x4.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sf_vqmacc_4x8x4.ll
@@ -7,10 +7,10 @@
 declare <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv8i8(
   <vscale x 2 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 8 x i8>,
+  <vscale x 4 x i8>,
   iXLen, iXLen);
 
-define <vscale x 2 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
+define <vscale x 2 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_tu_i32m1:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
@@ -20,13 +20,13 @@ entry:
   %a = call <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv8i8(
     <vscale x 2 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 8 x i8> %2,
+    <vscale x 4 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 2 x i32> %a
 }
 
-define <vscale x 2 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
+define <vscale x 2 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_ta_i32m1:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
@@ -36,7 +36,7 @@ entry:
   %a = call <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv8i8(
     <vscale x 2 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 8 x i8> %2,
+    <vscale x 4 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 2 x i32> %a
@@ -45,36 +45,36 @@ entry:
 declare <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv16i8(
   <vscale x 4 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 16 x i8>,
+  <vscale x 8 x i8>,
   iXLen, iXLen);
 
-define <vscale x 4 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
+define <vscale x 4 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_tu_i32m2:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
-; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v10, v12
+; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv16i8(
     <vscale x 4 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 16 x i8> %2,
+    <vscale x 8 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 4 x i32> %a
 }
 
-define <vscale x 4 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
+define <vscale x 4 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_ta_i32m2:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
-; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v10, v12
+; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv16i8(
     <vscale x 4 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 16 x i8> %2,
+    <vscale x 8 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 4 x i32> %a
@@ -83,36 +83,36 @@ entry:
 declare <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv32i8(
   <vscale x 8 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 32 x i8>,
+  <vscale x 16 x i8>,
   iXLen, iXLen);
 
-define <vscale x 8 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
+define <vscale x 8 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_tu_i32m4:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
-; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v12, v16
+; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv32i8(
     <vscale x 8 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 32 x i8> %2,
+    <vscale x 16 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 8 x i32> %a
 }
 
-define <vscale x 8 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
+define <vscale x 8 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_ta_i32m4:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v12, v16
+; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv32i8(
     <vscale x 8 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 32 x i8> %2,
+    <vscale x 16 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 8 x i32> %a
@@ -121,38 +121,36 @@ entry:
 declare <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv64i8(
   <vscale x 16 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 64 x i8>,
+  <vscale x 32 x i8>,
   iXLen, iXLen);
 
-define <vscale x 16 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 64 x i8> %2, iXLen %3) nounwind {
+define <vscale x 16 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_tu_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vl8r.v v24, (a0)
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, tu, ma
-; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v16, v24
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, tu, ma
+; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv64i8(
     <vscale x 16 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 64 x i8> %2,
+    <vscale x 32 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 16 x i32> %a
 }
 
-define <vscale x 16 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 64 x i8> %2, iXLen %3) nounwind {
+define <vscale x 16 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_ta_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vl8r.v v24, (a0)
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
-; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v16, v24
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
+; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv64i8(
     <vscale x 16 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 64 x i8> %2,
+    <vscale x 32 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 16 x i32> %a
diff --git a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccsu_4x8x4.ll b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccsu_4x8x4.ll
index bfdab33965c13..02d7a377446f0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccsu_4x8x4.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccsu_4x8x4.ll
@@ -7,10 +7,10 @@
 declare <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv8i8(
   <vscale x 2 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 8 x i8>,
+  <vscale x 4 x i8>,
   iXLen, iXLen);
 
-define <vscale x 2 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
+define <vscale x 2 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_tu_i32m1:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
@@ -20,13 +20,13 @@ entry:
   %a = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv8i8(
     <vscale x 2 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 8 x i8> %2,
+    <vscale x 4 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 2 x i32> %a
 }
 
-define <vscale x 2 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
+define <vscale x 2 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_ta_i32m1:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
@@ -36,7 +36,7 @@ entry:
   %a = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv8i8(
     <vscale x 2 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 8 x i8> %2,
+    <vscale x 4 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 2 x i32> %a
@@ -45,36 +45,36 @@ entry:
 declare <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv16i8(
   <vscale x 4 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 16 x i8>,
+  <vscale x 8 x i8>,
   iXLen, iXLen);
 
-define <vscale x 4 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
+define <vscale x 4 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_tu_i32m2:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
-; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v10, v12
+; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv16i8(
     <vscale x 4 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 16 x i8> %2,
+    <vscale x 8 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 4 x i32> %a
 }
 
-define <vscale x 4 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
+define <vscale x 4 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_ta_i32m2:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
-; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v10, v12
+; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv16i8(
     <vscale x 4 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 16 x i8> %2,
+    <vscale x 8 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 4 x i32> %a
@@ -83,36 +83,36 @@ entry:
 declare <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv32i8(
   <vscale x 8 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 32 x i8>,
+  <vscale x 16 x i8>,
   iXLen, iXLen);
 
-define <vscale x 8 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
+define <vscale x 8 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_tu_i32m4:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
-; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v12, v16
+; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv32i8(
     <vscale x 8 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 32 x i8> %2,
+    <vscale x 16 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 8 x i32> %a
 }
 
-define <vscale x 8 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
+define <vscale x 8 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_ta_i32m4:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v12, v16
+; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv32i8(
     <vscale x 8 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 32 x i8> %2,
+    <vscale x 16 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 8 x i32> %a
@@ -121,38 +121,36 @@ entry:
 declare <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv64i8(
   <vscale x 16 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 64 x i8>,
+  <vscale x 32 x i8>,
   iXLen, iXLen);
 
-define <vscale x 16 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 64 x i8> %2, iXLen %3) nounwind {
+define <vscale x 16 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_tu_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vl8r.v v24, (a0)
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, tu, ma
-; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v16, v24
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, tu, ma
+; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv64i8(
     <vscale x 16 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 64 x i8> %2,
+    <vscale x 32 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 16 x i32> %a
 }
 
-define <vscale x 16 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 64 x i8> %2, iXLen %3) nounwind {
+define <vscale x 16 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_ta_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vl8r.v v24, (a0)
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
-; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v16, v24
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
+; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv64i8(
     <vscale x 16 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 64 x i8> %2,
+    <vscale x 32 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 16 x i32> %a
diff --git a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccu_4x8x4.ll b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccu_4x8x4.ll
index d1565fb9a634f..51f66cd4774a0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccu_4x8x4.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccu_4x8x4.ll
@@ -7,10 +7,10 @@
 declare <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv8i8(
   <vscale x 2 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 8 x i8>,
+  <vscale x 4 x i8>,
   iXLen, iXLen);
 
-define <vscale x 2 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
+define <vscale x 2 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_tu_i32m1:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
@@ -20,13 +20,13 @@ entry:
   %a = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv8i8(
     <vscale x 2 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 8 x i8> %2,
+    <vscale x 4 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 2 x i32> %a
 }
 
-define <vscale x 2 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
+define <vscale x 2 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_ta_i32m1:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
@@ -36,7 +36,7 @@ entry:
   %a = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv8i8(
     <vscale x 2 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 8 x i8> %2,
+    <vscale x 4 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 2 x i32> %a
@@ -45,36 +45,36 @@ entry:
 declare <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv16i8(
   <vscale x 4 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 16 x i8>,
+  <vscale x 8 x i8>,
   iXLen, iXLen);
 
-define <vscale x 4 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
+define <vscale x 4 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_tu_i32m2:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
-; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v10, v12
+; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv16i8(
     <vscale x 4 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 16 x i8> %2,
+    <vscale x 8 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 4 x i32> %a
 }
 
-define <vscale x 4 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
+define <vscale x 4 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_ta_i32m2:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
-; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v10, v12
+; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv16i8(
     <vscale x 4 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 16 x i8> %2,
+    <vscale x 8 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 4 x i32> %a
@@ -83,36 +83,36 @@ entry:
 declare <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv32i8(
   <vscale x 8 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 32 x i8>,
+  <vscale x 16 x i8>,
   iXLen, iXLen);
 
-define <vscale x 8 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
+define <vscale x 8 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_tu_i32m4:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
-; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v12, v16
+; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv32i8(
     <vscale x 8 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 32 x i8> %2,
+    <vscale x 16 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 8 x i32> %a
 }
 
-define <vscale x 8 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
+define <vscale x 8 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_ta_i32m4:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v12, v16
+; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv32i8(
     <vscale x 8 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 32 x i8> %2,
+    <vscale x 16 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 8 x i32> %a
@@ -121,38 +121,36 @@ entry:
 declare <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv64i8(
   <vscale x 16 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 64 x i8>,
+  <vscale x 32 x i8>,
   iXLen, iXLen);
 
-define <vscale x 16 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 64 x i8> %2, iXLen %3) nounwind {
+define <vscale x 16 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_tu_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vl8r.v v24, (a0)
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, tu, ma
-; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v16, v24
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, tu, ma
+; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv64i8(
     <vscale x 16 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 64 x i8> %2,
+    <vscale x 32 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 16 x i32> %a
 }
 
-define <vscale x 16 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 64 x i8> %2, iXLen %3) nounwind {
+define <vscale x 16 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_ta_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vl8r.v v24, (a0)
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
-; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v16, v24
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
+; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv64i8(
     <vscale x 16 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 64 x i8> %2,
+    <vscale x 32 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 16 x i32> %a
diff --git a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccus_4x8x4.ll b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccus_4x8x4.ll
index c6d2a048c5cbc..4d6ef90a6c096 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccus_4x8x4.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccus_4x8x4.ll
@@ -7,10 +7,10 @@
 declare <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv8i8(
   <vscale x 2 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 8 x i8>,
+  <vscale x 4 x i8>,
   iXLen, iXLen);
 
-define <vscale x 2 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
+define <vscale x 2 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_tu_i32m1:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
@@ -20,13 +20,13 @@ entry:
   %a = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv8i8(
     <vscale x 2 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 8 x i8> %2,
+    <vscale x 4 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 2 x i32> %a
 }
 
-define <vscale x 2 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
+define <vscale x 2 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_ta_i32m1:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
@@ -36,7 +36,7 @@ entry:
   %a = call <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv8i8(
     <vscale x 2 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 8 x i8> %2,
+    <vscale x 4 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 2 x i32> %a
@@ -45,36 +45,36 @@ entry:
 declare <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv16i8(
   <vscale x 4 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 16 x i8>,
+  <vscale x 8 x i8>,
   iXLen, iXLen);
 
-define <vscale x 4 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
+define <vscale x 4 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_tu_i32m2:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
-; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v10, v12
+; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv16i8(
     <vscale x 4 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 16 x i8> %2,
+    <vscale x 8 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 4 x i32> %a
 }
 
-define <vscale x 4 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
+define <vscale x 4 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_ta_i32m2:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
-; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v10, v12
+; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv16i8(
     <vscale x 4 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 16 x i8> %2,
+    <vscale x 8 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 4 x i32> %a
@@ -83,36 +83,36 @@ entry:
 declare <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv32i8(
   <vscale x 8 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 32 x i8>,
+  <vscale x 16 x i8>,
   iXLen, iXLen);
 
-define <vscale x 8 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
+define <vscale x 8 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_tu_i32m4:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
-; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v12, v16
+; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv32i8(
     <vscale x 8 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 32 x i8> %2,
+    <vscale x 16 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 8 x i32> %a
 }
 
-define <vscale x 8 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
+define <vscale x 8 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_ta_i32m4:
 ; CHECK:       # %bb.0: # %entry
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
-; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v12, v16
+; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv32i8(
     <vscale x 8 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 32 x i8> %2,
+    <vscale x 16 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 8 x i32> %a
@@ -121,38 +121,36 @@ entry:
 declare <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv64i8(
   <vscale x 16 x i32>,
   <vscale x 8 x i8>,
-  <vscale x 64 x i8>,
+  <vscale x 32 x i8>,
   iXLen, iXLen);
 
-define <vscale x 16 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 64 x i8> %2, iXLen %3) nounwind {
+define <vscale x 16 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_tu_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vl8r.v v24, (a0)
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, tu, ma
-; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v16, v24
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, tu, ma
+; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv64i8(
     <vscale x 16 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 64 x i8> %2,
+    <vscale x 32 x i8> %2,
     iXLen %3, iXLen 2)
 
   ret <vscale x 16 x i32> %a
 }
 
-define <vscale x 16 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 64 x i8> %2, iXLen %3) nounwind {
+define <vscale x 16 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_ta_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vl8r.v v24, (a0)
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
-; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v16, v24
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
+; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
   %a = call <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv64i8(
     <vscale x 16 x i32> %0,
     <vscale x 8 x i8> %1,
-    <vscale x 64 x i8> %2,
+    <vscale x 32 x i8> %2,
     iXLen %3, iXLen 3)
 
   ret <vscale x 16 x i32> %a

From 07cd8ad7a74df01cd028fe9ff74f757d737c064e Mon Sep 17 00:00:00 2001
From: Brandon Wu <brandon.wu@sifive.com>
Date: Wed, 20 Dec 2023 21:29:36 -0800
Subject: [PATCH 2/4] fixup! Recommit [RISCV] Update the interface of sifive
 vqmaccqoq (#74284)

---
 llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td    |  9 +++++----
 .../CodeGen/RISCV/rvv/sf_vfwmacc_4x4x4.ll     | 20 +++++++++----------
 .../test/CodeGen/RISCV/rvv/sf_vqmacc_4x8x4.ll | 16 +++++++--------
 .../CodeGen/RISCV/rvv/sf_vqmaccsu_4x8x4.ll    | 16 +++++++--------
 .../CodeGen/RISCV/rvv/sf_vqmaccu_4x8x4.ll     | 16 +++++++--------
 .../CodeGen/RISCV/rvv/sf_vqmaccus_4x8x4.ll    | 16 +++++++--------
 6 files changed, 47 insertions(+), 46 deletions(-)

diff --git a/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td b/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
index 572eed0de0d29..6472875908db0 100644
--- a/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
+++ b/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
@@ -356,14 +356,15 @@ multiclass VPseudoSiFiveVQMACCDOD<string Constraint = ""> {
 }
 
 multiclass VPseudoSiFiveVQMACCQOQ<string Constraint = ""> {
-  foreach m = MxListVF8 in
-    let VLMul = m.value in
-    defm NAME : VPseudoSiFiveVMACC<m.MX, m.vrclass, m.f2vrclass, Constraint>;
+  foreach i = [0, 1, 2, 3] in
+    let VLMul = MxListVF4[i].value in
+    defm NAME : VPseudoSiFiveVMACC<MxListVF8[i].MX, MxListVF8[i].vrclass,
+                                   MxListVF4[i].vrclass, Constraint>;
 }
 
 multiclass VPseudoSiFiveVFWMACC<string Constraint = ""> {
   foreach i = [0, 1, 2, 3, 4] in
-    let VLMul = MxListVF4[i].value in
+    let VLMul = MxListVF2[i].value in
     defm NAME : VPseudoSiFiveVMACC<MxListVF4[i].MX, MxListFW[i].wvrclass,
                                    MxListFW[i].vrclass, Constraint>;
 }
diff --git a/llvm/test/CodeGen/RISCV/rvv/sf_vfwmacc_4x4x4.ll b/llvm/test/CodeGen/RISCV/rvv/sf_vfwmacc_4x4x4.ll
index 3417beee4ef6d..e0da3e846759f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sf_vfwmacc_4x4x4.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sf_vfwmacc_4x4x4.ll
@@ -13,7 +13,7 @@ declare <vscale x 1 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv1f32.nxv4bf16.nxv1b
 define <vscale x 1 x float> @intrinsic_vfwmacc_4x4x4_tu_f32mf2(<vscale x 1 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 1 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32mf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, tu, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -29,7 +29,7 @@ entry:
 define <vscale x 1 x float> @intrinsic_vfwmacc_4x4x4_ta_f32mf2(<vscale x 1 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 1 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32mf2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -51,7 +51,7 @@ declare <vscale x 2 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv2f32.nxv4bf16.nxv2b
 define <vscale x 2 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m1(<vscale x 2 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 2 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, tu, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -67,7 +67,7 @@ entry:
 define <vscale x 2 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m1(<vscale x 2 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 2 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -89,7 +89,7 @@ declare <vscale x 4 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv4f32.nxv4bf16.nxv4b
 define <vscale x 4 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m2(<vscale x 4 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 4 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, tu, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -105,7 +105,7 @@ entry:
 define <vscale x 4 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m2(<vscale x 4 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 4 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -127,7 +127,7 @@ declare <vscale x 8 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv8f32.nxv4bf16.nxv8b
 define <vscale x 8 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m4(<vscale x 8 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 8 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, tu, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -143,7 +143,7 @@ entry:
 define <vscale x 8 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m4(<vscale x 8 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 8 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -165,7 +165,7 @@ declare <vscale x 16 x float> @llvm.riscv.sf.vfwmacc.4x4x4.nxv16f32.nxv4bf16.nxv
 define <vscale x 16 x float> @intrinsic_vfwmacc_4x4x4_tu_f32m8(<vscale x 16 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 16 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_tu_f32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m8, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, tu, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
@@ -181,7 +181,7 @@ entry:
 define <vscale x 16 x float> @intrinsic_vfwmacc_4x4x4_ta_f32m8(<vscale x 16 x float> %0, <vscale x 4 x bfloat> %1, <vscale x 16 x bfloat> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vfwmacc_4x4x4_ta_f32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
 ; CHECK-NEXT:    sf.vfwmacc.4x4x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/sf_vqmacc_4x8x4.ll b/llvm/test/CodeGen/RISCV/rvv/sf_vqmacc_4x8x4.ll
index 3c86acc7e233c..eebc51619480b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sf_vqmacc_4x8x4.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sf_vqmacc_4x8x4.ll
@@ -13,7 +13,7 @@ declare <vscale x 2 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv2i32.nxv8i8.nxv8i8(
 define <vscale x 2 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_tu_i32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, tu, ma
 ; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -29,7 +29,7 @@ entry:
 define <vscale x 2 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_ta_i32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
 ; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -51,7 +51,7 @@ declare <vscale x 4 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv4i32.nxv8i8.nxv16i8(
 define <vscale x 4 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_tu_i32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
 ; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -67,7 +67,7 @@ entry:
 define <vscale x 4 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_ta_i32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
 ; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -89,7 +89,7 @@ declare <vscale x 8 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv8i32.nxv8i8.nxv32i8(
 define <vscale x 8 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_tu_i32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
 ; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -105,7 +105,7 @@ entry:
 define <vscale x 8 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_ta_i32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
 ; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -127,7 +127,7 @@ declare <vscale x 16 x i32> @llvm.riscv.sf.vqmacc.4x8x4.nxv16i32.nxv8i8.nxv64i8(
 define <vscale x 16 x i32> @intrinsic_vqmacc_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_tu_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
 ; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
@@ -143,7 +143,7 @@ entry:
 define <vscale x 16 x i32> @intrinsic_vqmacc_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmacc_4x8x4_ta_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
 ; CHECK-NEXT:    sf.vqmacc.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccsu_4x8x4.ll b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccsu_4x8x4.ll
index 02d7a377446f0..0d7052356e558 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccsu_4x8x4.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccsu_4x8x4.ll
@@ -13,7 +13,7 @@ declare <vscale x 2 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv2i32.nxv8i8.nxv8i8(
 define <vscale x 2 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_tu_i32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, tu, ma
 ; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -29,7 +29,7 @@ entry:
 define <vscale x 2 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_ta_i32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
 ; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -51,7 +51,7 @@ declare <vscale x 4 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv4i32.nxv8i8.nxv16i8(
 define <vscale x 4 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_tu_i32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
 ; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -67,7 +67,7 @@ entry:
 define <vscale x 4 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_ta_i32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
 ; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -89,7 +89,7 @@ declare <vscale x 8 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv8i32.nxv8i8.nxv32i8(
 define <vscale x 8 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_tu_i32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
 ; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -105,7 +105,7 @@ entry:
 define <vscale x 8 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_ta_i32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
 ; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -127,7 +127,7 @@ declare <vscale x 16 x i32> @llvm.riscv.sf.vqmaccsu.4x8x4.nxv16i32.nxv8i8.nxv64i
 define <vscale x 16 x i32> @intrinsic_vqmaccsu_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_tu_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
 ; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
@@ -143,7 +143,7 @@ entry:
 define <vscale x 16 x i32> @intrinsic_vqmaccsu_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccsu_4x8x4_ta_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
 ; CHECK-NEXT:    sf.vqmaccsu.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccu_4x8x4.ll b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccu_4x8x4.ll
index 51f66cd4774a0..3332390f71e01 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccu_4x8x4.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccu_4x8x4.ll
@@ -13,7 +13,7 @@ declare <vscale x 2 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv2i32.nxv8i8.nxv8i8(
 define <vscale x 2 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_tu_i32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, tu, ma
 ; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -29,7 +29,7 @@ entry:
 define <vscale x 2 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_ta_i32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
 ; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -51,7 +51,7 @@ declare <vscale x 4 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv4i32.nxv8i8.nxv16i8(
 define <vscale x 4 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_tu_i32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
 ; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -67,7 +67,7 @@ entry:
 define <vscale x 4 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_ta_i32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
 ; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -89,7 +89,7 @@ declare <vscale x 8 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv8i32.nxv8i8.nxv32i8(
 define <vscale x 8 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_tu_i32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
 ; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -105,7 +105,7 @@ entry:
 define <vscale x 8 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_ta_i32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
 ; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -127,7 +127,7 @@ declare <vscale x 16 x i32> @llvm.riscv.sf.vqmaccu.4x8x4.nxv16i32.nxv8i8.nxv64i8
 define <vscale x 16 x i32> @intrinsic_vqmaccu_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_tu_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
 ; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
@@ -143,7 +143,7 @@ entry:
 define <vscale x 16 x i32> @intrinsic_vqmaccu_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccu_4x8x4_ta_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
 ; CHECK-NEXT:    sf.vqmaccu.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccus_4x8x4.ll b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccus_4x8x4.ll
index 4d6ef90a6c096..74fb66f5bf351 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccus_4x8x4.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sf_vqmaccus_4x8x4.ll
@@ -13,7 +13,7 @@ declare <vscale x 2 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv2i32.nxv8i8.nxv8i8(
 define <vscale x 2 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_tu_i32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, tu, ma
 ; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -29,7 +29,7 @@ entry:
 define <vscale x 2 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m1(<vscale x 2 x i32> %0, <vscale x 8 x i8> %1, <vscale x 4 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_ta_i32m1:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
 ; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v9, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -51,7 +51,7 @@ declare <vscale x 4 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv4i32.nxv8i8.nxv16i8(
 define <vscale x 4 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_tu_i32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, tu, ma
 ; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -67,7 +67,7 @@ entry:
 define <vscale x 4 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m2(<vscale x 4 x i32> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_ta_i32m2:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
 ; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v10, v11
 ; CHECK-NEXT:    ret
 entry:
@@ -89,7 +89,7 @@ declare <vscale x 8 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv8i32.nxv8i8.nxv32i8(
 define <vscale x 8 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_tu_i32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, tu, ma
 ; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -105,7 +105,7 @@ entry:
 define <vscale x 8 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m4(<vscale x 8 x i32> %0, <vscale x 8 x i8> %1, <vscale x 16 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_ta_i32m4:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
 ; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v12, v14
 ; CHECK-NEXT:    ret
 entry:
@@ -127,7 +127,7 @@ declare <vscale x 16 x i32> @llvm.riscv.sf.vqmaccus.4x8x4.nxv16i32.nxv8i8.nxv64i
 define <vscale x 16 x i32> @intrinsic_vqmaccus_4x8x4_tu_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_tu_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, tu, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, tu, ma
 ; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:
@@ -143,7 +143,7 @@ entry:
 define <vscale x 16 x i32> @intrinsic_vqmaccus_4x8x4_ta_i32m8(<vscale x 16 x i32> %0, <vscale x 8 x i8> %1, <vscale x 32 x i8> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vqmaccus_4x8x4_ta_i32m8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
 ; CHECK-NEXT:    sf.vqmaccus.4x8x4 v8, v16, v20
 ; CHECK-NEXT:    ret
 entry:

From 6efc6f1c5fc6a53d33fb624786d17a1672b9eedf Mon Sep 17 00:00:00 2001
From: Brandon Wu <brandon.wu@sifive.com>
Date: Thu, 21 Dec 2023 21:21:27 -0800
Subject: [PATCH 3/4] fixup! Recommit [RISCV] Update the interface of sifive
 vqmaccqoq (#74284)

---
 llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td | 17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)

diff --git a/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td b/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
index 6472875908db0..2397de44a11ff 100644
--- a/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
+++ b/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
@@ -358,15 +358,14 @@ multiclass VPseudoSiFiveVQMACCDOD<string Constraint = ""> {
 multiclass VPseudoSiFiveVQMACCQOQ<string Constraint = ""> {
   foreach i = [0, 1, 2, 3] in
     let VLMul = MxListVF4[i].value in
-    defm NAME : VPseudoSiFiveVMACC<MxListVF8[i].MX, MxListVF8[i].vrclass,
+    defm NAME : VPseudoSiFiveVMACC<MxListVF4[i].MX, MxListVF8[i].vrclass,
                                    MxListVF4[i].vrclass, Constraint>;
 }
 
 multiclass VPseudoSiFiveVFWMACC<string Constraint = ""> {
-  foreach i = [0, 1, 2, 3, 4] in
-    let VLMul = MxListVF2[i].value in
-    defm NAME : VPseudoSiFiveVMACC<MxListVF4[i].MX, MxListFW[i].wvrclass,
-                                   MxListFW[i].vrclass, Constraint>;
+  foreach m = MxListVF2 in
+    let VLMul = m.value in
+    defm NAME : VPseudoSiFiveVMACC<m.MX, m.wvrclass, m.vrclass, Constraint>;
 }
 
 multiclass VPseudoSiFiveVFNRCLIP<string Constraint = "@earlyclobber $rd"> {
@@ -561,15 +560,17 @@ class GetFTypeInfo<int Sew> {
 }
 
 multiclass VPatVMACC<string intrinsic, string instruction, string kind,
-                     list<VTypeInfoToWide> info_pairs, ValueType vec_m1> {
+                     list<VTypeInfoToWide> info_pairs, ValueType vec_m1,
+                     bit lmul_follows_vd = 0> {
   foreach pair = info_pairs in {
     defvar VdInfo = pair.Wti;
     defvar Vs2Info = pair.Vti;
+    defvar LMul = !if(!eq(lmul_follows_vd, 1), VdInfo.LMul, Vs2Info.LMul);
     let Predicates = [HasVInstructions] in
     def : VPatTernaryNoMaskWithPolicy<"int_riscv_sf_" # intrinsic,
                                       "Pseudo" # instruction, kind, VdInfo.Vector,
                                       vec_m1, Vs2Info.Vector,
-                                      Vs2Info.Log2SEW, VdInfo.LMul,
+                                      Vs2Info.Log2SEW, LMul,
                                       VdInfo.RegClass, VR, Vs2Info.RegClass>;
   }
 }
@@ -589,7 +590,7 @@ defset list<VTypeInfoToWide> VQMACCQOQInfoPairs = {
 }
 
 multiclass VPatVQMACCDOD<string intrinsic, string instruction, string kind>
-    : VPatVMACC<intrinsic, instruction, kind, VQMACCDODInfoPairs, vint8m1_t>;
+    : VPatVMACC<intrinsic, instruction, kind, VQMACCDODInfoPairs, vint8m1_t, 1>;
 
 multiclass VPatVQMACCQOQ<string intrinsic, string instruction, string kind>
     : VPatVMACC<intrinsic, instruction, kind, VQMACCQOQInfoPairs, vint8m1_t>;

From bf1a33e2cdf833fbdd53e52a804ab997363dc539 Mon Sep 17 00:00:00 2001
From: Brandon Wu <brandon.wu@sifive.com>
Date: Mon, 25 Dec 2023 19:34:19 -0800
Subject: [PATCH 4/4] fixup! Recommit [RISCV] Update the interface of sifive
 vqmaccqoq (#74284)

---
 llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)

diff --git a/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td b/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
index 2397de44a11ff..9f1ce851c1570 100644
--- a/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
+++ b/llvm/lib/Target/RISCV/RISCVInstrInfoXSf.td
@@ -356,7 +356,7 @@ multiclass VPseudoSiFiveVQMACCDOD<string Constraint = ""> {
 }
 
 multiclass VPseudoSiFiveVQMACCQOQ<string Constraint = ""> {
-  foreach i = [0, 1, 2, 3] in
+  foreach i = 0-3 in
     let VLMul = MxListVF4[i].value in
     defm NAME : VPseudoSiFiveVMACC<MxListVF4[i].MX, MxListVF8[i].vrclass,
                                    MxListVF4[i].vrclass, Constraint>;
@@ -560,17 +560,15 @@ class GetFTypeInfo<int Sew> {
 }
 
 multiclass VPatVMACC<string intrinsic, string instruction, string kind,
-                     list<VTypeInfoToWide> info_pairs, ValueType vec_m1,
-                     bit lmul_follows_vd = 0> {
+                     list<VTypeInfoToWide> info_pairs, ValueType vec_m1> {
   foreach pair = info_pairs in {
     defvar VdInfo = pair.Wti;
     defvar Vs2Info = pair.Vti;
-    defvar LMul = !if(!eq(lmul_follows_vd, 1), VdInfo.LMul, Vs2Info.LMul);
     let Predicates = [HasVInstructions] in
     def : VPatTernaryNoMaskWithPolicy<"int_riscv_sf_" # intrinsic,
                                       "Pseudo" # instruction, kind, VdInfo.Vector,
                                       vec_m1, Vs2Info.Vector,
-                                      Vs2Info.Log2SEW, LMul,
+                                      Vs2Info.Log2SEW, Vs2Info.LMul,
                                       VdInfo.RegClass, VR, Vs2Info.RegClass>;
   }
 }
@@ -590,7 +588,7 @@ defset list<VTypeInfoToWide> VQMACCQOQInfoPairs = {
 }
 
 multiclass VPatVQMACCDOD<string intrinsic, string instruction, string kind>
-    : VPatVMACC<intrinsic, instruction, kind, VQMACCDODInfoPairs, vint8m1_t, 1>;
+    : VPatVMACC<intrinsic, instruction, kind, VQMACCDODInfoPairs, vint8m1_t>;
 
 multiclass VPatVQMACCQOQ<string intrinsic, string instruction, string kind>
     : VPatVMACC<intrinsic, instruction, kind, VQMACCQOQInfoPairs, vint8m1_t>;