pub const all_features = blk: {
const len = @typeInfo(Feature).@"enum".field_names.len;
std.debug.assert(len <= CpuFeature.Set.needed_bit_count);
var result: [len]CpuFeature = undefined;
result[@backingInt(Feature.@"16bit_mode")] = .{
.llvm_name = "16bit-mode",
.description = "16-bit mode (i8086)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.@"32bit_mode")] = .{
.llvm_name = "32bit-mode",
.description = "32-bit mode (80386)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.@"3dnow")] = .{
.llvm_name = null,
.description = "Enable 3DNow! instructions",
.dependencies = featureSet(&[_]Feature{
.mmx,
}),
};
result[@backingInt(Feature.@"3dnowa")] = .{
.llvm_name = null,
.description = "Enable 3DNow! Athlon instructions",
.dependencies = featureSet(&[_]Feature{
.@"3dnow",
}),
};
result[@backingInt(Feature.@"64bit")] = .{
.llvm_name = "64bit",
.description = "Support 64-bit instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.adx)] = .{
.llvm_name = "adx",
.description = "Support ADX instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.aes)] = .{
.llvm_name = "aes",
.description = "Enable AES instructions",
.dependencies = featureSet(&[_]Feature{
.sse2,
}),
};
result[@backingInt(Feature.allow_light_256_bit)] = .{
.llvm_name = "allow-light-256-bit",
.description = "Enable generation of 256-bit load/stores even if we prefer 128-bit",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.amx_avx512)] = .{
.llvm_name = "amx-avx512",
.description = "Support AMX-AVX512 instructions",
.dependencies = featureSet(&[_]Feature{
.amx_tile,
}),
};
result[@backingInt(Feature.amx_bf16)] = .{
.llvm_name = "amx-bf16",
.description = "Support AMX-BF16 instructions",
.dependencies = featureSet(&[_]Feature{
.amx_tile,
}),
};
result[@backingInt(Feature.amx_complex)] = .{
.llvm_name = "amx-complex",
.description = "Support AMX-COMPLEX instructions",
.dependencies = featureSet(&[_]Feature{
.amx_tile,
}),
};
result[@backingInt(Feature.amx_fp16)] = .{
.llvm_name = "amx-fp16",
.description = "Support AMX amx-fp16 instructions",
.dependencies = featureSet(&[_]Feature{
.amx_tile,
}),
};
result[@backingInt(Feature.amx_fp8)] = .{
.llvm_name = "amx-fp8",
.description = "Support AMX-FP8 instructions",
.dependencies = featureSet(&[_]Feature{
.amx_tile,
}),
};
result[@backingInt(Feature.amx_int8)] = .{
.llvm_name = "amx-int8",
.description = "Support AMX-INT8 instructions",
.dependencies = featureSet(&[_]Feature{
.amx_tile,
}),
};
result[@backingInt(Feature.amx_movrs)] = .{
.llvm_name = "amx-movrs",
.description = "Support AMX-MOVRS instructions",
.dependencies = featureSet(&[_]Feature{
.amx_tile,
}),
};
result[@backingInt(Feature.amx_tf32)] = .{
.llvm_name = "amx-tf32",
.description = "Support AMX-TF32 instructions",
.dependencies = featureSet(&[_]Feature{
.amx_tile,
}),
};
result[@backingInt(Feature.amx_tile)] = .{
.llvm_name = "amx-tile",
.description = "Support AMX-TILE instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.avx)] = .{
.llvm_name = "avx",
.description = "Enable AVX instructions",
.dependencies = featureSet(&[_]Feature{
.sse4_2,
}),
};
result[@backingInt(Feature.avx10_1)] = .{
.llvm_name = "avx10.1",
.description = "Support AVX10.1 instruction",
.dependencies = featureSet(&[_]Feature{
.avx512bf16,
.avx512bitalg,
.avx512cd,
.avx512dq,
.avx512fp16,
.avx512ifma,
.avx512vbmi,
.avx512vbmi2,
.avx512vl,
.avx512vnni,
.avx512vpopcntdq,
}),
};
result[@backingInt(Feature.avx10_2)] = .{
.llvm_name = "avx10.2",
.description = "Support AVX10.2 instruction",
.dependencies = featureSet(&[_]Feature{
.avx10_1,
}),
};
result[@backingInt(Feature.avx2)] = .{
.llvm_name = "avx2",
.description = "Enable AVX2 instructions",
.dependencies = featureSet(&[_]Feature{
.avx,
}),
};
result[@backingInt(Feature.avx512bf16)] = .{
.llvm_name = "avx512bf16",
.description = "Support bfloat16 floating point",
.dependencies = featureSet(&[_]Feature{
.avx512bw,
}),
};
result[@backingInt(Feature.avx512bitalg)] = .{
.llvm_name = "avx512bitalg",
.description = "Enable AVX-512 Bit Algorithms",
.dependencies = featureSet(&[_]Feature{
.avx512bw,
}),
};
result[@backingInt(Feature.avx512bw)] = .{
.llvm_name = "avx512bw",
.description = "Enable AVX-512 Byte and Word Instructions",
.dependencies = featureSet(&[_]Feature{
.avx512f,
}),
};
result[@backingInt(Feature.avx512cd)] = .{
.llvm_name = "avx512cd",
.description = "Enable AVX-512 Conflict Detection Instructions",
.dependencies = featureSet(&[_]Feature{
.avx512f,
}),
};
result[@backingInt(Feature.avx512dq)] = .{
.llvm_name = "avx512dq",
.description = "Enable AVX-512 Doubleword and Quadword Instructions",
.dependencies = featureSet(&[_]Feature{
.avx512f,
}),
};
result[@backingInt(Feature.avx512er)] = .{
.llvm_name = null,
.description = "Enable AVX-512 Exponential and Reciprocal Instructions",
.dependencies = featureSet(&[_]Feature{
.avx512f,
}),
};
result[@backingInt(Feature.avx512f)] = .{
.llvm_name = "avx512f",
.description = "Enable AVX-512 instructions",
.dependencies = featureSet(&[_]Feature{
.avx2,
.f16c,
.fma,
}),
};
result[@backingInt(Feature.avx512fp16)] = .{
.llvm_name = "avx512fp16",
.description = "Support 16-bit floating point",
.dependencies = featureSet(&[_]Feature{
.avx512bw,
}),
};
result[@backingInt(Feature.avx512ifma)] = .{
.llvm_name = "avx512ifma",
.description = "Enable AVX-512 Integer Fused Multiple-Add",
.dependencies = featureSet(&[_]Feature{
.avx512f,
}),
};
result[@backingInt(Feature.avx512pf)] = .{
.llvm_name = null,
.description = "Enable AVX-512 PreFetch Instructions",
.dependencies = featureSet(&[_]Feature{
.avx512f,
}),
};
result[@backingInt(Feature.avx512vbmi)] = .{
.llvm_name = "avx512vbmi",
.description = "Enable AVX-512 Vector Byte Manipulation Instructions",
.dependencies = featureSet(&[_]Feature{
.avx512bw,
}),
};
result[@backingInt(Feature.avx512vbmi2)] = .{
.llvm_name = "avx512vbmi2",
.description = "Enable AVX-512 further Vector Byte Manipulation Instructions",
.dependencies = featureSet(&[_]Feature{
.avx512bw,
}),
};
result[@backingInt(Feature.avx512vl)] = .{
.llvm_name = "avx512vl",
.description = "Enable AVX-512 Vector Length eXtensions",
.dependencies = featureSet(&[_]Feature{
.avx512f,
}),
};
result[@backingInt(Feature.avx512vnni)] = .{
.llvm_name = "avx512vnni",
.description = "Enable AVX-512 Vector Neural Network Instructions",
.dependencies = featureSet(&[_]Feature{
.avx512f,
}),
};
result[@backingInt(Feature.avx512vp2intersect)] = .{
.llvm_name = "avx512vp2intersect",
.description = "Enable AVX-512 vp2intersect",
.dependencies = featureSet(&[_]Feature{
.avx512f,
}),
};
result[@backingInt(Feature.avx512vpopcntdq)] = .{
.llvm_name = "avx512vpopcntdq",
.description = "Enable AVX-512 Population Count Instructions",
.dependencies = featureSet(&[_]Feature{
.avx512f,
}),
};
result[@backingInt(Feature.avxifma)] = .{
.llvm_name = "avxifma",
.description = "Enable AVX-IFMA",
.dependencies = featureSet(&[_]Feature{
.avx2,
}),
};
result[@backingInt(Feature.avxneconvert)] = .{
.llvm_name = "avxneconvert",
.description = "Support AVX-NE-CONVERT instructions",
.dependencies = featureSet(&[_]Feature{
.avx2,
}),
};
result[@backingInt(Feature.avxvnni)] = .{
.llvm_name = "avxvnni",
.description = "Support AVX_VNNI encoding",
.dependencies = featureSet(&[_]Feature{
.avx2,
}),
};
result[@backingInt(Feature.avxvnniint16)] = .{
.llvm_name = "avxvnniint16",
.description = "Enable AVX-VNNI-INT16",
.dependencies = featureSet(&[_]Feature{
.avx2,
}),
};
result[@backingInt(Feature.avxvnniint8)] = .{
.llvm_name = "avxvnniint8",
.description = "Enable AVX-VNNI-INT8",
.dependencies = featureSet(&[_]Feature{
.avx2,
}),
};
result[@backingInt(Feature.bmi)] = .{
.llvm_name = "bmi",
.description = "Support BMI instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.bmi2)] = .{
.llvm_name = "bmi2",
.description = "Support BMI2 instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.branch_hint)] = .{
.llvm_name = "branch-hint",
.description = "Target has branch hint feature",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.branchfusion)] = .{
.llvm_name = "branchfusion",
.description = "CMP/TEST can be fused with conditional branches",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.bsf_bsr_0_clobbers_result)] = .{
.llvm_name = null,
.description = "BSF/BSR may clobber the lower 32-bits of the result register when the source is zero",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.ccmp)] = .{
.llvm_name = "ccmp",
.description = "Support conditional cmp & test instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.cf)] = .{
.llvm_name = "cf",
.description = "Support conditional faulting",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.cldemote)] = .{
.llvm_name = "cldemote",
.description = "Enable Cache Line Demote",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.clflushopt)] = .{
.llvm_name = "clflushopt",
.description = "Flush A Cache Line Optimized",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.clwb)] = .{
.llvm_name = "clwb",
.description = "Cache Line Write Back",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.clzero)] = .{
.llvm_name = "clzero",
.description = "Enable Cache Line Zero",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.cmov)] = .{
.llvm_name = "cmov",
.description = "Enable conditional move instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.cmpccxadd)] = .{
.llvm_name = "cmpccxadd",
.description = "Support CMPCCXADD instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.crc32)] = .{
.llvm_name = "crc32",
.description = "Enable SSE 4.2 CRC32 instruction (used when SSE4.2 is supported but function is GPR only)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.cx16)] = .{
.llvm_name = "cx16",
.description = "64-bit with cmpxchg16b (this is true for most x86-64 chips, but not the first AMD chips)",
.dependencies = featureSet(&[_]Feature{
.cx8,
}),
};
result[@backingInt(Feature.cx8)] = .{
.llvm_name = "cx8",
.description = "Support CMPXCHG8B instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.egpr)] = .{
.llvm_name = "egpr",
.description = "Support extended general purpose register",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.enqcmd)] = .{
.llvm_name = "enqcmd",
.description = "Has ENQCMD instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.ermsb)] = .{
.llvm_name = "ermsb",
.description = "REP MOVS/STOS are fast",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.f16c)] = .{
.llvm_name = "f16c",
.description = "Support 16-bit floating point conversion instructions",
.dependencies = featureSet(&[_]Feature{
.avx,
}),
};
result[@backingInt(Feature.false_deps_getmant)] = .{
.llvm_name = "false-deps-getmant",
.description = "VGETMANTSS/SD/SH and VGETMANDPS/PD(memory version) has a false dependency on dest register",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.false_deps_lzcnt_tzcnt)] = .{
.llvm_name = "false-deps-lzcnt-tzcnt",
.description = "LZCNT/TZCNT have a false dependency on dest register",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.false_deps_mulc)] = .{
.llvm_name = "false-deps-mulc",
.description = "VF[C]MULCPH/SH has a false dependency on dest register",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.false_deps_mullq)] = .{
.llvm_name = "false-deps-mullq",
.description = "VPMULLQ has a false dependency on dest register",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.false_deps_perm)] = .{
.llvm_name = "false-deps-perm",
.description = "VPERMD/Q/PS/PD has a false dependency on dest register",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.false_deps_popcnt)] = .{
.llvm_name = "false-deps-popcnt",
.description = "POPCNT has a false dependency on dest register",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.false_deps_range)] = .{
.llvm_name = "false-deps-range",
.description = "VRANGEPD/PS/SD/SS has a false dependency on dest register",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_11bytenop)] = .{
.llvm_name = "fast-11bytenop",
.description = "Target can quickly decode up to 11 byte NOPs",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_15bytenop)] = .{
.llvm_name = "fast-15bytenop",
.description = "Target can quickly decode up to 15 byte NOPs",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_7bytenop)] = .{
.llvm_name = "fast-7bytenop",
.description = "Target can quickly decode up to 7 byte NOPs",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_bextr)] = .{
.llvm_name = "fast-bextr",
.description = "Indicates that the BEXTR instruction is implemented as a single uop with good throughput",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_dpwssd)] = .{
.llvm_name = "fast-dpwssd",
.description = "Prefer vpdpwssd instruction over vpmaddwd+vpaddd instruction sequence",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_gather)] = .{
.llvm_name = "fast-gather",
.description = "Indicates if gather is reasonably fast (this is true for Skylake client and all AVX-512 CPUs)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_hops)] = .{
.llvm_name = "fast-hops",
.description = "Prefer horizontal vector math instructions (haddp, phsub, etc.) over normal vector instructions with shuffles",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_imm16)] = .{
.llvm_name = "fast-imm16",
.description = "Prefer a i16 instruction with i16 immediate over extension to i32",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_lzcnt)] = .{
.llvm_name = "fast-lzcnt",
.description = "LZCNT instructions are as fast as most simple integer ops",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_movbe)] = .{
.llvm_name = "fast-movbe",
.description = "Prefer a movbe over a single-use load + bswap / single-use bswap + store",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_scalar_fsqrt)] = .{
.llvm_name = "fast-scalar-fsqrt",
.description = "Scalar SQRT is fast (disable Newton-Raphson)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_scalar_shift_masks)] = .{
.llvm_name = "fast-scalar-shift-masks",
.description = "Prefer a left/right scalar logical shift pair over a shift+and pair",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_shld_rotate)] = .{
.llvm_name = "fast-shld-rotate",
.description = "SHLD can be used as a faster rotate",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_variable_crosslane_shuffle)] = .{
.llvm_name = "fast-variable-crosslane-shuffle",
.description = "Cross-lane shuffles with variable masks are fast",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_variable_perlane_shuffle)] = .{
.llvm_name = "fast-variable-perlane-shuffle",
.description = "Per-lane shuffles with variable masks are fast",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_vector_fsqrt)] = .{
.llvm_name = "fast-vector-fsqrt",
.description = "Vector SQRT is fast (disable Newton-Raphson)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fast_vector_shift_masks)] = .{
.llvm_name = "fast-vector-shift-masks",
.description = "Prefer a left/right vector logical shift pair over a shift+and pair",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.faster_shift_than_shuffle)] = .{
.llvm_name = "faster-shift-than-shuffle",
.description = "Shifts are faster (or as fast) as shuffle",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fma)] = .{
.llvm_name = "fma",
.description = "Enable three-operand fused multiple-add",
.dependencies = featureSet(&[_]Feature{
.avx,
}),
};
result[@backingInt(Feature.fma4)] = .{
.llvm_name = "fma4",
.description = "Enable four-operand fused multiple-add",
.dependencies = featureSet(&[_]Feature{
.avx,
.sse4a,
}),
};
result[@backingInt(Feature.fsgsbase)] = .{
.llvm_name = "fsgsbase",
.description = "Support FS/GS Base instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fsrm)] = .{
.llvm_name = "fsrm",
.description = "REP MOVSB of short lengths is faster",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.fxsr)] = .{
.llvm_name = "fxsr",
.description = "Support fxsave/fxrestore instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.gfni)] = .{
.llvm_name = "gfni",
.description = "Enable Galois Field Arithmetic Instructions",
.dependencies = featureSet(&[_]Feature{
.sse2,
}),
};
result[@backingInt(Feature.harden_sls_ijmp)] = .{
.llvm_name = "harden-sls-ijmp",
.description = "Harden against straight line speculation across indirect JMP instructions.",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.harden_sls_ret)] = .{
.llvm_name = "harden-sls-ret",
.description = "Harden against straight line speculation across RET instructions.",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.hreset)] = .{
.llvm_name = "hreset",
.description = "Has hreset instruction",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.idivl_to_divb)] = .{
.llvm_name = "idivl-to-divb",
.description = "Use 8-bit divide for positive values less than 256",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.idivq_to_divl)] = .{
.llvm_name = "idivq-to-divl",
.description = "Use 32-bit divide for positive values less than 2^32",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.inline_asm_use_gpr32)] = .{
.llvm_name = "inline-asm-use-gpr32",
.description = "Enable use of GPR32 in inline assembly for APX",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.invpcid)] = .{
.llvm_name = "invpcid",
.description = "Invalidate Process-Context Identifier",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.kl)] = .{
.llvm_name = "kl",
.description = "Support Key Locker kl Instructions",
.dependencies = featureSet(&[_]Feature{
.sse2,
}),
};
result[@backingInt(Feature.lea_sp)] = .{
.llvm_name = "lea-sp",
.description = "Use LEA for adjusting the stack pointer (this is an optimization for Intel Atom processors)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.lea_uses_ag)] = .{
.llvm_name = "lea-uses-ag",
.description = "LEA instruction needs inputs at AG stage",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.lvi_cfi)] = .{
.llvm_name = "lvi-cfi",
.description = "Prevent indirect calls/branches from using a memory operand, and precede all indirect calls/branches from a register with an LFENCE instruction to serialize control flow. Also decompose RET instructions into a POP+LFENCE+JMP sequence.",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.lvi_load_hardening)] = .{
.llvm_name = "lvi-load-hardening",
.description = "Insert LFENCE instructions to prevent data speculatively injected into loads from being used maliciously.",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.lwp)] = .{
.llvm_name = "lwp",
.description = "Enable LWP instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.lzcnt)] = .{
.llvm_name = "lzcnt",
.description = "Support LZCNT instruction",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.macrofusion)] = .{
.llvm_name = "macrofusion",
.description = "Various instructions can be fused with conditional branches",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.mmx)] = .{
.llvm_name = "mmx",
.description = "Enable MMX instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.movbe)] = .{
.llvm_name = "movbe",
.description = "Support MOVBE instruction",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.movdir64b)] = .{
.llvm_name = "movdir64b",
.description = "Support movdir64b instruction (direct store 64 bytes)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.movdiri)] = .{
.llvm_name = "movdiri",
.description = "Support movdiri instruction (direct store integer)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.movrs)] = .{
.llvm_name = "movrs",
.description = "Enable MOVRS",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.mwaitx)] = .{
.llvm_name = "mwaitx",
.description = "Enable MONITORX/MWAITX timer functionality",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.ndd)] = .{
.llvm_name = "ndd",
.description = "Support non-destructive destination",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.nf)] = .{
.llvm_name = "nf",
.description = "Support status flags update suppression",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.no_bypass_delay)] = .{
.llvm_name = "no-bypass-delay",
.description = "Has no bypass delay when using the 'wrong' domain",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.no_bypass_delay_blend)] = .{
.llvm_name = "no-bypass-delay-blend",
.description = "Has no bypass delay when using the 'wrong' blend type",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.no_bypass_delay_mov)] = .{
.llvm_name = "no-bypass-delay-mov",
.description = "Has no bypass delay when using the 'wrong' mov type",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.no_bypass_delay_shuffle)] = .{
.llvm_name = "no-bypass-delay-shuffle",
.description = "Has no bypass delay when using the 'wrong' shuffle type",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.nopl)] = .{
.llvm_name = "nopl",
.description = "Enable NOPL instruction (generally pentium pro+)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.pad_short_functions)] = .{
.llvm_name = "pad-short-functions",
.description = "Pad short functions (to prevent a stall when returning too early)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.pclmul)] = .{
.llvm_name = "pclmul",
.description = "Enable packed carry-less multiplication instructions",
.dependencies = featureSet(&[_]Feature{
.sse2,
}),
};
result[@backingInt(Feature.pconfig)] = .{
.llvm_name = "pconfig",
.description = "platform configuration instruction",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.pku)] = .{
.llvm_name = "pku",
.description = "Enable protection keys",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.popcnt)] = .{
.llvm_name = "popcnt",
.description = "Support POPCNT instruction",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.ppx)] = .{
.llvm_name = "ppx",
.description = "Support Push-Pop Acceleration",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.prefer_128_bit)] = .{
.llvm_name = "prefer-128-bit",
.description = "Prefer 128-bit AVX instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.prefer_256_bit)] = .{
.llvm_name = "prefer-256-bit",
.description = "Prefer 256-bit AVX instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.prefer_legacy_setcc)] = .{
.llvm_name = "prefer-legacy-setcc",
.description = "Prefer to emit legacy SetCC.",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.prefer_mask_registers)] = .{
.llvm_name = "prefer-mask-registers",
.description = "Prefer AVX512 mask registers over PTEST/MOVMSK",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.prefer_movmsk_over_vtest)] = .{
.llvm_name = "prefer-movmsk-over-vtest",
.description = "Prefer movmsk over vtest instruction",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.prefer_no_gather)] = .{
.llvm_name = "prefer-no-gather",
.description = "Prefer no gather instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.prefer_no_scatter)] = .{
.llvm_name = "prefer-no-scatter",
.description = "Prefer no scatter instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.prefetchi)] = .{
.llvm_name = "prefetchi",
.description = "Prefetch instruction with T0 or T1 Hint",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.prefetchwt1)] = .{
.llvm_name = null,
.description = "Prefetch with Intent to Write and T1 Hint",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.prfchw)] = .{
.llvm_name = "prfchw",
.description = "Support PRFCHW instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.ptwrite)] = .{
.llvm_name = "ptwrite",
.description = "Support ptwrite instruction",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.push2pop2)] = .{
.llvm_name = "push2pop2",
.description = "Support PUSH2/POP2 instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.raoint)] = .{
.llvm_name = "raoint",
.description = "Support RAO-INT instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.rdpid)] = .{
.llvm_name = "rdpid",
.description = "Support RDPID instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.rdpru)] = .{
.llvm_name = "rdpru",
.description = "Support RDPRU instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.rdrnd)] = .{
.llvm_name = "rdrnd",
.description = "Support RDRAND instruction",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.rdseed)] = .{
.llvm_name = "rdseed",
.description = "Support RDSEED instruction",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.retpoline)] = .{
.llvm_name = "retpoline",
.description = "Remove speculation of indirect branches from the generated code, either by avoiding them entirely or lowering them with a speculation blocking construct",
.dependencies = featureSet(&[_]Feature{
.retpoline_indirect_branches,
.retpoline_indirect_calls,
}),
};
result[@backingInt(Feature.retpoline_external_thunk)] = .{
.llvm_name = "retpoline-external-thunk",
.description = "When lowering an indirect call or branch using a `retpoline`, rely on the specified user provided thunk rather than emitting one ourselves. Only has effect when combined with some other retpoline feature",
.dependencies = featureSet(&[_]Feature{
.retpoline_indirect_calls,
}),
};
result[@backingInt(Feature.retpoline_indirect_branches)] = .{
.llvm_name = "retpoline-indirect-branches",
.description = "Remove speculation of indirect branches from the generated code",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.retpoline_indirect_calls)] = .{
.llvm_name = "retpoline-indirect-calls",
.description = "Remove speculation of indirect calls from the generated code",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.rtm)] = .{
.llvm_name = "rtm",
.description = "Support RTM instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.sahf)] = .{
.llvm_name = "sahf",
.description = "Support LAHF and SAHF instructions in 64-bit mode",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.sbb_dep_breaking)] = .{
.llvm_name = "sbb-dep-breaking",
.description = "SBB with same register has no source dependency",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.serialize)] = .{
.llvm_name = "serialize",
.description = "Has serialize instruction",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.seses)] = .{
.llvm_name = "seses",
.description = "Prevent speculative execution side channel timing attacks by inserting a speculation barrier before memory reads, memory writes, and conditional branches. Implies LVI Control Flow integrity.",
.dependencies = featureSet(&[_]Feature{
.lvi_cfi,
}),
};
result[@backingInt(Feature.sgx)] = .{
.llvm_name = "sgx",
.description = "Enable Software Guard Extensions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.sha)] = .{
.llvm_name = "sha",
.description = "Enable SHA instructions",
.dependencies = featureSet(&[_]Feature{
.sse2,
}),
};
result[@backingInt(Feature.sha512)] = .{
.llvm_name = "sha512",
.description = "Support SHA512 instructions",
.dependencies = featureSet(&[_]Feature{
.avx2,
}),
};
result[@backingInt(Feature.shstk)] = .{
.llvm_name = "shstk",
.description = "Support CET Shadow-Stack instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.slow_3ops_lea)] = .{
.llvm_name = "slow-3ops-lea",
.description = "LEA instruction with 3 ops or certain registers is slow",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.slow_incdec)] = .{
.llvm_name = "slow-incdec",
.description = "INC and DEC instructions are slower than ADD and SUB",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.slow_lea)] = .{
.llvm_name = "slow-lea",
.description = "LEA instruction with certain arguments is slow",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.slow_pmaddwd)] = .{
.llvm_name = "slow-pmaddwd",
.description = "PMADDWD is slower than PMULLD",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.slow_pmulld)] = .{
.llvm_name = "slow-pmulld",
.description = "PMULLD instruction is slow (compared to PMULLW/PMULHW and PMULUDQ)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.slow_pmullq)] = .{
.llvm_name = "slow-pmullq",
.description = "PMULLQ instruction is slow",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.slow_shld)] = .{
.llvm_name = "slow-shld",
.description = "SHLD instruction is slow",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.slow_two_mem_ops)] = .{
.llvm_name = "slow-two-mem-ops",
.description = "Two memory operand instructions are slow",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.slow_unaligned_mem_16)] = .{
.llvm_name = "slow-unaligned-mem-16",
.description = "Slow unaligned 16-byte memory access",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.slow_unaligned_mem_32)] = .{
.llvm_name = "slow-unaligned-mem-32",
.description = "Slow unaligned 32-byte memory access",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.sm3)] = .{
.llvm_name = "sm3",
.description = "Support SM3 instructions",
.dependencies = featureSet(&[_]Feature{
.avx,
}),
};
result[@backingInt(Feature.sm4)] = .{
.llvm_name = "sm4",
.description = "Support SM4 instructions",
.dependencies = featureSet(&[_]Feature{
.avx2,
}),
};
result[@backingInt(Feature.smap)] = .{
.llvm_name = null,
.description = "Enable Supervisor Mode Access Prevention",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.smep)] = .{
.llvm_name = null,
.description = "Enable Supervisor Mode Execution Prevention",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.soft_float)] = .{
.llvm_name = "soft-float",
.description = "Use software floating point features",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.sse)] = .{
.llvm_name = "sse",
.description = "Enable SSE instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.sse2)] = .{
.llvm_name = "sse2",
.description = "Enable SSE2 instructions",
.dependencies = featureSet(&[_]Feature{
.sse,
}),
};
result[@backingInt(Feature.sse3)] = .{
.llvm_name = "sse3",
.description = "Enable SSE3 instructions",
.dependencies = featureSet(&[_]Feature{
.sse2,
}),
};
result[@backingInt(Feature.sse4_1)] = .{
.llvm_name = "sse4.1",
.description = "Enable SSE 4.1 instructions",
.dependencies = featureSet(&[_]Feature{
.ssse3,
}),
};
result[@backingInt(Feature.sse4_2)] = .{
.llvm_name = "sse4.2",
.description = "Enable SSE 4.2 instructions",
.dependencies = featureSet(&[_]Feature{
.crc32,
.sse4_1,
}),
};
result[@backingInt(Feature.sse4a)] = .{
.llvm_name = "sse4a",
.description = "Support SSE 4a instructions",
.dependencies = featureSet(&[_]Feature{
.sse3,
}),
};
result[@backingInt(Feature.sse_unaligned_mem)] = .{
.llvm_name = "sse-unaligned-mem",
.description = "Allow unaligned memory operands with SSE instructions (this may require setting a configuration bit in the processor)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.ssse3)] = .{
.llvm_name = "ssse3",
.description = "Enable SSSE3 instructions",
.dependencies = featureSet(&[_]Feature{
.sse3,
}),
};
result[@backingInt(Feature.tagged_globals)] = .{
.llvm_name = "tagged-globals",
.description = "Use an instruction sequence for taking the address of a global that allows a memory tag in the upper address bits.",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.tbm)] = .{
.llvm_name = "tbm",
.description = "Enable TBM instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.tsxldtrk)] = .{
.llvm_name = "tsxldtrk",
.description = "Support TSXLDTRK instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.tuning_fast_imm_vector_shift)] = .{
.llvm_name = "tuning-fast-imm-vector-shift",
.description = "Vector shifts are fast (2/cycle) as opposed to slow (1/cycle)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.uintr)] = .{
.llvm_name = "uintr",
.description = "Has UINTR Instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.use_glm_div_sqrt_costs)] = .{
.llvm_name = "use-glm-div-sqrt-costs",
.description = "Use Goldmont specific floating point div/sqrt costs",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.use_slm_arith_costs)] = .{
.llvm_name = "use-slm-arith-costs",
.description = "Use Silvermont specific arithmetic costs",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.usermsr)] = .{
.llvm_name = "usermsr",
.description = "Support USERMSR instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.vaes)] = .{
.llvm_name = "vaes",
.description = "Promote selected AES instructions to AVX512/AVX registers",
.dependencies = featureSet(&[_]Feature{
.aes,
.avx2,
}),
};
result[@backingInt(Feature.vpclmulqdq)] = .{
.llvm_name = "vpclmulqdq",
.description = "Enable vpclmulqdq instructions",
.dependencies = featureSet(&[_]Feature{
.avx,
.pclmul,
}),
};
result[@backingInt(Feature.vzeroupper)] = .{
.llvm_name = "vzeroupper",
.description = "Should insert vzeroupper instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.waitpkg)] = .{
.llvm_name = "waitpkg",
.description = "Wait and pause enhancements",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.wbnoinvd)] = .{
.llvm_name = "wbnoinvd",
.description = "Write Back No Invalidate",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.widekl)] = .{
.llvm_name = "widekl",
.description = "Support Key Locker wide Instructions",
.dependencies = featureSet(&[_]Feature{
.kl,
}),
};
result[@backingInt(Feature.x32)] = .{
.llvm_name = "x32",
.description = "64-bit with ILP32 programming model (e.g. x32 ABI)",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.x87)] = .{
.llvm_name = "x87",
.description = "Enable X87 float instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.xop)] = .{
.llvm_name = "xop",
.description = "Enable XOP instructions",
.dependencies = featureSet(&[_]Feature{
.fma4,
}),
};
result[@backingInt(Feature.xsave)] = .{
.llvm_name = "xsave",
.description = "Support xsave instructions",
.dependencies = featureSet(&[_]Feature{}),
};
result[@backingInt(Feature.xsavec)] = .{
.llvm_name = "xsavec",
.description = "Support xsavec instructions",
.dependencies = featureSet(&[_]Feature{
.xsave,
}),
};
result[@backingInt(Feature.xsaveopt)] = .{
.llvm_name = "xsaveopt",
.description = "Support xsaveopt instructions",
.dependencies = featureSet(&[_]Feature{
.xsave,
}),
};
result[@backingInt(Feature.xsaves)] = .{
.llvm_name = "xsaves",
.description = "Support xsaves instructions",
.dependencies = featureSet(&[_]Feature{
.xsave,
}),
};
result[@backingInt(Feature.zu)] = .{
.llvm_name = "zu",
.description = "Support zero-upper SETcc/IMUL",
.dependencies = featureSet(&[_]Feature{}),
};
const ti = @typeInfo(Feature);
for (&result, 0..) |*elem, i| {
elem.index = i;
elem.name = ti.@"enum".field_names[i];
}
break :blk result;
}
