[llvm] [ARM][AArch64] Split processor and feature definitions out of the ARM.td/AArch64.td [NFC] (PR #88282)
Lucas Duarte Prates via llvm-commits
llvm-commits at lists.llvm.org
Mon Apr 15 01:51:53 PDT 2024
================
@@ -0,0 +1,1328 @@
+//=- AArch64Features.td - Describe AArch64 SubtargetFeatures -*- tablegen -*-=//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+//
+//===----------------------------------------------------------------------===//
+
+// Each SubtargetFeature which corresponds to an Arm Architecture feature should
+// be annotated with the respective FEAT_ feature name from the Architecture
+// Reference Manual. If a SubtargetFeature enables instructions from multiple
+// Arm Architecture Features, it should list all the relevant features. Not all
+// FEAT_ features have a corresponding SubtargetFeature.
+
+def FeatureFPARMv8 : SubtargetFeature<"fp-armv8", "HasFPARMv8", "true",
+ "Enable ARMv8 FP (FEAT_FP)">;
+
+def FeatureNEON : SubtargetFeature<"neon", "HasNEON", "true",
+ "Enable Advanced SIMD instructions (FEAT_AdvSIMD)", [FeatureFPARMv8]>;
+
+def FeatureSM4 : SubtargetFeature<
+ "sm4", "HasSM4", "true",
+ "Enable SM3 and SM4 support (FEAT_SM4, FEAT_SM3)", [FeatureNEON]>;
+
+def FeatureSHA2 : SubtargetFeature<
+ "sha2", "HasSHA2", "true",
+ "Enable SHA1 and SHA256 support (FEAT_SHA1, FEAT_SHA256)", [FeatureNEON]>;
+
+def FeatureSHA3 : SubtargetFeature<
+ "sha3", "HasSHA3", "true",
+ "Enable SHA512 and SHA3 support (FEAT_SHA3, FEAT_SHA512)", [FeatureNEON, FeatureSHA2]>;
+
+def FeatureAES : SubtargetFeature<
+ "aes", "HasAES", "true",
+ "Enable AES support (FEAT_AES, FEAT_PMULL)", [FeatureNEON]>;
+
+// Crypto has been split up and any combination is now valid (see the
+// crypto definitions above). Also, crypto is now context sensitive:
+// it has a different meaning for e.g. Armv8.4 than it has for Armv8.2.
+// Therefore, we rely on Clang, the user interfacing tool, to pass on the
+// appropriate crypto options. But here in the backend, crypto has very little
+// meaning anymore. We kept the Crypto definition here for backward
+// compatibility, and now imply features SHA2 and AES, which was the
+// "traditional" meaning of Crypto.
+def FeatureCrypto : SubtargetFeature<"crypto", "HasCrypto", "true",
+ "Enable cryptographic instructions", [FeatureNEON, FeatureSHA2, FeatureAES]>;
+
+def FeatureCRC : SubtargetFeature<"crc", "HasCRC", "true",
+ "Enable ARMv8 CRC-32 checksum instructions (FEAT_CRC32)">;
+
+def FeatureRAS : SubtargetFeature<"ras", "HasRAS", "true",
+ "Enable ARMv8 Reliability, Availability and Serviceability Extensions (FEAT_RAS, FEAT_RASv1p1)">;
+
+def FeatureRASv2 : SubtargetFeature<"rasv2", "HasRASv2", "true",
+ "Enable ARMv8.9-A Reliability, Availability and Serviceability Extensions (FEAT_RASv2)",
+ [FeatureRAS]>;
+
+def FeatureLSE : SubtargetFeature<"lse", "HasLSE", "true",
+ "Enable ARMv8.1 Large System Extension (LSE) atomic instructions (FEAT_LSE)">;
+
+def FeatureLSE2 : SubtargetFeature<"lse2", "HasLSE2", "true",
+ "Enable ARMv8.4 Large System Extension 2 (LSE2) atomicity rules (FEAT_LSE2)">;
+
+def FeatureOutlineAtomics : SubtargetFeature<"outline-atomics", "OutlineAtomics", "true",
+ "Enable out of line atomics to support LSE instructions">;
+
+def FeatureFMV : SubtargetFeature<"fmv", "HasFMV", "true",
+ "Enable Function Multi Versioning support.">;
+
+def FeatureRDM : SubtargetFeature<"rdm", "HasRDM", "true",
+ "Enable ARMv8.1 Rounding Double Multiply Add/Subtract instructions (FEAT_RDM)",
+ [FeatureNEON]>;
+
+def FeaturePAN : SubtargetFeature<
+ "pan", "HasPAN", "true",
+ "Enables ARM v8.1 Privileged Access-Never extension (FEAT_PAN)">;
+
+def FeatureLOR : SubtargetFeature<
+ "lor", "HasLOR", "true",
+ "Enables ARM v8.1 Limited Ordering Regions extension (FEAT_LOR)">;
+
+def FeatureCONTEXTIDREL2 : SubtargetFeature<"CONTEXTIDREL2", "HasCONTEXTIDREL2",
+ "true", "Enable RW operand CONTEXTIDR_EL2" >;
+
+def FeatureVH : SubtargetFeature<"vh", "HasVH", "true",
+ "Enables ARM v8.1 Virtual Host extension (FEAT_VHE)", [FeatureCONTEXTIDREL2] >;
+
+// This SubtargetFeature is special. It controls only whether codegen will turn
+// `llvm.readcyclecounter()` into an access to a PMUv3 System Register. The
+// `FEAT_PMUv3*` system registers are always available for assembly/disassembly.
+def FeaturePerfMon : SubtargetFeature<"perfmon", "HasPerfMon", "true",
+ "Enable Code Generation for ARMv8 PMUv3 Performance Monitors extension (FEAT_PMUv3)">;
+
+def FeatureFullFP16 : SubtargetFeature<"fullfp16", "HasFullFP16", "true",
+ "Full FP16 (FEAT_FP16)", [FeatureFPARMv8]>;
+
+def FeatureFP16FML : SubtargetFeature<"fp16fml", "HasFP16FML", "true",
+ "Enable FP16 FML instructions (FEAT_FHM)", [FeatureFullFP16]>;
+
+def FeatureSPE : SubtargetFeature<"spe", "HasSPE", "true",
+ "Enable Statistical Profiling extension (FEAT_SPE)">;
+
+def FeaturePAN_RWV : SubtargetFeature<
+ "pan-rwv", "HasPAN_RWV", "true",
+ "Enable v8.2 PAN s1e1R and s1e1W Variants (FEAT_PAN2)",
+ [FeaturePAN]>;
+
+// UAO PState
+def FeaturePsUAO : SubtargetFeature< "uaops", "HasPsUAO", "true",
+ "Enable v8.2 UAO PState (FEAT_UAO)">;
+
+def FeatureCCPP : SubtargetFeature<"ccpp", "HasCCPP",
+ "true", "Enable v8.2 data Cache Clean to Point of Persistence (FEAT_DPB)" >;
+
+def FeatureSVE : SubtargetFeature<"sve", "HasSVE", "true",
+ "Enable Scalable Vector Extension (SVE) instructions (FEAT_SVE)", [FeatureFullFP16]>;
+
+def FeatureFPMR : SubtargetFeature<"fpmr", "HasFPMR", "true",
+ "Enable FPMR Register (FEAT_FPMR)">;
+
+def FeatureFP8 : SubtargetFeature<"fp8", "HasFP8", "true",
+ "Enable FP8 instructions (FEAT_FP8)">;
+
+// This flag is currently still labeled as Experimental, but when fully
+// implemented this should tell the compiler to use the zeroing pseudos to
+// benefit from the reverse instructions (e.g. SUB vs SUBR) if the inactive
+// lanes are known to be zero. The pseudos will then be expanded using the
+// MOVPRFX instruction to zero the inactive lanes. This feature should only be
+// enabled if MOVPRFX instructions are known to merge with the destructive
+// operations they prefix.
+//
+// This feature could similarly be extended to support cheap merging of _any_
+// value into the inactive lanes using the MOVPRFX instruction that uses
+// merging-predication.
+def FeatureExperimentalZeroingPseudos
+ : SubtargetFeature<"use-experimental-zeroing-pseudos",
+ "UseExperimentalZeroingPseudos", "true",
+ "Hint to the compiler that the MOVPRFX instruction is "
+ "merged with destructive operations",
+ []>;
+
+def FeatureUseScalarIncVL : SubtargetFeature<"use-scalar-inc-vl",
+ "UseScalarIncVL", "true", "Prefer inc/dec over add+cnt">;
+
+def FeatureBF16 : SubtargetFeature<"bf16", "HasBF16",
+ "true", "Enable BFloat16 Extension (FEAT_BF16)" >;
+
+def FeatureNoSVEFPLD1R : SubtargetFeature<"no-sve-fp-ld1r",
+ "NoSVEFPLD1R", "true", "Avoid using LD1RX instructions for FP">;
+
+def FeatureSVE2 : SubtargetFeature<"sve2", "HasSVE2", "true",
+ "Enable Scalable Vector Extension 2 (SVE2) instructions (FEAT_SVE2)",
+ [FeatureSVE, FeatureUseScalarIncVL]>;
+
+def FeatureSVE2AES : SubtargetFeature<"sve2-aes", "HasSVE2AES", "true",
+ "Enable AES SVE2 instructions (FEAT_SVE_AES, FEAT_SVE_PMULL128)",
+ [FeatureSVE2, FeatureAES]>;
+
+def FeatureSVE2SM4 : SubtargetFeature<"sve2-sm4", "HasSVE2SM4", "true",
+ "Enable SM4 SVE2 instructions (FEAT_SVE_SM4)", [FeatureSVE2, FeatureSM4]>;
+
+def FeatureSVE2SHA3 : SubtargetFeature<"sve2-sha3", "HasSVE2SHA3", "true",
+ "Enable SHA3 SVE2 instructions (FEAT_SVE_SHA3)", [FeatureSVE2, FeatureSHA3]>;
+
+def FeatureSVE2BitPerm : SubtargetFeature<"sve2-bitperm", "HasSVE2BitPerm", "true",
+ "Enable bit permutation SVE2 instructions (FEAT_SVE_BitPerm)", [FeatureSVE2]>;
+
+def FeatureSVE2p1: SubtargetFeature<"sve2p1", "HasSVE2p1", "true",
+ "Enable Scalable Vector Extension 2.1 instructions", [FeatureSVE2]>;
+
+def FeatureB16B16 : SubtargetFeature<"b16b16", "HasB16B16", "true",
+ "Enable SVE2.1 or SME2.1 non-widening BFloat16 to BFloat16 instructions (FEAT_B16B16)", [FeatureBF16]>;
+
+def FeatureZCRegMove : SubtargetFeature<"zcm", "HasZeroCycleRegMove", "true",
+ "Has zero-cycle register moves">;
+
+def FeatureZCZeroingGP : SubtargetFeature<"zcz-gp", "HasZeroCycleZeroingGP", "true",
+ "Has zero-cycle zeroing instructions for generic registers">;
+
+// It is generally beneficial to rewrite "fmov s0, wzr" to "movi d0, #0".
+// as movi is more efficient across all cores. Newer cores can eliminate
+// fmovs early and there is no difference with movi, but this not true for
+// all implementations.
+def FeatureNoZCZeroingFP : SubtargetFeature<"no-zcz-fp", "HasZeroCycleZeroingFP", "false",
+ "Has no zero-cycle zeroing instructions for FP registers">;
+
+def FeatureZCZeroing : SubtargetFeature<"zcz", "HasZeroCycleZeroing", "true",
+ "Has zero-cycle zeroing instructions",
+ [FeatureZCZeroingGP]>;
+
+/// ... but the floating-point version doesn't quite work in rare cases on older
+/// CPUs.
+def FeatureZCZeroingFPWorkaround : SubtargetFeature<"zcz-fp-workaround",
+ "HasZeroCycleZeroingFPWorkaround", "true",
+ "The zero-cycle floating-point zeroing instruction has a bug">;
+
+def FeatureStrictAlign : SubtargetFeature<"strict-align",
+ "RequiresStrictAlign", "true",
+ "Disallow all unaligned memory "
+ "access">;
+
+foreach i = {1-7,9-15,18,20-28} in
+ def FeatureReserveX#i : SubtargetFeature<"reserve-x"#i, "ReserveXRegister["#i#"]", "true",
+ "Reserve X"#i#", making it unavailable "
+ "as a GPR">;
+
+foreach i = {8-15,18} in
+ def FeatureCallSavedX#i : SubtargetFeature<"call-saved-x"#i,
+ "CustomCallSavedXRegs["#i#"]", "true", "Make X"#i#" callee saved.">;
+
+def FeatureBalanceFPOps : SubtargetFeature<"balance-fp-ops", "BalanceFPOps",
+ "true",
+ "balance mix of odd and even D-registers for fp multiply(-accumulate) ops">;
+
+def FeaturePredictableSelectIsExpensive : SubtargetFeature<
+ "predictable-select-expensive", "PredictableSelectIsExpensive", "true",
+ "Prefer likely predicted branches over selects">;
+
+def FeatureEnableSelectOptimize : SubtargetFeature<
+ "enable-select-opt", "EnableSelectOptimize", "true",
+ "Enable the select optimize pass for select loop heuristics">;
+
+def FeatureExynosCheapAsMoveHandling : SubtargetFeature<"exynos-cheap-as-move",
+ "HasExynosCheapAsMoveHandling", "true",
+ "Use Exynos specific handling of cheap instructions">;
+
+def FeaturePostRAScheduler : SubtargetFeature<"use-postra-scheduler",
+ "UsePostRAScheduler", "true", "Schedule again after register allocation">;
+
+def FeatureSlowMisaligned128Store : SubtargetFeature<"slow-misaligned-128store",
+ "IsMisaligned128StoreSlow", "true", "Misaligned 128 bit stores are slow">;
+
+def FeatureSlowPaired128 : SubtargetFeature<"slow-paired-128",
+ "IsPaired128Slow", "true", "Paired 128 bit loads and stores are slow">;
+
+def FeatureAscendStoreAddress : SubtargetFeature<"ascend-store-address",
+ "IsStoreAddressAscend", "true",
+ "Schedule vector stores by ascending address">;
+
+def FeatureSlowSTRQro : SubtargetFeature<"slow-strqro-store", "IsSTRQroSlow",
+ "true", "STR of Q register with register offset is slow">;
+
+def FeatureAlternateSExtLoadCVTF32Pattern : SubtargetFeature<
+ "alternate-sextload-cvt-f32-pattern", "UseAlternateSExtLoadCVTF32Pattern",
+ "true", "Use alternative pattern for sextload convert to f32">;
+
+def FeatureArithmeticBccFusion : SubtargetFeature<
+ "arith-bcc-fusion", "HasArithmeticBccFusion", "true",
+ "CPU fuses arithmetic+bcc operations">;
+
+def FeatureArithmeticCbzFusion : SubtargetFeature<
+ "arith-cbz-fusion", "HasArithmeticCbzFusion", "true",
+ "CPU fuses arithmetic + cbz/cbnz operations">;
+
+def FeatureCmpBccFusion : SubtargetFeature<
+ "cmp-bcc-fusion", "HasCmpBccFusion", "true",
+ "CPU fuses cmp+bcc operations">;
+
+def FeatureFuseAddress : SubtargetFeature<
+ "fuse-address", "HasFuseAddress", "true",
+ "CPU fuses address generation and memory operations">;
+
+def FeatureFuseAES : SubtargetFeature<
+ "fuse-aes", "HasFuseAES", "true",
+ "CPU fuses AES crypto operations">;
+
+def FeatureFuseArithmeticLogic : SubtargetFeature<
+ "fuse-arith-logic", "HasFuseArithmeticLogic", "true",
+ "CPU fuses arithmetic and logic operations">;
+
+def FeatureFuseCCSelect : SubtargetFeature<
+ "fuse-csel", "HasFuseCCSelect", "true",
+ "CPU fuses conditional select operations">;
+
+def FeatureFuseCryptoEOR : SubtargetFeature<
+ "fuse-crypto-eor", "HasFuseCryptoEOR", "true",
+ "CPU fuses AES/PMULL and EOR operations">;
+
+def FeatureFuseAdrpAdd : SubtargetFeature<
+ "fuse-adrp-add", "HasFuseAdrpAdd", "true",
+ "CPU fuses adrp+add operations">;
+
+def FeatureFuseLiterals : SubtargetFeature<
+ "fuse-literals", "HasFuseLiterals", "true",
+ "CPU fuses literal generation operations">;
+
+def FeatureFuseAddSub2RegAndConstOne : SubtargetFeature<
+ "fuse-addsub-2reg-const1", "HasFuseAddSub2RegAndConstOne", "true",
+ "CPU fuses (a + b + 1) and (a - b - 1)">;
+
+def FeatureDisableLatencySchedHeuristic : SubtargetFeature<
+ "disable-latency-sched-heuristic", "DisableLatencySchedHeuristic", "true",
+ "Disable latency scheduling heuristic">;
+
+def FeatureStorePairSuppress : SubtargetFeature<
+ "store-pair-suppress", "EnableStorePairSuppress", "true",
+ "Enable Store Pair Suppression heuristics">;
+
+def FeatureForce32BitJumpTables
+ : SubtargetFeature<"force-32bit-jump-tables", "Force32BitJumpTables", "true",
+ "Force jump table entries to be 32-bits wide except at MinSize">;
+
+def FeatureRCPC : SubtargetFeature<"rcpc", "HasRCPC", "true",
+ "Enable support for RCPC extension (FEAT_LRCPC)">;
+
+def FeatureUseRSqrt : SubtargetFeature<
+ "use-reciprocal-square-root", "UseRSqrt", "true",
+ "Use the reciprocal square root approximation">;
+
+def FeatureDotProd : SubtargetFeature<
+ "dotprod", "HasDotProd", "true",
+ "Enable dot product support (FEAT_DotProd)", [FeatureNEON]>;
+
+def FeaturePAuth : SubtargetFeature<
+ "pauth", "HasPAuth", "true",
+ "Enable v8.3-A Pointer Authentication extension (FEAT_PAuth)">;
+
+def FeatureJS : SubtargetFeature<
+ "jsconv", "HasJS", "true",
+ "Enable v8.3-A JavaScript FP conversion instructions (FEAT_JSCVT)",
+ [FeatureFPARMv8]>;
+
+def FeatureCCIDX : SubtargetFeature<
+ "ccidx", "HasCCIDX", "true",
+ "Enable v8.3-A Extend of the CCSIDR number of sets (FEAT_CCIDX)">;
+
+def FeatureComplxNum : SubtargetFeature<
+ "complxnum", "HasComplxNum", "true",
+ "Enable v8.3-A Floating-point complex number support (FEAT_FCMA)",
+ [FeatureNEON]>;
+
+def FeatureNV : SubtargetFeature<
+ "nv", "HasNV", "true",
+ "Enable v8.4-A Nested Virtualization Enchancement (FEAT_NV, FEAT_NV2)">;
+
+def FeatureMPAM : SubtargetFeature<
+ "mpam", "HasMPAM", "true",
+ "Enable v8.4-A Memory system Partitioning and Monitoring extension (FEAT_MPAM)">;
+
+def FeatureDIT : SubtargetFeature<
+ "dit", "HasDIT", "true",
+ "Enable v8.4-A Data Independent Timing instructions (FEAT_DIT)">;
+
+def FeatureTRACEV8_4 : SubtargetFeature<
+ "tracev8.4", "HasTRACEV8_4", "true",
+ "Enable v8.4-A Trace extension (FEAT_TRF)">;
+
+def FeatureAM : SubtargetFeature<
+ "am", "HasAM", "true",
+ "Enable v8.4-A Activity Monitors extension (FEAT_AMUv1)">;
+
+def FeatureAMVS : SubtargetFeature<
+ "amvs", "HasAMVS", "true",
+ "Enable v8.6-A Activity Monitors Virtualization support (FEAT_AMUv1p1)",
+ [FeatureAM]>;
+
+def FeatureSEL2 : SubtargetFeature<
+ "sel2", "HasSEL2", "true",
+ "Enable v8.4-A Secure Exception Level 2 extension (FEAT_SEL2)">;
+
+def FeatureTLB_RMI : SubtargetFeature<
+ "tlb-rmi", "HasTLB_RMI", "true",
+ "Enable v8.4-A TLB Range and Maintenance Instructions (FEAT_TLBIOS, FEAT_TLBIRANGE)">;
+
+def FeatureFlagM : SubtargetFeature<
+ "flagm", "HasFlagM", "true",
+ "Enable v8.4-A Flag Manipulation Instructions (FEAT_FlagM)">;
+
+// 8.4 RCPC enchancements: LDAPR & STLR instructions with Immediate Offset
+def FeatureRCPC_IMMO : SubtargetFeature<"rcpc-immo", "HasRCPC_IMMO", "true",
+ "Enable v8.4-A RCPC instructions with Immediate Offsets (FEAT_LRCPC2)",
+ [FeatureRCPC]>;
+
+def FeatureNoNegativeImmediates : SubtargetFeature<"no-neg-immediates",
+ "NegativeImmediates", "false",
+ "Convert immediates and instructions "
+ "to their negated or complemented "
+ "equivalent when the immediate does "
+ "not fit in the encoding.">;
+
+// Address operands with shift amount 2 or 3 are fast on all Arm chips except
+// some old Apple cores (A7-A10?) which handle all shifts slowly. Cortex-A57
+// and derived designs through Cortex-X1 take an extra micro-op for shifts
+// of 1 or 4. Other Arm chips handle all shifted operands at the same speed
+// as unshifted operands.
+//
+// We don't try to model the behavior of the old Apple cores because new code
+// targeting A7 is very unlikely to actually run on an A7. The Cortex cores
+// are modeled by FeatureAddrLSLSlow14.
+def FeatureAddrLSLSlow14 : SubtargetFeature<
+ "addr-lsl-slow-14", "HasAddrLSLSlow14", "true",
+ "Address operands with shift amount of 1 or 4 are slow">;
+
+def FeatureALULSLFast : SubtargetFeature<
+ "alu-lsl-fast", "HasALULSLFast", "true",
+ "Add/Sub operations with lsl shift <= 4 are cheap">;
+
+def FeatureAggressiveFMA :
+ SubtargetFeature<"aggressive-fma",
+ "HasAggressiveFMA",
+ "true",
+ "Enable Aggressive FMA for floating-point.">;
+
+def FeatureAltFPCmp : SubtargetFeature<"altnzcv", "HasAlternativeNZCV", "true",
+ "Enable alternative NZCV format for floating point comparisons (FEAT_FlagM2)">;
+
+def FeatureFRInt3264 : SubtargetFeature<"fptoint", "HasFRInt3264", "true",
+ "Enable FRInt[32|64][Z|X] instructions that round a floating-point number to "
+ "an integer (in FP format) forcing it to fit into a 32- or 64-bit int (FEAT_FRINTTS)" >;
+
+def FeatureSpecRestrict : SubtargetFeature<"specrestrict", "HasSpecRestrict",
+ "true", "Enable architectural speculation restriction (FEAT_CSV2_2)">;
+
+def FeatureSB : SubtargetFeature<"sb", "HasSB",
+ "true", "Enable v8.5 Speculation Barrier (FEAT_SB)" >;
+
+def FeatureSSBS : SubtargetFeature<"ssbs", "HasSSBS",
+ "true", "Enable Speculative Store Bypass Safe bit (FEAT_SSBS, FEAT_SSBS2)" >;
+
+def FeaturePredRes : SubtargetFeature<"predres", "HasPredRes", "true",
+ "Enable v8.5a execution and data prediction invalidation instructions (FEAT_SPECRES)" >;
+
+def FeatureCacheDeepPersist : SubtargetFeature<"ccdp", "HasCCDP",
+ "true", "Enable v8.5 Cache Clean to Point of Deep Persistence (FEAT_DPB2)" >;
+
+def FeatureBranchTargetId : SubtargetFeature<"bti", "HasBTI",
+ "true", "Enable Branch Target Identification (FEAT_BTI)" >;
+
+def FeatureRandGen : SubtargetFeature<"rand", "HasRandGen",
+ "true", "Enable Random Number generation instructions (FEAT_RNG)" >;
+
+def FeatureMTE : SubtargetFeature<"mte", "HasMTE",
+ "true", "Enable Memory Tagging Extension (FEAT_MTE, FEAT_MTE2)" >;
+
+def FeatureTRBE : SubtargetFeature<"trbe", "HasTRBE",
+ "true", "Enable Trace Buffer Extension (FEAT_TRBE)">;
+
+def FeatureETE : SubtargetFeature<"ete", "HasETE",
+ "true", "Enable Embedded Trace Extension (FEAT_ETE)",
+ [FeatureTRBE]>;
+
+def FeatureTME : SubtargetFeature<"tme", "HasTME",
+ "true", "Enable Transactional Memory Extension (FEAT_TME)" >;
+
+def FeatureTaggedGlobals : SubtargetFeature<"tagged-globals",
+ "AllowTaggedGlobals",
+ "true", "Use an instruction sequence for taking the address of a global "
+ "that allows a memory tag in the upper address bits">;
+
+def FeatureMatMulInt8 : SubtargetFeature<"i8mm", "HasMatMulInt8",
+ "true", "Enable Matrix Multiply Int8 Extension (FEAT_I8MM)">;
+
+def FeatureMatMulFP32 : SubtargetFeature<"f32mm", "HasMatMulFP32",
+ "true", "Enable Matrix Multiply FP32 Extension (FEAT_F32MM)", [FeatureSVE]>;
+
+def FeatureMatMulFP64 : SubtargetFeature<"f64mm", "HasMatMulFP64",
+ "true", "Enable Matrix Multiply FP64 Extension (FEAT_F64MM)", [FeatureSVE]>;
+
+def FeatureXS : SubtargetFeature<"xs", "HasXS",
+ "true", "Enable Armv8.7-A limited-TLB-maintenance instruction (FEAT_XS)">;
+
+def FeatureWFxT : SubtargetFeature<"wfxt", "HasWFxT",
+ "true", "Enable Armv8.7-A WFET and WFIT instruction (FEAT_WFxT)">;
+
+def FeatureHCX : SubtargetFeature<
+ "hcx", "HasHCX", "true", "Enable Armv8.7-A HCRX_EL2 system register (FEAT_HCX)">;
+
+def FeatureLS64 : SubtargetFeature<"ls64", "HasLS64",
+ "true", "Enable Armv8.7-A LD64B/ST64B Accelerator Extension (FEAT_LS64, FEAT_LS64_V, FEAT_LS64_ACCDATA)">;
+
+def FeatureHBC : SubtargetFeature<"hbc", "HasHBC",
+ "true", "Enable Armv8.8-A Hinted Conditional Branches Extension (FEAT_HBC)">;
+
+def FeatureMOPS : SubtargetFeature<"mops", "HasMOPS",
+ "true", "Enable Armv8.8-A memcpy and memset acceleration instructions (FEAT_MOPS)">;
+
+def FeatureNMI : SubtargetFeature<"nmi", "HasNMI",
+ "true", "Enable Armv8.8-A Non-maskable Interrupts (FEAT_NMI, FEAT_GICv3_NMI)">;
+
+def FeatureBRBE : SubtargetFeature<"brbe", "HasBRBE",
+ "true", "Enable Branch Record Buffer Extension (FEAT_BRBE)">;
+
+def FeatureSPE_EEF : SubtargetFeature<"spe-eef", "HasSPE_EEF",
+ "true", "Enable extra register in the Statistical Profiling Extension (FEAT_SPEv1p2)">;
+
+def FeatureFineGrainedTraps : SubtargetFeature<"fgt", "HasFineGrainedTraps",
+ "true", "Enable fine grained virtualization traps extension (FEAT_FGT)">;
+
+def FeatureEnhancedCounterVirtualization :
+ SubtargetFeature<"ecv", "HasEnhancedCounterVirtualization",
+ "true", "Enable enhanced counter virtualization extension (FEAT_ECV)">;
+
+def FeatureRME : SubtargetFeature<"rme", "HasRME",
+ "true", "Enable Realm Management Extension (FEAT_RME)">;
+
+def FeatureSME : SubtargetFeature<"sme", "HasSME", "true",
+ "Enable Scalable Matrix Extension (SME) (FEAT_SME)", [FeatureBF16, FeatureUseScalarIncVL]>;
+
+def FeatureSMEF64F64 : SubtargetFeature<"sme-f64f64", "HasSMEF64F64", "true",
+ "Enable Scalable Matrix Extension (SME) F64F64 instructions (FEAT_SME_F64F64)", [FeatureSME]>;
+
+def FeatureSMEI16I64 : SubtargetFeature<"sme-i16i64", "HasSMEI16I64", "true",
+ "Enable Scalable Matrix Extension (SME) I16I64 instructions (FEAT_SME_I16I64)", [FeatureSME]>;
+
+def FeatureSMEF16F16 : SubtargetFeature<"sme-f16f16", "HasSMEF16F16", "true",
+ "Enable SME2.1 non-widening Float16 instructions (FEAT_SME_F16F16)", []>;
+
+def FeatureSMEFA64 : SubtargetFeature<"sme-fa64", "HasSMEFA64", "true",
+ "Enable the full A64 instruction set in streaming SVE mode (FEAT_SME_FA64)", [FeatureSME, FeatureSVE2]>;
+
+def FeatureSME2 : SubtargetFeature<"sme2", "HasSME2", "true",
+ "Enable Scalable Matrix Extension 2 (SME2) instructions", [FeatureSME]>;
+
+def FeatureSME2p1 : SubtargetFeature<"sme2p1", "HasSME2p1", "true",
+ "Enable Scalable Matrix Extension 2.1 (FEAT_SME2p1) instructions", [FeatureSME2]>;
+
+def FeatureFAMINMAX: SubtargetFeature<"faminmax", "HasFAMINMAX", "true",
+ "Enable FAMIN and FAMAX instructions (FEAT_FAMINMAX)">;
+
+def FeatureFP8FMA : SubtargetFeature<"fp8fma", "HasFP8FMA", "true",
+ "Enable fp8 multiply-add instructions (FEAT_FP8FMA)">;
+
+def FeatureSSVE_FP8FMA : SubtargetFeature<"ssve-fp8fma", "HasSSVE_FP8FMA", "true",
+ "Enable SVE2 fp8 multiply-add instructions (FEAT_SSVE_FP8FMA)", [FeatureSME2]>;
+
+def FeatureFP8DOT2: SubtargetFeature<"fp8dot2", "HasFP8DOT2", "true",
+ "Enable fp8 2-way dot instructions (FEAT_FP8DOT2)">;
+
+def FeatureSSVE_FP8DOT2 : SubtargetFeature<"ssve-fp8dot2", "HasSSVE_FP8DOT2", "true",
+ "Enable SVE2 fp8 2-way dot product instructions (FEAT_SSVE_FP8DOT2)", [FeatureSME2]>;
+
+def FeatureFP8DOT4: SubtargetFeature<"fp8dot4", "HasFP8DOT4", "true",
+ "Enable fp8 4-way dot instructions (FEAT_FP8DOT4)">;
+
+def FeatureSSVE_FP8DOT4 : SubtargetFeature<"ssve-fp8dot4", "HasSSVE_FP8DOT4", "true",
+ "Enable SVE2 fp8 4-way dot product instructions (FEAT_SSVE_FP8DOT4)", [FeatureSME2]>;
+def FeatureLUT: SubtargetFeature<"lut", "HasLUT", "true",
+ "Enable Lookup Table instructions (FEAT_LUT)">;
+
+def FeatureSME_LUTv2 : SubtargetFeature<"sme-lutv2", "HasSME_LUTv2", "true",
+ "Enable Scalable Matrix Extension (SME) LUTv2 instructions (FEAT_SME_LUTv2)">;
+
+def FeatureSMEF8F16 : SubtargetFeature<"sme-f8f16", "HasSMEF8F16", "true",
+ "Enable Scalable Matrix Extension (SME) F8F16 instructions(FEAT_SME_F8F16)", [FeatureSME2, FeatureFP8]>;
+
+def FeatureSMEF8F32 : SubtargetFeature<"sme-f8f32", "HasSMEF8F32", "true",
+ "Enable Scalable Matrix Extension (SME) F8F32 instructions (FEAT_SME_F8F32)", [FeatureSME2, FeatureFP8]>;
+
+def FeatureAppleA7SysReg : SubtargetFeature<"apple-a7-sysreg", "HasAppleA7SysReg", "true",
+ "Apple A7 (the CPU formerly known as Cyclone)">;
+
+def FeatureEL2VMSA : SubtargetFeature<"el2vmsa", "HasEL2VMSA", "true",
+ "Enable Exception Level 2 Virtual Memory System Architecture">;
+
+def FeatureEL3 : SubtargetFeature<"el3", "HasEL3", "true",
+ "Enable Exception Level 3">;
+
+def FeatureCSSC : SubtargetFeature<"cssc", "HasCSSC", "true",
+ "Enable Common Short Sequence Compression (CSSC) instructions (FEAT_CSSC)">;
+
+def FeatureFixCortexA53_835769 : SubtargetFeature<"fix-cortex-a53-835769",
+ "FixCortexA53_835769", "true", "Mitigate Cortex-A53 Erratum 835769">;
+
+def FeatureNoBTIAtReturnTwice : SubtargetFeature<"no-bti-at-return-twice",
+ "NoBTIAtReturnTwice", "true",
+ "Don't place a BTI instruction "
+ "after a return-twice">;
+
+def FeatureCHK : SubtargetFeature<"chk", "HasCHK",
+ "true", "Enable Armv8.0-A Check Feature Status Extension (FEAT_CHK)">;
+
+def FeatureGCS : SubtargetFeature<"gcs", "HasGCS",
+ "true", "Enable Armv9.4-A Guarded Call Stack Extension", [FeatureCHK]>;
+
+def FeatureCLRBHB : SubtargetFeature<"clrbhb", "HasCLRBHB",
+ "true", "Enable Clear BHB instruction (FEAT_CLRBHB)">;
+
+def FeaturePRFM_SLC : SubtargetFeature<"prfm-slc-target", "HasPRFM_SLC",
+ "true", "Enable SLC target for PRFM instruction">;
+
+def FeatureSPECRES2 : SubtargetFeature<"specres2", "HasSPECRES2",
+ "true", "Enable Speculation Restriction Instruction (FEAT_SPECRES2)",
+ [FeaturePredRes]>;
+
+def FeatureMEC : SubtargetFeature<"mec", "HasMEC",
+ "true", "Enable Memory Encryption Contexts Extension", [FeatureRME]>;
+
+def FeatureITE : SubtargetFeature<"ite", "HasITE",
+ "true", "Enable Armv9.4-A Instrumentation Extension FEAT_ITE", [FeatureETE,
+ FeatureTRBE]>;
+
+def FeatureRCPC3 : SubtargetFeature<"rcpc3", "HasRCPC3",
+ "true", "Enable Armv8.9-A RCPC instructions for A64 and Advanced SIMD and floating-point instruction set (FEAT_LRCPC3)",
+ [FeatureRCPC_IMMO]>;
+
+def FeatureTHE : SubtargetFeature<"the", "HasTHE",
+ "true", "Enable Armv8.9-A Translation Hardening Extension (FEAT_THE)">;
+
+def FeatureLSE128 : SubtargetFeature<"lse128", "HasLSE128",
+ "true", "Enable Armv9.4-A 128-bit Atomic Instructions (FEAT_LSE128)",
+ [FeatureLSE]>;
+
+// FEAT_D128, FEAT_LVA3, FEAT_SYSREG128, and FEAT_SYSINSTR128 are mutually implicit.
+// Therefore group them all under a single feature flag, d128:
+def FeatureD128 : SubtargetFeature<"d128", "HasD128",
+ "true", "Enable Armv9.4-A 128-bit Page Table Descriptors, System Registers "
+ "and Instructions (FEAT_D128, FEAT_LVA3, FEAT_SYSREG128, FEAT_SYSINSTR128)",
+ [FeatureLSE128]>;
+
+def FeatureDisableLdp : SubtargetFeature<"disable-ldp", "HasDisableLdp",
+ "true", "Do not emit ldp">;
+
+def FeatureDisableStp : SubtargetFeature<"disable-stp", "HasDisableStp",
+ "true", "Do not emit stp">;
+
+def FeatureLdpAlignedOnly : SubtargetFeature<"ldp-aligned-only", "HasLdpAlignedOnly",
+ "true", "In order to emit ldp, first check if the load will be aligned to 2 * element_size">;
+
+def FeatureStpAlignedOnly : SubtargetFeature<"stp-aligned-only", "HasStpAlignedOnly",
+ "true", "In order to emit stp, first check if the store will be aligned to 2 * element_size">;
+
+// AArch64 2023 Architecture Extensions (v9.5-A)
+
+def FeatureCPA : SubtargetFeature<"cpa", "HasCPA", "true",
+ "Enable Armv9.5-A Checked Pointer Arithmetic (FEAT_CPA)">;
+
+def FeaturePAuthLR : SubtargetFeature<"pauth-lr", "HasPAuthLR",
+ "true", "Enable Armv9.5-A PAC enhancements (FEAT_PAuth_LR)">;
+
+def FeatureTLBIW : SubtargetFeature<"tlbiw", "HasTLBIW", "true",
+ "Enable ARMv9.5-A TLBI VMALL for Dirty State (FEAT_TLBIW)">;
+
+
+//===----------------------------------------------------------------------===//
+// Architectures.
+//
+def HasV8_0aOps : SubtargetFeature<"v8a", "HasV8_0aOps", "true",
+ "Support ARM v8.0a instructions", [FeatureEL2VMSA, FeatureEL3]>;
+
+def HasV8_1aOps : SubtargetFeature<"v8.1a", "HasV8_1aOps", "true",
+ "Support ARM v8.1a instructions", [HasV8_0aOps, FeatureCRC, FeatureLSE,
+ FeatureRDM, FeaturePAN, FeatureLOR, FeatureVH]>;
+
+def HasV8_2aOps : SubtargetFeature<"v8.2a", "HasV8_2aOps", "true",
+ "Support ARM v8.2a instructions", [HasV8_1aOps, FeaturePsUAO,
+ FeaturePAN_RWV, FeatureRAS, FeatureCCPP]>;
+
+def HasV8_3aOps : SubtargetFeature<"v8.3a", "HasV8_3aOps", "true",
+ "Support ARM v8.3a instructions", [HasV8_2aOps, FeatureRCPC, FeaturePAuth,
+ FeatureJS, FeatureCCIDX, FeatureComplxNum]>;
+
+def HasV8_4aOps : SubtargetFeature<"v8.4a", "HasV8_4aOps", "true",
+ "Support ARM v8.4a instructions", [HasV8_3aOps, FeatureDotProd,
+ FeatureNV, FeatureMPAM, FeatureDIT,
+ FeatureTRACEV8_4, FeatureAM, FeatureSEL2, FeatureTLB_RMI,
+ FeatureFlagM, FeatureRCPC_IMMO, FeatureLSE2]>;
+
+def HasV8_5aOps : SubtargetFeature<
+ "v8.5a", "HasV8_5aOps", "true", "Support ARM v8.5a instructions",
+ [HasV8_4aOps, FeatureAltFPCmp, FeatureFRInt3264, FeatureSpecRestrict,
+ FeatureSSBS, FeatureSB, FeaturePredRes, FeatureCacheDeepPersist,
+ FeatureBranchTargetId]>;
+
+def HasV8_6aOps : SubtargetFeature<
+ "v8.6a", "HasV8_6aOps", "true", "Support ARM v8.6a instructions",
+ [HasV8_5aOps, FeatureAMVS, FeatureBF16, FeatureFineGrainedTraps,
+ FeatureEnhancedCounterVirtualization, FeatureMatMulInt8]>;
+
+def HasV8_7aOps : SubtargetFeature<
+ "v8.7a", "HasV8_7aOps", "true", "Support ARM v8.7a instructions",
+ [HasV8_6aOps, FeatureXS, FeatureWFxT, FeatureHCX]>;
+
+def HasV8_8aOps : SubtargetFeature<
+ "v8.8a", "HasV8_8aOps", "true", "Support ARM v8.8a instructions",
+ [HasV8_7aOps, FeatureHBC, FeatureMOPS, FeatureNMI]>;
+
+def HasV8_9aOps : SubtargetFeature<
+ "v8.9a", "HasV8_9aOps", "true", "Support ARM v8.9a instructions",
+ [HasV8_8aOps, FeatureCLRBHB, FeaturePRFM_SLC, FeatureSPECRES2,
+ FeatureCSSC, FeatureRASv2, FeatureCHK]>;
+
+def HasV9_0aOps : SubtargetFeature<
+ "v9a", "HasV9_0aOps", "true", "Support ARM v9a instructions",
+ [HasV8_5aOps, FeatureMEC, FeatureSVE2]>;
+
+def HasV9_1aOps : SubtargetFeature<
+ "v9.1a", "HasV9_1aOps", "true", "Support ARM v9.1a instructions",
+ [HasV8_6aOps, HasV9_0aOps]>;
+
+def HasV9_2aOps : SubtargetFeature<
+ "v9.2a", "HasV9_2aOps", "true", "Support ARM v9.2a instructions",
+ [HasV8_7aOps, HasV9_1aOps]>;
+
+def HasV9_3aOps : SubtargetFeature<
+ "v9.3a", "HasV9_3aOps", "true", "Support ARM v9.3a instructions",
+ [HasV8_8aOps, HasV9_2aOps]>;
+
+def HasV9_4aOps : SubtargetFeature<
+ "v9.4a", "HasV9_4aOps", "true", "Support ARM v9.4a instructions",
+ [HasV8_9aOps, HasV9_3aOps]>;
+
+def HasV9_5aOps : SubtargetFeature<
+ "v9.5a", "HasV9_5aOps", "true", "Support ARM v9.5a instructions",
+ [HasV9_4aOps, FeatureCPA]>;
+
+def HasV8_0rOps : SubtargetFeature<
+ "v8r", "HasV8_0rOps", "true", "Support ARM v8r instructions",
+ [//v8.1
+ FeatureCRC, FeaturePAN, FeatureLSE, FeatureCONTEXTIDREL2,
+ //v8.2
+ FeatureRAS, FeaturePsUAO, FeatureCCPP, FeaturePAN_RWV,
+ //v8.3
+ FeatureCCIDX, FeaturePAuth, FeatureRCPC,
+ //v8.4
+ FeatureTRACEV8_4, FeatureTLB_RMI, FeatureFlagM, FeatureDIT, FeatureSEL2,
+ FeatureRCPC_IMMO,
+ // Not mandatory in v8.0-R, but included here on the grounds that it
+ // only enables names of system registers
+ FeatureSpecRestrict
+ ]>;
+
+//===----------------------------------------------------------------------===//
+// Access to privileged registers
+//===----------------------------------------------------------------------===//
+
+foreach i = 1-3 in
+def FeatureUseEL#i#ForTP : SubtargetFeature<"tpidr-el"#i, "UseEL"#i#"ForTP",
+ "true", "Permit use of TPIDR_EL"#i#" for the TLS base">;
+def FeatureUseROEL0ForTP : SubtargetFeature<"tpidrro-el0", "UseROEL0ForTP",
+ "true", "Permit use of TPIDRRO_EL0 for the TLS base">;
+
+//===----------------------------------------------------------------------===//
+// Control codegen mitigation against Straight Line Speculation vulnerability.
+//===----------------------------------------------------------------------===//
+
+def FeatureHardenSlsRetBr : SubtargetFeature<"harden-sls-retbr",
+ "HardenSlsRetBr", "true",
+ "Harden against straight line speculation across RET and BR instructions">;
+def FeatureHardenSlsBlr : SubtargetFeature<"harden-sls-blr",
+ "HardenSlsBlr", "true",
+ "Harden against straight line speculation across BLR instructions">;
+def FeatureHardenSlsNoComdat : SubtargetFeature<"harden-sls-nocomdat",
+ "HardenSlsNoComdat", "true",
+ "Generate thunk code for SLS mitigation in the normal text section">;
+
+
+//===----------------------------------------------------------------------===//
+// AArch64 Processor subtarget features.
+//===----------------------------------------------------------------------===//
+
+
+def TuneA35 : SubtargetFeature<"a35", "ARMProcFamily", "CortexA35",
----------------
pratlucas wrote:
I agree with @davemgreen. It'd be clearer to have the content split based on the logical concepts rather than on the structural class layout.
https://github.com/llvm/llvm-project/pull/88282
More information about the llvm-commits
mailing list