[PATCH] x86: Update -mtune=alderlake
Uros Bizjak
ubizjak@gmail.com
Wed Nov 10 12:25:48 GMT 2021
On Wed, Nov 10, 2021 at 10:09 AM Cui,Lili <lili.cui@intel.com> wrote:
>
> Hi Uros,
>
> This patch is to update mtune for alderlake.
>
> Bootstrap is ok, and no regressions for i386/x86-64 testsuite.
>
> OK for master?
>
> Update mtune for alderlake, Alder Lake Intel Hybrid Technology will not support
> Intel® AVX-512. ISA features such as Intel® AVX, AVX-VNNI, Intel® AVX2, and
> UMONITOR/UMWAIT/TPAUSE are supported.
>
> gcc/ChangeLog
>
> * config/i386/i386-options.c (m_CORE_AVX2): Remove Alderlake
> from m_CORE_AVX2.
> (processor_cost_table): Use alderlake_cost for Alderlake.
> * config/i386/i386.c (ix86_sched_init_global): Handle Alderlake.
> * config/i386/x86-tune-costs.h (struct processor_costs): Add alderlake
> cost.
> * config/i386/x86-tune-sched.c (ix86_issue_rate): Change Alderlake
> issue rate to 4.
> (ix86_adjust_cost): Handle Alderlake.
> * config/i386/x86-tune.def (X86_TUNE_SCHEDULE): Enable for Alderlake.
> (X86_TUNE_PARTIAL_REG_DEPENDENCY): Likewise.
> (X86_TUNE_SSE_PARTIAL_REG_DEPENDENCY): Likewise.
> (X86_TUNE_SSE_PARTIAL_REG_FP_CONVERTS_DEPENDENCY): Likewise.
> (X86_TUNE_SSE_PARTIAL_REG_CONVERTS_DEPENDENCY): Likewise.
> (X86_TUNE_MEMORY_MISMATCH_STALL): Likewise.
> (X86_TUNE_USE_LEAVE): Likewise.
> (X86_TUNE_PUSH_MEMORY): Likewise.
> (X86_TUNE_USE_INCDEC): Likewise.
> (X86_TUNE_INTEGER_DFMODE_MOVES): Likewise.
> (X86_TUNE_MISALIGNED_MOVE_STRING_PRO_EPILOGUES): Likewise.
> (X86_TUNE_USE_SAHF): Likewise.
> (X86_TUNE_USE_BT): Likewise.
> (X86_TUNE_AVOID_FALSE_DEP_FOR_BMI): Likewise.
> (X86_TUNE_ONE_IF_CONV_INSN): Likewise.
> (X86_TUNE_AVOID_MFENCE): Likewise.
> (X86_TUNE_USE_SIMODE_FIOP): Likewise.
> (X86_TUNE_EXT_80387_CONSTANTS): Likewise.
> (X86_TUNE_SSE_UNALIGNED_LOAD_OPTIMAL): Likewise.
> (X86_TUNE_SSE_UNALIGNED_STORE_OPTIMAL): Likewise.
> (X86_TUNE_SSE_TYPELESS_STORES): Likewise.
> (X86_TUNE_SSE_LOAD0_BY_PXOR): Likewise.
> (X86_TUNE_AVOID_4BYTE_PREFIXES): Likewise.
> (X86_TUNE_USE_GATHER): Disable for Alderlake.
> (X86_TUNE_AVX256_MOVE_BY_PIECES): Likewise.
> (X86_TUNE_AVX256_STORE_BY_PIECES): Likewise.
OK.
Thanks,
Uros.
> ---
> gcc/config/i386/i386-options.c | 4 +-
> gcc/config/i386/i386.c | 1 +
> gcc/config/i386/x86-tune-costs.h | 120 +++++++++++++++++++++++++++++++
> gcc/config/i386/x86-tune-sched.c | 2 +
> gcc/config/i386/x86-tune.def | 58 +++++++--------
> 5 files changed, 155 insertions(+), 30 deletions(-)
>
> diff --git a/gcc/config/i386/i386-options.c b/gcc/config/i386/i386-options.c
> index e7a3bd4aaea..a8cc0664f11 100644
> --- a/gcc/config/i386/i386-options.c
> +++ b/gcc/config/i386/i386-options.c
> @@ -131,7 +131,7 @@ along with GCC; see the file COPYING3. If not see
> | m_ICELAKE_CLIENT | m_ICELAKE_SERVER | m_CASCADELAKE \
> | m_TIGERLAKE | m_COOPERLAKE | m_SAPPHIRERAPIDS \
> | m_ROCKETLAKE)
> -#define m_CORE_AVX2 (m_HASWELL | m_SKYLAKE | m_ALDERLAKE | m_CORE_AVX512)
> +#define m_CORE_AVX2 (m_HASWELL | m_SKYLAKE | m_CORE_AVX512)
> #define m_CORE_ALL (m_CORE2 | m_NEHALEM | m_SANDYBRIDGE | m_CORE_AVX2)
> #define m_GOLDMONT (HOST_WIDE_INT_1U<<PROCESSOR_GOLDMONT)
> #define m_GOLDMONT_PLUS (HOST_WIDE_INT_1U<<PROCESSOR_GOLDMONT_PLUS)
> @@ -736,7 +736,7 @@ static const struct processor_costs *processor_cost_table[] =
> &icelake_cost,
> &skylake_cost,
> &icelake_cost,
> - &icelake_cost,
> + &alderlake_cost,
> &icelake_cost,
> &intel_cost,
> &geode_cost,
> diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c
> index e94efdf39fb..73c4d5115bb 100644
> --- a/gcc/config/i386/i386.c
> +++ b/gcc/config/i386/i386.c
> @@ -17014,6 +17014,7 @@ ix86_sched_init_global (FILE *, int, int)
> case PROCESSOR_SANDYBRIDGE:
> case PROCESSOR_HASWELL:
> case PROCESSOR_TREMONT:
> + case PROCESSOR_ALDERLAKE:
> case PROCESSOR_GENERIC:
> /* Do not perform multipass scheduling for pre-reload schedule
> to save compile time. */
> diff --git a/gcc/config/i386/x86-tune-costs.h b/gcc/config/i386/x86-tune-costs.h
> index 93644be9cb3..dd5563d2e64 100644
> --- a/gcc/config/i386/x86-tune-costs.h
> +++ b/gcc/config/i386/x86-tune-costs.h
> @@ -2070,6 +2070,126 @@ struct processor_costs icelake_cost = {
> "16", /* Func alignment. */
> };
>
> +/* alderlake_cost should produce code tuned for alderlake family of CPUs. */
> +static stringop_algs alderlake_memcpy[2] = {
> + {libcall,
> + {{256, rep_prefix_1_byte, true},
> + {256, loop, false},
> + {-1, libcall, false}}},
> + {libcall,
> + {{256, rep_prefix_1_byte, true},
> + {256, loop, false},
> + {-1, libcall, false}}}};
> +static stringop_algs alderlake_memset[2] = {
> + {libcall,
> + {{256, rep_prefix_1_byte, true},
> + {256, loop, false},
> + {-1, libcall, false}}},
> + {libcall,
> + {{256, rep_prefix_1_byte, true},
> + {256, loop, false},
> + {-1, libcall, false}}}};
> +static const
> +struct processor_costs alderlake_cost = {
> + {
> + /* Start of register allocator costs. integer->integer move cost is 2. */
> + 6, /* cost for loading QImode using movzbl */
> + {6, 6, 6}, /* cost of loading integer registers
> + in QImode, HImode and SImode.
> + Relative to reg-reg move (2). */
> + {6, 6, 6}, /* cost of storing integer registers */
> + 4, /* cost of reg,reg fld/fst */
> + {6, 6, 12}, /* cost of loading fp registers
> + in SFmode, DFmode and XFmode */
> + {6, 6, 12}, /* cost of storing fp registers
> + in SFmode, DFmode and XFmode */
> + 2, /* cost of moving MMX register */
> + {6, 6}, /* cost of loading MMX registers
> + in SImode and DImode */
> + {6, 6}, /* cost of storing MMX registers
> + in SImode and DImode */
> + 2, 3, 4, /* cost of moving XMM,YMM,ZMM register */
> + {6, 6, 6, 10, 15}, /* cost of loading SSE registers
> + in 32,64,128,256 and 512-bit */
> + {6, 6, 6, 10, 15}, /* cost of storing SSE registers
> + in 32,64,128,256 and 512-bit */
> + 6, 6, /* SSE->integer and integer->SSE moves */
> + 6, 6, /* mask->integer and integer->mask moves */
> + {6, 6, 6}, /* cost of loading mask register
> + in QImode, HImode, SImode. */
> + {6, 6, 6}, /* cost if storing mask register
> + in QImode, HImode, SImode. */
> + 2, /* cost of moving mask register. */
> + /* End of register allocator costs. */
> + },
> +
> + COSTS_N_INSNS (1), /* cost of an add instruction */
> + COSTS_N_INSNS (1) + 1, /* cost of a lea instruction */
> + COSTS_N_INSNS (1), /* variable shift costs */
> + COSTS_N_INSNS (1), /* constant shift costs */
> + {COSTS_N_INSNS (3), /* cost of starting multiply for QI */
> + COSTS_N_INSNS (4), /* HI */
> + COSTS_N_INSNS (3), /* SI */
> + COSTS_N_INSNS (4), /* DI */
> + COSTS_N_INSNS (4)}, /* other */
> + 0, /* cost of multiply per each bit set */
> + {COSTS_N_INSNS (16), /* cost of a divide/mod for QI */
> + COSTS_N_INSNS (22), /* HI */
> + COSTS_N_INSNS (30), /* SI */
> + COSTS_N_INSNS (74), /* DI */
> + COSTS_N_INSNS (74)}, /* other */
> + COSTS_N_INSNS (1), /* cost of movsx */
> + COSTS_N_INSNS (1), /* cost of movzx */
> + 8, /* "large" insn */
> + 17, /* MOVE_RATIO */
> + 17, /* CLEAR_RATIO */
> + {6, 6, 6}, /* cost of loading integer registers
> + in QImode, HImode and SImode.
> + Relative to reg-reg move (2). */
> + {6, 6, 6}, /* cost of storing integer registers */
> + {6, 6, 6, 10, 15}, /* cost of loading SSE register
> + in 32bit, 64bit, 128bit, 256bit and 512bit */
> + {6, 6, 6, 10, 15}, /* cost of storing SSE register
> + in 32bit, 64bit, 128bit, 256bit and 512bit */
> + {6, 6, 6, 10, 15}, /* cost of unaligned loads. */
> + {6, 6, 6, 10, 15}, /* cost of unaligned storess. */
> + 2, 3, 4, /* cost of moving XMM,YMM,ZMM register */
> + 6, /* cost of moving SSE register to integer. */
> + 18, 6, /* Gather load static, per_elt. */
> + 18, 6, /* Gather store static, per_elt. */
> + 32, /* size of l1 cache. */
> + 512, /* size of l2 cache. */
> + 64, /* size of prefetch block */
> + 6, /* number of parallel prefetches */
> + 3, /* Branch cost */
> + COSTS_N_INSNS (3), /* cost of FADD and FSUB insns. */
> + COSTS_N_INSNS (5), /* cost of FMUL instruction. */
> + COSTS_N_INSNS (17), /* cost of FDIV instruction. */
> + COSTS_N_INSNS (1), /* cost of FABS instruction. */
> + COSTS_N_INSNS (1), /* cost of FCHS instruction. */
> + COSTS_N_INSNS (14), /* cost of FSQRT instruction. */
> +
> + COSTS_N_INSNS (1), /* cost of cheap SSE instruction. */
> + COSTS_N_INSNS (3), /* cost of ADDSS/SD SUBSS/SD insns. */
> + COSTS_N_INSNS (4), /* cost of MULSS instruction. */
> + COSTS_N_INSNS (5), /* cost of MULSD instruction. */
> + COSTS_N_INSNS (5), /* cost of FMA SS instruction. */
> + COSTS_N_INSNS (5), /* cost of FMA SD instruction. */
> + COSTS_N_INSNS (13), /* cost of DIVSS instruction. */
> + COSTS_N_INSNS (17), /* cost of DIVSD instruction. */
> + COSTS_N_INSNS (14), /* cost of SQRTSS instruction. */
> + COSTS_N_INSNS (18), /* cost of SQRTSD instruction. */
> + 1, 4, 3, 3, /* reassoc int, fp, vec_int, vec_fp. */
> + alderlake_memcpy,
> + alderlake_memset,
> + COSTS_N_INSNS (4), /* cond_taken_branch_cost. */
> + COSTS_N_INSNS (2), /* cond_not_taken_branch_cost. */
> + "16:11:8", /* Loop alignment. */
> + "16:11:8", /* Jump alignment. */
> + "0:0:8", /* Label alignment. */
> + "16", /* Func alignment. */
> +};
> +
> /* BTVER1 has optimized REP instruction for medium sized blocks, but for
> very small blocks it is better to use loop. For large blocks, libcall can
> do nontemporary accesses and beat inline considerably. */
> diff --git a/gcc/config/i386/x86-tune-sched.c b/gcc/config/i386/x86-tune-sched.c
> index 56ada99a450..0c149a09531 100644
> --- a/gcc/config/i386/x86-tune-sched.c
> +++ b/gcc/config/i386/x86-tune-sched.c
> @@ -72,6 +72,7 @@ ix86_issue_rate (void)
> case PROCESSOR_SANDYBRIDGE:
> case PROCESSOR_HASWELL:
> case PROCESSOR_TREMONT:
> + case PROCESSOR_ALDERLAKE:
> case PROCESSOR_GENERIC:
> return 4;
>
> @@ -431,6 +432,7 @@ ix86_adjust_cost (rtx_insn *insn, int dep_type, rtx_insn *dep_insn, int cost,
> case PROCESSOR_SANDYBRIDGE:
> case PROCESSOR_HASWELL:
> case PROCESSOR_TREMONT:
> + case PROCESSOR_ALDERLAKE:
> case PROCESSOR_GENERIC:
> /* Stack engine allows to execute push&pop instructions in parall. */
> if ((insn_type == TYPE_PUSH || insn_type == TYPE_POP)
> diff --git a/gcc/config/i386/x86-tune.def b/gcc/config/i386/x86-tune.def
> index 58e8ead56b4..4ae0b569841 100644
> --- a/gcc/config/i386/x86-tune.def
> +++ b/gcc/config/i386/x86-tune.def
> @@ -42,7 +42,7 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
> DEF_TUNE (X86_TUNE_SCHEDULE, "schedule",
> m_PENT | m_LAKEMONT | m_PPRO | m_CORE_ALL | m_BONNELL | m_SILVERMONT
> | m_INTEL | m_KNL | m_KNM | m_K6_GEODE | m_AMD_MULTIPLE | m_GOLDMONT
> - | m_GOLDMONT_PLUS | m_TREMONT | m_GENERIC)
> + | m_GOLDMONT_PLUS | m_TREMONT | m_ALDERLAKE |m_GENERIC)
>
> /* X86_TUNE_PARTIAL_REG_DEPENDENCY: Enable more register renaming
> on modern chips. Preffer stores affecting whole integer register
> @@ -51,7 +51,7 @@ DEF_TUNE (X86_TUNE_SCHEDULE, "schedule",
> DEF_TUNE (X86_TUNE_PARTIAL_REG_DEPENDENCY, "partial_reg_dependency",
> m_P4_NOCONA | m_CORE2 | m_NEHALEM | m_SANDYBRIDGE | m_CORE_AVX2
> | m_BONNELL | m_SILVERMONT | m_GOLDMONT | m_GOLDMONT_PLUS | m_INTEL
> - | m_KNL | m_KNM | m_AMD_MULTIPLE | m_TREMONT
> + | m_KNL | m_KNM | m_AMD_MULTIPLE | m_TREMONT | m_ALDERLAKE
> | m_GENERIC)
>
> /* X86_TUNE_SSE_PARTIAL_REG_DEPENDENCY: This knob promotes all store
> @@ -62,7 +62,7 @@ DEF_TUNE (X86_TUNE_PARTIAL_REG_DEPENDENCY, "partial_reg_dependency",
> that can be partly masked by careful scheduling of moves. */
> DEF_TUNE (X86_TUNE_SSE_PARTIAL_REG_DEPENDENCY, "sse_partial_reg_dependency",
> m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_BONNELL | m_AMDFAM10
> - | m_BDVER | m_ZNVER | m_TREMONT | m_GENERIC)
> + | m_BDVER | m_ZNVER | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_SSE_PARTIAL_REG_FP_CONVERTS_DEPENDENCY: This knob avoids
> partial write to the destination in scalar SSE conversion from FP
> @@ -70,14 +70,14 @@ DEF_TUNE (X86_TUNE_SSE_PARTIAL_REG_DEPENDENCY, "sse_partial_reg_dependency",
> DEF_TUNE (X86_TUNE_SSE_PARTIAL_REG_FP_CONVERTS_DEPENDENCY,
> "sse_partial_reg_fp_converts_dependency",
> m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_BONNELL | m_AMDFAM10
> - | m_BDVER | m_ZNVER | m_GENERIC)
> + | m_BDVER | m_ZNVER | m_ALDERLAKE| m_GENERIC)
>
> /* X86_TUNE_SSE_PARTIAL_REG_CONVERTS_DEPENDENCY: This knob avoids partial
> write to the destination in scalar SSE conversion from integer to FP. */
> DEF_TUNE (X86_TUNE_SSE_PARTIAL_REG_CONVERTS_DEPENDENCY,
> "sse_partial_reg_converts_dependency",
> m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_BONNELL | m_AMDFAM10
> - | m_BDVER | m_ZNVER | m_GENERIC)
> + | m_BDVER | m_ZNVER | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_SSE_SPLIT_REGS: Set for machines where the type and dependencies
> are resolved on SSE register parts instead of whole registers, so we may
> @@ -103,14 +103,14 @@ DEF_TUNE (X86_TUNE_MOVX, "movx",
> m_PPRO | m_P4_NOCONA | m_CORE2 | m_NEHALEM | m_SANDYBRIDGE
> | m_BONNELL | m_SILVERMONT | m_GOLDMONT | m_KNL | m_KNM | m_INTEL
> | m_GOLDMONT_PLUS | m_GEODE | m_AMD_MULTIPLE
> - | m_CORE_AVX2 | m_TREMONT | m_GENERIC)
> + | m_CORE_AVX2 | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_MEMORY_MISMATCH_STALL: Avoid partial stores that are followed by
> full sized loads. */
> DEF_TUNE (X86_TUNE_MEMORY_MISMATCH_STALL, "memory_mismatch_stall",
> m_P4_NOCONA | m_CORE_ALL | m_BONNELL | m_SILVERMONT | m_INTEL
> | m_KNL | m_KNM | m_GOLDMONT | m_GOLDMONT_PLUS | m_AMD_MULTIPLE
> - | m_TREMONT | m_GENERIC)
> + | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_FUSE_CMP_AND_BRANCH_32: Fuse compare with a subsequent
> conditional jump instruction for 32 bit TARGET. */
> @@ -166,14 +166,14 @@ DEF_TUNE (X86_TUNE_EPILOGUE_USING_MOVE, "epilogue_using_move",
> /* X86_TUNE_USE_LEAVE: Use "leave" instruction in epilogues where it fits. */
> DEF_TUNE (X86_TUNE_USE_LEAVE, "use_leave",
> m_386 | m_CORE_ALL | m_K6_GEODE | m_AMD_MULTIPLE | m_TREMONT
> - | m_GENERIC)
> + | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_PUSH_MEMORY: Enable generation of "push mem" instructions.
> Some chips, like 486 and Pentium works faster with separate load
> and push instructions. */
> DEF_TUNE (X86_TUNE_PUSH_MEMORY, "push_memory",
> m_386 | m_P4_NOCONA | m_CORE_ALL | m_K6_GEODE | m_AMD_MULTIPLE
> - | m_TREMONT | m_GENERIC)
> + | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_SINGLE_PUSH: Enable if single push insn is preferred
> over esp subtraction. */
> @@ -243,14 +243,14 @@ DEF_TUNE (X86_TUNE_READ_MODIFY, "read_modify", ~(m_PENT | m_LAKEMONT | m_PPRO))
> DEF_TUNE (X86_TUNE_USE_INCDEC, "use_incdec",
> ~(m_P4_NOCONA | m_CORE2 | m_NEHALEM | m_SANDYBRIDGE
> | m_BONNELL | m_SILVERMONT | m_INTEL | m_KNL | m_KNM | m_GOLDMONT
> - | m_GOLDMONT_PLUS | m_TREMONT | m_GENERIC))
> + | m_GOLDMONT_PLUS | m_TREMONT | m_ALDERLAKE | m_GENERIC))
>
> /* X86_TUNE_INTEGER_DFMODE_MOVES: Enable if integer moves are preferred
> for DFmode copies */
> DEF_TUNE (X86_TUNE_INTEGER_DFMODE_MOVES, "integer_dfmode_moves",
> ~(m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_BONNELL | m_SILVERMONT
> | m_KNL | m_KNM | m_INTEL | m_GEODE | m_AMD_MULTIPLE | m_GOLDMONT
> - | m_GOLDMONT_PLUS | m_TREMONT | m_GENERIC))
> + | m_GOLDMONT_PLUS | m_TREMONT | m_ALDERLAKE | m_GENERIC))
>
> /* X86_TUNE_OPT_AGU: Optimize for Address Generation Unit. This flag
> will impact LEA instruction selection. */
> @@ -298,14 +298,14 @@ DEF_TUNE (X86_TUNE_PREFER_KNOWN_REP_MOVSB_STOSB,
> DEF_TUNE (X86_TUNE_MISALIGNED_MOVE_STRING_PRO_EPILOGUES,
> "misaligned_move_string_pro_epilogues",
> m_386 | m_486 | m_CORE_ALL | m_AMD_MULTIPLE | m_TREMONT
> - | m_GENERIC)
> + | m_ALDERLAKE |m_GENERIC)
>
> /* X86_TUNE_USE_SAHF: Controls use of SAHF. */
> DEF_TUNE (X86_TUNE_USE_SAHF, "use_sahf",
> m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_BONNELL | m_SILVERMONT
> | m_KNL | m_KNM | m_INTEL | m_K6_GEODE | m_K8 | m_AMDFAM10 | m_BDVER
> | m_BTVER | m_ZNVER | m_GOLDMONT | m_GOLDMONT_PLUS | m_TREMONT
> - | m_GENERIC)
> + | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_USE_CLTD: Controls use of CLTD and CTQO instructions. */
> DEF_TUNE (X86_TUNE_USE_CLTD, "use_cltd",
> @@ -316,12 +316,12 @@ DEF_TUNE (X86_TUNE_USE_CLTD, "use_cltd",
> DEF_TUNE (X86_TUNE_USE_BT, "use_bt",
> m_CORE_ALL | m_BONNELL | m_SILVERMONT | m_KNL | m_KNM | m_INTEL
> | m_LAKEMONT | m_AMD_MULTIPLE | m_GOLDMONT | m_GOLDMONT_PLUS
> - | m_TREMONT | m_GENERIC)
> + | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_AVOID_FALSE_DEP_FOR_BMI: Avoid false dependency
> for bit-manipulation instructions. */
> DEF_TUNE (X86_TUNE_AVOID_FALSE_DEP_FOR_BMI, "avoid_false_dep_for_bmi",
> - m_SANDYBRIDGE | m_CORE_AVX2 | m_TREMONT | m_GENERIC)
> + m_SANDYBRIDGE | m_CORE_AVX2 | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_ADJUST_UNROLL: This enables adjusting the unroll factor based
> on hardware capabilities. Bdver3 hardware has a loop buffer which makes
> @@ -333,11 +333,11 @@ DEF_TUNE (X86_TUNE_ADJUST_UNROLL, "adjust_unroll_factor", m_BDVER3 | m_BDVER4)
> if-converted sequence to one. */
> DEF_TUNE (X86_TUNE_ONE_IF_CONV_INSN, "one_if_conv_insn",
> m_SILVERMONT | m_KNL | m_KNM | m_INTEL | m_CORE_ALL | m_GOLDMONT
> - | m_GOLDMONT_PLUS | m_TREMONT | m_GENERIC)
> + | m_GOLDMONT_PLUS | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_AVOID_MFENCE: Use lock prefixed instructions instead of mfence. */
> DEF_TUNE (X86_TUNE_AVOID_MFENCE, "avoid_mfence",
> - m_CORE_ALL | m_BDVER | m_ZNVER | m_TREMONT | m_GENERIC)
> + m_CORE_ALL | m_BDVER | m_ZNVER | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_EXPAND_ABS: This enables a new abs pattern by
> generating instructions for abs (x) = (((signed) x >> (W-1) ^ x) -
> @@ -361,7 +361,8 @@ DEF_TUNE (X86_TUNE_USE_HIMODE_FIOP, "use_himode_fiop",
> DEF_TUNE (X86_TUNE_USE_SIMODE_FIOP, "use_simode_fiop",
> ~(m_PENT | m_LAKEMONT | m_PPRO | m_CORE_ALL | m_BONNELL
> | m_SILVERMONT | m_KNL | m_KNM | m_INTEL | m_AMD_MULTIPLE
> - | m_GOLDMONT | m_GOLDMONT_PLUS | m_TREMONT | m_GENERIC))
> + | m_GOLDMONT | m_GOLDMONT_PLUS | m_TREMONT | m_ALDERLAKE
> + | m_GENERIC))
>
> /* X86_TUNE_USE_FFREEP: Use freep instruction instead of fstp. */
> DEF_TUNE (X86_TUNE_USE_FFREEP, "use_ffreep", m_AMD_MULTIPLE)
> @@ -370,7 +371,7 @@ DEF_TUNE (X86_TUNE_USE_FFREEP, "use_ffreep", m_AMD_MULTIPLE)
> DEF_TUNE (X86_TUNE_EXT_80387_CONSTANTS, "ext_80387_constants",
> m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_BONNELL | m_SILVERMONT
> | m_KNL | m_KNM | m_INTEL | m_K6_GEODE | m_ATHLON_K8 | m_GOLDMONT
> - | m_GOLDMONT_PLUS | m_TREMONT | m_GENERIC)
> + | m_GOLDMONT_PLUS | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /*****************************************************************************/
> /* SSE instruction selection tuning */
> @@ -385,15 +386,15 @@ DEF_TUNE (X86_TUNE_GENERAL_REGS_SSE_SPILL, "general_regs_sse_spill",
> of a sequence loading registers by parts. */
> DEF_TUNE (X86_TUNE_SSE_UNALIGNED_LOAD_OPTIMAL, "sse_unaligned_load_optimal",
> m_NEHALEM | m_SANDYBRIDGE | m_CORE_AVX2 | m_SILVERMONT | m_KNL | m_KNM
> - | m_INTEL | m_GOLDMONT | m_GOLDMONT_PLUS
> - | m_TREMONT | m_AMDFAM10 | m_BDVER | m_BTVER | m_ZNVER | m_GENERIC)
> + | m_INTEL | m_GOLDMONT | m_GOLDMONT_PLUS | m_TREMONT | m_ALDERLAKE
> + | m_AMDFAM10 | m_BDVER | m_BTVER | m_ZNVER | m_GENERIC)
>
> /* X86_TUNE_SSE_UNALIGNED_STORE_OPTIMAL: Use movups for misaligned stores
> instead of a sequence loading registers by parts. */
> DEF_TUNE (X86_TUNE_SSE_UNALIGNED_STORE_OPTIMAL, "sse_unaligned_store_optimal",
> m_NEHALEM | m_SANDYBRIDGE | m_CORE_AVX2 | m_SILVERMONT | m_KNL | m_KNM
> | m_INTEL | m_GOLDMONT | m_GOLDMONT_PLUS
> - | m_TREMONT | m_BDVER | m_ZNVER | m_GENERIC)
> + | m_TREMONT | m_ALDERLAKE | m_BDVER | m_ZNVER | m_GENERIC)
>
> /* X86_TUNE_SSE_PACKED_SINGLE_INSN_OPTIMAL: Use packed single
> precision 128bit instructions instead of double where possible. */
> @@ -402,13 +403,13 @@ DEF_TUNE (X86_TUNE_SSE_PACKED_SINGLE_INSN_OPTIMAL, "sse_packed_single_insn_optim
>
> /* X86_TUNE_SSE_TYPELESS_STORES: Always movaps/movups for 128bit stores. */
> DEF_TUNE (X86_TUNE_SSE_TYPELESS_STORES, "sse_typeless_stores",
> - m_AMD_MULTIPLE | m_CORE_ALL | m_TREMONT | m_GENERIC)
> + m_AMD_MULTIPLE | m_CORE_ALL | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_SSE_LOAD0_BY_PXOR: Always use pxor to load0 as opposed to
> xorps/xorpd and other variants. */
> DEF_TUNE (X86_TUNE_SSE_LOAD0_BY_PXOR, "sse_load0_by_pxor",
> m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_BDVER | m_BTVER | m_ZNVER
> - | m_TREMONT | m_GENERIC)
> + | m_TREMONT | m_ALDERLAKE | m_GENERIC)
>
> /* X86_TUNE_INTER_UNIT_MOVES_TO_VEC: Enable moves in from integer
> to SSE registers. If disabled, the moves will be done by storing
> @@ -454,11 +455,12 @@ DEF_TUNE (X86_TUNE_SLOW_PSHUFB, "slow_pshufb",
>
> /* X86_TUNE_AVOID_4BYTE_PREFIXES: Avoid instructions requiring 4+ bytes of prefixes. */
> DEF_TUNE (X86_TUNE_AVOID_4BYTE_PREFIXES, "avoid_4byte_prefixes",
> - m_SILVERMONT | m_GOLDMONT | m_GOLDMONT_PLUS | m_TREMONT | m_INTEL)
> + m_SILVERMONT | m_GOLDMONT | m_GOLDMONT_PLUS | m_TREMONT | m_ALDERLAKE
> + | m_INTEL)
>
> /* X86_TUNE_USE_GATHER: Use gather instructions. */
> DEF_TUNE (X86_TUNE_USE_GATHER, "use_gather",
> - ~(m_ZNVER1 | m_ZNVER2 | m_GENERIC))
> + ~(m_ZNVER1 | m_ZNVER2 | m_GENERIC | m_ALDERLAKE))
>
> /* X86_TUNE_AVOID_128FMA_CHAINS: Avoid creating loops with tight 128bit or
> smaller FMA chain. */
> @@ -503,12 +505,12 @@ DEF_TUNE (X86_TUNE_AVX256_OPTIMAL, "avx256_optimal", m_CORE_AVX512)
> /* X86_TUNE_AVX256_MOVE_BY_PIECES: Optimize move_by_pieces with 256-bit
> AVX instructions. */
> DEF_TUNE (X86_TUNE_AVX256_MOVE_BY_PIECES, "avx256_move_by_pieces",
> - m_ALDERLAKE | m_CORE_AVX512)
> + m_CORE_AVX512)
>
> /* X86_TUNE_AVX256_STORE_BY_PIECES: Optimize store_by_pieces with 256-bit
> AVX instructions. */
> DEF_TUNE (X86_TUNE_AVX256_STORE_BY_PIECES, "avx256_store_by_pieces",
> - m_ALDERLAKE | m_CORE_AVX512)
> + m_CORE_AVX512)
>
> /*****************************************************************************/
> /*****************************************************************************/
> --
> 2.17.1
>
> Thanks,
> Lili.
More information about the Gcc-patches
mailing list