[clang] 0eefcaf - [Clang][SME] Add IsStreamingOrSVE2p1 (#76975)
via cfe-commits
cfe-commits at lists.llvm.org
Fri Jan 5 01:55:54 PST 2024
Author: Sam Tebbs
Date: 2024-01-05T09:55:50Z
New Revision: 0eefcaf96d2900a4f3009026a3673ed3b7793fcc
URL: https://github.com/llvm/llvm-project/commit/0eefcaf96d2900a4f3009026a3673ed3b7793fcc
DIFF: https://github.com/llvm/llvm-project/commit/0eefcaf96d2900a4f3009026a3673ed3b7793fcc.diff
LOG: [Clang][SME] Add IsStreamingOrSVE2p1 (#76975)
This patch adds IsStreamingOrSVE2p1 to the applicable builtins and a
warning for when those builtins are not used in a streaming or sve2p1
function.
Added:
clang/test/Sema/aarch64-sme2-sve2p1-diagnostics.c
Modified:
clang/include/clang/Basic/arm_sve.td
clang/include/clang/Basic/arm_sve_sme_incl.td
clang/lib/Basic/Targets/AArch64.h
clang/lib/Sema/SemaChecking.cpp
clang/test/CodeGen/aarch64-sve2-intrinsics/acle_sve2_revd.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_bfmlsl.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_cntp.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_fclamp.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_ld1.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pext.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pfalse.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_sclamp.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_stnt1.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_uclamp.c
clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_while_pn.c
clang/test/Sema/aarch64-sme2-intrinsics/acle_sme2_imm.cpp
clang/utils/TableGen/SveEmitter.cpp
Removed:
################################################################################
diff --git a/clang/include/clang/Basic/arm_sve.td b/clang/include/clang/Basic/arm_sve.td
index 91f62c4c76339d..7f80fb0386cc77 100644
--- a/clang/include/clang/Basic/arm_sve.td
+++ b/clang/include/clang/Basic/arm_sve.td
@@ -1976,39 +1976,37 @@ def SVFMINQV: SInst<"svminqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_fminq
}
let TargetGuard = "sve2p1|sme2" in {
-//FIXME: Replace IsStreamingCompatible with IsStreamingOrHasSVE2p1 when available
-def SVPEXT_SINGLE : SInst<"svpext_lane_{d}", "P}i", "QcQsQiQl", MergeNone, "aarch64_sve_pext", [IsStreamingCompatible], [ImmCheck<1, ImmCheck0_3>]>;
-def SVPEXT_X2 : SInst<"svpext_lane_{d}_x2", "2.P}i", "QcQsQiQl", MergeNone, "aarch64_sve_pext_x2", [IsStreamingCompatible], [ImmCheck<1, ImmCheck0_1>]>;
+def SVPEXT_SINGLE : SInst<"svpext_lane_{d}", "P}i", "QcQsQiQl", MergeNone, "aarch64_sve_pext", [IsStreamingOrSVE2p1], [ImmCheck<1, ImmCheck0_3>]>;
+def SVPEXT_X2 : SInst<"svpext_lane_{d}_x2", "2.P}i", "QcQsQiQl", MergeNone, "aarch64_sve_pext_x2", [IsStreamingOrSVE2p1], [ImmCheck<1, ImmCheck0_1>]>;
-def SVWHILEGE_COUNT : SInst<"svwhilege_{d}[_{1}]", "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilege_{d}", [IsOverloadNone], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
-def SVWHILEGT_COUNT : SInst<"svwhilegt_{d}[_{1}]", "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilegt_{d}", [IsOverloadNone], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
-def SVWHILELE_COUNT : SInst<"svwhilele_{d}[_{1}]", "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilele_{d}", [IsOverloadNone], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
-def SVWHILELT_COUNT : SInst<"svwhilelt_{d}[_{1}]", "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilelt_{d}", [IsOverloadNone], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
-def SVWHILELO_COUNT : SInst<"svwhilelt_{d}[_{1}]", "}nni", "QcQsQiQl", MergeNone, "aarch64_sve_whilelo_{d}", [IsOverloadNone], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
-def SVWHILELS_COUNT : SInst<"svwhilele_{d}[_{1}]", "}nni", "QcQsQiQl", MergeNone, "aarch64_sve_whilels_{d}", [IsOverloadNone], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
-def SVWHILEHI_COUNT : SInst<"svwhilegt_{d}[_{1}]", "}nni", "QcQsQiQl", MergeNone, "aarch64_sve_whilehi_{d}", [IsOverloadNone], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
-def SVWHILEHS_COUNT : SInst<"svwhilege_{d}[_{1}]", "}nni", "QcQsQiQl", MergeNone, "aarch64_sve_whilehs_{d}", [IsOverloadNone], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
+def SVWHILEGE_COUNT : SInst<"svwhilege_{d}[_{1}]", "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilege_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
+def SVWHILEGT_COUNT : SInst<"svwhilegt_{d}[_{1}]", "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilegt_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
+def SVWHILELE_COUNT : SInst<"svwhilele_{d}[_{1}]", "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilele_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
+def SVWHILELT_COUNT : SInst<"svwhilelt_{d}[_{1}]", "}lli", "QcQsQiQl", MergeNone, "aarch64_sve_whilelt_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
+def SVWHILELO_COUNT : SInst<"svwhilelt_{d}[_{1}]", "}nni", "QcQsQiQl", MergeNone, "aarch64_sve_whilelo_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
+def SVWHILELS_COUNT : SInst<"svwhilele_{d}[_{1}]", "}nni", "QcQsQiQl", MergeNone, "aarch64_sve_whilels_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
+def SVWHILEHI_COUNT : SInst<"svwhilegt_{d}[_{1}]", "}nni", "QcQsQiQl", MergeNone, "aarch64_sve_whilehi_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
+def SVWHILEHS_COUNT : SInst<"svwhilege_{d}[_{1}]", "}nni", "QcQsQiQl", MergeNone, "aarch64_sve_whilehs_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<2, ImmCheck2_4_Mul2>]>;
}
multiclass MultiVecLoad<string i> {
- // FIXME: Replace IsStreamingCompatible with IsStreamingOrHasSVE2p1 when available (SME2 requires __arm_streaming)
- def SV # NAME # B_X2 : MInst<"sv" # i # "[_{2}]_x2", "2}c", "cUc", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # H_X2 : MInst<"sv" # i # "[_{2}]_x2", "2}c", "sUshb", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # W_X2 : MInst<"sv" # i # "[_{2}]_x2", "2}c", "iUif", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # D_X2 : MInst<"sv" # i # "[_{2}]_x2", "2}c", "lUld", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # B_X4 : MInst<"sv" # i # "[_{2}]_x4", "4}c", "cUc", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # H_X4 : MInst<"sv" # i # "[_{2}]_x4", "4}c", "sUshb", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # W_X4 : MInst<"sv" # i # "[_{2}]_x4", "4}c", "iUif", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # D_X4 : MInst<"sv" # i # "[_{2}]_x4", "4}c", "lUld", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
-
- def SV # NAME # B_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}]_x2", "2}cl", "cUc", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # H_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}]_x2", "2}cl", "sUshb", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # W_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}]_x2", "2}cl", "iUif", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # D_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}]_x2", "2}cl", "lUld", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # B_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}]_x4", "4}cl", "cUc", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # H_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}]_x4", "4}cl", "sUshb", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # W_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}]_x4", "4}cl", "iUif", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # D_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}]_x4", "4}cl", "lUld", [IsStructLoad, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # B_X2 : MInst<"sv" # i # "[_{2}]_x2", "2}c", "cUc", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # H_X2 : MInst<"sv" # i # "[_{2}]_x2", "2}c", "sUshb", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # W_X2 : MInst<"sv" # i # "[_{2}]_x2", "2}c", "iUif", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # D_X2 : MInst<"sv" # i # "[_{2}]_x2", "2}c", "lUld", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # B_X4 : MInst<"sv" # i # "[_{2}]_x4", "4}c", "cUc", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # H_X4 : MInst<"sv" # i # "[_{2}]_x4", "4}c", "sUshb", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # W_X4 : MInst<"sv" # i # "[_{2}]_x4", "4}c", "iUif", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # D_X4 : MInst<"sv" # i # "[_{2}]_x4", "4}c", "lUld", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+
+ def SV # NAME # B_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}]_x2", "2}cl", "cUc", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # H_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}]_x2", "2}cl", "sUshb", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # W_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}]_x2", "2}cl", "iUif", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # D_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}]_x2", "2}cl", "lUld", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # B_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}]_x4", "4}cl", "cUc", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # H_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}]_x4", "4}cl", "sUshb", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # W_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}]_x4", "4}cl", "iUif", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # D_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}]_x4", "4}cl", "lUld", [IsStructLoad, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
}
let TargetGuard = "sve2p1|sme2" in {
@@ -2017,24 +2015,23 @@ let TargetGuard = "sve2p1|sme2" in {
}
multiclass MultiVecStore<string i> {
- // FIXME: Replace IsStreamingCompatible with IsStreamingOrHasSVE2p1 when available (SME2 requires __arm_streaming)
- def SV # NAME # B_X2 : MInst<"sv" # i # "[_{2}_x2]", "v}p2", "cUc", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # H_X2 : MInst<"sv" # i # "[_{2}_x2]", "v}p2", "sUshb", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # W_X2 : MInst<"sv" # i # "[_{2}_x2]", "v}p2", "iUif", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # D_X2 : MInst<"sv" # i # "[_{2}_x2]", "v}p2", "lUld", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # B_X4 : MInst<"sv" # i # "[_{2}_x4]", "v}p4", "cUc", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # H_X4 : MInst<"sv" # i # "[_{2}_x4]", "v}p4", "sUshb", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # W_X4 : MInst<"sv" # i # "[_{2}_x4]", "v}p4", "iUif", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # D_X4 : MInst<"sv" # i # "[_{2}_x4]", "v}p4", "lUld", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
-
- def SV # NAME # B_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}_x2]", "v}pl2", "cUc", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # H_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}_x2]", "v}pl2", "sUshb", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # W_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}_x2]", "v}pl2", "iUif", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # D_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}_x2]", "v}pl2", "lUld", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
- def SV # NAME # B_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}_x4]", "v}pl4", "cUc", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # H_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}_x4]", "v}pl4", "sUshb", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # W_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}_x4]", "v}pl4", "iUif", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
- def SV # NAME # D_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}_x4]", "v}pl4", "lUld", [IsStructStore, IsStreamingCompatible], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # B_X2 : MInst<"sv" # i # "[_{2}_x2]", "v}p2", "cUc", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # H_X2 : MInst<"sv" # i # "[_{2}_x2]", "v}p2", "sUshb", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # W_X2 : MInst<"sv" # i # "[_{2}_x2]", "v}p2", "iUif", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # D_X2 : MInst<"sv" # i # "[_{2}_x2]", "v}p2", "lUld", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # B_X4 : MInst<"sv" # i # "[_{2}_x4]", "v}p4", "cUc", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # H_X4 : MInst<"sv" # i # "[_{2}_x4]", "v}p4", "sUshb", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # W_X4 : MInst<"sv" # i # "[_{2}_x4]", "v}p4", "iUif", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # D_X4 : MInst<"sv" # i # "[_{2}_x4]", "v}p4", "lUld", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+
+ def SV # NAME # B_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}_x2]", "v}pl2", "cUc", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # H_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}_x2]", "v}pl2", "sUshb", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # W_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}_x2]", "v}pl2", "iUif", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # D_VNUM_X2 : MInst<"sv" # i # "_vnum" # "[_{2}_x2]", "v}pl2", "lUld", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x2">;
+ def SV # NAME # B_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}_x4]", "v}pl4", "cUc", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # H_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}_x4]", "v}pl4", "sUshb", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # W_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}_x4]", "v}pl4", "iUif", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
+ def SV # NAME # D_VNUM_X4 : MInst<"sv" # i # "_vnum" # "[_{2}_x4]", "v}pl4", "lUld", [IsStructStore, IsStreamingOrSVE2p1], MemEltTyDefault, "aarch64_sve_" # i # "_pn_x4">;
}
let TargetGuard = "sve2p1|sme2" in {
@@ -2051,21 +2048,20 @@ def SVDOT_LANE_X2_U : SInst<"svdot_lane[_{d}_{2}_{3}]", "ddhhi", "Ui", MergeNone
def SVDOT_LANE_X2_F : SInst<"svdot_lane[_{d}_{2}_{3}]", "ddhhi", "f", MergeNone, "aarch64_sve_fdot_lane_x2", [], [ImmCheck<3, ImmCheck0_3>]>;
}
-let TargetGuard = "sve2p1|sme" in {
-def SVSCLAMP : SInst<"svclamp[_{d}]", "dddd", "csil", MergeNone, "aarch64_sve_sclamp", [], []>;
-def SVUCLAMP : SInst<"svclamp[_{d}]", "dddd", "UcUsUiUl", MergeNone, "aarch64_sve_uclamp", [], []>;
+let TargetGuard = "sve2p1|sme2" in {
+def SVSCLAMP : SInst<"svclamp[_{d}]", "dddd", "csil", MergeNone, "aarch64_sve_sclamp", [IsStreamingOrSVE2p1], []>;
+def SVUCLAMP : SInst<"svclamp[_{d}]", "dddd", "UcUsUiUl", MergeNone, "aarch64_sve_uclamp", [IsStreamingOrSVE2p1], []>;
defm SVREVD : SInstZPZ<"svrevd", "csilUcUsUiUlbhfd", "aarch64_sve_revd">;
}
let TargetGuard = "sve2p1|sme2" in {
- //FIXME: Replace IsStreamingCompatible with IsStreamingOrHasSVE2p1 when available
- def SVPTRUE_COUNT : SInst<"svptrue_{d}", "}v", "QcQsQiQl", MergeNone, "aarch64_sve_ptrue_{d}", [IsOverloadNone, IsStreamingCompatible], []>;
+ def SVPTRUE_COUNT : SInst<"svptrue_{d}", "}v", "QcQsQiQl", MergeNone, "aarch64_sve_ptrue_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], []>;
- def SVPFALSE_COUNT_ALIAS : SInst<"svpfalse_c", "}v", "", MergeNone, "", [IsOverloadNone, IsStreamingCompatible]>;
+ def SVPFALSE_COUNT_ALIAS : SInst<"svpfalse_c", "}v", "", MergeNone, "", [IsOverloadNone, IsStreamingOrSVE2p1]>;
- def SVFCLAMP : SInst<"svclamp[_{d}]", "dddd", "hfd", MergeNone, "aarch64_sve_fclamp", [IsStreamingCompatible], []>;
- def SVCNTP_COUNT : SInst<"svcntp_{d}", "n}i", "QcQsQiQl", MergeNone, "aarch64_sve_cntp_{d}", [IsOverloadNone, IsStreamingCompatible], [ImmCheck<1, ImmCheck2_4_Mul2>]>;
+ def SVFCLAMP : SInst<"svclamp[_{d}]", "dddd", "hfd", MergeNone, "aarch64_sve_fclamp", [IsStreamingOrSVE2p1], []>;
+ def SVCNTP_COUNT : SInst<"svcntp_{d}", "n}i", "QcQsQiQl", MergeNone, "aarch64_sve_cntp_{d}", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<1, ImmCheck2_4_Mul2>]>;
}
let TargetGuard = "(sve2|sme2),b16b16" in {
@@ -2326,10 +2322,9 @@ let TargetGuard = "sme2" in {
let TargetGuard = "sve2p1|sme2" in {
// == BFloat16 multiply-subtract ==
-// FIXME: Make all of these IsStreamingOrSVE2p1 once that is added
- def SVBFMLSLB : SInst<"svbfmlslb[_{d}]", "dd$$", "f", MergeNone, "aarch64_sve_bfmlslb", [IsOverloadNone, IsStreamingCompatible], []>;
- def SVBFMLSLT : SInst<"svbfmlslt[_{d}]", "dd$$", "f", MergeNone, "aarch64_sve_bfmlslt", [IsOverloadNone, IsStreamingCompatible], []>;
+ def SVBFMLSLB : SInst<"svbfmlslb[_{d}]", "dd$$", "f", MergeNone, "aarch64_sve_bfmlslb", [IsOverloadNone, IsStreamingOrSVE2p1], []>;
+ def SVBFMLSLT : SInst<"svbfmlslt[_{d}]", "dd$$", "f", MergeNone, "aarch64_sve_bfmlslt", [IsOverloadNone, IsStreamingOrSVE2p1], []>;
- def SVBFMLSLB_LANE : SInst<"svbfmlslb_lane[_{d}]", "dd$$i", "f", MergeNone, "aarch64_sve_bfmlslb_lane", [IsOverloadNone, IsStreamingCompatible], [ImmCheck<3, ImmCheck0_7>]>;
- def SVBFMLSLT_LANE : SInst<"svbfmlslt_lane[_{d}]", "dd$$i", "f", MergeNone, "aarch64_sve_bfmlslt_lane", [IsOverloadNone, IsStreamingCompatible], [ImmCheck<3, ImmCheck0_7>]>;
+ def SVBFMLSLB_LANE : SInst<"svbfmlslb_lane[_{d}]", "dd$$i", "f", MergeNone, "aarch64_sve_bfmlslb_lane", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<3, ImmCheck0_7>]>;
+ def SVBFMLSLT_LANE : SInst<"svbfmlslt_lane[_{d}]", "dd$$i", "f", MergeNone, "aarch64_sve_bfmlslt_lane", [IsOverloadNone, IsStreamingOrSVE2p1], [ImmCheck<3, ImmCheck0_7>]>;
}
diff --git a/clang/include/clang/Basic/arm_sve_sme_incl.td b/clang/include/clang/Basic/arm_sve_sme_incl.td
index 0dba8493bad2d6..ad29864440c96f 100644
--- a/clang/include/clang/Basic/arm_sve_sme_incl.td
+++ b/clang/include/clang/Basic/arm_sve_sme_incl.td
@@ -227,6 +227,7 @@ def IsPreservesZA : FlagType<0x10000000000>;
def IsReadZA : FlagType<0x20000000000>;
def IsWriteZA : FlagType<0x40000000000>;
def IsReductionQV : FlagType<0x80000000000>;
+def IsStreamingOrSVE2p1 : FlagType<0x80000000000>; // Use for intrinsics that are common between sme/sme2 and sve2p1.
// These must be kept in sync with the flags in include/clang/Basic/TargetBuiltins.h
class ImmCheckType<int val> {
diff --git a/clang/lib/Basic/Targets/AArch64.h b/clang/lib/Basic/Targets/AArch64.h
index f0e0782e7abe97..e38fa5af56598d 100644
--- a/clang/lib/Basic/Targets/AArch64.h
+++ b/clang/lib/Basic/Targets/AArch64.h
@@ -50,6 +50,7 @@ class LLVM_LIBRARY_VISIBILITY AArch64TargetInfo : public TargetInfo {
bool HasMatMul = false;
bool HasBFloat16 = false;
bool HasSVE2 = false;
+ bool HasSVE2p1 = false;
bool HasSVE2AES = false;
bool HasSVE2SHA3 = false;
bool HasSVE2SM4 = false;
diff --git a/clang/lib/Sema/SemaChecking.cpp b/clang/lib/Sema/SemaChecking.cpp
index 3168d38dd66c36..f13164dc063866 100644
--- a/clang/lib/Sema/SemaChecking.cpp
+++ b/clang/lib/Sema/SemaChecking.cpp
@@ -2998,7 +2998,12 @@ static QualType getNeonEltType(NeonTypeFlags Flags, ASTContext &Context,
llvm_unreachable("Invalid NeonTypeFlag!");
}
-enum ArmStreamingType { ArmNonStreaming, ArmStreaming, ArmStreamingCompatible };
+enum ArmStreamingType {
+ ArmNonStreaming,
+ ArmStreaming,
+ ArmStreamingCompatible,
+ ArmStreamingOrSVE2p1
+};
bool Sema::ParseSVEImmChecks(
CallExpr *TheCall, SmallVector<std::tuple<int, int, int>, 3> &ImmChecks) {
@@ -3156,6 +3161,16 @@ static void checkArmStreamingBuiltin(Sema &S, CallExpr *TheCall,
const FunctionDecl *FD,
ArmStreamingType BuiltinType) {
ArmStreamingType FnType = getArmStreamingFnType(FD);
+ if (BuiltinType == ArmStreamingOrSVE2p1) {
+ // Check intrinsics that are available in [sve2p1 or sme/sme2].
+ llvm::StringMap<bool> CallerFeatureMap;
+ S.Context.getFunctionFeatureMap(CallerFeatureMap, FD);
+ if (Builtin::evaluateRequiredTargetFeatures("sve2p1", CallerFeatureMap))
+ BuiltinType = ArmStreamingCompatible;
+ else
+ BuiltinType = ArmStreaming;
+ }
+
if (FnType == ArmStreaming && BuiltinType == ArmNonStreaming) {
S.Diag(TheCall->getBeginLoc(), diag::warn_attribute_arm_sm_incompat_builtin)
<< TheCall->getSourceRange() << "streaming";
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 74a90583a173a5..d82d69442b8ffb 100644
--- a/clang/test/CodeGen/aarch64-sve2-intrinsics/acle_sve2_revd.c
+++ b/clang/test/CodeGen/aarch64-sve2-intrinsics/acle_sve2_revd.c
@@ -1,7 +1,7 @@
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
// REQUIRES: aarch64-registered-target
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu \
-// RUN: -target-feature +sme -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
+// RUN: -target-feature +sme2 -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu \
// RUN: -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu \
@@ -9,7 +9,7 @@
// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu \
// RUN: -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -target-feature +bf16 -S -disable-O0-optnone -Werror -o /dev/null %s
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme -target-feature +bf16 -S -disable-O0-optnone -Werror -o /dev/null %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -target-feature +bf16 -S -disable-O0-optnone -Werror -o /dev/null %s
#include <arm_sve.h>
#ifdef SVE_OVERLOADED_FORMS
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_bfmlsl.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_bfmlsl.c
index c1d14e16ad17b6..22d951c069bc8a 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_bfmlsl.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_bfmlsl.c
@@ -2,20 +2,20 @@
// REQUIRES: aarch64-registered-target
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
-// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 -target-feature +sve -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 -target-feature +sve -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
-// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -target-feature -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
-// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 -target-feature +sve -target-feature -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
#include <arm_sve.h>
#ifndef TEST_SME2
#define ATTR
#else
-#define ATTR __arm_streaming_compatible
+#define ATTR __arm_streaming
#endif
#ifdef SVE_OVERLOADED_FORMS
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_cntp.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_cntp.c
index 56b1d992622145..9bf55eaa6a08bf 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_cntp.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_cntp.c
@@ -3,10 +3,19 @@
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -O1 -Werror -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
#include <arm_sve.h>
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
// CHECK-LABEL: @test_svcntp_c8_vlx2(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call i64 @llvm.aarch64.sve.cntp.c8(target("aarch64.svcount") [[PNN:%.*]], i32 2)
@@ -17,7 +26,7 @@
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call i64 @llvm.aarch64.sve.cntp.c8(target("aarch64.svcount") [[PNN:%.*]], i32 2)
// CPP-CHECK-NEXT: ret i64 [[TMP0]]
//
-uint64_t test_svcntp_c8_vlx2(svcount_t pnn) {
+uint64_t test_svcntp_c8_vlx2(svcount_t pnn) ATTR {
return svcntp_c8(pnn, 2);
}
@@ -31,7 +40,7 @@ uint64_t test_svcntp_c8_vlx2(svcount_t pnn) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call i64 @llvm.aarch64.sve.cntp.c8(target("aarch64.svcount") [[PNN:%.*]], i32 4)
// CPP-CHECK-NEXT: ret i64 [[TMP0]]
//
-uint64_t test_svcntp_c8_vlx4(svcount_t pnn) {
+uint64_t test_svcntp_c8_vlx4(svcount_t pnn) ATTR {
return svcntp_c8(pnn, 4);
}
@@ -45,7 +54,7 @@ uint64_t test_svcntp_c8_vlx4(svcount_t pnn) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call i64 @llvm.aarch64.sve.cntp.c16(target("aarch64.svcount") [[PNN:%.*]], i32 2)
// CPP-CHECK-NEXT: ret i64 [[TMP0]]
//
-uint64_t test_svcntp_c16_vlx2(svcount_t pnn) {
+uint64_t test_svcntp_c16_vlx2(svcount_t pnn) ATTR {
return svcntp_c16(pnn, 2);
}
@@ -59,7 +68,7 @@ uint64_t test_svcntp_c16_vlx2(svcount_t pnn) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call i64 @llvm.aarch64.sve.cntp.c16(target("aarch64.svcount") [[PNN:%.*]], i32 4)
// CPP-CHECK-NEXT: ret i64 [[TMP0]]
//
-uint64_t test_svcntp_c16_vlx4(svcount_t pnn) {
+uint64_t test_svcntp_c16_vlx4(svcount_t pnn) ATTR {
return svcntp_c16(pnn, 4);
}
@@ -73,7 +82,7 @@ uint64_t test_svcntp_c16_vlx4(svcount_t pnn) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call i64 @llvm.aarch64.sve.cntp.c32(target("aarch64.svcount") [[PNN:%.*]], i32 2)
// CPP-CHECK-NEXT: ret i64 [[TMP0]]
//
-uint64_t test_svcntp_c32_vlx2(svcount_t pnn) {
+uint64_t test_svcntp_c32_vlx2(svcount_t pnn) ATTR {
return svcntp_c32(pnn, 2);
}
@@ -87,7 +96,7 @@ uint64_t test_svcntp_c32_vlx2(svcount_t pnn) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call i64 @llvm.aarch64.sve.cntp.c32(target("aarch64.svcount") [[PNN:%.*]], i32 4)
// CPP-CHECK-NEXT: ret i64 [[TMP0]]
//
-uint64_t test_svcntp_c32_vlx4(svcount_t pnn) {
+uint64_t test_svcntp_c32_vlx4(svcount_t pnn) ATTR {
return svcntp_c32(pnn, 4);
}
@@ -101,7 +110,7 @@ uint64_t test_svcntp_c32_vlx4(svcount_t pnn) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call i64 @llvm.aarch64.sve.cntp.c64(target("aarch64.svcount") [[PNN:%.*]], i32 2)
// CPP-CHECK-NEXT: ret i64 [[TMP0]]
//
-uint64_t test_svcntp_c64_vlx2(svcount_t pnn) {
+uint64_t test_svcntp_c64_vlx2(svcount_t pnn) ATTR {
return svcntp_c64(pnn, 2);
}
@@ -115,6 +124,6 @@ uint64_t test_svcntp_c64_vlx2(svcount_t pnn) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call i64 @llvm.aarch64.sve.cntp.c64(target("aarch64.svcount") [[PNN:%.*]], i32 4)
// CPP-CHECK-NEXT: ret i64 [[TMP0]]
//
-uint64_t test_svcntp_c64_vlx4(svcount_t pnn) {
+uint64_t test_svcntp_c64_vlx4(svcount_t pnn) ATTR {
return svcntp_c64(pnn, 4);
}
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_fclamp.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_fclamp.c
index 5d8c5b7b8a18c6..7687257701a6e2 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_fclamp.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_fclamp.c
@@ -11,10 +11,16 @@
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 \
// RUN: -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 -target-feature +sve \
-// RUN: -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
#include <arm_sve.h>
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
#ifdef SVE_OVERLOADED_FORMS
// A simple used,unused... macro, long enough to represent any SVE builtin.
#define SVE_ACLE_FUNC(A1, A2_UNUSED, A3, A4_UNUSED) A1##A3
@@ -32,7 +38,7 @@
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fclamp.nxv8f16(<vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 8 x half> [[TMP0]]
//
-svfloat16_t test_svclamp_f16(svfloat16_t op1, svfloat16_t op2, svfloat16_t op3) {
+svfloat16_t test_svclamp_f16(svfloat16_t op1, svfloat16_t op2, svfloat16_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _f16, , )(op1, op2, op3);
}
@@ -46,7 +52,7 @@ svfloat16_t test_svclamp_f16(svfloat16_t op1, svfloat16_t op2, svfloat16_t op3)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fclamp.nxv4f32(<vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 4 x float> [[TMP0]]
//
-svfloat32_t test_svclamp_f32(svfloat32_t op1, svfloat32_t op2, svfloat32_t op3) {
+svfloat32_t test_svclamp_f32(svfloat32_t op1, svfloat32_t op2, svfloat32_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _f32, , )(op1, op2, op3);
}
@@ -60,7 +66,7 @@ svfloat32_t test_svclamp_f32(svfloat32_t op1, svfloat32_t op2, svfloat32_t op3)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fclamp.nxv2f64(<vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 2 x double> [[TMP0]]
//
-svfloat64_t test_svclamp_f64(svfloat64_t op1, svfloat64_t op2, svfloat64_t op3) {
+svfloat64_t test_svclamp_f64(svfloat64_t op1, svfloat64_t op2, svfloat64_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _f64, , )(op1, op2, op3);
}
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_ld1.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_ld1.c
index 6f1231e776aa35..7657165d8b3f6f 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_ld1.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_ld1.c
@@ -1,6 +1,8 @@
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
// REQUIRES: aarch64-registered-target
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -DTEST_SME2 -S -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -Wno-unknown-attributes -Wall -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pext.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pext.c
index a3206029019c3d..8f08b32618b050 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pext.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pext.c
@@ -1,8 +1,11 @@
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
// REQUIRES: aarch64-registered-target
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -DTEST_SME2 -target-feature +sve -target-feature +sme2 -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -DTEST_SME2 -O1 -Werror -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme -target-feature +sme2 -S -DTEST_SME2 -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
#include <arm_sve.h>
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pfalse.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pfalse.c
index 19993e5418128e..afdb038fb9312a 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pfalse.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pfalse.c
@@ -1,14 +1,20 @@
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
// REQUIRES: aarch64-registered-target
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
#include <arm_sve.h>
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
// CHECK-LABEL: @test_svpfalse_c(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt(<vscale x 16 x i1> zeroinitializer)
@@ -19,7 +25,7 @@
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt(<vscale x 16 x i1> zeroinitializer)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svpfalse_c(void) __arm_streaming_compatible
+svcount_t test_svpfalse_c(void) ATTR
{
return svpfalse_c();
}
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel.c
index 73b7b0347dd970..de3f6a9a57bfeb 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_psel.c
@@ -10,9 +10,19 @@
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu \
// RUN: -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu \
+// RUN: -target-feature +sme2 -S -DTEST_SME2 -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
#include <arm_sve.h>
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming_compatible
+#endif
+
// CHECK-LABEL: @test_svpsel_lane_b8(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[ADD:%.*]] = add i32 [[IDX:%.*]], 15
@@ -25,7 +35,7 @@
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i1> @llvm.aarch64.sve.psel.nxv16i1(<vscale x 16 x i1> [[P1:%.*]], <vscale x 16 x i1> [[P2:%.*]], i32 [[ADD]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i1> [[TMP0]]
//
-svbool_t test_svpsel_lane_b8(svbool_t p1, svbool_t p2, uint32_t idx) __arm_streaming_compatible {
+svbool_t test_svpsel_lane_b8(svbool_t p1, svbool_t p2, uint32_t idx) ATTR {
return svpsel_lane_b8(p1, p2, idx + 15);
}
@@ -43,7 +53,7 @@ svbool_t test_svpsel_lane_b8(svbool_t p1, svbool_t p2, uint32_t idx) __arm_strea
// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 16 x i1> @llvm.aarch64.sve.psel.nxv8i1(<vscale x 16 x i1> [[P1:%.*]], <vscale x 8 x i1> [[TMP0]], i32 [[ADD]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i1> [[TMP1]]
//
-svbool_t test_svpsel_lane_b16(svbool_t p1, svbool_t p2, uint32_t idx) __arm_streaming_compatible {
+svbool_t test_svpsel_lane_b16(svbool_t p1, svbool_t p2, uint32_t idx) ATTR {
return svpsel_lane_b16(p1, p2, idx + 7);
}
@@ -61,7 +71,7 @@ svbool_t test_svpsel_lane_b16(svbool_t p1, svbool_t p2, uint32_t idx) __arm_stre
// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 16 x i1> @llvm.aarch64.sve.psel.nxv4i1(<vscale x 16 x i1> [[P1:%.*]], <vscale x 4 x i1> [[TMP0]], i32 [[ADD]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i1> [[TMP1]]
//
-svbool_t test_svpsel_lane_b32(svbool_t p1, svbool_t p2, uint32_t idx) __arm_streaming_compatible {
+svbool_t test_svpsel_lane_b32(svbool_t p1, svbool_t p2, uint32_t idx) ATTR {
return svpsel_lane_b32(p1, p2, idx + 3);
}
@@ -79,7 +89,7 @@ svbool_t test_svpsel_lane_b32(svbool_t p1, svbool_t p2, uint32_t idx) __arm_stre
// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 16 x i1> @llvm.aarch64.sve.psel.nxv2i1(<vscale x 16 x i1> [[P1:%.*]], <vscale x 2 x i1> [[TMP0]], i32 [[ADD]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i1> [[TMP1]]
//
-svbool_t test_svpsel_lane_b64(svbool_t p1, svbool_t p2, uint32_t idx) __arm_streaming_compatible {
+svbool_t test_svpsel_lane_b64(svbool_t p1, svbool_t p2, uint32_t idx) ATTR {
return svpsel_lane_b64(p1, p2, idx + 1);
}
@@ -99,7 +109,7 @@ svbool_t test_svpsel_lane_b64(svbool_t p1, svbool_t p2, uint32_t idx) __arm_stre
// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt(<vscale x 16 x i1> [[TMP1]])
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP2]]
//
-svcount_t test_svpsel_lane_c8(svcount_t p1, svbool_t p2, uint32_t idx) __arm_streaming_compatible {
+svcount_t test_svpsel_lane_c8(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
return svpsel_lane_c8(p1, p2, idx + 15);
}
@@ -121,7 +131,7 @@ svcount_t test_svpsel_lane_c8(svcount_t p1, svbool_t p2, uint32_t idx) __arm_str
// CPP-CHECK-NEXT: [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt(<vscale x 16 x i1> [[TMP2]])
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP3]]
//
-svcount_t test_svpsel_lane_c16(svcount_t p1, svbool_t p2, uint32_t idx) __arm_streaming_compatible {
+svcount_t test_svpsel_lane_c16(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
return svpsel_lane_c16(p1, p2, idx + 7);
}
@@ -143,7 +153,7 @@ svcount_t test_svpsel_lane_c16(svcount_t p1, svbool_t p2, uint32_t idx) __arm_st
// CPP-CHECK-NEXT: [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt(<vscale x 16 x i1> [[TMP2]])
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP3]]
//
-svcount_t test_svpsel_lane_c32(svcount_t p1, svbool_t p2, uint32_t idx) __arm_streaming_compatible {
+svcount_t test_svpsel_lane_c32(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
return svpsel_lane_c32(p1, p2, idx + 3);
}
@@ -165,6 +175,6 @@ svcount_t test_svpsel_lane_c32(svcount_t p1, svbool_t p2, uint32_t idx) __arm_st
// CPP-CHECK-NEXT: [[TMP3:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt(<vscale x 16 x i1> [[TMP2]])
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP3]]
//
-svcount_t test_svpsel_lane_c64(svcount_t p1, svbool_t p2, uint32_t idx) __arm_streaming_compatible {
+svcount_t test_svpsel_lane_c64(svcount_t p1, svbool_t p2, uint32_t idx) ATTR {
return svpsel_lane_c64(p1, p2, idx + 1);
}
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_sclamp.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_sclamp.c
index 8c63a7455c79f4..04869fd550ec1f 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_sclamp.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_sclamp.c
@@ -10,9 +10,21 @@
// RUN: -S -Werror -emit-llvm -disable-O0-optnone -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 \
// RUN: -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 \
+// RUN: -S -DTEST_SME2 -Werror -emit-llvm -disable-O0-optnone -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 \
+// RUN: -S -DTEST_SME2 -Werror -emit-llvm -disable-O0-optnone -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 \
+// RUN: -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
#include <arm_sve.h>
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
#ifdef SVE_OVERLOADED_FORMS
// A simple used,unused... macro, long enough to represent any SVE builtin.
#define SVE_ACLE_FUNC(A1, A2_UNUSED, A3, A4_UNUSED) A1##A3
@@ -30,7 +42,7 @@
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sclamp.nxv16i8(<vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i8> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
-svint8_t test_svclamp_s8(svint8_t op1, svint8_t op2, svint8_t op3) {
+svint8_t test_svclamp_s8(svint8_t op1, svint8_t op2, svint8_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _s8, , )(op1, op2, op3);
}
@@ -44,7 +56,7 @@ svint8_t test_svclamp_s8(svint8_t op1, svint8_t op2, svint8_t op3) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sclamp.nxv8i16(<vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i16> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP0]]
//
-svint16_t test_svclamp_s16(svint16_t op1, svint16_t op2, svint16_t op3) {
+svint16_t test_svclamp_s16(svint16_t op1, svint16_t op2, svint16_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _s16, , )(op1, op2, op3);
}
@@ -58,7 +70,7 @@ svint16_t test_svclamp_s16(svint16_t op1, svint16_t op2, svint16_t op3) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sclamp.nxv4i32(<vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i32> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP0]]
//
-svint32_t test_svclamp_s32(svint32_t op1, svint32_t op2, svint32_t op3) {
+svint32_t test_svclamp_s32(svint32_t op1, svint32_t op2, svint32_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _s32, , )(op1, op2, op3);
}
@@ -72,7 +84,7 @@ svint32_t test_svclamp_s32(svint32_t op1, svint32_t op2, svint32_t op3) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.sclamp.nxv2i64(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i64> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP0]]
//
-svint64_t test_svclamp_s64(svint64_t op1, svint64_t op2, svint64_t op3) {
+svint64_t test_svclamp_s64(svint64_t op1, svint64_t op2, svint64_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _s64, , )(op1, op2, op3);
}
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_stnt1.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_stnt1.c
index 0d8696a7634a70..b1ca27b7b68a12 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_stnt1.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_stnt1.c
@@ -505,11 +505,9 @@ void test_svstnt1_f64_x4(svcount_t pn, float64_t *base, svfloat64x4_t v) ATTR
return SVE_ACLE_FUNC(svstnt1,_f64_x4,,)(pn, base, v);
}
-
// == VNUM variants ==
-
// CHECK-LABEL: @test_svstnt1_vnum_u8_x2(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[V:%.*]], i64 0)
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_uclamp.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_uclamp.c
index b8789862488777..37bfd4265a43ad 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_uclamp.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_uclamp.c
@@ -10,9 +10,21 @@
// RUN: -S -Werror -emit-llvm -disable-O0-optnone -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sve2p1 \
// RUN: -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 \
+// RUN: -S -DTEST_SME2 -Werror -emit-llvm -disable-O0-optnone -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 \
+// RUN: -S -DTEST_SME2 -Werror -emit-llvm -disable-O0-optnone -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64-none-linux-gnu -target-feature +sme2 \
+// RUN: -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
#include <arm_sve.h>
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
#ifdef SVE_OVERLOADED_FORMS
// A simple used,unused... macro, long enough to represent any SVE builtin.
#define SVE_ACLE_FUNC(A1, A2_UNUSED, A3, A4_UNUSED) A1##A3
@@ -30,7 +42,7 @@
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.uclamp.nxv16i8(<vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], <vscale x 16 x i8> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
-svuint8_t test_svclamp_u8(svuint8_t op1, svuint8_t op2, svuint8_t op3) {
+svuint8_t test_svclamp_u8(svuint8_t op1, svuint8_t op2, svuint8_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _u8, , )(op1, op2, op3);
}
@@ -44,7 +56,7 @@ svuint8_t test_svclamp_u8(svuint8_t op1, svuint8_t op2, svuint8_t op3) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.uclamp.nxv8i16(<vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]], <vscale x 8 x i16> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP0]]
//
-svuint16_t test_svclamp_u16(svuint16_t op1, svuint16_t op2, svuint16_t op3) {
+svuint16_t test_svclamp_u16(svuint16_t op1, svuint16_t op2, svuint16_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _u16, , )(op1, op2, op3);
}
@@ -58,7 +70,7 @@ svuint16_t test_svclamp_u16(svuint16_t op1, svuint16_t op2, svuint16_t op3) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.uclamp.nxv4i32(<vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]], <vscale x 4 x i32> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP0]]
//
-svuint32_t test_svclamp_u32(svuint32_t op1, svuint32_t op2, svuint32_t op3) {
+svuint32_t test_svclamp_u32(svuint32_t op1, svuint32_t op2, svuint32_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _u32, , )(op1, op2, op3);
}
@@ -72,7 +84,7 @@ svuint32_t test_svclamp_u32(svuint32_t op1, svuint32_t op2, svuint32_t op3) {
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.uclamp.nxv2i64(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i64> [[OP3:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP0]]
//
-svuint64_t test_svclamp_u64(svuint64_t op1, svuint64_t op2, svuint64_t op3) {
+svuint64_t test_svclamp_u64(svuint64_t op1, svuint64_t op2, svuint64_t op3) ATTR {
return SVE_ACLE_FUNC(svclamp, _u64, , )(op1, op2, op3);
}
diff --git a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_while_pn.c b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_while_pn.c
index 143a43b4a92198..11ebec9e7cbf11 100644
--- a/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_while_pn.c
+++ b/clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_while_pn.c
@@ -1,10 +1,10 @@
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
-// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -DTEST_SME2 -disable-O0-optnone -Werror -Wall -o /dev/null %s
// REQUIRES: aarch64-registered-target
@@ -16,6 +16,11 @@
#define SVE_ACLE_FUNC(A1, A2) A1##A2
#endif
+#ifdef TEST_SME2
+#define ATTR __arm_streaming
+#else
+#define ATTR
+#endif
// WHILEGE
@@ -29,7 +34,7 @@
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilege.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilege_c8_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilege_c8_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c8,_s64)(op1, op2, 2);
}
@@ -44,7 +49,7 @@ svcount_t test_svwhilege_c8_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilege.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilege_c8_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilege_c8_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c8,_s64)(op1, op2, 4);
}
@@ -59,7 +64,7 @@ svcount_t test_svwhilege_c8_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilege.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilege_c16_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilege_c16_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c16,_s64)(op1, op2, 2);
}
@@ -74,7 +79,7 @@ svcount_t test_svwhilege_c16_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilege.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilege_c16_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilege_c16_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c16,_s64)(op1, op2, 4);
}
@@ -89,7 +94,7 @@ svcount_t test_svwhilege_c16_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilege.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilege_c32_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilege_c32_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c32,_s64)(op1, op2, 2);
}
@@ -104,7 +109,7 @@ svcount_t test_svwhilege_c32_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilege.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilege_c32_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilege_c32_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c32,_s64)(op1, op2, 4);
}
@@ -119,7 +124,7 @@ svcount_t test_svwhilege_c32_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilege.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilege_c64_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilege_c64_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c64,_s64)(op1, op2, 2);
}
@@ -134,7 +139,7 @@ svcount_t test_svwhilege_c64_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilege.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilege_c64_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilege_c64_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c64,_s64)(op1, op2, 4);
}
@@ -152,7 +157,7 @@ svcount_t test_svwhilege_c64_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilegt.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilegt_c8_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilegt_c8_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c8,_s64)(op1, op2, 2);
}
@@ -167,7 +172,7 @@ svcount_t test_svwhilegt_c8_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilegt.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilegt_c8_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilegt_c8_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c8,_s64)(op1, op2, 4);
}
@@ -182,7 +187,7 @@ svcount_t test_svwhilegt_c8_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilegt.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilegt_c16_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilegt_c16_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c16,_s64)(op1, op2, 2);
}
@@ -197,7 +202,7 @@ svcount_t test_svwhilegt_c16_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilegt.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilegt_c16_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilegt_c16_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c16,_s64)(op1, op2, 4);
}
@@ -212,7 +217,7 @@ svcount_t test_svwhilegt_c16_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilegt.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilegt_c32_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilegt_c32_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c32,_s64)(op1, op2, 2);
}
@@ -227,7 +232,7 @@ svcount_t test_svwhilegt_c32_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilegt.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilegt_c32_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilegt_c32_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c32,_s64)(op1, op2, 4);
}
@@ -242,7 +247,7 @@ svcount_t test_svwhilegt_c32_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilegt.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilegt_c64_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilegt_c64_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c64,_s64)(op1, op2, 2);
}
@@ -257,7 +262,7 @@ svcount_t test_svwhilegt_c64_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilegt.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilegt_c64_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilegt_c64_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c64,_s64)(op1, op2, 4);
}
@@ -275,7 +280,7 @@ svcount_t test_svwhilegt_c64_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehi.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehi_c8_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehi_c8_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c8,_u64)(op1, op2, 2);
}
@@ -290,7 +295,7 @@ svcount_t test_svwhilehi_c8_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehi.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehi_c8_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehi_c8_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c8,_u64)(op1, op2, 4);
}
@@ -305,7 +310,7 @@ svcount_t test_svwhilehi_c8_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehi.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehi_c16_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehi_c16_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c16,_u64)(op1, op2, 2);
}
@@ -320,7 +325,7 @@ svcount_t test_svwhilehi_c16_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehi.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehi_c16_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehi_c16_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c16,_u64)(op1, op2, 4);
}
@@ -335,7 +340,7 @@ svcount_t test_svwhilehi_c16_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehi.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehi_c32_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehi_c32_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c32,_u64)(op1, op2, 2);
}
@@ -350,7 +355,7 @@ svcount_t test_svwhilehi_c32_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehi.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehi_c32_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehi_c32_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c32,_u64)(op1, op2, 4);
}
@@ -365,7 +370,7 @@ svcount_t test_svwhilehi_c32_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehi.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehi_c64_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehi_c64_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c64,_u64)(op1, op2, 2);
}
@@ -380,7 +385,7 @@ svcount_t test_svwhilehi_c64_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehi.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehi_c64_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehi_c64_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilegt_c64,_u64)(op1, op2, 4);
}
@@ -398,7 +403,7 @@ svcount_t test_svwhilehi_c64_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehs.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehs_c8_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehs_c8_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c8,_u64)(op1, op2, 2);
}
@@ -413,7 +418,7 @@ svcount_t test_svwhilehs_c8_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehs.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehs_c8_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehs_c8_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c8,_u64)(op1, op2, 4);
}
@@ -428,7 +433,7 @@ svcount_t test_svwhilehs_c8_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehs.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehs_c16_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehs_c16_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c16,_u64)(op1, op2, 2);
}
@@ -443,7 +448,7 @@ svcount_t test_svwhilehs_c16_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehs.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehs_c16_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehs_c16_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c16,_u64)(op1, op2, 4);
}
@@ -458,7 +463,7 @@ svcount_t test_svwhilehs_c16_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehs.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehs_c32_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehs_c32_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c32,_u64)(op1, op2, 2);
}
@@ -473,7 +478,7 @@ svcount_t test_svwhilehs_c32_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehs.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehs_c32_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehs_c32_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c32,_u64)(op1, op2, 4);
}
@@ -488,7 +493,7 @@ svcount_t test_svwhilehs_c32_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehs.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehs_c64_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehs_c64_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c64,_u64)(op1, op2, 2);
}
@@ -503,7 +508,7 @@ svcount_t test_svwhilehs_c64_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilehs.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilehs_c64_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilehs_c64_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilege_c64,_u64)(op1, op2, 4);
}
@@ -521,7 +526,7 @@ svcount_t test_svwhilehs_c64_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilele.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilele_c8_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilele_c8_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c8,_s64)(op1, op2, 2);
}
@@ -536,7 +541,7 @@ svcount_t test_svwhilele_c8_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilele.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilele_c8_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilele_c8_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c8,_s64)(op1, op2, 4);
}
@@ -551,7 +556,7 @@ svcount_t test_svwhilele_c8_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilele.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilele_c16_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilele_c16_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c16,_s64)(op1, op2, 2);
}
@@ -566,7 +571,7 @@ svcount_t test_svwhilele_c16_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilele.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilele_c16_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilele_c16_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c16,_s64)(op1, op2, 4);
}
@@ -581,7 +586,7 @@ svcount_t test_svwhilele_c16_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilele.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilele_c32_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilele_c32_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c32,_s64)(op1, op2, 2);
}
@@ -596,7 +601,7 @@ svcount_t test_svwhilele_c32_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilele.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilele_c32_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilele_c32_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c32,_s64)(op1, op2, 4);
}
@@ -611,7 +616,7 @@ svcount_t test_svwhilele_c32_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilele.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilele_c64_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilele_c64_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c64,_s64)(op1, op2, 2);
}
@@ -626,7 +631,7 @@ svcount_t test_svwhilele_c64_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilele.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilele_c64_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilele_c64_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c64,_s64)(op1, op2, 4);
}
@@ -644,7 +649,7 @@ svcount_t test_svwhilele_c64_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelo.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelo_c8_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilelo_c8_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c8,_u64)(op1, op2, 2);
}
@@ -659,7 +664,7 @@ svcount_t test_svwhilelo_c8_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelo.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelo_c8_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilelo_c8_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c8,_u64)(op1, op2, 4);
}
@@ -674,7 +679,7 @@ svcount_t test_svwhilelo_c8_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelo.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelo_c16_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilelo_c16_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c16,_u64)(op1, op2, 2);
}
@@ -689,7 +694,7 @@ svcount_t test_svwhilelo_c16_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelo.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelo_c16_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilelo_c16_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c16,_u64)(op1, op2, 4);
}
@@ -704,7 +709,7 @@ svcount_t test_svwhilelo_c16_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelo.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelo_c32_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilelo_c32_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c32,_u64)(op1, op2, 2);
}
@@ -719,7 +724,7 @@ svcount_t test_svwhilelo_c32_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelo.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelo_c32_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilelo_c32_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c32,_u64)(op1, op2, 4);
}
@@ -734,7 +739,7 @@ svcount_t test_svwhilelo_c32_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelo.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelo_c64_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilelo_c64_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c64,_u64)(op1, op2, 2);
}
@@ -749,7 +754,7 @@ svcount_t test_svwhilelo_c64_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelo.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelo_c64_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilelo_c64_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c64,_u64)(op1, op2, 4);
}
@@ -767,7 +772,7 @@ svcount_t test_svwhilelo_c64_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilels.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilels_c8_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilels_c8_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c8,_u64)(op1, op2, 2);
}
@@ -782,7 +787,7 @@ svcount_t test_svwhilels_c8_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilels.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilels_c8_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilels_c8_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c8,_u64)(op1, op2, 4);
}
@@ -797,7 +802,7 @@ svcount_t test_svwhilels_c8_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilels.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilels_c16_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilels_c16_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c16,_u64)(op1, op2, 2);
}
@@ -812,7 +817,7 @@ svcount_t test_svwhilels_c16_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilels.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilels_c16_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilels_c16_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c16,_u64)(op1, op2, 4);
}
@@ -827,7 +832,7 @@ svcount_t test_svwhilels_c16_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilels.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilels_c32_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilels_c32_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c32,_u64)(op1, op2, 2);
}
@@ -842,7 +847,7 @@ svcount_t test_svwhilels_c32_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilels.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilels_c32_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilels_c32_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c32,_u64)(op1, op2, 4);
}
@@ -857,7 +862,7 @@ svcount_t test_svwhilels_c32_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilels.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilels_c64_vl2(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilels_c64_vl2(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c64,_u64)(op1, op2, 2);
}
@@ -872,7 +877,7 @@ svcount_t test_svwhilels_c64_vl2(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilels.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilels_c64_vl4(uint64_t op1, uint64_t op2)
+svcount_t test_svwhilels_c64_vl4(uint64_t op1, uint64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilele_c64,_u64)(op1, op2, 4);
}
@@ -890,7 +895,7 @@ svcount_t test_svwhilels_c64_vl4(uint64_t op1, uint64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelt.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelt_c8_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilelt_c8_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c8,_s64)(op1, op2, 2);
}
@@ -905,7 +910,7 @@ svcount_t test_svwhilelt_c8_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelt.c8(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelt_c8_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilelt_c8_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c8,_s64)(op1, op2, 4);
}
@@ -920,7 +925,7 @@ svcount_t test_svwhilelt_c8_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelt.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelt_c16_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilelt_c16_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c16,_s64)(op1, op2, 2);
}
@@ -935,7 +940,7 @@ svcount_t test_svwhilelt_c16_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelt.c16(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelt_c16_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilelt_c16_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c16,_s64)(op1, op2, 4);
}
@@ -950,7 +955,7 @@ svcount_t test_svwhilelt_c16_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelt.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelt_c32_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilelt_c32_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c32,_s64)(op1, op2, 2);
}
@@ -965,7 +970,7 @@ svcount_t test_svwhilelt_c32_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelt.c32(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelt_c32_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilelt_c32_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c32,_s64)(op1, op2, 4);
}
@@ -980,7 +985,7 @@ svcount_t test_svwhilelt_c32_vl4(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelt.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 2)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelt_c64_vl2(int64_t op1, int64_t op2)
+svcount_t test_svwhilelt_c64_vl2(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c64,_s64)(op1, op2, 2);
}
@@ -995,7 +1000,7 @@ svcount_t test_svwhilelt_c64_vl2(int64_t op1, int64_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.whilelt.c64(i64 [[OP1:%.*]], i64 [[OP2:%.*]], i32 4)
// CPP-CHECK-NEXT: ret target("aarch64.svcount") [[TMP0]]
//
-svcount_t test_svwhilelt_c64_vl4(int64_t op1, int64_t op2)
+svcount_t test_svwhilelt_c64_vl4(int64_t op1, int64_t op2) ATTR
{
return SVE_ACLE_FUNC(svwhilelt_c64,_s64)(op1, op2, 4);
}
diff --git a/clang/test/Sema/aarch64-sme2-intrinsics/acle_sme2_imm.cpp b/clang/test/Sema/aarch64-sme2-intrinsics/acle_sme2_imm.cpp
index 6a6370bf99b108..5118f743174c25 100644
--- a/clang/test/Sema/aarch64-sme2-intrinsics/acle_sme2_imm.cpp
+++ b/clang/test/Sema/aarch64-sme2-intrinsics/acle_sme2_imm.cpp
@@ -237,7 +237,7 @@ void test_svluti4_lane_zt_x2(svuint8_t zn_u8) __arm_streaming __arm_shared_za __
svluti4_lane_zt_f32_x2(0, zn_u8, 4); // expected-error {{argument value 4 is outside the valid range [0, 3]}}
}
-void test_bfmlslb_bad_lane(svfloat32_t zda, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming_compatible {
+void test_bfmlslb_bad_lane(svfloat32_t zda, svbfloat16_t zn, svbfloat16_t zm) __arm_streaming {
svbfmlslb_lane_f32(zda, zn, zm, 8); // expected-error {{argument value 8 is outside the valid range [0, 7]}}
svbfmlslt_lane_f32(zda, zn, zm, 8); // expected-error {{argument value 8 is outside the valid range [0, 7]}}
}
diff --git a/clang/test/Sema/aarch64-sme2-sve2p1-diagnostics.c b/clang/test/Sema/aarch64-sme2-sve2p1-diagnostics.c
new file mode 100644
index 00000000000000..4debc14190aa8a
--- /dev/null
+++ b/clang/test/Sema/aarch64-sme2-sve2p1-diagnostics.c
@@ -0,0 +1,37 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme -fsyntax-only -verify %s
+
+// REQUIRES: aarch64-registered-target
+#include "arm_sve.h"
+
+//svldnt1:
+
+__attribute__((target("+sme2")))
+svuint8x2_t sme2_or_sve2p1_intrinsic_test_sme2_invalid(svcount_t png, const uint8_t *rn) {
+ // expected-warning at +1 {{builtin call has undefined behaviour when called from a non-streaming function}}
+ return svldnt1_u8_x2(png, rn);
+}
+
+__attribute__((target("+sme2")))
+svint16x4_t sme2_or_sve2p1_intrinsic_test_sme2(svcount_t png, const int16_t *rn) __arm_streaming {
+ // expected-no-warning
+ return svldnt1_s16_x4(png, rn);
+}
+
+__attribute__((target("+sve2p1")))
+svuint32x2_t sme2_or_sve2p1_intrinsic_test_sve2p1(svcount_t png, const uint32_t *rn) {
+ // expected-no-warning
+ return svldnt1_u32_x2(png, rn);
+}
+
+__attribute__((target("+sme2,+sve2p1")))
+svint64x4_t sme2_or_sve2p1_intrinsic_test_both_arm_streaming(svcount_t png, const int64_t *rn) __arm_streaming {
+ // expected-no-warning
+ return svldnt1_s64_x4(png, rn);
+}
+
+__attribute__((target("+sme2,+sve2p1")))
+svint64x4_t sme2_or_sve2p1_intrinsic_test_both_no_arm_streaming(svcount_t png, const int64_t *rn) {
+ // expected-no-warning
+ return svldnt1_s64_x4(png, rn);
+}
diff --git a/clang/utils/TableGen/SveEmitter.cpp b/clang/utils/TableGen/SveEmitter.cpp
index 6c302da106a2cf..5de2223e71b046 100644
--- a/clang/utils/TableGen/SveEmitter.cpp
+++ b/clang/utils/TableGen/SveEmitter.cpp
@@ -1773,11 +1773,14 @@ void SVEEmitter::createStreamingAttrs(raw_ostream &OS, ACLEKind Kind) {
llvm::StringMap<std::set<std::string>> StreamingMap;
uint64_t IsStreamingFlag = getEnumValueForFlag("IsStreaming");
+ uint64_t IsStreamingOrSVE2p1Flag = getEnumValueForFlag("IsStreamingOrSVE2p1");
uint64_t IsStreamingCompatibleFlag =
getEnumValueForFlag("IsStreamingCompatible");
for (auto &Def : Defs) {
if (Def->isFlagSet(IsStreamingFlag))
StreamingMap["ArmStreaming"].insert(Def->getMangledName());
+ else if (Def->isFlagSet(IsStreamingOrSVE2p1Flag))
+ StreamingMap["ArmStreamingOrSVE2p1"].insert(Def->getMangledName());
else if (Def->isFlagSet(IsStreamingCompatibleFlag))
StreamingMap["ArmStreamingCompatible"].insert(Def->getMangledName());
else
More information about the cfe-commits
mailing list