[clang] [AArch64][Clang] Update untyped sme intrinsics with fp8 variants (PR #124543)
via cfe-commits
cfe-commits at lists.llvm.org
Tue Feb 4 06:55:07 PST 2025
https://github.com/Lukacma updated https://github.com/llvm/llvm-project/pull/124543
>From a6d2a91865c001db4ed9d82b126cca7ab6b6395f Mon Sep 17 00:00:00 2001
From: Marian Lukac <Marian.Lukac at arm.com>
Date: Mon, 27 Jan 2025 13:36:19 +0000
Subject: [PATCH 1/2] [AArch64][Clang] Update untyped sme intrinsics with fp8
variants
---
clang/include/clang/Basic/arm_sme.td | 32 ++---
clang/include/clang/Basic/arm_sve.td | 24 ++--
.../AArch64/sme-intrinsics/acle_sme_read.c | 134 +++++++++++++++++
.../AArch64/sme-intrinsics/acle_sme_write.c | 135 +++++++++++++++++
.../sme2-intrinsics/acle_sme2_luti2_lane_zt.c | 13 ++
.../acle_sme2_luti2_lane_zt_x2.c | 14 ++
.../acle_sme2_luti2_lane_zt_x4.c | 14 ++
.../sme2-intrinsics/acle_sme2_luti4_lane_zt.c | 14 ++
.../acle_sme2_luti4_lane_zt_x2.c | 14 ++
.../AArch64/sme2-intrinsics/acle_sme2_read.c | 84 ++++++++++-
.../sme2-intrinsics/acle_sme2_vector_selx2.c | 13 ++
.../sme2-intrinsics/acle_sme2_vector_selx4.c | 100 +++++++------
.../sme2-intrinsics/acle_sme2_vector_uzpx2.c | 28 ++++
.../sme2-intrinsics/acle_sme2_vector_uzpx4.c | 28 ++++
.../sme2-intrinsics/acle_sme2_vector_zipx2.c | 28 ++++
.../sme2-intrinsics/acle_sme2_vector_zipx4.c | 28 ++++
.../AArch64/sme2-intrinsics/acle_sme2_write.c | 78 +++++++++-
.../sme2p1-intrinsics/acle_sme2p1_movaz.c | 136 ++++++++++++++++++
.../AArch64/sve2-intrinsics/acle_sve2_revd.c | 42 ++++++
19 files changed, 883 insertions(+), 76 deletions(-)
diff --git a/clang/include/clang/Basic/arm_sme.td b/clang/include/clang/Basic/arm_sme.td
index 891ed9874bb3d0..51cab572ce357d 100644
--- a/clang/include/clang/Basic/arm_sme.td
+++ b/clang/include/clang/Basic/arm_sme.td
@@ -110,11 +110,11 @@ multiclass ZARead<string n_suffix, string t, string i_prefix, list<ImmCheck> ch>
}
}
-defm SVREAD_ZA8 : ZARead<"za8", "cUc", "aarch64_sme_read", [ImmCheck<2, ImmCheck0_0>]>;
+defm SVREAD_ZA8 : ZARead<"za8", "cUcm", "aarch64_sme_read", [ImmCheck<2, ImmCheck0_0>]>;
defm SVREAD_ZA16 : ZARead<"za16", "sUshb", "aarch64_sme_read", [ImmCheck<2, ImmCheck0_1>]>;
defm SVREAD_ZA32 : ZARead<"za32", "iUif", "aarch64_sme_read", [ImmCheck<2, ImmCheck0_3>]>;
defm SVREAD_ZA64 : ZARead<"za64", "lUld", "aarch64_sme_read", [ImmCheck<2, ImmCheck0_7>]>;
-defm SVREAD_ZA128 : ZARead<"za128", "csilUcUsUiUlhbfd", "aarch64_sme_readq", [ImmCheck<2, ImmCheck0_15>]>;
+defm SVREAD_ZA128 : ZARead<"za128", "csilUcUsUiUlmhbfd", "aarch64_sme_readq", [ImmCheck<2, ImmCheck0_15>]>;
////////////////////////////////////////////////////////////////////////////////
// Write horizontal/vertical ZA slices
@@ -131,11 +131,11 @@ multiclass ZAWrite<string n_suffix, string t, string i_prefix, list<ImmCheck> ch
}
}
-defm SVWRITE_ZA8 : ZAWrite<"za8", "cUc", "aarch64_sme_write", [ImmCheck<0, ImmCheck0_0>]>;
+defm SVWRITE_ZA8 : ZAWrite<"za8", "cUcm", "aarch64_sme_write", [ImmCheck<0, ImmCheck0_0>]>;
defm SVWRITE_ZA16 : ZAWrite<"za16", "sUshb", "aarch64_sme_write", [ImmCheck<0, ImmCheck0_1>]>;
defm SVWRITE_ZA32 : ZAWrite<"za32", "iUif", "aarch64_sme_write", [ImmCheck<0, ImmCheck0_3>]>;
defm SVWRITE_ZA64 : ZAWrite<"za64", "lUld", "aarch64_sme_write", [ImmCheck<0, ImmCheck0_7>]>;
-defm SVWRITE_ZA128 : ZAWrite<"za128", "csilUcUsUiUlhbfd", "aarch64_sme_writeq", [ImmCheck<0, ImmCheck0_15>]>;
+defm SVWRITE_ZA128 : ZAWrite<"za128", "csilUcUsUiUlmhbfd", "aarch64_sme_writeq", [ImmCheck<0, ImmCheck0_15>]>;
////////////////////////////////////////////////////////////////////////////////
// SME - Zero
@@ -350,7 +350,7 @@ multiclass ZAWrite_VG<string n, string t, string i, list<ImmCheck> checks> {
}
let SMETargetGuard = "sme2" in {
- defm SVWRITE_ZA8 : ZAWrite_VG<"za8", "cUc", "aarch64_sme_write", [ImmCheck<0, ImmCheck0_0>]>;
+ defm SVWRITE_ZA8 : ZAWrite_VG<"za8", "cUcm", "aarch64_sme_write", [ImmCheck<0, ImmCheck0_0>]>;
defm SVWRITE_ZA16 : ZAWrite_VG<"za16", "sUshb", "aarch64_sme_write", [ImmCheck<0, ImmCheck0_1>]>;
defm SVWRITE_ZA32 : ZAWrite_VG<"za32", "iUif", "aarch64_sme_write", [ImmCheck<0, ImmCheck0_3>]>;
defm SVWRITE_ZA64 : ZAWrite_VG<"za64", "lUld", "aarch64_sme_write", [ImmCheck<0, ImmCheck0_7>]>;
@@ -366,7 +366,7 @@ multiclass ZARead_VG<string n, string t, string i, list<ImmCheck> checks> {
}
let SMETargetGuard = "sme2" in {
- defm SVREAD_ZA8 : ZARead_VG<"za8", "cUc", "aarch64_sme_read", [ImmCheck<0, ImmCheck0_0>]>;
+ defm SVREAD_ZA8 : ZARead_VG<"za8", "cUcm", "aarch64_sme_read", [ImmCheck<0, ImmCheck0_0>]>;
defm SVREAD_ZA16 : ZARead_VG<"za16", "sUshb", "aarch64_sme_read", [ImmCheck<0, ImmCheck0_1>]>;
defm SVREAD_ZA32 : ZARead_VG<"za32", "iUif", "aarch64_sme_read", [ImmCheck<0, ImmCheck0_3>]>;
defm SVREAD_ZA64 : ZARead_VG<"za64", "lUld", "aarch64_sme_read", [ImmCheck<0, ImmCheck0_7>]>;
@@ -722,7 +722,7 @@ def IN_STREAMING_MODE : Inst<"__arm_in_streaming_mode", "sv", "Pc", MergeNone,
// lookup table expand four contiguous registers
//
let SMETargetGuard = "sme2" in {
- def SVLUTI2_LANE_ZT_X4 : Inst<"svluti2_lane_zt_{d}_x4", "4.di[i", "cUcsUsiUibhf", MergeNone, "aarch64_sme_luti2_lane_zt_x4", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_3>]>;
+ def SVLUTI2_LANE_ZT_X4 : Inst<"svluti2_lane_zt_{d}_x4", "4.di[i", "cUcsUsiUimbhf", MergeNone, "aarch64_sme_luti2_lane_zt_x4", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_3>]>;
def SVLUTI4_LANE_ZT_X4 : Inst<"svluti4_lane_zt_{d}_x4", "4.di[i", "sUsiUibhf", MergeNone, "aarch64_sme_luti4_lane_zt_x4", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_1>]>;
}
@@ -730,16 +730,16 @@ let SMETargetGuard = "sme2" in {
// lookup table expand one register
//
let SMETargetGuard = "sme2" in {
- def SVLUTI2_LANE_ZT : Inst<"svluti2_lane_zt_{d}", "di[i", "cUcsUsiUibhf", MergeNone, "aarch64_sme_luti2_lane_zt", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_15>]>;
- def SVLUTI4_LANE_ZT : Inst<"svluti4_lane_zt_{d}", "di[i", "cUcsUsiUibhf", MergeNone, "aarch64_sme_luti4_lane_zt", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_7>]>;
+ def SVLUTI2_LANE_ZT : Inst<"svluti2_lane_zt_{d}", "di[i", "cUcsUsiUimbhf", MergeNone, "aarch64_sme_luti2_lane_zt", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_15>]>;
+ def SVLUTI4_LANE_ZT : Inst<"svluti4_lane_zt_{d}", "di[i", "cUcsUsiUimbhf", MergeNone, "aarch64_sme_luti4_lane_zt", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_7>]>;
}
//
// lookup table expand two contiguous registers
//
let SMETargetGuard = "sme2" in {
- def SVLUTI2_LANE_ZT_X2 : Inst<"svluti2_lane_zt_{d}_x2", "2.di[i", "cUcsUsiUibhf", MergeNone, "aarch64_sme_luti2_lane_zt_x2", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_7>]>;
- def SVLUTI4_LANE_ZT_X2 : Inst<"svluti4_lane_zt_{d}_x2", "2.di[i", "cUcsUsiUibhf", MergeNone, "aarch64_sme_luti4_lane_zt_x2", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_3>]>;
+ def SVLUTI2_LANE_ZT_X2 : Inst<"svluti2_lane_zt_{d}_x2", "2.di[i", "cUcsUsiUimbhf", MergeNone, "aarch64_sme_luti2_lane_zt_x2", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_7>]>;
+ def SVLUTI4_LANE_ZT_X2 : Inst<"svluti4_lane_zt_{d}_x2", "2.di[i", "cUcsUsiUimbhf", MergeNone, "aarch64_sme_luti4_lane_zt_x2", [IsStreaming, IsInZT0], [ImmCheck<0, ImmCheck0_0>, ImmCheck<2, ImmCheck0_3>]>;
}
//
@@ -811,12 +811,12 @@ multiclass ZAReadz<string n_suffix, string vg_num, string t, string i_prefix, li
}
}
-defm SVREADZ_ZA8_X2 : ZAReadz<"za8", "2", "cUc", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_0>]>;
+defm SVREADZ_ZA8_X2 : ZAReadz<"za8", "2", "cUcm", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_0>]>;
defm SVREADZ_ZA16_X2 : ZAReadz<"za16", "2", "sUshb", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_1>]>;
defm SVREADZ_ZA32_X2 : ZAReadz<"za32", "2", "iUif", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_3>]>;
defm SVREADZ_ZA64_X2 : ZAReadz<"za64", "2", "lUld", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_7>]>;
-defm SVREADZ_ZA8_X4 : ZAReadz<"za8", "4", "cUc", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_0>]>;
+defm SVREADZ_ZA8_X4 : ZAReadz<"za8", "4", "cUcm", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_0>]>;
defm SVREADZ_ZA16_X4 : ZAReadz<"za16", "4", "sUshb", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_1>]>;
defm SVREADZ_ZA32_X4 : ZAReadz<"za32", "4", "iUif", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_3>]>;
defm SVREADZ_ZA64_X4 : ZAReadz<"za64", "4", "lUld", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_7>]>;
@@ -834,15 +834,15 @@ multiclass ZAReadzSingle<string n_suffix, string t, string i_prefix, list<ImmChe
}
}
-defm SVREADZ_ZA8 : ZAReadzSingle<"za8", "cUc", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_0>]>;
+defm SVREADZ_ZA8 : ZAReadzSingle<"za8", "cUcm", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_0>]>;
defm SVREADZ_ZA16 : ZAReadzSingle<"za16", "sUshb", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_1>]>;
defm SVREADZ_ZA32 : ZAReadzSingle<"za32", "iUif", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_3>]>;
defm SVREADZ_ZA64 : ZAReadzSingle<"za64", "lUld", "aarch64_sme_readz", [ImmCheck<0, ImmCheck0_7>]>;
-defm SVREADZ_ZA128 : ZAReadzSingle<"za128", "csilUcUiUsUlbhfd", "aarch64_sme_readz_q", [ImmCheck<0, ImmCheck0_15>]>;
+defm SVREADZ_ZA128 : ZAReadzSingle<"za128", "csilUcUiUsUlmbhfd", "aarch64_sme_readz_q", [ImmCheck<0, ImmCheck0_15>]>;
multiclass ZAReadzArray<string vg_num>{
let SMETargetGuard = "sme2p1" in {
- def NAME # _B : SInst<"svreadz_za8_{d}_vg1x" # vg_num, vg_num # "m", "cUc", MergeNone, "aarch64_sme_readz_x" # vg_num, [IsStreaming, IsInOutZA]>;
+ def NAME # _B : SInst<"svreadz_za8_{d}_vg1x" # vg_num, vg_num # "m", "cUcm", MergeNone, "aarch64_sme_readz_x" # vg_num, [IsStreaming, IsInOutZA]>;
def NAME # _H : SInst<"svreadz_za16_{d}_vg1x" # vg_num, vg_num # "m", "sUsbh", MergeNone, "aarch64_sme_readz_x" # vg_num, [IsStreaming, IsInOutZA]>;
def NAME # _S : SInst<"svreadz_za32_{d}_vg1x" # vg_num, vg_num # "m", "iUif", MergeNone, "aarch64_sme_readz_x" # vg_num, [IsStreaming, IsInOutZA]>;
def NAME # _D : SInst<"svreadz_za64_{d}_vg1x" # vg_num, vg_num # "m", "lUld", MergeNone, "aarch64_sme_readz_x" # vg_num, [IsStreaming, IsInOutZA]>;
diff --git a/clang/include/clang/Basic/arm_sve.td b/clang/include/clang/Basic/arm_sve.td
index e7001bac450e89..12cca992a21ab7 100644
--- a/clang/include/clang/Basic/arm_sve.td
+++ b/clang/include/clang/Basic/arm_sve.td
@@ -2104,7 +2104,7 @@ let SVETargetGuard = "sve2p1", SMETargetGuard = "sme" in {
def SVSCLAMP : SInst<"svclamp[_{d}]", "dddd", "csil", MergeNone, "aarch64_sve_sclamp", [VerifyRuntimeMode], []>;
def SVUCLAMP : SInst<"svclamp[_{d}]", "dddd", "UcUsUiUl", MergeNone, "aarch64_sve_uclamp", [VerifyRuntimeMode], []>;
-defm SVREVD : SInstZPZ<"svrevd", "csilUcUsUiUlbhfd", "aarch64_sve_revd">;
+defm SVREVD : SInstZPZ<"svrevd", "csilUcUsUiUlmbhfd", "aarch64_sve_revd">;
}
let SVETargetGuard = "sve2p1", SMETargetGuard = "sme2" in {
@@ -2223,8 +2223,8 @@ let SVETargetGuard = InvalidMode, SMETargetGuard = "sme2" in {
def SVADD_SINGLE_X4 : SInst<"svadd[_single_{d}_x4]", "44d", "cUcsUsiUilUl", MergeNone, "aarch64_sve_add_single_x4", [IsStreaming], []>;
// 2-way and 4-way selects
- def SVSEL_X2 : SInst<"svsel[_{d}_x2]", "2}22", "cUcsUsiUilUlbhfd", MergeNone, "aarch64_sve_sel_x2", [IsStreaming], []>;
- def SVSEL_X4 : SInst<"svsel[_{d}_x4]", "4}44", "cUcsUsiUilUlbhfd", MergeNone, "aarch64_sve_sel_x4", [IsStreaming], []>;
+ def SVSEL_X2 : SInst<"svsel[_{d}_x2]", "2}22", "cUcsUsiUilUlmbhfd", MergeNone, "aarch64_sve_sel_x2", [IsStreaming], []>;
+ def SVSEL_X4 : SInst<"svsel[_{d}_x4]", "4}44", "cUcsUsiUilUlmbhfd", MergeNone, "aarch64_sve_sel_x4", [IsStreaming], []>;
// SRSHL / URSHL
def SVSRSHL_SINGLE_X2 : SInst<"svrshl[_single_{d}_x2]", "22d", "csil", MergeNone, "aarch64_sve_srshl_single_x2", [IsStreaming], []>;
@@ -2402,15 +2402,15 @@ let SVETargetGuard = InvalidMode, SMETargetGuard = "sme2" in {
//
let SVETargetGuard = InvalidMode, SMETargetGuard = "sme2" in {
- def SVZIP_X2 : SInst<"svzip[_{d}_x2]", "22", "cUcsUsiUilUlbhfd", MergeNone, "aarch64_sve_zip_x2", [IsStreaming], []>;
- def SVZIPQ_X2 : SInst<"svzipq[_{d}_x2]", "22", "cUcsUsiUilUlbhfd", MergeNone, "aarch64_sve_zipq_x2", [IsStreaming], []>;
- def SVZIP_X4 : SInst<"svzip[_{d}_x4]", "44", "cUcsUsiUilUlbhfd", MergeNone, "aarch64_sve_zip_x4", [IsStreaming], []>;
- def SVZIPQ_X4 : SInst<"svzipq[_{d}_x4]", "44", "cUcsUsiUilUlbhfd", MergeNone, "aarch64_sve_zipq_x4", [IsStreaming], []>;
-
- def SVUZP_X2 : SInst<"svuzp[_{d}_x2]", "22", "cUcsUsiUilUlbhfd", MergeNone, "aarch64_sve_uzp_x2", [IsStreaming], []>;
- def SVUZPQ_X2 : SInst<"svuzpq[_{d}_x2]", "22", "cUcsUsiUilUlbhfd", MergeNone, "aarch64_sve_uzpq_x2", [IsStreaming], []>;
- def SVUZP_X4 : SInst<"svuzp[_{d}_x4]", "44", "cUcsUsiUilUlbhfd", MergeNone, "aarch64_sve_uzp_x4", [IsStreaming], []>;
- def SVUZPQ_X4 : SInst<"svuzpq[_{d}_x4]", "44", "cUcsUsiUilUlbhfd", MergeNone, "aarch64_sve_uzpq_x4", [IsStreaming], []>;
+ def SVZIP_X2 : SInst<"svzip[_{d}_x2]", "22", "cUcsUsiUilUlmbhfd", MergeNone, "aarch64_sve_zip_x2", [IsStreaming], []>;
+ def SVZIPQ_X2 : SInst<"svzipq[_{d}_x2]", "22", "cUcsUsiUilUlmbhfd", MergeNone, "aarch64_sve_zipq_x2", [IsStreaming], []>;
+ def SVZIP_X4 : SInst<"svzip[_{d}_x4]", "44", "cUcsUsiUilUlmbhfd", MergeNone, "aarch64_sve_zip_x4", [IsStreaming], []>;
+ def SVZIPQ_X4 : SInst<"svzipq[_{d}_x4]", "44", "cUcsUsiUilUlmbhfd", MergeNone, "aarch64_sve_zipq_x4", [IsStreaming], []>;
+
+ def SVUZP_X2 : SInst<"svuzp[_{d}_x2]", "22", "cUcsUsiUilUlmbhfd", MergeNone, "aarch64_sve_uzp_x2", [IsStreaming], []>;
+ def SVUZPQ_X2 : SInst<"svuzpq[_{d}_x2]", "22", "cUcsUsiUilUlmbhfd", MergeNone, "aarch64_sve_uzpq_x2", [IsStreaming], []>;
+ def SVUZP_X4 : SInst<"svuzp[_{d}_x4]", "44", "cUcsUsiUilUlmbhfd", MergeNone, "aarch64_sve_uzp_x4", [IsStreaming], []>;
+ def SVUZPQ_X4 : SInst<"svuzpq[_{d}_x4]", "44", "cUcsUsiUilUlmbhfd", MergeNone, "aarch64_sve_uzpq_x4", [IsStreaming], []>;
}
//
diff --git a/clang/test/CodeGen/AArch64/sme-intrinsics/acle_sme_read.c b/clang/test/CodeGen/AArch64/sme-intrinsics/acle_sme_read.c
index 508fad09ea715b..0605f9eef036fd 100644
--- a/clang/test/CodeGen/AArch64/sme-intrinsics/acle_sme_read.c
+++ b/clang/test/CodeGen/AArch64/sme-intrinsics/acle_sme_read.c
@@ -318,6 +318,41 @@ svuint64_t test_svread_hor_za64_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice
return SME_ACLE_FUNC(svread_hor_za64, _u64, _m)(zd, pg, 7, slice);
}
+// CHECK-C-LABEL: define dso_local <vscale x 16 x i8> @test_svread_hor_za8_mf8(
+// CHECK-C-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.read.horiz.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[SLICE_BASE]])
+// CHECK-C-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CHECK-CXX-LABEL: define dso_local <vscale x 16 x i8> @_Z23test_svread_hor_za8_mf8u13__SVMfloat8_tu10__SVBool_tj(
+// CHECK-CXX-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.read.horiz.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[SLICE_BASE]])
+// CHECK-CXX-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svread_hor_za8_mf8(svmfloat8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
+ return SME_ACLE_FUNC(svread_hor_za8, _mf8, _m)(zd, pg, 0, slice_base);
+}
+
+// CHECK-C-LABEL: define dso_local <vscale x 16 x i8> @test_svread_hor_za8_mf8_1(
+// CHECK-C-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: [[ADD:%.*]] = add i32 [[SLICE_BASE]], 15
+// CHECK-C-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.read.horiz.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[ADD]])
+// CHECK-C-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CHECK-CXX-LABEL: define dso_local <vscale x 16 x i8> @_Z25test_svread_hor_za8_mf8_1u13__SVMfloat8_tu10__SVBool_tj(
+// CHECK-CXX-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: [[ADD:%.*]] = add i32 [[SLICE_BASE]], 15
+// CHECK-CXX-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.read.horiz.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[ADD]])
+// CHECK-CXX-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svread_hor_za8_mf8_1(svmfloat8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
+ uint32_t slice = slice_base + 15;
+ return SME_ACLE_FUNC(svread_hor_za8, _mf8, _m)(zd, pg, 0, slice);
+}
+
// CHECK-C-LABEL: define dso_local <vscale x 8 x half> @test_svread_hor_za16_f16(
// CHECK-C-SAME: <vscale x 8 x half> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
// CHECK-C-NEXT: entry:
@@ -754,6 +789,38 @@ svuint64_t test_svread_hor_za128_u64_1(svuint64_t zd, svbool_t pg, uint32_t slic
return SME_ACLE_FUNC(svread_hor_za128, _u64, _m)(zd, pg, 15, slice_base);
}
+// CHECK-C-LABEL: define dso_local <vscale x 16 x i8> @test_svread_hor_za128_mf8(
+// CHECK-C-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readq.horiz.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[SLICE_BASE]])
+// CHECK-C-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CHECK-CXX-LABEL: define dso_local <vscale x 16 x i8> @_Z25test_svread_hor_za128_mf8u13__SVMfloat8_tu10__SVBool_tj(
+// CHECK-CXX-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readq.horiz.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[SLICE_BASE]])
+// CHECK-CXX-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svread_hor_za128_mf8(svmfloat8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
+ return SME_ACLE_FUNC(svread_hor_za128, _mf8, _m)(zd, pg, 0, slice_base);
+}
+
+// CHECK-C-LABEL: define dso_local <vscale x 16 x i8> @test_svread_hor_za128_mf8_1(
+// CHECK-C-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readq.horiz.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 15, i32 [[SLICE_BASE]])
+// CHECK-C-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CHECK-CXX-LABEL: define dso_local <vscale x 16 x i8> @_Z27test_svread_hor_za128_mf8_1u13__SVMfloat8_tu10__SVBool_tj(
+// CHECK-CXX-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readq.horiz.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 15, i32 [[SLICE_BASE]])
+// CHECK-CXX-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svread_hor_za128_mf8_1(svmfloat8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
+ return SME_ACLE_FUNC(svread_hor_za128, _mf8, _m)(zd, pg, 15, slice_base);
+}
+
// CHECK-C-LABEL: define dso_local <vscale x 8 x half> @test_svread_hor_za128_f16(
// CHECK-C-SAME: <vscale x 8 x half> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
// CHECK-C-NEXT: entry:
@@ -1202,6 +1269,41 @@ svuint64_t test_svread_ver_za64_u64_1(svuint64_t zd, svbool_t pg, uint32_t slice
return SME_ACLE_FUNC(svread_ver_za64, _u64, _m)(zd, pg, 7, slice);
}
+// CHECK-C-LABEL: define dso_local <vscale x 16 x i8> @test_svread_ver_za8_mf8(
+// CHECK-C-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.read.vert.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[SLICE_BASE]])
+// CHECK-C-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CHECK-CXX-LABEL: define dso_local <vscale x 16 x i8> @_Z23test_svread_ver_za8_mf8u13__SVMfloat8_tu10__SVBool_tj(
+// CHECK-CXX-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.read.vert.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[SLICE_BASE]])
+// CHECK-CXX-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svread_ver_za8_mf8(svmfloat8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
+ return SME_ACLE_FUNC(svread_ver_za8, _mf8, _m)(zd, pg, 0, slice_base);
+}
+
+// CHECK-C-LABEL: define dso_local <vscale x 16 x i8> @test_svread_ver_za8_mf8_1(
+// CHECK-C-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: [[ADD:%.*]] = add i32 [[SLICE_BASE]], 15
+// CHECK-C-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.read.vert.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[ADD]])
+// CHECK-C-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CHECK-CXX-LABEL: define dso_local <vscale x 16 x i8> @_Z25test_svread_ver_za8_mf8_1u13__SVMfloat8_tu10__SVBool_tj(
+// CHECK-CXX-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: [[ADD:%.*]] = add i32 [[SLICE_BASE]], 15
+// CHECK-CXX-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.read.vert.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[ADD]])
+// CHECK-CXX-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svread_ver_za8_mf8_1(svmfloat8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
+ uint32_t slice = slice_base + 15;
+ return SME_ACLE_FUNC(svread_ver_za8, _mf8, _m)(zd, pg, 0, slice);
+}
+
// CHECK-C-LABEL: define dso_local <vscale x 8 x half> @test_svread_ver_za16_f16(
// CHECK-C-SAME: <vscale x 8 x half> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
// CHECK-C-NEXT: entry:
@@ -1638,6 +1740,38 @@ svuint64_t test_svread_ver_za128_u64_1(svuint64_t zd, svbool_t pg, uint32_t slic
return SME_ACLE_FUNC(svread_ver_za128, _u64, _m)(zd, pg, 15, slice_base);
}
+// CHECK-C-LABEL: define dso_local <vscale x 16 x i8> @test_svread_ver_za128_mf8(
+// CHECK-C-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readq.vert.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[SLICE_BASE]])
+// CHECK-C-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CHECK-CXX-LABEL: define dso_local <vscale x 16 x i8> @_Z25test_svread_ver_za128_mf8u13__SVMfloat8_tu10__SVBool_tj(
+// CHECK-CXX-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readq.vert.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 0, i32 [[SLICE_BASE]])
+// CHECK-CXX-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svread_ver_za128_mf8(svmfloat8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
+ return SME_ACLE_FUNC(svread_ver_za128, _mf8, _m)(zd, pg, 0, slice_base);
+}
+
+// CHECK-C-LABEL: define dso_local <vscale x 16 x i8> @test_svread_ver_za128_mf8_1(
+// CHECK-C-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readq.vert.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 15, i32 [[SLICE_BASE]])
+// CHECK-C-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CHECK-CXX-LABEL: define dso_local <vscale x 16 x i8> @_Z27test_svread_ver_za128_mf8_1u13__SVMfloat8_tu10__SVBool_tj(
+// CHECK-CXX-SAME: <vscale x 16 x i8> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readq.vert.nxv16i8(<vscale x 16 x i8> [[ZD]], <vscale x 16 x i1> [[PG]], i32 15, i32 [[SLICE_BASE]])
+// CHECK-CXX-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svread_ver_za128_mf8_1(svmfloat8_t zd, svbool_t pg, uint32_t slice_base) __arm_streaming __arm_in("za") {
+ return SME_ACLE_FUNC(svread_ver_za128, _mf8, _m)(zd, pg, 15, slice_base);
+}
+
// CHECK-C-LABEL: define dso_local <vscale x 8 x half> @test_svread_ver_za128_f16(
// CHECK-C-SAME: <vscale x 8 x half> [[ZD:%.*]], <vscale x 16 x i1> [[PG:%.*]], i32 noundef [[SLICE_BASE:%.*]]) local_unnamed_addr #[[ATTR0]] {
// CHECK-C-NEXT: entry:
diff --git a/clang/test/CodeGen/AArch64/sme-intrinsics/acle_sme_write.c b/clang/test/CodeGen/AArch64/sme-intrinsics/acle_sme_write.c
index 86a691f14623f9..99b96ee87553c7 100644
--- a/clang/test/CodeGen/AArch64/sme-intrinsics/acle_sme_write.c
+++ b/clang/test/CodeGen/AArch64/sme-intrinsics/acle_sme_write.c
@@ -318,6 +318,41 @@ void test_svwrite_hor_za64_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn
SME_ACLE_FUNC(svwrite_hor_za64, _u64, _m)(7, slice, pg, zn);
}
+// CHECK-C-LABEL: define dso_local void @test_svwrite_hor_za8_mf8(
+// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: tail call void @llvm.aarch64.sme.write.horiz.nxv16i8(i32 0, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-C-NEXT: ret void
+//
+// CHECK-CXX-LABEL: define dso_local void @_Z24test_svwrite_hor_za8_mf8ju10__SVBool_tu13__SVMfloat8_t(
+// CHECK-CXX-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: tail call void @llvm.aarch64.sme.write.horiz.nxv16i8(i32 0, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-CXX-NEXT: ret void
+//
+void test_svwrite_hor_za8_mf8(uint32_t slice_base, svbool_t pg, svmfloat8_t zn) __arm_streaming __arm_inout("za") {
+ SME_ACLE_FUNC(svwrite_hor_za8, _mf8, _m)(0, slice_base, pg, zn);
+}
+
+// CHECK-C-LABEL: define dso_local void @test_svwrite_hor_za8_mf8_1(
+// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: [[ADD:%.*]] = add i32 [[SLICE_BASE]], 15
+// CHECK-C-NEXT: tail call void @llvm.aarch64.sme.write.horiz.nxv16i8(i32 0, i32 [[ADD]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-C-NEXT: ret void
+//
+// CHECK-CXX-LABEL: define dso_local void @_Z26test_svwrite_hor_za8_mf8_1ju10__SVBool_tu13__SVMfloat8_t(
+// CHECK-CXX-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: [[ADD:%.*]] = add i32 [[SLICE_BASE]], 15
+// CHECK-CXX-NEXT: tail call void @llvm.aarch64.sme.write.horiz.nxv16i8(i32 0, i32 [[ADD]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-CXX-NEXT: ret void
+//
+void test_svwrite_hor_za8_mf8_1(uint32_t slice_base, svbool_t pg, svmfloat8_t zn) __arm_streaming __arm_inout("za") {
+ uint32_t slice = slice_base + 15;
+ SME_ACLE_FUNC(svwrite_hor_za8, _mf8, _m)(0, slice, pg, zn);
+}
+
// CHECK-C-LABEL: define dso_local void @test_svwrite_hor_za16_f16(
// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 8 x half> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
// CHECK-C-NEXT: entry:
@@ -754,6 +789,39 @@ void test_svwrite_hor_za128_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t z
SME_ACLE_FUNC(svwrite_hor_za128, _u64, _m)(15, slice_base, pg, zn);
}
+
+// CHECK-C-LABEL: define dso_local void @test_svwrite_hor_za128_mf8(
+// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: tail call void @llvm.aarch64.sme.writeq.horiz.nxv16i8(i32 0, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-C-NEXT: ret void
+//
+// CHECK-CXX-LABEL: define dso_local void @_Z26test_svwrite_hor_za128_mf8ju10__SVBool_tu13__SVMfloat8_t(
+// CHECK-CXX-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: tail call void @llvm.aarch64.sme.writeq.horiz.nxv16i8(i32 0, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-CXX-NEXT: ret void
+//
+void test_svwrite_hor_za128_mf8(uint32_t slice_base, svbool_t pg, svmfloat8_t zn) __arm_streaming __arm_inout("za") {
+ SME_ACLE_FUNC(svwrite_hor_za128, _mf8, _m)(0, slice_base, pg, zn);
+}
+
+// CHECK-C-LABEL: define dso_local void @test_svwrite_hor_za128_mf8_1(
+// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: tail call void @llvm.aarch64.sme.writeq.horiz.nxv16i8(i32 15, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-C-NEXT: ret void
+//
+// CHECK-CXX-LABEL: define dso_local void @_Z28test_svwrite_hor_za128_mf8_1ju10__SVBool_tu13__SVMfloat8_t(
+// CHECK-CXX-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: tail call void @llvm.aarch64.sme.writeq.horiz.nxv16i8(i32 15, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-CXX-NEXT: ret void
+//
+void test_svwrite_hor_za128_mf8_1(uint32_t slice_base, svbool_t pg, svmfloat8_t zn) __arm_streaming __arm_inout("za") {
+ SME_ACLE_FUNC(svwrite_hor_za128, _mf8, _m)(15, slice_base, pg, zn);
+}
+
// CHECK-C-LABEL: define dso_local void @test_svwrite_hor_za128_f16(
// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 8 x half> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
// CHECK-C-NEXT: entry:
@@ -1202,6 +1270,41 @@ void test_svwrite_ver_za64_u64_1(uint32_t slice_base, svbool_t pg, svuint64_t zn
SME_ACLE_FUNC(svwrite_ver_za64, _u64, _m)(7, slice, pg, zn);
}
+// CHECK-C-LABEL: define dso_local void @test_svwrite_ver_za8_mf8(
+// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: tail call void @llvm.aarch64.sme.write.vert.nxv16i8(i32 0, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-C-NEXT: ret void
+//
+// CHECK-CXX-LABEL: define dso_local void @_Z24test_svwrite_ver_za8_mf8ju10__SVBool_tu13__SVMfloat8_t(
+// CHECK-CXX-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: tail call void @llvm.aarch64.sme.write.vert.nxv16i8(i32 0, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-CXX-NEXT: ret void
+//
+void test_svwrite_ver_za8_mf8(uint32_t slice_base, svbool_t pg, svmfloat8_t zn) __arm_streaming __arm_inout("za") {
+ SME_ACLE_FUNC(svwrite_ver_za8, _mf8, _m)(0, slice_base, pg, zn);
+}
+
+// CHECK-C-LABEL: define dso_local void @test_svwrite_ver_za8_mf8_1(
+// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: [[ADD:%.*]] = add i32 [[SLICE_BASE]], 15
+// CHECK-C-NEXT: tail call void @llvm.aarch64.sme.write.vert.nxv16i8(i32 0, i32 [[ADD]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-C-NEXT: ret void
+//
+// CHECK-CXX-LABEL: define dso_local void @_Z26test_svwrite_ver_za8_mf8_1ju10__SVBool_tu13__SVMfloat8_t(
+// CHECK-CXX-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: [[ADD:%.*]] = add i32 [[SLICE_BASE]], 15
+// CHECK-CXX-NEXT: tail call void @llvm.aarch64.sme.write.vert.nxv16i8(i32 0, i32 [[ADD]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-CXX-NEXT: ret void
+//
+void test_svwrite_ver_za8_mf8_1(uint32_t slice_base, svbool_t pg, svmfloat8_t zn) __arm_streaming __arm_inout("za") {
+ uint32_t slice = slice_base + 15;
+ SME_ACLE_FUNC(svwrite_ver_za8, _mf8, _m)(0, slice, pg, zn);
+}
+
// CHECK-C-LABEL: define dso_local void @test_svwrite_ver_za16_f16(
// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 8 x half> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
// CHECK-C-NEXT: entry:
@@ -1656,6 +1759,38 @@ void test_svwrite_ver_za128_f16(uint32_t slice_base, svbool_t pg, svfloat16_t zn
SME_ACLE_FUNC(svwrite_ver_za128, _f16, _m)(0, slice_base, pg, zn);
}
+// CHECK-C-LABEL: define dso_local void @test_svwrite_ver_za128_mf8(
+// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: tail call void @llvm.aarch64.sme.writeq.vert.nxv16i8(i32 0, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-C-NEXT: ret void
+//
+// CHECK-CXX-LABEL: define dso_local void @_Z26test_svwrite_ver_za128_mf8ju10__SVBool_tu13__SVMfloat8_t(
+// CHECK-CXX-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: tail call void @llvm.aarch64.sme.writeq.vert.nxv16i8(i32 0, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-CXX-NEXT: ret void
+//
+void test_svwrite_ver_za128_mf8(uint32_t slice_base, svbool_t pg, svmfloat8_t zn) __arm_streaming __arm_inout("za") {
+ SME_ACLE_FUNC(svwrite_ver_za128, _mf8, _m)(0, slice_base, pg, zn);
+}
+
+// CHECK-C-LABEL: define dso_local void @test_svwrite_ver_za128_mf8_1(
+// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-C-NEXT: entry:
+// CHECK-C-NEXT: tail call void @llvm.aarch64.sme.writeq.vert.nxv16i8(i32 15, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-C-NEXT: ret void
+//
+// CHECK-CXX-LABEL: define dso_local void @_Z28test_svwrite_ver_za128_mf8_1ju10__SVBool_tu13__SVMfloat8_t(
+// CHECK-CXX-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
+// CHECK-CXX-NEXT: entry:
+// CHECK-CXX-NEXT: tail call void @llvm.aarch64.sme.writeq.vert.nxv16i8(i32 15, i32 [[SLICE_BASE]], <vscale x 16 x i1> [[PG]], <vscale x 16 x i8> [[ZN]])
+// CHECK-CXX-NEXT: ret void
+//
+void test_svwrite_ver_za128_mf8_1(uint32_t slice_base, svbool_t pg, svmfloat8_t zn) __arm_streaming __arm_inout("za") {
+ SME_ACLE_FUNC(svwrite_ver_za128, _mf8, _m)(15, slice_base, pg, zn);
+}
+
// CHECK-C-LABEL: define dso_local void @test_svwrite_ver_za128_f16_1(
// CHECK-C-SAME: i32 noundef [[SLICE_BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 8 x half> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
// CHECK-C-NEXT: entry:
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt.c
index 8de82abe05324b..1ab02afbe09042 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt.c
@@ -52,6 +52,19 @@ svuint16_t test_svluti2_lane_zt_u16(svuint8_t zn) __arm_streaming __arm_in("zt0"
return svluti2_lane_zt_u16(0, zn, 15);
}
+// CHECK-LABEL: @test_svluti2_lane_zt_mf8(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.luti2.lane.zt.nxv16i8(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 15)
+// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z24test_svluti2_lane_zt_mf8u11__SVUint8_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.luti2.lane.zt.nxv16i8(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 15)
+// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svluti2_lane_zt_mf8(svuint8_t zn) __arm_streaming __arm_in("zt0") {
+ return svluti2_lane_zt_mf8(0, zn, 15);
+}
// CHECK-LABEL: @test_svluti2_lane_zt_s16(
// CHECK-NEXT: entry:
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt_x2.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt_x2.c
index 3b17c6d9edb198..e97075703b1852 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt_x2.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt_x2.c
@@ -37,6 +37,20 @@ svint8x2_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_in("zt0")
return svluti2_lane_zt_s8_x2(0, zn, 7);
}
+// CHECK-LABEL: @test_svluti2_lane_zt_mf8(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.luti2.lane.zt.x2.nxv16i8(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 7)
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z24test_svluti2_lane_zt_mf8u11__SVUint8_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.luti2.lane.zt.x2.nxv16i8(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 7)
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svluti2_lane_zt_mf8(svuint8_t zn) __arm_streaming __arm_in("zt0") {
+ return svluti2_lane_zt_mf8_x2(0, zn, 7);
+}
+
// CHECK-LABEL: @test_svluti2_lane_zt_u16(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sme.luti2.lane.zt.x2.nxv8i16(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 7)
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt_x4.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt_x4.c
index 38059019737f8a..0730812b1f06f2 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt_x4.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti2_lane_zt_x4.c
@@ -37,6 +37,20 @@ svint8x4_t test_svluti2_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_in("zt0")
return svluti2_lane_zt_s8_x4(0, zn, 3);
}
+// CHECK-LABEL: @test_svluti2_lane_zt_mf8(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.luti2.lane.zt.x4.nxv16i8(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 3)
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z24test_svluti2_lane_zt_mf8u11__SVUint8_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.luti2.lane.zt.x4.nxv16i8(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 3)
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svluti2_lane_zt_mf8(svuint8_t zn) __arm_streaming __arm_in("zt0") {
+ return svluti2_lane_zt_mf8_x4(0, zn, 3);
+}
+
// CHECK-LABEL: @test_svluti2_lane_zt_u16(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sme.luti2.lane.zt.x4.nxv8i16(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 3)
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti4_lane_zt.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti4_lane_zt.c
index 16e66d1c212224..b687b580b15a64 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti4_lane_zt.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti4_lane_zt.c
@@ -38,6 +38,20 @@ svint8_t test_svluti4_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_in("zt0") {
return svluti4_lane_zt_s8(0, zn, 7);
}
+// CHECK-LABEL: @test_svluti4_lane_zt_mf8(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.luti4.lane.zt.nxv16i8(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 7)
+// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z24test_svluti4_lane_zt_mf8u11__SVUint8_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.luti4.lane.zt.nxv16i8(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 7)
+// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svluti4_lane_zt_mf8(svuint8_t zn) __arm_streaming __arm_in("zt0") {
+ return svluti4_lane_zt_mf8(0, zn, 7);
+}
+
// CHECK-LABEL: @test_svluti4_lane_zt_u16(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sme.luti4.lane.zt.nxv8i16(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 7)
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti4_lane_zt_x2.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti4_lane_zt_x2.c
index db615b3cd1c243..1a9e9d84c6359f 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti4_lane_zt_x2.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_luti4_lane_zt_x2.c
@@ -37,6 +37,20 @@ svint8x2_t test_svluti4_lane_zt_s8(svuint8_t zn) __arm_streaming __arm_in("zt0")
return svluti4_lane_zt_s8_x2(0, zn, 3);
}
+// CHECK-LABEL: @test_svluti4_lane_zt_mf8(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.luti4.lane.zt.x2.nxv16i8(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 3)
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z24test_svluti4_lane_zt_mf8u11__SVUint8_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.luti4.lane.zt.x2.nxv16i8(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 3)
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svluti4_lane_zt_mf8(svuint8_t zn) __arm_streaming __arm_in("zt0") {
+ return svluti4_lane_zt_mf8_x2(0, zn, 3);
+}
+
// CHECK-LABEL: @test_svluti4_lane_zt_u16(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sme.luti4.lane.zt.x2.nxv8i16(i32 0, <vscale x 16 x i8> [[ZN:%.*]], i32 3)
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_read.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_read.c
index b8cd1e1653ea98..c9d532d5fce45d 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_read.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_read.c
@@ -35,6 +35,20 @@ svint8x2_t test_svread_ver_za8_s8_vg2(uint32_t base) __arm_streaming __arm_in("z
return svread_ver_za8_s8_vg2(0, base);
}
+// CHECK-LABEL: @test_svread_ver_za8_mf8_vg2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.ver.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z27test_svread_ver_za8_mf8_vg2j(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.ver.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svread_ver_za8_mf8_vg2(uint32_t base) __arm_streaming __arm_in("za") {
+ return svread_ver_za8_mf8_vg2(0, base);
+}
+
// CHECK-LABEL: @test_svread_hor_za8_u8_vg2(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.hor.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]])
@@ -63,6 +77,20 @@ svint8x2_t test_svread_hor_za8_s8_vg2(uint32_t base) __arm_streaming __arm_in("z
return svread_hor_za8_s8_vg2(0, base);
}
+// CHECK-LABEL: @test_svread_hor_za8_mf8_vg2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.hor.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z27test_svread_hor_za8_mf8_vg2j(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.hor.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svread_hor_za8_mf8_vg2(uint32_t base) __arm_streaming __arm_in("za") {
+ return svread_hor_za8_mf8_vg2(0, base);
+}
+
// CHECK-LABEL: @test_svread_hor_za8_u8_vg4(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.hor.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]])
@@ -91,6 +119,20 @@ svint8x4_t test_svread_hor_za8_s8_vg4(uint32_t base) __arm_streaming __arm_in("z
return svread_hor_za8_s8_vg4(0, base);
}
+// CHECK-LABEL: @test_svread_hor_za8_mf8_vg4(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.hor.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z27test_svread_hor_za8_mf8_vg4j(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.hor.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svread_hor_za8_mf8_vg4(uint32_t base) __arm_streaming __arm_in("za") {
+ return svread_hor_za8_mf8_vg4(0, base);
+}
+
// CHECK-LABEL: @test_svread_ver_za8_u8_vg4(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.ver.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]])
@@ -119,6 +161,20 @@ svint8x4_t test_svread_ver_za8_s8_vg4(uint32_t base) __arm_streaming __arm_in("z
return svread_ver_za8_s8_vg4(0, base);
}
+// CHECK-LABEL: @test_svread_ver_za8_mf8_vg4(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.ver.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z27test_svread_ver_za8_mf8_vg4j(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.ver.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svread_ver_za8_mf8_vg4(uint32_t base) __arm_streaming __arm_in("za") {
+ return svread_ver_za8_mf8_vg4(0, base);
+}
+
// CHECK-LABEL: @test_svread_hor_za16_u16_vg2(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sme.read.hor.vg2.nxv8i16(i32 1, i32 [[BASE:%.*]])
@@ -581,7 +637,6 @@ svfloat64x2_t test_svread_ver_za64_f64_vg2(uint32_t base) __arm_streaming __arm_
return svread_ver_za64_f64_vg2(7, base);
}
-//
// CHECK-LABEL: @test_svread_ver_za64_s64_vg2(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sme.read.ver.vg2.nxv2i64(i32 7, i32 [[BASE:%.*]])
@@ -708,6 +763,19 @@ svuint8x2_t test_svread_za8_u8_vg1x2(uint32_t base) __arm_streaming __arm_in("za
return svread_za8_u8_vg1x2(base);
}
+// CHECK-LABEL: @test_svread_za8_mf8_vg1x2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.vg1x2.nxv16i8(i32 [[BASE:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z25test_svread_za8_mf8_vg1x2j(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.vg1x2.nxv16i8(i32 [[BASE:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svread_za8_mf8_vg1x2(uint32_t base) __arm_streaming __arm_in("za") {
+ return svread_za8_mf8_vg1x2(base);
+}
// CHECK-LABEL: @test_svread_za16_s16_vg1x2(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sme.read.vg1x2.nxv8i16(i32 [[BASE:%.*]])
@@ -876,6 +944,20 @@ svuint8x4_t test_svread_za8_u8_vg1x4(uint32_t base) __arm_streaming __arm_in("za
return svread_za8_u8_vg1x4(base);
}
+// CHECK-LABEL: @test_svread_za8_mf8_vg1x4(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.vg1x4.nxv16i8(i32 [[BASE:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z25test_svread_za8_mf8_vg1x4j(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.read.vg1x4.nxv16i8(i32 [[BASE:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svread_za8_mf8_vg1x4(uint32_t base) __arm_streaming __arm_in("za") {
+ return svread_za8_mf8_vg1x4(base);
+}
+
// CHECK-LABEL: @test_svread_za16_s16_vg1x4(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sme.read.vg1x4.nxv8i16(i32 [[BASE:%.*]])
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_selx2.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_selx2.c
index 4047b2fbd19652..893cc7519a1d0d 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_selx2.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_selx2.c
@@ -45,6 +45,19 @@ svuint8x2_t test_svsel_u8_x2(svcount_t pn, svuint8x2_t zn, svuint8x2_t zm) __arm
return SVE_ACLE_FUNC(svsel,_u8_x2)(pn, zn, zm);
}
+// CHECK-LABEL: @test_svsel_mf8_x2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.sel.x2.nxv16i8(target("aarch64.svcount") [[PN:%.*]], <vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]], <vscale x 16 x i8> [[ZM_COERCE0:%.*]], <vscale x 16 x i8> [[ZM_COERCE1:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svsel_mf8_x2u11__SVCount_t13svmfloat8x2_tS0_(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.sel.x2.nxv16i8(target("aarch64.svcount") [[PN:%.*]], <vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]], <vscale x 16 x i8> [[ZM_COERCE0:%.*]], <vscale x 16 x i8> [[ZM_COERCE1:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svsel_mf8_x2(svcount_t pn, svmfloat8x2_t zn, svmfloat8x2_t zm) __arm_streaming {
+ return SVE_ACLE_FUNC(svsel,_mf8_x2)(pn, zn, zm);
+}
// 16-bit SELs
// CHECK-LABEL: @test_svsel_s16_x2(
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_selx4.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_selx4.c
index 871d70943c9df4..d4e77d998e3c2a 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_selx4.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_selx4.c
@@ -33,29 +33,43 @@ svint8x4_t test_svsel_s8_x4(svcount_t pn, svint8x4_t zn1, svint8x4_t zn2) __arm_
// CHECK-LABEL: @test_svsel_u8_x4(
// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.sel.x4.nxv16i8(target("aarch64.svcount") [[PN:%.*]], <vscale x 16 x i8> [[ZN1_COERCE0:%.*]], <vscale x 16 x i8> [[ZN1_COERCE1:%.*]], <vscale x 16 x i8> [[ZN1_COERCE2:%.*]], <vscale x 16 x i8> [[ZN1_COERCE3:%.*]], <vscale x 16 x i8> [[ZN2_COERCE0:%.*]], <vscale x 16 x i8> [[ZN2_COERCE1:%.*]], <vscale x 16 x i8> [[ZN2_COERCE2:%.*]], <vscale x 16 x i8> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP2]]
+//
+// CPP-CHECK-LABEL: @_Z16test_svsel_u8_x4u11__SVCount_t11svuint8x4_tS0_S0_S0_(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.sel.x4.nxv16i8(target("aarch64.svcount") [[PN:%.*]], <vscale x 16 x i8> [[ZN1_COERCE0:%.*]], <vscale x 16 x i8> [[ZN1_COERCE1:%.*]], <vscale x 16 x i8> [[ZN1_COERCE2:%.*]], <vscale x 16 x i8> [[ZN1_COERCE3:%.*]], <vscale x 16 x i8> [[ZN2_COERCE0:%.*]], <vscale x 16 x i8> [[ZN2_COERCE1:%.*]], <vscale x 16 x i8> [[ZN2_COERCE2:%.*]], <vscale x 16 x i8> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP2]]
+//
+svuint8x4_t test_svsel_u8_x4(svcount_t pn, svuint8x4_t zn1, svuint8x4_t zn2, svuint8x4_t zn3, svuint8x4_t zn4) __arm_streaming {
+ return SVE_ACLE_FUNC(svsel,_u8_x4)(pn, zn1, zn2);
+}
+
+// CHECK-LABEL: @test_svsel_mf8_x4(
+// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.sel.x4.nxv16i8(target("aarch64.svcount") [[PN:%.*]], <vscale x 16 x i8> [[ZN1_COERCE0:%.*]], <vscale x 16 x i8> [[ZN1_COERCE1:%.*]], <vscale x 16 x i8> [[ZN1_COERCE2:%.*]], <vscale x 16 x i8> [[ZN1_COERCE3:%.*]], <vscale x 16 x i8> [[ZN2_COERCE0:%.*]], <vscale x 16 x i8> [[ZN2_COERCE1:%.*]], <vscale x 16 x i8> [[ZN2_COERCE2:%.*]], <vscale x 16 x i8> [[ZN2_COERCE3:%.*]])
// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
//
-// CPP-CHECK-LABEL: @_Z16test_svsel_u8_x4u11__SVCount_t11svuint8x4_tS0_S0_S0_(
+// CPP-CHECK-LABEL: @_Z17test_svsel_mf8_x4u11__SVCount_t13svmfloat8x4_tS0_(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.sel.x4.nxv16i8(target("aarch64.svcount") [[PN:%.*]], <vscale x 16 x i8> [[ZN1_COERCE0:%.*]], <vscale x 16 x i8> [[ZN1_COERCE1:%.*]], <vscale x 16 x i8> [[ZN1_COERCE2:%.*]], <vscale x 16 x i8> [[ZN1_COERCE3:%.*]], <vscale x 16 x i8> [[ZN2_COERCE0:%.*]], <vscale x 16 x i8> [[ZN2_COERCE1:%.*]], <vscale x 16 x i8> [[ZN2_COERCE2:%.*]], <vscale x 16 x i8> [[ZN2_COERCE3:%.*]])
// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
//
-svuint8x4_t test_svsel_u8_x4(svcount_t pn, svuint8x4_t zn1, svuint8x4_t zn2, svuint8x4_t zn3, svuint8x4_t zn4) __arm_streaming {
- return SVE_ACLE_FUNC(svsel,_u8_x4)(pn, zn1, zn2);
+svmfloat8x4_t test_svsel_mf8_x4(svcount_t pn, svmfloat8x4_t zn1, svmfloat8x4_t zn2) __arm_streaming {
+ return SVE_ACLE_FUNC(svsel,_mf8_x4)(pn, zn1, zn2);
}
// 16-bit SELs
// CHECK-LABEL: @test_svsel_s16_x4(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.sel.x4.nxv8i16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x i16> [[ZN1_COERCE0:%.*]], <vscale x 8 x i16> [[ZN1_COERCE1:%.*]], <vscale x 8 x i16> [[ZN1_COERCE2:%.*]], <vscale x 8 x i16> [[ZN1_COERCE3:%.*]], <vscale x 8 x i16> [[ZN2_COERCE0:%.*]], <vscale x 8 x i16> [[ZN2_COERCE1:%.*]], <vscale x 8 x i16> [[ZN2_COERCE2:%.*]], <vscale x 8 x i16> [[ZN2_COERCE3:%.*]])
-// CHECK-NEXT: ret { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP0]]
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.sel.x4.nxv8i16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x i16> [[ZN1_COERCE0:%.*]], <vscale x 8 x i16> [[ZN1_COERCE1:%.*]], <vscale x 8 x i16> [[ZN1_COERCE2:%.*]], <vscale x 8 x i16> [[ZN1_COERCE3:%.*]], <vscale x 8 x i16> [[ZN2_COERCE0:%.*]], <vscale x 8 x i16> [[ZN2_COERCE1:%.*]], <vscale x 8 x i16> [[ZN2_COERCE2:%.*]], <vscale x 8 x i16> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z17test_svsel_s16_x4u11__SVCount_t11svint16x4_tS0_S0_S0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.sel.x4.nxv8i16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x i16> [[ZN1_COERCE0:%.*]], <vscale x 8 x i16> [[ZN1_COERCE1:%.*]], <vscale x 8 x i16> [[ZN1_COERCE2:%.*]], <vscale x 8 x i16> [[ZN1_COERCE3:%.*]], <vscale x 8 x i16> [[ZN2_COERCE0:%.*]], <vscale x 8 x i16> [[ZN2_COERCE1:%.*]], <vscale x 8 x i16> [[ZN2_COERCE2:%.*]], <vscale x 8 x i16> [[ZN2_COERCE3:%.*]])
-// CPP-CHECK-NEXT: ret { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP0]]
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.sel.x4.nxv8i16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x i16> [[ZN1_COERCE0:%.*]], <vscale x 8 x i16> [[ZN1_COERCE1:%.*]], <vscale x 8 x i16> [[ZN1_COERCE2:%.*]], <vscale x 8 x i16> [[ZN1_COERCE3:%.*]], <vscale x 8 x i16> [[ZN2_COERCE0:%.*]], <vscale x 8 x i16> [[ZN2_COERCE1:%.*]], <vscale x 8 x i16> [[ZN2_COERCE2:%.*]], <vscale x 8 x i16> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]]
//
svint16x4_t test_svsel_s16_x4(svcount_t pn, svint16x4_t zn1, svint16x4_t zn2, svint16x4_t zn3, svint16x4_t zn4) __arm_streaming {
return SVE_ACLE_FUNC(svsel,_s16_x4)(pn, zn1, zn2);
@@ -63,13 +77,13 @@ svint16x4_t test_svsel_s16_x4(svcount_t pn, svint16x4_t zn1, svint16x4_t zn2, sv
// CHECK-LABEL: @test_svsel_u16_x4(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.sel.x4.nxv8i16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x i16> [[ZN1_COERCE0:%.*]], <vscale x 8 x i16> [[ZN1_COERCE1:%.*]], <vscale x 8 x i16> [[ZN1_COERCE2:%.*]], <vscale x 8 x i16> [[ZN1_COERCE3:%.*]], <vscale x 8 x i16> [[ZN2_COERCE0:%.*]], <vscale x 8 x i16> [[ZN2_COERCE1:%.*]], <vscale x 8 x i16> [[ZN2_COERCE2:%.*]], <vscale x 8 x i16> [[ZN2_COERCE3:%.*]])
-// CHECK-NEXT: ret { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP0]]
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.sel.x4.nxv8i16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x i16> [[ZN1_COERCE0:%.*]], <vscale x 8 x i16> [[ZN1_COERCE1:%.*]], <vscale x 8 x i16> [[ZN1_COERCE2:%.*]], <vscale x 8 x i16> [[ZN1_COERCE3:%.*]], <vscale x 8 x i16> [[ZN2_COERCE0:%.*]], <vscale x 8 x i16> [[ZN2_COERCE1:%.*]], <vscale x 8 x i16> [[ZN2_COERCE2:%.*]], <vscale x 8 x i16> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z17test_svsel_u16_x4u11__SVCount_t12svuint16x4_tS0_S0_S0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.sel.x4.nxv8i16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x i16> [[ZN1_COERCE0:%.*]], <vscale x 8 x i16> [[ZN1_COERCE1:%.*]], <vscale x 8 x i16> [[ZN1_COERCE2:%.*]], <vscale x 8 x i16> [[ZN1_COERCE3:%.*]], <vscale x 8 x i16> [[ZN2_COERCE0:%.*]], <vscale x 8 x i16> [[ZN2_COERCE1:%.*]], <vscale x 8 x i16> [[ZN2_COERCE2:%.*]], <vscale x 8 x i16> [[ZN2_COERCE3:%.*]])
-// CPP-CHECK-NEXT: ret { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP0]]
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.sel.x4.nxv8i16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x i16> [[ZN1_COERCE0:%.*]], <vscale x 8 x i16> [[ZN1_COERCE1:%.*]], <vscale x 8 x i16> [[ZN1_COERCE2:%.*]], <vscale x 8 x i16> [[ZN1_COERCE3:%.*]], <vscale x 8 x i16> [[ZN2_COERCE0:%.*]], <vscale x 8 x i16> [[ZN2_COERCE1:%.*]], <vscale x 8 x i16> [[ZN2_COERCE2:%.*]], <vscale x 8 x i16> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]]
//
svuint16x4_t test_svsel_u16_x4(svcount_t pn, svuint16x4_t zn1, svuint16x4_t zn2, svuint16x4_t zn3, svuint16x4_t zn4) __arm_streaming {
return SVE_ACLE_FUNC(svsel,_u16_x4)(pn, zn1, zn2);
@@ -77,13 +91,13 @@ svuint16x4_t test_svsel_u16_x4(svcount_t pn, svuint16x4_t zn1, svuint16x4_t zn2,
// CHECK-LABEL: @test_svsel_f16_x4(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } @llvm.aarch64.sve.sel.x4.nxv8f16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x half> [[ZN1_COERCE0:%.*]], <vscale x 8 x half> [[ZN1_COERCE1:%.*]], <vscale x 8 x half> [[ZN1_COERCE2:%.*]], <vscale x 8 x half> [[ZN1_COERCE3:%.*]], <vscale x 8 x half> [[ZN2_COERCE0:%.*]], <vscale x 8 x half> [[ZN2_COERCE1:%.*]], <vscale x 8 x half> [[ZN2_COERCE2:%.*]], <vscale x 8 x half> [[ZN2_COERCE3:%.*]])
-// CHECK-NEXT: ret { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } [[TMP0]]
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } @llvm.aarch64.sve.sel.x4.nxv8f16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x half> [[ZN1_COERCE0:%.*]], <vscale x 8 x half> [[ZN1_COERCE1:%.*]], <vscale x 8 x half> [[ZN1_COERCE2:%.*]], <vscale x 8 x half> [[ZN1_COERCE3:%.*]], <vscale x 8 x half> [[ZN2_COERCE0:%.*]], <vscale x 8 x half> [[ZN2_COERCE1:%.*]], <vscale x 8 x half> [[ZN2_COERCE2:%.*]], <vscale x 8 x half> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z17test_svsel_f16_x4u11__SVCount_t13svfloat16x4_tS0_S0_S0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } @llvm.aarch64.sve.sel.x4.nxv8f16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x half> [[ZN1_COERCE0:%.*]], <vscale x 8 x half> [[ZN1_COERCE1:%.*]], <vscale x 8 x half> [[ZN1_COERCE2:%.*]], <vscale x 8 x half> [[ZN1_COERCE3:%.*]], <vscale x 8 x half> [[ZN2_COERCE0:%.*]], <vscale x 8 x half> [[ZN2_COERCE1:%.*]], <vscale x 8 x half> [[ZN2_COERCE2:%.*]], <vscale x 8 x half> [[ZN2_COERCE3:%.*]])
-// CPP-CHECK-NEXT: ret { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } [[TMP0]]
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } @llvm.aarch64.sve.sel.x4.nxv8f16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x half> [[ZN1_COERCE0:%.*]], <vscale x 8 x half> [[ZN1_COERCE1:%.*]], <vscale x 8 x half> [[ZN1_COERCE2:%.*]], <vscale x 8 x half> [[ZN1_COERCE3:%.*]], <vscale x 8 x half> [[ZN2_COERCE0:%.*]], <vscale x 8 x half> [[ZN2_COERCE1:%.*]], <vscale x 8 x half> [[ZN2_COERCE2:%.*]], <vscale x 8 x half> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half> } [[TMP2]]
//
svfloat16x4_t test_svsel_f16_x4(svcount_t pn, svfloat16x4_t zn1, svfloat16x4_t zn2, svfloat16x4_t zn3, svfloat16x4_t zn4) __arm_streaming {
return SVE_ACLE_FUNC(svsel,_f16_x4)(pn, zn1, zn2);
@@ -91,13 +105,13 @@ svfloat16x4_t test_svsel_f16_x4(svcount_t pn, svfloat16x4_t zn1, svfloat16x4_t z
// CHECK-LABEL: @test_svsel_bf16_x4(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @llvm.aarch64.sve.sel.x4.nxv8bf16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE0:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE1:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE2:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE3:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE0:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE1:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE2:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE3:%.*]])
-// CHECK-NEXT: ret { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } [[TMP0]]
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @llvm.aarch64.sve.sel.x4.nxv8bf16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE0:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE1:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE2:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE3:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE0:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE1:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE2:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z18test_svsel_bf16_x4u11__SVCount_t14svbfloat16x4_tS0_S0_S0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @llvm.aarch64.sve.sel.x4.nxv8bf16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE0:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE1:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE2:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE3:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE0:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE1:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE2:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE3:%.*]])
-// CPP-CHECK-NEXT: ret { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } [[TMP0]]
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } @llvm.aarch64.sve.sel.x4.nxv8bf16(target("aarch64.svcount") [[PN:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE0:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE1:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE2:%.*]], <vscale x 8 x bfloat> [[ZN1_COERCE3:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE0:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE1:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE2:%.*]], <vscale x 8 x bfloat> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat>, <vscale x 8 x bfloat> } [[TMP2]]
//
svbfloat16x4_t test_svsel_bf16_x4(svcount_t pn, svbfloat16x4_t zn1, svbfloat16x4_t zn2, svbfloat16x4_t zn3, svbfloat16x4_t zn4) __arm_streaming {
return SVE_ACLE_FUNC(svsel,_bf16_x4)(pn, zn1, zn2);
@@ -107,13 +121,13 @@ svbfloat16x4_t test_svsel_bf16_x4(svcount_t pn, svbfloat16x4_t zn1, svbfloat16x4
// CHECK-LABEL: @test_svsel_s32_x4(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.sel.x4.nxv4i32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x i32> [[ZN1_COERCE0:%.*]], <vscale x 4 x i32> [[ZN1_COERCE1:%.*]], <vscale x 4 x i32> [[ZN1_COERCE2:%.*]], <vscale x 4 x i32> [[ZN1_COERCE3:%.*]], <vscale x 4 x i32> [[ZN2_COERCE0:%.*]], <vscale x 4 x i32> [[ZN2_COERCE1:%.*]], <vscale x 4 x i32> [[ZN2_COERCE2:%.*]], <vscale x 4 x i32> [[ZN2_COERCE3:%.*]])
-// CHECK-NEXT: ret { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP0]]
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.sel.x4.nxv4i32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x i32> [[ZN1_COERCE0:%.*]], <vscale x 4 x i32> [[ZN1_COERCE1:%.*]], <vscale x 4 x i32> [[ZN1_COERCE2:%.*]], <vscale x 4 x i32> [[ZN1_COERCE3:%.*]], <vscale x 4 x i32> [[ZN2_COERCE0:%.*]], <vscale x 4 x i32> [[ZN2_COERCE1:%.*]], <vscale x 4 x i32> [[ZN2_COERCE2:%.*]], <vscale x 4 x i32> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z17test_svsel_s32_x4u11__SVCount_t11svint32x4_tS0_S0_S0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.sel.x4.nxv4i32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x i32> [[ZN1_COERCE0:%.*]], <vscale x 4 x i32> [[ZN1_COERCE1:%.*]], <vscale x 4 x i32> [[ZN1_COERCE2:%.*]], <vscale x 4 x i32> [[ZN1_COERCE3:%.*]], <vscale x 4 x i32> [[ZN2_COERCE0:%.*]], <vscale x 4 x i32> [[ZN2_COERCE1:%.*]], <vscale x 4 x i32> [[ZN2_COERCE2:%.*]], <vscale x 4 x i32> [[ZN2_COERCE3:%.*]])
-// CPP-CHECK-NEXT: ret { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP0]]
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.sel.x4.nxv4i32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x i32> [[ZN1_COERCE0:%.*]], <vscale x 4 x i32> [[ZN1_COERCE1:%.*]], <vscale x 4 x i32> [[ZN1_COERCE2:%.*]], <vscale x 4 x i32> [[ZN1_COERCE3:%.*]], <vscale x 4 x i32> [[ZN2_COERCE0:%.*]], <vscale x 4 x i32> [[ZN2_COERCE1:%.*]], <vscale x 4 x i32> [[ZN2_COERCE2:%.*]], <vscale x 4 x i32> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]]
//
svint32x4_t test_svsel_s32_x4(svcount_t pn, svint32x4_t zn1, svint32x4_t zn2, svint32x4_t zn3, svint32x4_t zn4) __arm_streaming {
return SVE_ACLE_FUNC(svsel,_s32_x4)(pn, zn1, zn2);
@@ -121,13 +135,13 @@ svint32x4_t test_svsel_s32_x4(svcount_t pn, svint32x4_t zn1, svint32x4_t zn2, sv
// CHECK-LABEL: @test_svsel_u32_x4(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.sel.x4.nxv4i32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x i32> [[ZN1_COERCE0:%.*]], <vscale x 4 x i32> [[ZN1_COERCE1:%.*]], <vscale x 4 x i32> [[ZN1_COERCE2:%.*]], <vscale x 4 x i32> [[ZN1_COERCE3:%.*]], <vscale x 4 x i32> [[ZN2_COERCE0:%.*]], <vscale x 4 x i32> [[ZN2_COERCE1:%.*]], <vscale x 4 x i32> [[ZN2_COERCE2:%.*]], <vscale x 4 x i32> [[ZN2_COERCE3:%.*]])
-// CHECK-NEXT: ret { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP0]]
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.sel.x4.nxv4i32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x i32> [[ZN1_COERCE0:%.*]], <vscale x 4 x i32> [[ZN1_COERCE1:%.*]], <vscale x 4 x i32> [[ZN1_COERCE2:%.*]], <vscale x 4 x i32> [[ZN1_COERCE3:%.*]], <vscale x 4 x i32> [[ZN2_COERCE0:%.*]], <vscale x 4 x i32> [[ZN2_COERCE1:%.*]], <vscale x 4 x i32> [[ZN2_COERCE2:%.*]], <vscale x 4 x i32> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z17test_svsel_u32_x4u11__SVCount_t12svuint32x4_tS0_S0_S0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.sel.x4.nxv4i32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x i32> [[ZN1_COERCE0:%.*]], <vscale x 4 x i32> [[ZN1_COERCE1:%.*]], <vscale x 4 x i32> [[ZN1_COERCE2:%.*]], <vscale x 4 x i32> [[ZN1_COERCE3:%.*]], <vscale x 4 x i32> [[ZN2_COERCE0:%.*]], <vscale x 4 x i32> [[ZN2_COERCE1:%.*]], <vscale x 4 x i32> [[ZN2_COERCE2:%.*]], <vscale x 4 x i32> [[ZN2_COERCE3:%.*]])
-// CPP-CHECK-NEXT: ret { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP0]]
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.sel.x4.nxv4i32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x i32> [[ZN1_COERCE0:%.*]], <vscale x 4 x i32> [[ZN1_COERCE1:%.*]], <vscale x 4 x i32> [[ZN1_COERCE2:%.*]], <vscale x 4 x i32> [[ZN1_COERCE3:%.*]], <vscale x 4 x i32> [[ZN2_COERCE0:%.*]], <vscale x 4 x i32> [[ZN2_COERCE1:%.*]], <vscale x 4 x i32> [[ZN2_COERCE2:%.*]], <vscale x 4 x i32> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]]
//
svuint32x4_t test_svsel_u32_x4(svcount_t pn, svuint32x4_t zn1, svuint32x4_t zn2, svuint32x4_t zn3, svuint32x4_t zn4) __arm_streaming {
return SVE_ACLE_FUNC(svsel,_u32_x4)(pn, zn1, zn2);
@@ -135,13 +149,13 @@ svuint32x4_t test_svsel_u32_x4(svcount_t pn, svuint32x4_t zn1, svuint32x4_t zn2,
// CHECK-LABEL: @test_svsel_f32_x4(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } @llvm.aarch64.sve.sel.x4.nxv4f32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x float> [[ZN1_COERCE0:%.*]], <vscale x 4 x float> [[ZN1_COERCE1:%.*]], <vscale x 4 x float> [[ZN1_COERCE2:%.*]], <vscale x 4 x float> [[ZN1_COERCE3:%.*]], <vscale x 4 x float> [[ZN2_COERCE0:%.*]], <vscale x 4 x float> [[ZN2_COERCE1:%.*]], <vscale x 4 x float> [[ZN2_COERCE2:%.*]], <vscale x 4 x float> [[ZN2_COERCE3:%.*]])
-// CHECK-NEXT: ret { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } [[TMP0]]
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } @llvm.aarch64.sve.sel.x4.nxv4f32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x float> [[ZN1_COERCE0:%.*]], <vscale x 4 x float> [[ZN1_COERCE1:%.*]], <vscale x 4 x float> [[ZN1_COERCE2:%.*]], <vscale x 4 x float> [[ZN1_COERCE3:%.*]], <vscale x 4 x float> [[ZN2_COERCE0:%.*]], <vscale x 4 x float> [[ZN2_COERCE1:%.*]], <vscale x 4 x float> [[ZN2_COERCE2:%.*]], <vscale x 4 x float> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z17test_svsel_f32_x4u11__SVCount_t13svfloat32x4_tS0_S0_S0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } @llvm.aarch64.sve.sel.x4.nxv4f32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x float> [[ZN1_COERCE0:%.*]], <vscale x 4 x float> [[ZN1_COERCE1:%.*]], <vscale x 4 x float> [[ZN1_COERCE2:%.*]], <vscale x 4 x float> [[ZN1_COERCE3:%.*]], <vscale x 4 x float> [[ZN2_COERCE0:%.*]], <vscale x 4 x float> [[ZN2_COERCE1:%.*]], <vscale x 4 x float> [[ZN2_COERCE2:%.*]], <vscale x 4 x float> [[ZN2_COERCE3:%.*]])
-// CPP-CHECK-NEXT: ret { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } [[TMP0]]
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } @llvm.aarch64.sve.sel.x4.nxv4f32(target("aarch64.svcount") [[PN:%.*]], <vscale x 4 x float> [[ZN1_COERCE0:%.*]], <vscale x 4 x float> [[ZN1_COERCE1:%.*]], <vscale x 4 x float> [[ZN1_COERCE2:%.*]], <vscale x 4 x float> [[ZN1_COERCE3:%.*]], <vscale x 4 x float> [[ZN2_COERCE0:%.*]], <vscale x 4 x float> [[ZN2_COERCE1:%.*]], <vscale x 4 x float> [[ZN2_COERCE2:%.*]], <vscale x 4 x float> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float> } [[TMP2]]
//
svfloat32x4_t test_svsel_f32_x4(svcount_t pn, svfloat32x4_t zn1, svfloat32x4_t zn2, svfloat32x4_t zn3, svfloat32x4_t zn4) __arm_streaming {
return SVE_ACLE_FUNC(svsel,_f32_x4)(pn, zn1, zn2);
@@ -151,13 +165,13 @@ svfloat32x4_t test_svsel_f32_x4(svcount_t pn, svfloat32x4_t zn1, svfloat32x4_t z
// CHECK-LABEL: @test_svsel_s64_x4(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.sel.x4.nxv2i64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x i64> [[ZN1_COERCE0:%.*]], <vscale x 2 x i64> [[ZN1_COERCE1:%.*]], <vscale x 2 x i64> [[ZN1_COERCE2:%.*]], <vscale x 2 x i64> [[ZN1_COERCE3:%.*]], <vscale x 2 x i64> [[ZN2_COERCE0:%.*]], <vscale x 2 x i64> [[ZN2_COERCE1:%.*]], <vscale x 2 x i64> [[ZN2_COERCE2:%.*]], <vscale x 2 x i64> [[ZN2_COERCE3:%.*]])
-// CHECK-NEXT: ret { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.sel.x4.nxv2i64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x i64> [[ZN1_COERCE0:%.*]], <vscale x 2 x i64> [[ZN1_COERCE1:%.*]], <vscale x 2 x i64> [[ZN1_COERCE2:%.*]], <vscale x 2 x i64> [[ZN1_COERCE3:%.*]], <vscale x 2 x i64> [[ZN2_COERCE0:%.*]], <vscale x 2 x i64> [[ZN2_COERCE1:%.*]], <vscale x 2 x i64> [[ZN2_COERCE2:%.*]], <vscale x 2 x i64> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z17test_svsel_s64_x4u11__SVCount_t11svint64x4_tS0_S0_S0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.sel.x4.nxv2i64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x i64> [[ZN1_COERCE0:%.*]], <vscale x 2 x i64> [[ZN1_COERCE1:%.*]], <vscale x 2 x i64> [[ZN1_COERCE2:%.*]], <vscale x 2 x i64> [[ZN1_COERCE3:%.*]], <vscale x 2 x i64> [[ZN2_COERCE0:%.*]], <vscale x 2 x i64> [[ZN2_COERCE1:%.*]], <vscale x 2 x i64> [[ZN2_COERCE2:%.*]], <vscale x 2 x i64> [[ZN2_COERCE3:%.*]])
-// CPP-CHECK-NEXT: ret { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.sel.x4.nxv2i64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x i64> [[ZN1_COERCE0:%.*]], <vscale x 2 x i64> [[ZN1_COERCE1:%.*]], <vscale x 2 x i64> [[ZN1_COERCE2:%.*]], <vscale x 2 x i64> [[ZN1_COERCE3:%.*]], <vscale x 2 x i64> [[ZN2_COERCE0:%.*]], <vscale x 2 x i64> [[ZN2_COERCE1:%.*]], <vscale x 2 x i64> [[ZN2_COERCE2:%.*]], <vscale x 2 x i64> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP2]]
//
svint64x4_t test_svsel_s64_x4(svcount_t pn, svint64x4_t zn1, svint64x4_t zn2, svint64x4_t zn3, svint64x4_t zn4) __arm_streaming {
return SVE_ACLE_FUNC(svsel,_s64_x4)(pn, zn1, zn2);
@@ -165,13 +179,13 @@ svint64x4_t test_svsel_s64_x4(svcount_t pn, svint64x4_t zn1, svint64x4_t zn2, sv
// CHECK-LABEL: @test_svsel_u64_x4(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.sel.x4.nxv2i64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x i64> [[ZN1_COERCE0:%.*]], <vscale x 2 x i64> [[ZN1_COERCE1:%.*]], <vscale x 2 x i64> [[ZN1_COERCE2:%.*]], <vscale x 2 x i64> [[ZN1_COERCE3:%.*]], <vscale x 2 x i64> [[ZN2_COERCE0:%.*]], <vscale x 2 x i64> [[ZN2_COERCE1:%.*]], <vscale x 2 x i64> [[ZN2_COERCE2:%.*]], <vscale x 2 x i64> [[ZN2_COERCE3:%.*]])
-// CHECK-NEXT: ret { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.sel.x4.nxv2i64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x i64> [[ZN1_COERCE0:%.*]], <vscale x 2 x i64> [[ZN1_COERCE1:%.*]], <vscale x 2 x i64> [[ZN1_COERCE2:%.*]], <vscale x 2 x i64> [[ZN1_COERCE3:%.*]], <vscale x 2 x i64> [[ZN2_COERCE0:%.*]], <vscale x 2 x i64> [[ZN2_COERCE1:%.*]], <vscale x 2 x i64> [[ZN2_COERCE2:%.*]], <vscale x 2 x i64> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z17test_svsel_u64_x4u11__SVCount_t12svuint64x4_tS0_S0_S0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.sel.x4.nxv2i64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x i64> [[ZN1_COERCE0:%.*]], <vscale x 2 x i64> [[ZN1_COERCE1:%.*]], <vscale x 2 x i64> [[ZN1_COERCE2:%.*]], <vscale x 2 x i64> [[ZN1_COERCE3:%.*]], <vscale x 2 x i64> [[ZN2_COERCE0:%.*]], <vscale x 2 x i64> [[ZN2_COERCE1:%.*]], <vscale x 2 x i64> [[ZN2_COERCE2:%.*]], <vscale x 2 x i64> [[ZN2_COERCE3:%.*]])
-// CPP-CHECK-NEXT: ret { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.sel.x4.nxv2i64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x i64> [[ZN1_COERCE0:%.*]], <vscale x 2 x i64> [[ZN1_COERCE1:%.*]], <vscale x 2 x i64> [[ZN1_COERCE2:%.*]], <vscale x 2 x i64> [[ZN1_COERCE3:%.*]], <vscale x 2 x i64> [[ZN2_COERCE0:%.*]], <vscale x 2 x i64> [[ZN2_COERCE1:%.*]], <vscale x 2 x i64> [[ZN2_COERCE2:%.*]], <vscale x 2 x i64> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP2]]
//
svuint64x4_t test_svsel_u64_x4(svcount_t pn, svuint64x4_t zn1, svuint64x4_t zn2, svuint64x4_t zn3, svuint64x4_t zn4) __arm_streaming {
return SVE_ACLE_FUNC(svsel,_u64_x4)(pn, zn1, zn2);
@@ -179,13 +193,13 @@ svuint64x4_t test_svsel_u64_x4(svcount_t pn, svuint64x4_t zn1, svuint64x4_t zn2,
// CHECK-LABEL: @test_svsel_f64_x4(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } @llvm.aarch64.sve.sel.x4.nxv2f64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x double> [[ZN1_COERCE0:%.*]], <vscale x 2 x double> [[ZN1_COERCE1:%.*]], <vscale x 2 x double> [[ZN1_COERCE2:%.*]], <vscale x 2 x double> [[ZN1_COERCE3:%.*]], <vscale x 2 x double> [[ZN2_COERCE0:%.*]], <vscale x 2 x double> [[ZN2_COERCE1:%.*]], <vscale x 2 x double> [[ZN2_COERCE2:%.*]], <vscale x 2 x double> [[ZN2_COERCE3:%.*]])
-// CHECK-NEXT: ret { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } [[TMP0]]
+// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } @llvm.aarch64.sve.sel.x4.nxv2f64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x double> [[ZN1_COERCE0:%.*]], <vscale x 2 x double> [[ZN1_COERCE1:%.*]], <vscale x 2 x double> [[ZN1_COERCE2:%.*]], <vscale x 2 x double> [[ZN1_COERCE3:%.*]], <vscale x 2 x double> [[ZN2_COERCE0:%.*]], <vscale x 2 x double> [[ZN2_COERCE1:%.*]], <vscale x 2 x double> [[ZN2_COERCE2:%.*]], <vscale x 2 x double> [[ZN2_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z17test_svsel_f64_x4u11__SVCount_t13svfloat64x4_tS0_S0_S0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } @llvm.aarch64.sve.sel.x4.nxv2f64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x double> [[ZN1_COERCE0:%.*]], <vscale x 2 x double> [[ZN1_COERCE1:%.*]], <vscale x 2 x double> [[ZN1_COERCE2:%.*]], <vscale x 2 x double> [[ZN1_COERCE3:%.*]], <vscale x 2 x double> [[ZN2_COERCE0:%.*]], <vscale x 2 x double> [[ZN2_COERCE1:%.*]], <vscale x 2 x double> [[ZN2_COERCE2:%.*]], <vscale x 2 x double> [[ZN2_COERCE3:%.*]])
-// CPP-CHECK-NEXT: ret { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } [[TMP0]]
+// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } @llvm.aarch64.sve.sel.x4.nxv2f64(target("aarch64.svcount") [[PN:%.*]], <vscale x 2 x double> [[ZN1_COERCE0:%.*]], <vscale x 2 x double> [[ZN1_COERCE1:%.*]], <vscale x 2 x double> [[ZN1_COERCE2:%.*]], <vscale x 2 x double> [[ZN1_COERCE3:%.*]], <vscale x 2 x double> [[ZN2_COERCE0:%.*]], <vscale x 2 x double> [[ZN2_COERCE1:%.*]], <vscale x 2 x double> [[ZN2_COERCE2:%.*]], <vscale x 2 x double> [[ZN2_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double> } [[TMP2]]
//
svfloat64x4_t test_svsel_f64_x4(svcount_t pn, svfloat64x4_t zn1, svfloat64x4_t zn2, svfloat64x4_t zn3, svfloat64x4_t zn4) __arm_streaming {
return SVE_ACLE_FUNC(svsel,_f64_x4)(pn, zn1, zn2);
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_uzpx2.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_uzpx2.c
index 9a66ee5262082a..ec3a2952b2ac6a 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_uzpx2.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_uzpx2.c
@@ -46,6 +46,20 @@ svuint8x2_t test_svuzp_u8_x2(svuint8x2_t zn) __arm_streaming {
return SVE_ACLE_FUNC(svuzp,_u8_x2)(zn);
}
+// CHECK-LABEL: @test_svuzp_mf8_x2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.uzp.x2.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svuzp_mf8_x213svmfloat8x2_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.uzp.x2.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svuzp_mf8_x2(svmfloat8x2_t zn) __arm_streaming {
+ return SVE_ACLE_FUNC(svuzp,_mf8_x2)(zn);
+}
+
// 16-bit UZPs
// CHECK-LABEL: @test_svuzp_s16_x2(
@@ -222,6 +236,20 @@ svuint8x2_t test_svuzpq_u8_x2(svuint8x2_t zn) __arm_streaming {
return SVE_ACLE_FUNC(svuzpq,_u8_x2)(zn);
}
+// CHECK-LABEL: @test_svuzpq_mf8_x2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.uzpq.x2.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z18test_svuzpq_mf8_x213svmfloat8x2_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.uzpq.x2.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svuzpq_mf8_x2(svmfloat8x2_t zn) __arm_streaming {
+ return SVE_ACLE_FUNC(svuzpq,_mf8_x2)(zn);
+}
+
// CHECK-LABEL: @test_svuzpq_s16_x2(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.uzpq.x2.nxv8i16(<vscale x 8 x i16> [[ZN_COERCE0:%.*]], <vscale x 8 x i16> [[ZN_COERCE1:%.*]])
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_uzpx4.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_uzpx4.c
index 131928615edcd7..aeac2ae78f6e6c 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_uzpx4.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_uzpx4.c
@@ -46,6 +46,20 @@ svuint8x4_t test_svuzp_u8_x4(svuint8x4_t zn) __arm_streaming {
return SVE_ACLE_FUNC(svuzp,_u8_x4)(zn);
}
+// CHECK-LABEL: @test_svuzp_mf8_x4(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.uzp.x4.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]], <vscale x 16 x i8> [[ZN_COERCE2:%.*]], <vscale x 16 x i8> [[ZN_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svuzp_mf8_x413svmfloat8x4_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.uzp.x4.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]], <vscale x 16 x i8> [[ZN_COERCE2:%.*]], <vscale x 16 x i8> [[ZN_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svuzp_mf8_x4(svmfloat8x4_t zn) __arm_streaming {
+ return SVE_ACLE_FUNC(svuzp,_mf8_x4)(zn);
+}
+
// 16-bit UZPs
// CHECK-LABEL: @test_svuzp_s16_x4(
@@ -222,6 +236,20 @@ svuint8x4_t test_svuzpq_u8_x4(svuint8x4_t zn) __arm_streaming {
return SVE_ACLE_FUNC(svuzpq,_u8_x4)(zn);
}
+// CHECK-LABEL: @test_svuzpq_mf8_x4(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.uzpq.x4.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]], <vscale x 16 x i8> [[ZN_COERCE2:%.*]], <vscale x 16 x i8> [[ZN_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z18test_svuzpq_mf8_x413svmfloat8x4_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.uzpq.x4.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]], <vscale x 16 x i8> [[ZN_COERCE2:%.*]], <vscale x 16 x i8> [[ZN_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svuzpq_mf8_x4(svmfloat8x4_t zn) __arm_streaming {
+ return SVE_ACLE_FUNC(svuzpq,_mf8_x4)(zn);
+}
+
// CHECK-LABEL: @test_svuzpq_s16_x4(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.uzpq.x4.nxv8i16(<vscale x 8 x i16> [[ZN_COERCE0:%.*]], <vscale x 8 x i16> [[ZN_COERCE1:%.*]], <vscale x 8 x i16> [[ZN_COERCE2:%.*]], <vscale x 8 x i16> [[ZN_COERCE3:%.*]])
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_zipx2.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_zipx2.c
index 787b7d0b3ea1a0..735b3697f150b3 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_zipx2.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_zipx2.c
@@ -45,6 +45,20 @@ svuint8x2_t test_svzip_u8_x2(svuint8x2_t zn) __arm_streaming {
return SVE_ACLE_FUNC(svzip,_u8_x2)(zn);
}
+// CHECK-LABEL: @test_svzip_mf8_x2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.zip.x2.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svzip_mf8_x213svmfloat8x2_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.zip.x2.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svzip_mf8_x2(svmfloat8x2_t zn) __arm_streaming {
+ return SVE_ACLE_FUNC(svzip,_mf8_x2)(zn);
+}
+
// 16-bit ZIPs
// CHECK-LABEL: @test_svzip_s16_x2(
@@ -221,6 +235,20 @@ svuint8x2_t test_svzipq_u8_x2(svuint8x2_t zn) __arm_streaming {
return SVE_ACLE_FUNC(svzipq,_u8_x2)(zn);
}
+// CHECK-LABEL: @test_svzipq_mf8_x2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.zipq.x2.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z18test_svzipq_mf8_x213svmfloat8x2_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.zipq.x2.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svzipq_mf8_x2(svmfloat8x2_t zn) __arm_streaming {
+ return SVE_ACLE_FUNC(svzipq,_mf8_x2)(zn);
+}
+
// CHECK-LABEL: @test_svzipq_s16_x2(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.zipq.x2.nxv8i16(<vscale x 8 x i16> [[ZN_COERCE0:%.*]], <vscale x 8 x i16> [[ZN_COERCE1:%.*]])
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_zipx4.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_zipx4.c
index 9bea471bc98375..341ae290e9b0e9 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_zipx4.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_vector_zipx4.c
@@ -45,6 +45,20 @@ svuint8x4_t test_svzip_u8_x4(svuint8x4_t zn) __arm_streaming {
return SVE_ACLE_FUNC(svzip,_u8_x4)(zn);
}
+// CHECK-LABEL: @test_svzip_mf8_x4(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.zip.x4.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]], <vscale x 16 x i8> [[ZN_COERCE2:%.*]], <vscale x 16 x i8> [[ZN_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svzip_mf8_x413svmfloat8x4_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.zip.x4.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]], <vscale x 16 x i8> [[ZN_COERCE2:%.*]], <vscale x 16 x i8> [[ZN_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svzip_mf8_x4(svmfloat8x4_t zn) __arm_streaming {
+ return SVE_ACLE_FUNC(svzip,_mf8_x4)(zn);
+}
+
// 16-bit ZIPs
// CHECK-LABEL: @test_svzip_s16_x4(
@@ -221,6 +235,20 @@ svuint8x4_t test_svzipq_u8_x4(svuint8x4_t zn) __arm_streaming {
return SVE_ACLE_FUNC(svzipq,_u8_x4)(zn);
}
+// CHECK-LABEL: @test_svzipq_mf8_x4(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.zipq.x4.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]], <vscale x 16 x i8> [[ZN_COERCE2:%.*]], <vscale x 16 x i8> [[ZN_COERCE3:%.*]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z18test_svzipq_mf8_x413svmfloat8x4_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.zipq.x4.nxv16i8(<vscale x 16 x i8> [[ZN_COERCE0:%.*]], <vscale x 16 x i8> [[ZN_COERCE1:%.*]], <vscale x 16 x i8> [[ZN_COERCE2:%.*]], <vscale x 16 x i8> [[ZN_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svzipq_mf8_x4(svmfloat8x4_t zn) __arm_streaming {
+ return SVE_ACLE_FUNC(svzipq,_mf8_x4)(zn);
+}
+
// CHECK-LABEL: @test_svzipq_s16_x4(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.zipq.x4.nxv8i16(<vscale x 8 x i16> [[ZN_COERCE0:%.*]], <vscale x 8 x i16> [[ZN_COERCE1:%.*]], <vscale x 8 x i16> [[ZN_COERCE2:%.*]], <vscale x 8 x i16> [[ZN_COERCE3:%.*]])
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_write.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_write.c
index 1bdca4a12bcbd9..73aaaf66fd4b1b 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_write.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_write.c
@@ -44,6 +44,20 @@ void test_svwrite_ver_za8_s8_vg2(uint32_t base, svint8x2_t val) __arm_streaming
SVE_ACLE_FUNC(svwrite_ver_za8,_s8,_vg2,)(0, base, val);
}
+// CHECK-LABEL: @test_svwrite_ver_za8_mf8_vg2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.ver.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]])
+// CHECK-NEXT: ret void
+//
+// CPP-CHECK-LABEL: @_Z28test_svwrite_ver_za8_mf8_vg2j13svmfloat8x2_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: tail call void @llvm.aarch64.sme.write.ver.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]])
+// CPP-CHECK-NEXT: ret void
+//
+void test_svwrite_ver_za8_mf8_vg2(uint32_t base, svmfloat8x2_t val) __arm_streaming __arm_inout("za") {
+ SVE_ACLE_FUNC(svwrite_ver_za8,_mf8,_vg2,)(0, base, val);
+}
+
// CHECK-LABEL: @test_svwrite_hor_za8_u8_vg2(
// CHECK-NEXT: entry:
// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.hor.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]])
@@ -72,6 +86,20 @@ void test_svwrite_hor_za8_s8_vg2(uint32_t base, svint8x2_t val) __arm_streaming
SVE_ACLE_FUNC(svwrite_hor_za8,_s8,_vg2,)(0, base, val);
}
+// CHECK-LABEL: @test_svwrite_hor_za8_mf8_vg2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.hor.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]])
+// CHECK-NEXT: ret void
+//
+// CPP-CHECK-LABEL: @_Z28test_svwrite_hor_za8_mf8_vg2j13svmfloat8x2_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: tail call void @llvm.aarch64.sme.write.hor.vg2.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]])
+// CPP-CHECK-NEXT: ret void
+//
+void test_svwrite_hor_za8_mf8_vg2(uint32_t base, svmfloat8x2_t val) __arm_streaming __arm_inout("za") {
+ SVE_ACLE_FUNC(svwrite_hor_za8,_mf8,_vg2,)(0, base, val);
+}
+
// CHECK-LABEL: @test_svwrite_hor_za8_u8_vg4(
// CHECK-NEXT: entry:
// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.hor.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
@@ -100,6 +128,20 @@ void test_svwrite_hor_za8_s8_vg4(uint32_t base, svint8x4_t val) __arm_streaming
SVE_ACLE_FUNC(svwrite_hor_za8,_s8,_vg4,)(0, base, val);
}
+// CHECK-LABEL: @test_svwrite_hor_za8_mf8_vg4(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.hor.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
+// CHECK-NEXT: ret void
+//
+// CPP-CHECK-LABEL: @_Z28test_svwrite_hor_za8_mf8_vg4j13svmfloat8x4_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: tail call void @llvm.aarch64.sme.write.hor.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret void
+//
+void test_svwrite_hor_za8_mf8_vg4(uint32_t base, svmfloat8x4_t val) __arm_streaming __arm_inout("za") {
+ SVE_ACLE_FUNC(svwrite_hor_za8,_mf8,_vg4,)(0, base, val);
+}
+
// CHECK-LABEL: @test_svwrite_ver_za8_u8_vg4(
// CHECK-NEXT: entry:
// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.ver.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
@@ -128,6 +170,20 @@ void test_svwrite_ver_za8_s8_vg4(uint32_t base, svint8x4_t val) __arm_streaming
SVE_ACLE_FUNC(svwrite_ver_za8,_s8,_vg4,)(0, base, val);
}
+// CHECK-LABEL: @test_svwrite_ver_za8_mf8_vg4(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.ver.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
+// CHECK-NEXT: ret void
+//
+// CPP-CHECK-LABEL: @_Z28test_svwrite_ver_za8_mf8_vg4j13svmfloat8x4_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: tail call void @llvm.aarch64.sme.write.ver.vg4.nxv16i8(i32 0, i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret void
+//
+void test_svwrite_ver_za8_mf8_vg4(uint32_t base, svmfloat8x4_t val) __arm_streaming __arm_inout("za") {
+ SVE_ACLE_FUNC(svwrite_ver_za8,_mf8,_vg4,)(0, base, val);
+}
+
// CHECK-LABEL: @test_svwrite_hor_za16_u16_vg2(
// CHECK-NEXT: entry:
// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.hor.vg2.nxv8i16(i32 1, i32 [[BASE:%.*]], <vscale x 8 x i16> [[VAL_COERCE0:%.*]], <vscale x 8 x i16> [[VAL_COERCE1:%.*]])
@@ -716,6 +772,20 @@ void test_svwrite_za8_u8_vg1x2(uint32_t base, svuint8x2_t val) __arm_streaming _
SVE_ACLE_FUNC(svwrite_za8,_u8,_vg1x2,)(base, val);
}
+// CHECK-LABEL: @test_svwrite_za8_mf8_vg1x2(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.vg1x2.nxv16i8(i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]])
+// CHECK-NEXT: ret void
+//
+// CPP-CHECK-LABEL: @_Z26test_svwrite_za8_mf8_vg1x2j13svmfloat8x2_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: tail call void @llvm.aarch64.sme.write.vg1x2.nxv16i8(i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]])
+// CPP-CHECK-NEXT: ret void
+//
+void test_svwrite_za8_mf8_vg1x2(uint32_t base, svmfloat8x2_t val) __arm_streaming __arm_inout("za") {
+ SVE_ACLE_FUNC(svwrite_za8,_mf8,_vg1x2,)(base, val);
+}
+
// CHECK-LABEL: @test_svwrite_za16_s16_vg1x2(
// CHECK-NEXT: entry:
// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.vg1x2.nxv8i16(i32 [[BASE:%.*]], <vscale x 8 x i16> [[VAL_COERCE0:%.*]], <vscale x 8 x i16> [[VAL_COERCE1:%.*]])
@@ -856,18 +926,18 @@ void test_svwrite_za64_s64_vg1x2(uint32_t base, svint64x2_t val) __arm_streaming
SVE_ACLE_FUNC(svwrite_za64,_s64,_vg1x2,)(base, val);
}
-// CHECK-LABEL: @test_svwrite_za8_s8_vg1x4(
+// CHECK-LABEL: @test_svwrite_za8_mf8_vg1x4(
// CHECK-NEXT: entry:
// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.vg1x4.nxv16i8(i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
// CHECK-NEXT: ret void
//
-// CPP-CHECK-LABEL: @_Z25test_svwrite_za8_s8_vg1x4j10svint8x4_t(
+// CPP-CHECK-LABEL: @_Z26test_svwrite_za8_mf8_vg1x4j13svmfloat8x4_t(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: tail call void @llvm.aarch64.sme.write.vg1x4.nxv16i8(i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
// CPP-CHECK-NEXT: ret void
//
-void test_svwrite_za8_s8_vg1x4(uint32_t base, svint8x4_t val) __arm_streaming __arm_inout("za") {
- SVE_ACLE_FUNC(svwrite_za8,_s8,_vg1x4,)(base, val);
+void test_svwrite_za8_mf8_vg1x4(uint32_t base, svmfloat8x4_t val) __arm_streaming __arm_inout("za") {
+ SVE_ACLE_FUNC(svwrite_za8,_mf8,_vg1x4,)(base, val);
}
// CHECK-LABEL: @test_svwrite_za8_u8_vg1x4(
diff --git a/clang/test/CodeGen/AArch64/sme2p1-intrinsics/acle_sme2p1_movaz.c b/clang/test/CodeGen/AArch64/sme2p1-intrinsics/acle_sme2p1_movaz.c
index 7fa2249827c4e0..98324e78b16bc7 100644
--- a/clang/test/CodeGen/AArch64/sme2p1-intrinsics/acle_sme2p1_movaz.c
+++ b/clang/test/CodeGen/AArch64/sme2p1-intrinsics/acle_sme2p1_movaz.c
@@ -42,6 +42,23 @@ svuint8x2_t test_svreadz_hor_za8_u8_x2(uint32_t slice) __arm_streaming __arm_ino
return svreadz_hor_za8_u8_vg2(0, slice);
}
+// CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8> } @test_svreadz_hor_za8_mf8_x2(
+// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.horiz.x2.nxv16i8(i32 0, i32 [[SLICE]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8> } @_Z27test_svreadz_hor_za8_mf8_x2j(
+// CPP-CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.horiz.x2.nxv16i8(i32 0, i32 [[SLICE]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svreadz_hor_za8_mf8_x2(uint32_t slice) __arm_streaming __arm_inout("za")
+{
+ return svreadz_hor_za8_mf8_vg2(0, slice);
+}
+
// CHECK-LABEL: define dso_local { <vscale x 8 x i16>, <vscale x 8 x i16> } @test_svreadz_hor_za16_s16_x2(
// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
// CHECK-NEXT: entry:
@@ -252,6 +269,23 @@ svuint8x2_t test_svreadz_ver_za8_u8_x2(uint32_t slice) __arm_streaming __arm_ino
return svreadz_ver_za8_u8_vg2(0, slice);
}
+// CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8> } @test_svreadz_ver_za8_mf8_x2(
+// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.vert.x2.nxv16i8(i32 0, i32 [[SLICE]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8> } @_Z27test_svreadz_ver_za8_mf8_x2j(
+// CPP-CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.vert.x2.nxv16i8(i32 0, i32 [[SLICE]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svreadz_ver_za8_mf8_x2(uint32_t slice) __arm_streaming __arm_inout("za")
+{
+ return svreadz_ver_za8_mf8_vg2(0, slice);
+}
+
// CHECK-LABEL: define dso_local { <vscale x 8 x i16>, <vscale x 8 x i16> } @test_svreadz_ver_za16_s16_x2(
// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
// CHECK-NEXT: entry:
@@ -460,6 +494,23 @@ svuint8x4_t test_svreadz_hor_za8_u8_x4(uint32_t slice) __arm_streaming __arm_ino
return svreadz_hor_za8_u8_vg4(0, slice);
}
+// CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @test_svreadz_hor_za8_mf8_x4(
+// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.horiz.x4.nxv16i8(i32 0, i32 [[SLICE]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @_Z27test_svreadz_hor_za8_mf8_x4j(
+// CPP-CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.horiz.x4.nxv16i8(i32 0, i32 [[SLICE]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svreadz_hor_za8_mf8_x4(uint32_t slice) __arm_streaming __arm_inout("za")
+{
+ return svreadz_hor_za8_mf8_vg4(0, slice);
+}
+
// CHECK-LABEL: define dso_local { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @test_svreadz_hor_za16_s16_x4(
// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
// CHECK-NEXT: entry:
@@ -667,6 +718,23 @@ svuint8x4_t test_svreadz_ver_za8_u8_x4(uint32_t slice) __arm_streaming __arm_ino
return svreadz_ver_za8_u8_vg4(0, slice);
}
+// CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @test_svreadz_ver_za8_mf8_x4(
+// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.vert.x4.nxv16i8(i32 0, i32 [[SLICE]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @_Z27test_svreadz_ver_za8_mf8_x4j(
+// CPP-CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.vert.x4.nxv16i8(i32 0, i32 [[SLICE]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svreadz_ver_za8_mf8_x4(uint32_t slice) __arm_streaming __arm_inout("za")
+{
+ return svreadz_ver_za8_mf8_vg4(0, slice);
+}
+
// CHECK-LABEL: define dso_local { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @test_svreadz_ver_za16_s16_x4(
// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
// CHECK-NEXT: entry:
@@ -872,6 +940,23 @@ svuint8_t test_svreadz_hor_za8_u8(uint32_t slice) __arm_streaming __arm_inout("z
return svreadz_hor_za8_u8(0, slice);
}
+// CHECK-LABEL: define dso_local <vscale x 16 x i8> @test_svreadz_hor_za8_mf8(
+// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readz.horiz.nxv16i8(i32 0, i32 [[SLICE]])
+// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CPP-CHECK-LABEL: define dso_local <vscale x 16 x i8> @_Z24test_svreadz_hor_za8_mf8j(
+// CPP-CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readz.horiz.nxv16i8(i32 0, i32 [[SLICE]])
+// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svreadz_hor_za8_mf8(uint32_t slice) __arm_streaming __arm_inout("za")
+{
+ return svreadz_hor_za8_mf8(0, slice);
+}
+
// CHECK-LABEL: define dso_local <vscale x 8 x i16> @test_svreadz_hor_za16_s16(
// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
// CHECK-NEXT: entry:
@@ -1078,6 +1163,23 @@ svuint8_t test_svreadz_hor_za128_u8(uint32_t slice) __arm_streaming __arm_inout(
return svreadz_hor_za128_u8(1, slice);
}
+// CHECK-LABEL: define dso_local <vscale x 16 x i8> @test_svreadz_hor_za128_mf8(
+// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readz.q.horiz.nxv16i8(i32 0, i32 [[SLICE]])
+// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CPP-CHECK-LABEL: define dso_local <vscale x 16 x i8> @_Z26test_svreadz_hor_za128_mf8j(
+// CPP-CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sme.readz.q.horiz.nxv16i8(i32 0, i32 [[SLICE]])
+// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svreadz_hor_za128_mf8(uint32_t slice) __arm_streaming __arm_inout("za")
+{
+ return svreadz_hor_za128_mf8(0, slice);
+}
+
// CHECK-LABEL: define dso_local <vscale x 8 x i16> @test_svreadz_hor_za128_s16(
// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
// CHECK-NEXT: entry:
@@ -1287,6 +1389,23 @@ svuint8x2_t test_svreadz_za8_u8_x2(uint32_t slice) __arm_streaming __arm_inout("
return svreadz_za8_u8_vg1x2(slice);
}
+// CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8> } @test_svreadz_za8_mf8_x2(
+// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.x2.nxv16i8(i32 [[SLICE]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8> } @_Z23test_svreadz_za8_mf8_x2j(
+// CPP-CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.x2.nxv16i8(i32 [[SLICE]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x2_t test_svreadz_za8_mf8_x2(uint32_t slice) __arm_streaming __arm_inout("za")
+{
+ return svreadz_za8_mf8_vg1x2(slice);
+}
+
// CHECK-LABEL: define dso_local { <vscale x 8 x i16>, <vscale x 8 x i16> } @test_svreadz_za16_s16_x2(
// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
// CHECK-NEXT: entry:
@@ -1495,6 +1614,23 @@ svuint8x4_t test_svreadz_za8_u8_x4(uint32_t slice) __arm_streaming __arm_inout("
return svreadz_za8_u8_vg1x4(slice);
}
+// CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @test_svreadz_za8_mf8_x4(
+// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.x4.nxv16i8(i32 [[SLICE]])
+// CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: define dso_local { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @_Z23test_svreadz_za8_mf8_x4j(
+// CPP-CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sme.readz.x4.nxv16i8(i32 [[SLICE]])
+// CPP-CHECK-NEXT: ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svmfloat8x4_t test_svreadz_za8_mf8_x4(uint32_t slice) __arm_streaming __arm_inout("za")
+{
+ return svreadz_za8_mf8_vg1x4(slice);
+}
+
// CHECK-LABEL: define dso_local { <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } @test_svreadz_za16_s16_x4(
// CHECK-SAME: i32 noundef [[SLICE:%.*]]) #[[ATTR0]] {
// CHECK-NEXT: entry:
diff --git a/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_revd.c b/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_revd.c
index b1121eb46e2a22..d947cf78660ffc 100644
--- a/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_revd.c
+++ b/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_revd.c
@@ -397,6 +397,20 @@ svuint64_t test_svrevd_u64_x(svbool_t pg, svuint64_t op) MODE_ATTR {
}
+// CHECK-LABEL: @test_svrevd_mf8_z(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.revd.nxv16i8(<vscale x 16 x i8> zeroinitializer, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP:%.*]])
+// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svrevd_mf8_zu10__SVBool_tu13__SVMfloat8_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.revd.nxv16i8(<vscale x 16 x i8> zeroinitializer, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP:%.*]])
+// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svrevd_mf8_z(svbool_t pg, svmfloat8_t op) MODE_ATTR {
+ return SVE_ACLE_FUNC(svrevd, _mf8, _z, )(pg, op);
+}
+
// CHECK-LABEL: @test_svrevd_bf16_z(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
@@ -461,6 +475,20 @@ svfloat64_t test_svrevd_f64_z(svbool_t pg, svfloat64_t op) MODE_ATTR {
return SVE_ACLE_FUNC(svrevd, _f64, _z, )(pg, op);
}
+// CHECK-LABEL: @test_svrevd_mf8_m(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.revd.nxv16i8(<vscale x 16 x i8> [[INACTIVE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP:%.*]])
+// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svrevd_mf8_mu13__SVMfloat8_tu10__SVBool_tS_(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.revd.nxv16i8(<vscale x 16 x i8> [[INACTIVE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP:%.*]])
+// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svrevd_mf8_m(svmfloat8_t inactive, svbool_t pg, svmfloat8_t op) MODE_ATTR {
+ return SVE_ACLE_FUNC(svrevd, _mf8, _m, )(inactive, pg, op);
+}
+
// CHECK-LABEL: @test_svrevd_bf16_m(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
@@ -525,6 +553,20 @@ svfloat64_t test_svrevd_f64_m(svfloat64_t inactive, svbool_t pg, svfloat64_t op)
return SVE_ACLE_FUNC(svrevd, _f64, _m, )(inactive, pg, op);
}
+// CHECK-LABEL: @test_svrevd_mf8_x(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.revd.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP:%.*]])
+// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svrevd_mf8_xu10__SVBool_tu13__SVMfloat8_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.revd.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP:%.*]])
+// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
+//
+svmfloat8_t test_svrevd_mf8_x(svbool_t pg, svmfloat8_t op) MODE_ATTR {
+ return SVE_ACLE_FUNC(svrevd, _mf8, _x, )(pg, op);
+}
+
// CHECK-LABEL: @test_svrevd_bf16_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
>From 6d44b6f0a8152f8ec365526c906676387087605e Mon Sep 17 00:00:00 2001
From: Marian Lukac <Marian.Lukac at arm.com>
Date: Tue, 4 Feb 2025 14:54:04 +0000
Subject: [PATCH 2/2] Add previously removed test
---
.../AArch64/sme2-intrinsics/acle_sme2_write.c | 14 ++++++++++++++
1 file changed, 14 insertions(+)
diff --git a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_write.c b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_write.c
index 73aaaf66fd4b1b..7d0fbc9479a879 100644
--- a/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_write.c
+++ b/clang/test/CodeGen/AArch64/sme2-intrinsics/acle_sme2_write.c
@@ -940,6 +940,20 @@ void test_svwrite_za8_mf8_vg1x4(uint32_t base, svmfloat8x4_t val) __arm_streamin
SVE_ACLE_FUNC(svwrite_za8,_mf8,_vg1x4,)(base, val);
}
+// CHECK-LABEL: @test_svwrite_za8_s8_vg1x4(
+// CHECK-NEXT: entry:
+// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.vg1x4.nxv16i8(i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
+// CHECK-NEXT: ret void
+//
+// CPP-CHECK-LABEL: @_Z25test_svwrite_za8_s8_vg1x4j10svint8x4_t(
+// CPP-CHECK-NEXT: entry:
+// CPP-CHECK-NEXT: tail call void @llvm.aarch64.sme.write.vg1x4.nxv16i8(i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
+// CPP-CHECK-NEXT: ret void
+//
+void test_svwrite_za8_s8_vg1x4(uint32_t base, svint8x4_t val) __arm_streaming __arm_inout("za") {
+ SVE_ACLE_FUNC(svwrite_za8,_s8,_vg1x4,)(base, val);
+}
+
// CHECK-LABEL: @test_svwrite_za8_u8_vg1x4(
// CHECK-NEXT: entry:
// CHECK-NEXT: tail call void @llvm.aarch64.sme.write.vg1x4.nxv16i8(i32 [[BASE:%.*]], <vscale x 16 x i8> [[VAL_COERCE0:%.*]], <vscale x 16 x i8> [[VAL_COERCE1:%.*]], <vscale x 16 x i8> [[VAL_COERCE2:%.*]], <vscale x 16 x i8> [[VAL_COERCE3:%.*]])
More information about the cfe-commits
mailing list