Skip to content

Commit 1d2942d

Browse files
paulwalker-armgithub-actions[bot]
paulwalker-arm
authored and
github-actions[bot]
committed
Automerge: [Clang][AArch64] Add missing builtins for __ARM_FEATURE_SME2p1. (#147362)
The quadword vector instructions introduced by SVE2p1/SME2p1 but the builtins were not available to streaming mode. RAX1 is available in streaming mode when SME2p1 is available.
2 parents d428735 + a4d95c2 commit 1d2942d

19 files changed

+3677
-3450
lines changed

‎clang/include/clang/Basic/arm_sve.td

Lines changed: 59 additions & 55 deletions
Original file line numberDiff line numberDiff line change
@@ -320,23 +320,25 @@ let SVETargetGuard = "sve2p1", SMETargetGuard = InvalidMode in {
320320
// Load one vector (scalar base + vector offset)
321321
def SVLD1Q_GATHER_U64OFFSET : MInst<"svld1q_gather_[{3}]offset[_{d}]", "dPcg", "cUcsUsiUilUlfhdbm", [IsGatherLoad, IsByteIndexed], MemEltTyDefault, "aarch64_sve_ld1q_gather_vector_offset">;
322322

323-
// Load N-element structure into N vectors (scalar base)
324-
defm SVLD2Q : StructLoad<"svld2q[_{2}]", "2Pc", "aarch64_sve_ld2q_sret">;
325-
defm SVLD3Q : StructLoad<"svld3q[_{2}]", "3Pc", "aarch64_sve_ld3q_sret">;
326-
defm SVLD4Q : StructLoad<"svld4q[_{2}]", "4Pc", "aarch64_sve_ld4q_sret">;
327-
328-
// Load N-element structure into N vectors (scalar base, VL displacement)
329-
defm SVLD2Q_VNUM : StructLoad<"svld2q_vnum[_{2}]", "2Pcl", "aarch64_sve_ld2q_sret">;
330-
defm SVLD3Q_VNUM : StructLoad<"svld3q_vnum[_{2}]", "3Pcl", "aarch64_sve_ld3q_sret">;
331-
defm SVLD4Q_VNUM : StructLoad<"svld4q_vnum[_{2}]", "4Pcl", "aarch64_sve_ld4q_sret">;
332-
333323
// Load quadwords (scalar base + vector index)
334324
def SVLD1Q_GATHER_INDICES_U : MInst<"svld1q_gather_[{3}]index[_{d}]", "dPcg", "sUsiUilUlbhfd", [IsGatherLoad], MemEltTyDefault, "aarch64_sve_ld1q_gather_index">;
335325

336326
// Load quadwords (vector base + scalar index)
337327
def SVLD1Q_GATHER_INDEX_S : MInst<"svld1q_gather[_{2}base]_index_{d}", "dPgl", "sUsiUilUlbhfd", [IsGatherLoad], MemEltTyDefault, "aarch64_sve_ld1q_gather_scalar_offset">;
338328
}
339329

330+
let SVETargetGuard = "sve2p1|sme2p1", SMETargetGuard = "sve2p1|sme2p1" in {
331+
// Load N-element structure into N vectors (scalar base)
332+
defm SVLD2Q : StructLoad<"svld2q[_{2}]", "2Pc", "aarch64_sve_ld2q_sret", [VerifyRuntimeMode]>;
333+
defm SVLD3Q : StructLoad<"svld3q[_{2}]", "3Pc", "aarch64_sve_ld3q_sret", [VerifyRuntimeMode]>;
334+
defm SVLD4Q : StructLoad<"svld4q[_{2}]", "4Pc", "aarch64_sve_ld4q_sret", [VerifyRuntimeMode]>;
335+
336+
// Load N-element structure into N vectors (scalar base, VL displacement)
337+
defm SVLD2Q_VNUM : StructLoad<"svld2q_vnum[_{2}]", "2Pcl", "aarch64_sve_ld2q_sret", [VerifyRuntimeMode]>;
338+
defm SVLD3Q_VNUM : StructLoad<"svld3q_vnum[_{2}]", "3Pcl", "aarch64_sve_ld3q_sret", [VerifyRuntimeMode]>;
339+
defm SVLD4Q_VNUM : StructLoad<"svld4q_vnum[_{2}]", "4Pcl", "aarch64_sve_ld4q_sret", [VerifyRuntimeMode]>;
340+
}
341+
340342
////////////////////////////////////////////////////////////////////////////////
341343
// Stores
342344

@@ -477,16 +479,6 @@ let SVETargetGuard = "sve2p1", SMETargetGuard = InvalidMode in {
477479
def SVST1Q_SCATTER_OFFSETS_U : MInst<"svst1q_scatter_[{3}]offset[_{d}]", "vPpgd", "cUcsUsiUilUlfhdbm", [IsScatterStore, IsByteIndexed], MemEltTyDefault, "aarch64_sve_st1q_scatter_vector_offset">;
478480
def SVST1Q_SCATTER_OFFSETS_S : MInst<"svst1q_scatter_[{3}]offset[_{d}]", "vPp#d", "cUcsUsiUilUlfhdbm", [IsScatterStore, IsByteIndexed], MemEltTyDefault, "aarch64_sve_st1q_scatter_vector_offset">;
479481

480-
// Store N vectors into N-element structure (scalar base)
481-
defm SVST2Q : StructStore<"svst2q[_{d}]", "vPc2", "aarch64_sve_st2q">;
482-
defm SVST3Q : StructStore<"svst3q[_{d}]", "vPc3", "aarch64_sve_st3q">;
483-
defm SVST4Q : StructStore<"svst4q[_{d}]", "vPc4", "aarch64_sve_st4q">;
484-
485-
// Store N vectors into N-element structure (scalar base, VL displacement)
486-
defm SVST2Q_VNUM : StructStore<"svst2q_vnum[_{d}]", "vPcl2", "aarch64_sve_st2q">;
487-
defm SVST3Q_VNUM : StructStore<"svst3q_vnum[_{d}]", "vPcl3", "aarch64_sve_st3q">;
488-
defm SVST4Q_VNUM : StructStore<"svst4q_vnum[_{d}]", "vPcl4", "aarch64_sve_st4q">;
489-
490482
// Scatter store quadwords (scalar base + vector index)
491483
def SVST1Q_SCATTER_INDICES_U : MInst<"svst1q_scatter_[{3}]index[_{d}]", "vPpgd", "sUsiUilUlbhfd", [IsScatterStore], MemEltTyDefault, "aarch64_sve_st1q_scatter_index">;
492484
def SVST1Q_SCATTER_INDICES_S : MInst<"svst1q_scatter_[{3}]index[_{d}]", "vPp#d", "sUsiUilUlbhfd", [IsScatterStore], MemEltTyDefault, "aarch64_sve_st1q_scatter_index">;
@@ -495,6 +487,18 @@ let SVETargetGuard = "sve2p1", SMETargetGuard = InvalidMode in {
495487
def SVST1Q_SCATTER_INDEX_S : MInst<"svst1q_scatter[_{2}base]_index[_{d}]", "vPgld", "sUsiUilUlbhfd", [IsScatterStore], MemEltTyDefault, "aarch64_sve_st1q_scatter_scalar_offset">;
496488
}
497489

490+
let SVETargetGuard = "sve2p1|sme2p1", SMETargetGuard = "sve2p1|sme2p1" in {
491+
// Store N vectors into N-element structure (scalar base)
492+
defm SVST2Q : StructStore<"svst2q[_{d}]", "vPc2", "aarch64_sve_st2q", [VerifyRuntimeMode]>;
493+
defm SVST3Q : StructStore<"svst3q[_{d}]", "vPc3", "aarch64_sve_st3q", [VerifyRuntimeMode]>;
494+
defm SVST4Q : StructStore<"svst4q[_{d}]", "vPc4", "aarch64_sve_st4q", [VerifyRuntimeMode]>;
495+
496+
// Store N vectors into N-element structure (scalar base, VL displacement)
497+
defm SVST2Q_VNUM : StructStore<"svst2q_vnum[_{d}]", "vPcl2", "aarch64_sve_st2q", [VerifyRuntimeMode]>;
498+
defm SVST3Q_VNUM : StructStore<"svst3q_vnum[_{d}]", "vPcl3", "aarch64_sve_st3q", [VerifyRuntimeMode]>;
499+
defm SVST4Q_VNUM : StructStore<"svst4q_vnum[_{d}]", "vPcl4", "aarch64_sve_st4q", [VerifyRuntimeMode]>;
500+
}
501+
498502
////////////////////////////////////////////////////////////////////////////////
499503
// Prefetches
500504

@@ -1979,8 +1983,8 @@ def SVPMULLT_PAIR_U64 : SInst<"svpmullt_pair[_{d}]", "ddd", "Ul", MergeNone,
19791983
def SVPMULLT_PAIR_N_U64 : SInst<"svpmullt_pair[_n_{d}]", "dda", "Ul", MergeNone, "aarch64_sve_pmullt_pair">;
19801984
}
19811985

1982-
let SVETargetGuard = "sve2-sha3", SMETargetGuard = InvalidMode in { // FIXME: valid from sme2p1.
1983-
def SVRAX1 : SInst<"svrax1[_{d}]", "ddd", "lUl", MergeNone, "aarch64_sve_rax1", [IsOverloadNone]>;
1986+
let SVETargetGuard = "sve-sha3", SMETargetGuard = "sve-sha3,sme2p1" in {
1987+
def SVRAX1 : SInst<"svrax1[_{d}]", "ddd", "lUl", MergeNone, "aarch64_sve_rax1", [IsOverloadNone, VerifyRuntimeMode]>;
19841988
}
19851989

19861990
let SVETargetGuard = "sve2-sm4", SMETargetGuard = InvalidMode in {
@@ -2005,21 +2009,21 @@ def SVPSEL_D : SInst<"svpsel_lane_b64", "PPPm", "Pl", MergeNone, "", [VerifyRunt
20052009
}
20062010

20072011
// Standalone sve2.1 builtins
2008-
let SVETargetGuard = "sve2p1", SMETargetGuard = InvalidMode in {
2009-
def SVORQV : SInst<"svorqv[_{d}]", "{Pd", "csilUcUsUiUl", MergeNone, "aarch64_sve_orqv", [IsReductionQV]>;
2010-
def SVEORQV : SInst<"sveorqv[_{d}]", "{Pd", "csilUcUsUiUl", MergeNone, "aarch64_sve_eorqv", [IsReductionQV]>;
2011-
def SVADDQV : SInst<"svaddqv[_{d}]", "{Pd", "csilUcUsUiUl", MergeNone, "aarch64_sve_addqv", [IsReductionQV]>;
2012-
def SVANDQV : SInst<"svandqv[_{d}]", "{Pd", "csilUcUsUiUl", MergeNone, "aarch64_sve_andqv", [IsReductionQV]>;
2013-
def SVSMAXQV : SInst<"svmaxqv[_{d}]", "{Pd", "csil", MergeNone, "aarch64_sve_smaxqv", [IsReductionQV]>;
2014-
def SVUMAXQV : SInst<"svmaxqv[_{d}]", "{Pd", "UcUsUiUl", MergeNone, "aarch64_sve_umaxqv", [IsReductionQV]>;
2015-
def SVSMINQV : SInst<"svminqv[_{d}]", "{Pd", "csil", MergeNone, "aarch64_sve_sminqv", [IsReductionQV]>;
2016-
def SVUMINQV : SInst<"svminqv[_{d}]", "{Pd", "UcUsUiUl", MergeNone, "aarch64_sve_uminqv", [IsReductionQV]>;
2017-
2018-
def SVFADDQV : SInst<"svaddqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_faddqv", [IsReductionQV]>;
2019-
def SVFMAXNMQV : SInst<"svmaxnmqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_fmaxnmqv", [IsReductionQV]>;
2020-
def SVFMINNMQV : SInst<"svminnmqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_fminnmqv", [IsReductionQV]>;
2021-
def SVFMAXQV : SInst<"svmaxqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_fmaxqv", [IsReductionQV]>;
2022-
def SVFMINQV : SInst<"svminqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_fminqv", [IsReductionQV]>;
2012+
let SVETargetGuard = "sve2p1|sme2p1", SMETargetGuard = "sve2p1|sme2p1" in {
2013+
def SVORQV : SInst<"svorqv[_{d}]", "{Pd", "csilUcUsUiUl", MergeNone, "aarch64_sve_orqv", [IsReductionQV, VerifyRuntimeMode]>;
2014+
def SVEORQV : SInst<"sveorqv[_{d}]", "{Pd", "csilUcUsUiUl", MergeNone, "aarch64_sve_eorqv", [IsReductionQV, VerifyRuntimeMode]>;
2015+
def SVADDQV : SInst<"svaddqv[_{d}]", "{Pd", "csilUcUsUiUl", MergeNone, "aarch64_sve_addqv", [IsReductionQV, VerifyRuntimeMode]>;
2016+
def SVANDQV : SInst<"svandqv[_{d}]", "{Pd", "csilUcUsUiUl", MergeNone, "aarch64_sve_andqv", [IsReductionQV, VerifyRuntimeMode]>;
2017+
def SVSMAXQV : SInst<"svmaxqv[_{d}]", "{Pd", "csil", MergeNone, "aarch64_sve_smaxqv", [IsReductionQV, VerifyRuntimeMode]>;
2018+
def SVUMAXQV : SInst<"svmaxqv[_{d}]", "{Pd", "UcUsUiUl", MergeNone, "aarch64_sve_umaxqv", [IsReductionQV, VerifyRuntimeMode]>;
2019+
def SVSMINQV : SInst<"svminqv[_{d}]", "{Pd", "csil", MergeNone, "aarch64_sve_sminqv", [IsReductionQV, VerifyRuntimeMode]>;
2020+
def SVUMINQV : SInst<"svminqv[_{d}]", "{Pd", "UcUsUiUl", MergeNone, "aarch64_sve_uminqv", [IsReductionQV, VerifyRuntimeMode]>;
2021+
2022+
def SVFADDQV : SInst<"svaddqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_faddqv", [IsReductionQV, VerifyRuntimeMode]>;
2023+
def SVFMAXNMQV : SInst<"svmaxnmqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_fmaxnmqv", [IsReductionQV, VerifyRuntimeMode]>;
2024+
def SVFMINNMQV : SInst<"svminnmqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_fminnmqv", [IsReductionQV, VerifyRuntimeMode]>;
2025+
def SVFMAXQV : SInst<"svmaxqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_fmaxqv", [IsReductionQV, VerifyRuntimeMode]>;
2026+
def SVFMINQV : SInst<"svminqv[_{d}]", "{Pd", "hfd", MergeNone, "aarch64_sve_fminqv", [IsReductionQV, VerifyRuntimeMode]>;
20232027
}
20242028

20252029
let SVETargetGuard = "sve2p1", SMETargetGuard = "sme2" in {
@@ -2281,49 +2285,49 @@ let SVETargetGuard = "sve2p1|sme2", SMETargetGuard = "sve2p1|sme2" in {
22812285
def SVSQRSHRUN_X2 : SInst<"svqrshrun[_n]_{0}[_{d}_x2]", "e2i", "i", MergeNone, "aarch64_sve_sqrshrun_x2", [VerifyRuntimeMode], [ImmCheck<1, ImmCheck1_16>]>;
22822286
}
22832287

2284-
let SVETargetGuard = "sve2p1", SMETargetGuard = InvalidMode in {
2288+
let SVETargetGuard = "sve2p1|sme2p1", SMETargetGuard = "sve2p1|sme2p1" in {
22852289
// ZIPQ1, ZIPQ2, UZPQ1, UZPQ2
2286-
def SVZIPQ1 : SInst<"svzipq1[_{d}]", "ddd", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_zipq1", [], []>;
2287-
def SVZIPQ2 : SInst<"svzipq2[_{d}]", "ddd", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_zipq2", [], []>;
2288-
def SVUZPQ1 : SInst<"svuzpq1[_{d}]", "ddd", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_uzpq1", [], []>;
2289-
def SVUZPQ2 : SInst<"svuzpq2[_{d}]", "ddd", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_uzpq2", [], []>;
2290+
def SVZIPQ1 : SInst<"svzipq1[_{d}]", "ddd", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_zipq1", [VerifyRuntimeMode], []>;
2291+
def SVZIPQ2 : SInst<"svzipq2[_{d}]", "ddd", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_zipq2", [VerifyRuntimeMode], []>;
2292+
def SVUZPQ1 : SInst<"svuzpq1[_{d}]", "ddd", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_uzpq1", [VerifyRuntimeMode], []>;
2293+
def SVUZPQ2 : SInst<"svuzpq2[_{d}]", "ddd", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_uzpq2", [VerifyRuntimeMode], []>;
22902294
// TBLQ, TBXQ
2291-
def SVTBLQ : SInst<"svtblq[_{d}]", "ddu", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_tblq">;
2292-
def SVTBXQ : SInst<"svtbxq[_{d}]", "dddu", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_tbxq">;
2295+
def SVTBLQ : SInst<"svtblq[_{d}]", "ddu", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_tblq", [VerifyRuntimeMode]>;
2296+
def SVTBXQ : SInst<"svtbxq[_{d}]", "dddu", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_tbxq", [VerifyRuntimeMode]>;
22932297
// EXTQ
2294-
def EXTQ : SInst<"svextq[_{d}]", "dddk", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_extq", [], [ImmCheck<2, ImmCheckLaneIndex, 0>]>;
2298+
def EXTQ : SInst<"svextq[_{d}]", "dddk", "cUcsUsiUilUlbhfdm", MergeNone, "aarch64_sve_extq", [VerifyRuntimeMode], [ImmCheck<2, ImmCheckLaneIndex, 0>]>;
22952299

22962300
// PMOV
22972301
// Move to Pred
22982302
multiclass PMOV_TO_PRED<string name, string types, string intrinsic, list<FlagType> flags=[], ImmCheckType immCh > {
22992303
def _LANE : Inst<name # "_lane[_{d}]", "Pdi", types, MergeNone, intrinsic, flags, [ImmCheck<1, immCh>]>;
23002304
def _LANE_ZERO : SInst<name # "[_{d}]", "Pd", types, MergeNone, intrinsic # "_zero", flags, []>;
23012305
}
2302-
defm SVPMOV_B_TO_PRED : PMOV_TO_PRED<"svpmov", "cUc", "aarch64_sve_pmov_to_pred_lane", [], ImmCheck0_0>;
2303-
defm SVPMOV_H_TO_PRED : PMOV_TO_PRED<"svpmov", "sUs", "aarch64_sve_pmov_to_pred_lane", [], ImmCheck0_1>;
2304-
defm SVPMOV_S_TO_PRED : PMOV_TO_PRED<"svpmov", "iUi", "aarch64_sve_pmov_to_pred_lane", [], ImmCheck0_3>;
2305-
defm SVPMOV_D_TO_PRED : PMOV_TO_PRED<"svpmov", "lUl", "aarch64_sve_pmov_to_pred_lane", [], ImmCheck0_7>;
2306+
defm SVPMOV_B_TO_PRED : PMOV_TO_PRED<"svpmov", "cUc", "aarch64_sve_pmov_to_pred_lane", [VerifyRuntimeMode], ImmCheck0_0>;
2307+
defm SVPMOV_H_TO_PRED : PMOV_TO_PRED<"svpmov", "sUs", "aarch64_sve_pmov_to_pred_lane", [VerifyRuntimeMode], ImmCheck0_1>;
2308+
defm SVPMOV_S_TO_PRED : PMOV_TO_PRED<"svpmov", "iUi", "aarch64_sve_pmov_to_pred_lane", [VerifyRuntimeMode], ImmCheck0_3>;
2309+
defm SVPMOV_D_TO_PRED : PMOV_TO_PRED<"svpmov", "lUl", "aarch64_sve_pmov_to_pred_lane", [VerifyRuntimeMode], ImmCheck0_7>;
23062310

23072311
// Move to Vector
23082312
multiclass PMOV_TO_VEC<string name, string types, string intrinsic, list<FlagType> flags=[], ImmCheckType immCh > {
23092313
def _M : SInst<name # "_lane[_{d}]", "ddPi", types, MergeOp1, intrinsic # "_merging", flags, [ImmCheck<2, immCh>]>;
23102314
def _Z : SInst<name # "_{d}_z", "dP", types, MergeNone, intrinsic # "_zeroing", flags, []>;
23112315
}
2312-
def SVPMOV_TO_VEC_LANE_B : SInst<"svpmov_{d}_z", "dP", "cUc", MergeNone, "aarch64_sve_pmov_to_vector_lane_zeroing", [], []>;
2313-
defm SVPMOV_TO_VEC_LANE_H : PMOV_TO_VEC<"svpmov", "sUs", "aarch64_sve_pmov_to_vector_lane", [], ImmCheck1_1>;
2314-
defm SVPMOV_TO_VEC_LANE_S : PMOV_TO_VEC<"svpmov", "iUi", "aarch64_sve_pmov_to_vector_lane", [], ImmCheck1_3>;
2315-
defm SVPMOV_TO_VEC_LANE_D : PMOV_TO_VEC<"svpmov", "lUl", "aarch64_sve_pmov_to_vector_lane" ,[], ImmCheck1_7>;
2316+
def SVPMOV_TO_VEC_LANE_B : SInst<"svpmov_{d}_z", "dP", "cUc", MergeNone, "aarch64_sve_pmov_to_vector_lane_zeroing", [VerifyRuntimeMode], []>;
2317+
defm SVPMOV_TO_VEC_LANE_H : PMOV_TO_VEC<"svpmov", "sUs", "aarch64_sve_pmov_to_vector_lane", [VerifyRuntimeMode], ImmCheck1_1>;
2318+
defm SVPMOV_TO_VEC_LANE_S : PMOV_TO_VEC<"svpmov", "iUi", "aarch64_sve_pmov_to_vector_lane", [VerifyRuntimeMode], ImmCheck1_3>;
2319+
defm SVPMOV_TO_VEC_LANE_D : PMOV_TO_VEC<"svpmov", "lUl", "aarch64_sve_pmov_to_vector_lane" ,[VerifyRuntimeMode], ImmCheck1_7>;
23162320
}
23172321

2318-
let SVETargetGuard = "sve2p1", SMETargetGuard = "sme2p1" in {
2322+
let SVETargetGuard = "sve2p1|sme2p1", SMETargetGuard = "sve2p1|sme2p1" in {
23192323
// DUPQ
23202324
def SVDUP_LANEQ_B : SInst<"svdup_laneq[_{d}]", "ddi", "cUcm", MergeNone, "aarch64_sve_dup_laneq", [VerifyRuntimeMode], [ImmCheck<1, ImmCheck0_15>]>;
23212325
def SVDUP_LANEQ_H : SInst<"svdup_laneq[_{d}]", "ddi", "sUsh", MergeNone, "aarch64_sve_dup_laneq", [VerifyRuntimeMode], [ImmCheck<1, ImmCheck0_7>]>;
23222326
def SVDUP_LANEQ_S : SInst<"svdup_laneq[_{d}]", "ddi", "iUif", MergeNone, "aarch64_sve_dup_laneq", [VerifyRuntimeMode], [ImmCheck<1, ImmCheck0_3>]>;
23232327
def SVDUP_LANEQ_D : SInst<"svdup_laneq[_{d}]", "ddi", "lUld", MergeNone, "aarch64_sve_dup_laneq", [VerifyRuntimeMode], [ImmCheck<1, ImmCheck0_1>]>;
23242328
}
23252329

2326-
let SVETargetGuard = "sve2p1,bf16", SMETargetGuard = "sme2p1,bf16" in {
2330+
let SVETargetGuard = "(sve2p1|sme2p1),bf16", SMETargetGuard = "(sve2p1|sme2p1),bf16" in {
23272331
def SVDUP_LANEQ_BF16 : SInst<"svdup_laneq[_{d}]", "ddi", "b", MergeNone, "aarch64_sve_dup_laneq", [VerifyRuntimeMode], [ImmCheck<1, ImmCheck0_7>]>;
23282332
}
23292333

‎clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_rax1.c

Lines changed: 16 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -1,8 +1,10 @@
11
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
2-
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sve -target-feature +sve2 -target-feature +sve2-sha3 -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
3-
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sve -target-feature +sve2 -target-feature +sve2-sha3 -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
4-
// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64 -target-feature +sve -target-feature +sve2 -target-feature +sve2-sha3 -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
5-
// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64 -target-feature +sve -target-feature +sve2 -target-feature +sve2-sha3 -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
2+
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sve -target-feature +sve2 -target-feature +sve-sha3 -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
3+
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sme -target-feature +sme2 -target-feature +sme2p1 -target-feature +sve-sha3 -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
4+
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sve -target-feature +sve2 -target-feature +sme2p1 -target-feature +sve-sha3 -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
5+
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sve -target-feature +sve2 -target-feature +sve-sha3 -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
6+
// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64 -target-feature +sve -target-feature +sve2 -target-feature +sve-sha3 -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
7+
// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64 -target-feature +sve -target-feature +sve2 -target-feature +sve-sha3 -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
68

79
// REQUIRES: aarch64-registered-target
810

@@ -15,6 +17,14 @@
1517
#define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
1618
#endif
1719

20+
#if defined(__ARM_FEATURE_SME) && defined(__ARM_FEATURE_SVE)
21+
#define ATTR __arm_streaming_compatible
22+
#elif defined(__ARM_FEATURE_SME)
23+
#define ATTR __arm_streaming
24+
#else
25+
#define ATTR
26+
#endif
27+
1828
// CHECK-LABEL: @test_svrax1_s64(
1929
// CHECK-NEXT: entry:
2030
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.rax1(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
@@ -25,7 +35,7 @@
2535
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.rax1(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
2636
// CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP0]]
2737
//
28-
svint64_t test_svrax1_s64(svint64_t op1, svint64_t op2)
38+
svint64_t test_svrax1_s64(svint64_t op1, svint64_t op2) ATTR
2939
{
3040
return SVE_ACLE_FUNC(svrax1,_s64,,)(op1, op2);
3141
}
@@ -40,7 +50,7 @@ svint64_t test_svrax1_s64(svint64_t op1, svint64_t op2)
4050
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.rax1(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
4151
// CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP0]]
4252
//
43-
svuint64_t test_svrax1_u64(svuint64_t op1, svuint64_t op2)
53+
svuint64_t test_svrax1_u64(svuint64_t op1, svuint64_t op2) ATTR
4454
{
4555
return SVE_ACLE_FUNC(svrax1,_u64,,)(op1, op2);
4656
}

0 commit comments

Comments
 (0)