[Public] Hi all, PFA, the patch that enables support for the next generation AMD Zen5 CPU via -march=znver5 with basic znver5 scheduler Model. We may update the scheduler model going forward. Good for trunk? Thanks and Regards Karthiban Patch is inline here. From 6230938c1420604c8d0af27b0d080970d9b54ac5 Mon Sep 17 00:00:00 2001 From: karthiban Karthiban.Anbazhagan@amd.com Date: Fri, 9 Feb 2024 15:03:09 +0530 Subject: [PATCH] Add AMD znver5 processor enablement with scheduler model gcc/ChangeLog: * common/config/i386/cpuinfo.h (get_amd_cpu): Recognize znver5. * common/config/i386/i386-common.cc (processor_names): Add znver5. (processor_alias_table): Likewise. * common/config/i386/i386-cpuinfo.h (processor_types): Add new zen family. (processor_subtypes): Add znver5. * config.gcc (x86_64-*-* |...): Likewise. * config/i386/driver-i386.cc (host_detect_local_cpu): Let march=native detect znver5 cpu's. * config/i386/i386-c.cc (ix86_target_macros_internal): Add znver5. * config/i386/i386-options.cc (m_ZNVER5): New definition (processor_cost_table): Add znver5. * config/i386/i386.cc (ix86_reassociation_width): Likewise. * config/i386/i386.h (processor_type): Add PROCESSOR_ZNVER5 (PTA_ZNVER5): New definition. * config/i386/i386.md (define_attr "cpu"): Add znver5. (Scheduling descriptions) Add znver5.md. * config/i386/x86-tune-costs.h (znver5_cost): New definition. * config/i386/x86-tune-sched.cc (ix86_issue_rate): Add znver5. (ix86_adjust_cost): Likewise. * config/i386/x86-tune.def (avx512_move_by_pieces): Add m_ZNVER5. (avx512_store_by_pieces): Add m_ZNVER5. * doc/extend.texi: Add znver5. * doc/invoke.texi: Likewise. * config/i386/znver5.md: New. gcc/testsuite/ChangeLog: * g++.target/i386/mv29.C: Handle znver5 arch. * gcc.target/i386/funcspec-56.inc:Likewise. --- gcc/common/config/i386/cpuinfo.h | 16 + gcc/common/config/i386/i386-common.cc | 6 +- gcc/common/config/i386/i386-cpuinfo.h | 2 + gcc/config.gcc | 14 +- gcc/config/i386/driver-i386.cc | 5 + gcc/config/i386/i386-c.cc | 7 + gcc/config/i386/i386-options.cc | 6 +- gcc/config/i386/i386.cc | 3 +- gcc/config/i386/i386.h | 4 +- gcc/config/i386/i386.md | 4 +- gcc/config/i386/x86-tune-costs.h | 136 +++ gcc/config/i386/x86-tune-sched.cc | 2 + gcc/config/i386/x86-tune.def | 4 +- gcc/config/i386/znver5.md | 1081 +++++++++++++++++ gcc/doc/extend.texi | 3 + gcc/doc/invoke.texi | 10 + gcc/testsuite/g++.target/i386/mv29.C | 6 + gcc/testsuite/gcc.target/i386/funcspec-56.inc | 2 + 18 files changed, 1300 insertions(+), 11 deletions(-) create mode 100644 gcc/config/i386/znver5.md diff --git a/gcc/common/config/i386/cpuinfo.h b/gcc/common/config/i386/cpuinfo.h index a595ee537a8..017a952a5db 100644 --- a/gcc/common/config/i386/cpuinfo.h +++ b/gcc/common/config/i386/cpuinfo.h @@ -310,6 +310,22 @@ get_amd_cpu (struct __processor_model *cpu_model, cpu_model->__cpu_subtype = AMDFAM19H_ZNVER3; } break; + case 0x1a: + cpu_model->__cpu_type = AMDFAM1AH; + if (model <= 0x77) + { + cpu = "znver5"; + CHECK___builtin_cpu_is ("znver5"); + cpu_model->__cpu_subtype = AMDFAM1AH_ZNVER5; + } + else if (has_cpu_feature (cpu_model, cpu_features2, + FEATURE_AVX512VP2INTERSECT)) + { + cpu = "znver5"; + CHECK___builtin_cpu_is ("znver5"); + cpu_model->__cpu_subtype = AMDFAM1AH_ZNVER5; + } + break; default: break; } diff --git a/gcc/common/config/i386/i386-common.cc b/gcc/common/config/i386/i386-common.cc index c35191e6925..f814df8385b 100644 --- a/gcc/common/config/i386/i386-common.cc +++ b/gcc/common/config/i386/i386-common.cc @@ -2166,7 +2166,8 @@ const char *const processor_names[] = "znver1", "znver2", "znver3", - "znver4" + "znver4", + "znver5" }; /* Guarantee that the array is aligned with enum processor_type. */ @@ -2435,6 +2436,9 @@ const pta processor_alias_table[] = {"znver4", PROCESSOR_ZNVER4, CPU_ZNVER4, PTA_ZNVER4, M_CPU_SUBTYPE (AMDFAM19H_ZNVER4), P_PROC_AVX512F}, + {"znver5", PROCESSOR_ZNVER5, CPU_ZNVER5, + PTA_ZNVER5, + M_CPU_SUBTYPE (AMDFAM1AH_ZNVER5), P_PROC_AVX512F}, {"btver1", PROCESSOR_BTVER1, CPU_GENERIC, PTA_64BIT | PTA_MMX | PTA_SSE | PTA_SSE2 | PTA_SSE3 | PTA_SSSE3 | PTA_SSE4A | PTA_ABM | PTA_CX16 | PTA_PRFCHW diff --git a/gcc/common/config/i386/i386-cpuinfo.h b/gcc/common/config/i386/i386-cpuinfo.h index 2ee7470c8da..73131657eab 100644 --- a/gcc/common/config/i386/i386-cpuinfo.h +++ b/gcc/common/config/i386/i386-cpuinfo.h @@ -63,6 +63,7 @@ enum processor_types INTEL_SIERRAFOREST, INTEL_GRANDRIDGE, INTEL_CLEARWATERFOREST, + AMDFAM1AH, CPU_TYPE_MAX, BUILTIN_CPU_TYPE_MAX = CPU_TYPE_MAX }; @@ -104,6 +105,7 @@ enum processor_subtypes INTEL_COREI7_ARROWLAKE_S, INTEL_COREI7_PANTHERLAKE, ZHAOXIN_FAM7H_YONGFENG, + AMDFAM1AH_ZNVER5, CPU_SUBTYPE_MAX }; diff --git a/gcc/config.gcc b/gcc/config.gcc index a0f9c672308..39b14d2edd6 100644 --- a/gcc/config.gcc +++ b/gcc/config.gcc @@ -702,9 +702,9 @@ c7 esther" # 64-bit x86 processors supported by --with-arch=. Each processor # MUST be separated by exactly one space. x86_64_archs="amdfam10 athlon64 athlon64-sse3 barcelona bdver1 bdver2 \ -bdver3 bdver4 znver1 znver2 znver3 znver4 btver1 btver2 k8 k8-sse3 opteron \ -opteron-sse3 nocona core2 corei7 corei7-avx core-avx-i core-avx2 atom \ -slm nehalem westmere sandybridge ivybridge haswell broadwell bonnell \ +bdver3 bdver4 znver1 znver2 znver3 znver4 znver5 btver1 btver2 k8 k8-sse3 \ +opteron opteron-sse3 nocona core2 corei7 corei7-avx core-avx-i core-avx2 \ +atom slm nehalem westmere sandybridge ivybridge haswell broadwell bonnell \ silvermont knl knm skylake-avx512 cannonlake icelake-client icelake-server \ skylake goldmont goldmont-plus tremont cascadelake tigerlake cooperlake \ sapphirerapids alderlake rocketlake eden-x2 nano nano-1000 nano-2000 nano-3000 \ @@ -3755,6 +3755,10 @@ case ${target} in arch=znver4 cpu=znver4 ;; + znver5-*) + arch=znver5 + cpu=znver5 + ;; bdver4-*) arch=bdver4 cpu=bdver4 @@ -3892,6 +3896,10 @@ case ${target} in arch=znver4 cpu=znver4 ;; + znver5-*) + arch=znver5 + cpu=znver5 + ;; bdver4-*) arch=bdver4 cpu=bdver4 diff --git a/gcc/config/i386/driver-i386.cc b/gcc/config/i386/driver-i386.cc index 04f52396356..bb53af4b203 100644 --- a/gcc/config/i386/driver-i386.cc +++ b/gcc/config/i386/driver-i386.cc @@ -492,6 +492,8 @@ const char *host_detect_local_cpu (int argc, const char **argv) processor = PROCESSOR_GEODE; else if (has_feature (FEATURE_MOVBE) && family == 22) processor = PROCESSOR_BTVER2; + else if (has_feature (FEATURE_AVX512VP2INTERSECT)) + processor = PROCESSOR_ZNVER5; else if (has_feature (FEATURE_AVX512F)) processor = PROCESSOR_ZNVER4; else if (has_feature (FEATURE_VAES)) @@ -834,6 +836,9 @@ const char *host_detect_local_cpu (int argc, const char **argv) case PROCESSOR_ZNVER4: cpu = "znver4"; break; + case PROCESSOR_ZNVER5: + cpu = "znver5"; + break; case PROCESSOR_BTVER1: cpu = "btver1"; break; diff --git a/gcc/config/i386/i386-c.cc b/gcc/config/i386/i386-c.cc index 366b560158a..114908c7ec0 100644 --- a/gcc/config/i386/i386-c.cc +++ b/gcc/config/i386/i386-c.cc @@ -136,6 +136,10 @@ ix86_target_macros_internal (HOST_WIDE_INT isa_flag, def_or_undef (parse_in, "__znver4"); def_or_undef (parse_in, "__znver4__"); break; + case PROCESSOR_ZNVER5: + def_or_undef (parse_in, "__znver5"); + def_or_undef (parse_in, "__znver5__"); + break; case PROCESSOR_BTVER1: def_or_undef (parse_in, "__btver1"); def_or_undef (parse_in, "__btver1__"); @@ -374,6 +378,9 @@ ix86_target_macros_internal (HOST_WIDE_INT isa_flag, case PROCESSOR_ZNVER4: def_or_undef (parse_in, "__tune_znver4__"); break; + case PROCESSOR_ZNVER5: + def_or_undef (parse_in, "__tune_znver5__"); + break; case PROCESSOR_BTVER1: def_or_undef (parse_in, "__tune_btver1__"); break; diff --git a/gcc/config/i386/i386-options.cc b/gcc/config/i386/i386-options.cc index 8f5ce817630..b193dc3879e 100644 --- a/gcc/config/i386/i386-options.cc +++ b/gcc/config/i386/i386-options.cc @@ -172,11 +172,12 @@ along with GCC; see the file COPYING3. If not see #define m_ZNVER2 (HOST_WIDE_INT_1U<integer move cost is 2. */ + + /* reg-reg moves are done by renaming and thus they are even cheaper than + 1 cycle. Because reg-reg move cost is 2 and following tables correspond + to doubles of latencies, we do not model this correctly. It does not + seem to make practical difference to bump prices up even more. */ + 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). */ + {8, 8, 8}, /* cost of storing integer + registers. */ + 2, /* cost of reg,reg fld/fst. */ + {14, 14, 17}, /* cost of loading fp registers + in SFmode, DFmode and XFmode. */ + {12, 12, 16}, /* 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. */ + {8, 8}, /* cost of storing MMX registers + in SImode and DImode. */ + 2, 2, 3, /* cost of moving XMM,YMM,ZMM + register. */ + {6, 6, 10, 10, 12}, /* cost of loading SSE registers + in 32,64,128,256 and 512-bit. */ + {8, 8, 8, 12, 12}, /* cost of storing SSE registers + in 32,64,128,256 and 512-bit. */ + 6, 8, /* SSE->integer and integer->SSE + moves. */ + 8, 8, /* mask->integer and integer->mask moves */ + {6, 6, 6}, /* cost of loading mask register + in QImode, HImode, SImode. */ + {8, 8, 8}, /* 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. */ + /* TODO: Lea with 3 components has cost 2. */ + COSTS_N_INSNS (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 (3), /* HI. */ + COSTS_N_INSNS (3), /* SI. */ + COSTS_N_INSNS (3), /* DI. */ + COSTS_N_INSNS (3)}, /* other. */ + 0, /* cost of multiply per each bit + set. */ + {COSTS_N_INSNS (10), /* cost of a divide/mod for QI. */ + COSTS_N_INSNS (11), /* HI. */ + COSTS_N_INSNS (13), /* SI. */ + COSTS_N_INSNS (16), /* DI. */ + COSTS_N_INSNS (16)}, /* other. */ + COSTS_N_INSNS (1), /* cost of movsx. */ + COSTS_N_INSNS (1), /* cost of movzx. */ + 8, /* "large" insn. */ + 9, /* MOVE_RATIO. */ + 6, /* CLEAR_RATIO */ + {6, 6, 6}, /* cost of loading integer registers + in QImode, HImode and SImode. + Relative to reg-reg move (2). */ + {8, 8, 8}, /* cost of storing integer + registers. */ + {6, 6, 10, 10, 12}, /* cost of loading SSE registers + in 32bit, 64bit, 128bit, 256bit and 512bit */ + {8, 8, 8, 12, 12}, /* cost of storing SSE register + in 32bit, 64bit, 128bit, 256bit and 512bit */ + {6, 6, 6, 6, 6}, /* cost of unaligned loads. */ + {8, 8, 8, 8, 8}, /* cost of unaligned stores. */ + 2, 2, 2, /* cost of moving XMM,YMM,ZMM + register. */ + 6, /* cost of moving SSE register to integer. */ + /* VGATHERDPD is 17 uops and throughput is 4, VGATHERDPS is 24 uops, + throughput 5. Approx 7 uops do not depend on vector size and every load + is 5 uops. */ + 14, 10, /* Gather load static, per_elt. */ + 14, 20, /* Gather store static, per_elt. */ + 32, /* size of l1 cache. */ + 1024, /* size of l2 cache. */ + 64, /* size of prefetch block. */ + /* New AMD processors never drop prefetches; if they cannot be performed + immediately, they are queued. We set number of simultaneous prefetches + to a large constant to reflect this (it probably is not a good idea not + to limit number of prefetches at all, as their execution also takes some + time). */ + 100, /* number of parallel prefetches. */ + 3, /* Branch cost. */ + COSTS_N_INSNS (7), /* cost of FADD and FSUB insns. */ + COSTS_N_INSNS (7), /* cost of FMUL instruction. */ + /* Latency of fdiv is 8-15. */ + COSTS_N_INSNS (15), /* cost of FDIV instruction. */ + COSTS_N_INSNS (1), /* cost of FABS instruction. */ + COSTS_N_INSNS (1), /* cost of FCHS instruction. */ + /* Latency of fsqrt is 4-10. */ + COSTS_N_INSNS (25), /* 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 (3), /* cost of MULSS instruction. */ + COSTS_N_INSNS (3), /* cost of MULSD instruction. */ + COSTS_N_INSNS (4), /* cost of FMA SS instruction. */ + COSTS_N_INSNS (4), /* cost of FMA SD instruction. */ + COSTS_N_INSNS (10), /* cost of DIVSS instruction. */ + /* 9-13. */ + COSTS_N_INSNS (13), /* cost of DIVSD instruction. */ + COSTS_N_INSNS (14), /* cost of SQRTSS instruction. */ + COSTS_N_INSNS (20), /* cost of SQRTSD instruction. */ + /* Zen can execute 4 integer operations per cycle. FP operations + take 3 cycles and it can execute 2 integer additions and 2 + multiplications thus reassociation may make sense up to with of 6. + SPEC2k6 bencharks suggests + that 4 works better than 6 probably due to register pressure. + + Integer vector operations are taken by FP unit and execute 3 vector + plus/minus operations per cycle but only one multiply. This is adjusted + in ix86_reassociation_width. */ + 4, 4, 3, 6, /* reassoc int, fp, vec_int, vec_fp. */ + znver2_memcpy, + znver2_memset, + COSTS_N_INSNS (4), /* cond_taken_branch_cost. */ + COSTS_N_INSNS (2), /* cond_not_taken_branch_cost. */ + "16", /* Loop alignment. */ + "16", /* Jump alignment. */ + "0:0:8", /* Label alignment. */ + "16", /* Func alignment. */ + 4, /* Small unroll limit. */ + 2, /* Small unroll factor. */ +}; + /* skylake_cost should produce code tuned for Skylake familly of CPUs. */ static stringop_algs skylake_memcpy[2] = { {libcall, diff --git a/gcc/config/i386/x86-tune-sched.cc b/gcc/config/i386/x86-tune-sched.cc index 23a333714a6..578ba57e6b2 100644 --- a/gcc/config/i386/x86-tune-sched.cc +++ b/gcc/config/i386/x86-tune-sched.cc @@ -69,6 +69,7 @@ ix86_issue_rate (void) case PROCESSOR_ZNVER2: case PROCESSOR_ZNVER3: case PROCESSOR_ZNVER4: + case PROCESSOR_ZNVER5: case PROCESSOR_CORE2: case PROCESSOR_NEHALEM: case PROCESSOR_SANDYBRIDGE: @@ -417,6 +418,7 @@ ix86_adjust_cost (rtx_insn *insn, int dep_type, rtx_insn *dep_insn, int cost, case PROCESSOR_ZNVER2: case PROCESSOR_ZNVER3: case PROCESSOR_ZNVER4: + case PROCESSOR_ZNVER5: /* Stack engine allows to execute push&pop instructions in parall. */ if ((insn_type == TYPE_PUSH || insn_type == TYPE_POP) && (dep_insn_type == TYPE_PUSH || dep_insn_type == TYPE_POP)) diff --git a/gcc/config/i386/x86-tune.def b/gcc/config/i386/x86-tune.def index 8f855914316..ae2797b7cc2 100644 --- a/gcc/config/i386/x86-tune.def +++ b/gcc/config/i386/x86-tune.def @@ -575,12 +575,12 @@ DEF_TUNE (X86_TUNE_AVX256_STORE_BY_PIECES, "avx256_store_by_pieces", /* X86_TUNE_AVX512_MOVE_BY_PIECES: Optimize move_by_pieces with 512-bit AVX instructions. */ DEF_TUNE (X86_TUNE_AVX512_MOVE_BY_PIECES, "avx512_move_by_pieces", - m_SAPPHIRERAPIDS | m_ZNVER4) + m_SAPPHIRERAPIDS | m_ZNVER4 | m_ZNVER5) /* X86_TUNE_AVX512_STORE_BY_PIECES: Optimize store_by_pieces with 512-bit AVX instructions. */ DEF_TUNE (X86_TUNE_AVX512_STORE_BY_PIECES, "avx512_store_by_pieces", - m_SAPPHIRERAPIDS | m_ZNVER4) + m_SAPPHIRERAPIDS | m_ZNVER4 | m_ZNVER5) /*****************************************************************************/ /*****************************************************************************/ diff --git a/gcc/config/i386/znver5.md b/gcc/config/i386/znver5.md new file mode 100644 index 00000000000..9c9b69557b7 --- /dev/null +++ b/gcc/config/i386/znver5.md @@ -0,0 +1,1081 @@ +;; Copyright (C) 2012-2023 Free Software Foundation, Inc. +;; +;; This file is part of GCC. +;; +;; GCC is free software; you can redistribute it and/or modify +;; it under the terms of the GNU General Public License as published by +;; the Free Software Foundation; either version 3, or (at your option) +;; any later version. +;; +;; GCC is distributed in the hope that it will be useful, +;; but WITHOUT ANY WARRANTY; without even the implied warranty of +;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +;; GNU General Public License for more details. +;; +;; You should have received a copy of the GNU General Public License +;; along with GCC; see the file COPYING3. If not see +;; http://www.gnu.org/licenses/. +;; + + +(define_attr "znver5_decode" "direct,vector,double" + (const_string "direct")) + +;; AMD znver5 Scheduling +;; Modeling automatons for zen decoders, integer execution pipes, +;; AGU pipes, branch, floating point execution and fp store units. +(define_automaton "znver5, znver5_ieu, znver5_idiv, znver5_fdiv, znver5_agu, znver5_fpu, znver5_fp_store") + +;; Decoders unit has 4 decoders and all of them can decode fast path +;; and vector type instructions. +(define_cpu_unit "znver5-decode0" "znver5") +(define_cpu_unit "znver5-decode1" "znver5") +(define_cpu_unit "znver5-decode2" "znver5") +(define_cpu_unit "znver5-decode3" "znver5") + +;; Currently blocking all decoders for vector path instructions as +;; they are dispatched separetely as microcode sequence. +(define_reservation "znver5-vector" "znver5-decode0+znver5-decode1+znver5-decode2+znver5-decode3") + +;; Direct instructions can be issued to any of the four decoders. +(define_reservation "znver5-direct" "znver5-decode0|znver5-decode1|znver5-decode2|znver5-decode3") + +;; Fix me: Need to revisit this later to simulate fast path double behavior. +(define_reservation "znver5-double" "znver5-direct") + + +;; Integer unit 6 ALU pipes. +(define_cpu_unit "znver5-ieu0" "znver5_ieu") +(define_cpu_unit "znver5-ieu1" "znver5_ieu") +(define_cpu_unit "znver5-ieu2" "znver5_ieu") +(define_cpu_unit "znver5-ieu3" "znver5_ieu") +(define_cpu_unit "znver5-ieu4" "znver5_ieu") +(define_cpu_unit "znver5-ieu5" "znver5_ieu") + +;; As of now we have taken based on znver4, We need to revist once znver5 information +(define_cpu_unit "znver5-bru0" "znver5_ieu") +(define_reservation "znver5-ieu" "znver5-ieu0|znver5-ieu1|znver5-ieu2|znver5-ieu3|znver5-ieu4|znver5-ieu5") + +;; 4 AGU pipes in znver5 +(define_cpu_unit "znver5-agu0" "znver5_agu") +(define_cpu_unit "znver5-agu1" "znver5_agu") +(define_cpu_unit "znver5-agu2" "znver5_agu") +(define_cpu_unit "znver5-agu3" "znver5_agu") +(define_reservation "znver5-agu-reserve" "znver5-agu0|znver5-agu1|znver5-agu2|znver5-agu3") + +;; Load is 4 cycles. We do not model reservation of load unit. +(define_reservation "znver5-load" "znver5-agu-reserve") +(define_reservation "znver5-store" "znver5-agu-reserve") + +;; vectorpath (microcoded) instructions are single issue instructions. +;; So, they occupy all the integer units. +(define_reservation "znver5-ivector" "znver5-ieu0+znver5-ieu1 + +znver5-ieu2+znver5-ieu3+znver5-ieu4+znver5-ieu5+znver5-bru0 + +znver5-agu0+znver5-agu1+znver5-agu2+znver5-agu3") + +;; Floating point unit 4 FP pipes. +(define_cpu_unit "znver5-fpu0" "znver5_fpu") +(define_cpu_unit "znver5-fpu1" "znver5_fpu") +(define_cpu_unit "znver5-fpu2" "znver5_fpu") +(define_cpu_unit "znver5-fpu3" "znver5_fpu") + +(define_reservation "znver5-fpu" "znver5-fpu0|znver5-fpu1|znver5-fpu2|znver5-fpu3") + +;; Floating point store unit 2 FP pipes. +(define_cpu_unit "znver5-fp-store0" "znver5_fp_store") +(define_cpu_unit "znver5-fp-store1" "znver5_fp_store") + +(define_reservation "znver5-fvector" "znver5-fpu0+znver5-fpu1 + +znver5-fpu2+znver5-fpu3+znver5-fp-store0+znver5-fp-store1 + +znver5-agu0+znver5-agu1+znver5-agu2+znver5-agu3") + +(define_reservation "znver5-fp-store" "znver5-fp-store0|znver5-fp-store1") +(define_reservation "znver5-fp-store-512" "znver5-fp-store0+znver5-fp-store1") + +;; DIV units +(define_cpu_unit "znver5-idiv" "znver5_idiv") +(define_cpu_unit "znver5-fdiv" "znver5_fdiv") + +;; Integer Instructions +;; Move instructions +;; XCHG +(define_insn_reservation "znver5_imov_double" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "znver1_decode" "double") + (and (eq_attr "type" "imov") + (eq_attr "memory" "none")))) + "znver5-double,znver5-ieu") + +(define_insn_reservation "znver5_imov_double_load" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "znver1_decode" "double") + (and (eq_attr "type" "imov") + (eq_attr "memory" "load")))) + "znver5-double,znver5-load,znver5-ieu") + +;; imov, imovx +(define_insn_reservation "znver5_imov" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "imov,imovx") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-ieu") + +(define_insn_reservation "znver5_imov_load" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "imov,imovx") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-ieu") + +;; Push Instruction +(define_insn_reservation "znver5_push" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "push") + (eq_attr "memory" "store"))) + "znver5-direct,znver5-store") + +(define_insn_reservation "znver5_push_mem" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "push") + (eq_attr "memory" "both"))) + "znver5-direct,znver5-load,znver5-store") + +;; Pop instruction +(define_insn_reservation "znver5_pop" 4 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "pop") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load") + +(define_insn_reservation "znver5_pop_mem" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "pop") + (eq_attr "memory" "both"))) + "znver5-direct,znver5-load,znver5-store") + +;; Integer Instructions or General instructions +;; Multiplications +(define_insn_reservation "znver5_imul" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "imul") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-ieu1") + +(define_insn_reservation "znver5_imul_load" 7 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "imul") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-ieu1") + +;; Divisions +(define_insn_reservation "znver5_idiv_DI" 16 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "idiv") + (and (eq_attr "mode" "DI") + (eq_attr "memory" "none")))) + "znver5-double,znver5-idiv*10") + +(define_insn_reservation "znver5_idiv_SI" 13 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "idiv") + (and (eq_attr "mode" "SI") + (eq_attr "memory" "none")))) + "znver5-double,znver5-idiv*6") + +(define_insn_reservation "znver5_idiv_HI" 11 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "idiv") + (and (eq_attr "mode" "HI") + (eq_attr "memory" "none")))) + "znver5-double,znver5-idiv*4") + +(define_insn_reservation "znver5_idiv_QI" 10 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "idiv") + (and (eq_attr "mode" "QI") + (eq_attr "memory" "none")))) + "znver5-double,znver5-idiv*4") + +(define_insn_reservation "znver5_idiv_DI_load" 17 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "idiv") + (and (eq_attr "mode" "DI") + (eq_attr "memory" "load")))) + "znver5-double,znver5-load,znver5-idiv*10") + +(define_insn_reservation "znver5_idiv_SI_load" 17 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "idiv") + (and (eq_attr "mode" "SI") + (eq_attr "memory" "load")))) + "znver5-double,znver5-load,znver5-idiv*6") + +(define_insn_reservation "znver5_idiv_HI_load" 15 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "idiv") + (and (eq_attr "mode" "HI") + (eq_attr "memory" "load")))) + "znver5-double,znver5-load,znver5-idiv*4") + +(define_insn_reservation "znver5_idiv_QI_load" 14 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "idiv") + (and (eq_attr "mode" "QI") + (eq_attr "memory" "load")))) + "znver5-double,znver5-load,znver5-idiv*4") + +;; INTEGER/GENERAL Instructions +(define_insn_reservation "znver5_insn" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "alu,alu1,negnot,rotate1,ishift1,test,incdec,icmp") + (eq_attr "memory" "none,unknown"))) + "znver5-direct,znver5-ieu") + +(define_insn_reservation "znver5_insn_load" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "alu,alu1,negnot,rotate1,ishift1,test,incdec,icmp") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-ieu") + +(define_insn_reservation "znver5_insn2" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "icmov,setcc") + (eq_attr "memory" "none,unknown"))) + "znver5-direct,znver5-ieu") + +(define_insn_reservation "znver5_insn2_load" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "icmov,setcc") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-ieu") + +(define_insn_reservation "znver5_rotate" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "rotate") + (eq_attr "memory" "none,unknown"))) + "znver5-direct,znver5-ieu1|znver5-ieu2") + +(define_insn_reservation "znver5_rotate_load" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "rotate") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-ieu1|znver5-ieu2") + +(define_insn_reservation "znver5_insn_store" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "alu,alu1,negnot,rotate1,ishift1,test,incdec,icmp") + (eq_attr "memory" "store"))) + "znver5-direct,znver5-ieu,znver5-store") + +(define_insn_reservation "znver5_insn2_store" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "icmov,setcc") + (eq_attr "memory" "store"))) + "znver5-direct,znver5-ieu,znver5-store") + +(define_insn_reservation "znver5_rotate_store" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "rotate") + (eq_attr "memory" "store"))) + "znver5-direct,znver5-ieu1|znver5-ieu2,znver5-store") + +;; alu1 instructions +(define_insn_reservation "znver5_alu1_vector" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "znver1_decode" "vector") + (and (eq_attr "type" "alu1") + (eq_attr "memory" "none,unknown")))) + "znver5-vector,znver5-ivector*3") + +(define_insn_reservation "znver5_alu1_vector_load" 7 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "znver1_decode" "vector") + (and (eq_attr "type" "alu1") + (eq_attr "memory" "load")))) + "znver5-vector,znver5-load,znver5-ivector*3") + +;; Call Instruction +(define_insn_reservation "znver5_call" 1 + (and (eq_attr "cpu" "znver5") + (eq_attr "type" "call,callv")) + "znver5-double,znver5-ieu0|znver5-bru0,znver5-store") + +;; Branches +(define_insn_reservation "znver5_branch" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ibr") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-ieu0|znver5-bru0") + +(define_insn_reservation "znver5_branch_load" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ibr") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-ieu0|znver5-bru0") + +(define_insn_reservation "znver5_branch_vector" 2 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ibr") + (eq_attr "memory" "none,unknown"))) + "znver5-vector,znver5-ivector*2") + +(define_insn_reservation "znver5_branch_vector_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ibr") + (eq_attr "memory" "load"))) + "znver5-vector,znver5-load,znver5-ivector*2") + +;; LEA instruction with simple addressing +(define_insn_reservation "znver5_lea" 1 + (and (eq_attr "cpu" "znver5") + (eq_attr "type" "lea")) + "znver5-direct,znver5-ieu") + +;; Leave +(define_insn_reservation "znver5_leave" 1 + (and (eq_attr "cpu" "znver5") + (eq_attr "type" "leave")) + "znver5-double,znver5-ieu,znver5-store") + +;; STR and ISHIFT are microcoded. +(define_insn_reservation "znver5_str" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "str") + (eq_attr "memory" "none"))) + "znver5-vector,znver5-ivector*3") + +(define_insn_reservation "znver5_str_load" 7 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "str") + (eq_attr "memory" "load"))) + "znver5-vector,znver5-load,znver5-ivector*3") + +(define_insn_reservation "znver5_ishift" 2 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ishift") + (eq_attr "memory" "none"))) + "znver5-vector,znver5-ivector*2") + +(define_insn_reservation "znver5_ishift_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ishift") + (eq_attr "memory" "load"))) + "znver5-vector,znver5-load,znver5-ivector*2") + +;; Other vector type +(define_insn_reservation "znver5_ieu_vector" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "other,multi") + (eq_attr "memory" "none,unknown"))) + "znver5-vector,znver5-ivector*5") + +(define_insn_reservation "znver5_ieu_vector_load" 9 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "other,multi") + (eq_attr "memory" "load"))) + "znver5-vector,znver5-load,znver5-ivector*5") + +;; Floating Point +;; FP movs +(define_insn_reservation "znver5_fp_cmov" 4 + (and (eq_attr "cpu" "znver5") + (eq_attr "type" "fcmov")) + "znver5-vector,znver5-fvector*3") + +(define_insn_reservation "znver5_fp_mov_direct" 1 + (and (eq_attr "cpu" "znver5") + (eq_attr "type" "fmov")) + "znver5-direct,znver5-fpu0|znver5-fpu1") + +;;FLD +(define_insn_reservation "znver5_fp_mov_direct_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "znver1_decode" "direct") + (and (eq_attr "type" "fmov") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu1") + +;;FST +(define_insn_reservation "znver5_fp_mov_direct_store" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "znver1_decode" "direct") + (and (eq_attr "type" "fmov") + (eq_attr "memory" "store")))) + "znver5-direct,znver5-fpu0|znver5-fpu1,znver5-fp-store") + +;;FILD +(define_insn_reservation "znver5_fp_mov_double_load" 13 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "znver1_decode" "double") + (and (eq_attr "type" "fmov") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu1") + +;;FIST +(define_insn_reservation "znver5_fp_mov_double_store" 7 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "znver1_decode" "double") + (and (eq_attr "type" "fmov") + (eq_attr "memory" "store")))) + "znver5-double,znver5-fpu1,znver5-fp-store") + +;; FSQRT +(define_insn_reservation "znver5_fsqrt" 22 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "fpspc") + (and (eq_attr "mode" "XF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fdiv*10") + +;; FPSPC instructions +(define_insn_reservation "znver5_fp_spc" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "fpspc") + (eq_attr "memory" "none"))) + "znver5-vector,znver5-fvector*6") + +(define_insn_reservation "znver5_fp_insn_vector" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "znver1_decode" "vector") + (eq_attr "type" "mmxcvt,sselog1,ssemov"))) + "znver5-vector,znver5-fvector*6") + +;; FADD, FSUB, FMUL +(define_insn_reservation "znver5_fp_op_mul" 7 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "fop,fmul") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-fpu0") + +(define_insn_reservation "znver5_fp_op_mul_load" 12 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "fop,fmul") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-fpu0") + +;; FDIV +(define_insn_reservation "znver5_fp_div" 15 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "fdiv") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-fdiv*6") + +(define_insn_reservation "znver5_fp_div_load" 20 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "fdiv") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-fdiv*6") + +(define_insn_reservation "znver5_fp_idiv_load" 24 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "fdiv") + (and (eq_attr "fp_int_src" "true") + (eq_attr "memory" "load")))) + "znver5-double,znver5-load,znver5-fdiv*6") + +;; FABS, FCHS +(define_insn_reservation "znver5_fp_fsgn" 1 + (and (eq_attr "cpu" "znver5") + (eq_attr "type" "fsgn")) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +;; FCMP +(define_insn_reservation "znver5_fp_fcmp" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "fcmp") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-fpu1") + +(define_insn_reservation "znver5_fp_fcmp_double" 4 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "fcmp") + (and (eq_attr "znver1_decode" "double") + (eq_attr "memory" "none")))) + "znver5-double,znver5-fpu1,znver5-fp-store") + +;; MMX, SSE, SSEn.n instructions +(define_insn_reservation "znver5_fp_mmx " 1 + (and (eq_attr "cpu" "znver5") + (eq_attr "type" "mmx")) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_mmx_add_cmp" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "mmxadd,mmxcmp") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-fpu") + +(define_insn_reservation "znver5_mmx_add_cmp_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "mmxadd,mmxcmp") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-fpu") + +(define_insn_reservation "znver5_mmx_insn" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "mmxcvt,sseshuf,sseshuf1,mmxshft") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_mmx_insn_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "mmxcvt,sseshuf,sseshuf1,mmxshft") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu1|znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_mmx_mov" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "mmxmov") + (eq_attr "memory" "store"))) + "znver5-direct,znver5-fp-store") + +(define_insn_reservation "znver5_mmx_mov_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "mmxmov") + (eq_attr "memory" "both"))) + "znver5-direct,znver5-load,znver5-fp-store") + +(define_insn_reservation "znver5_mmx_mul" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "mmxmul") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-fpu0|znver5-fpu3") + +(define_insn_reservation "znver5_mmx_mul_load" 8 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "mmxmul") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu3") + +;; AVX instructions +(define_insn_reservation "znver5_sse_log" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sselog") + (and (eq_attr "mode" "V4SF,V8SF,V2DF,V4DF,QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu") + +(define_insn_reservation "znver5_sse_log_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sselog") + (and (eq_attr "mode" "V4SF,V8SF,V2DF,V4DF,QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu") + +(define_insn_reservation "znver5_sse_log1" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sselog1") + (and (eq_attr "mode" "V4SF,V8SF,V2DF,V4DF,QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "store")))) + "znver5-direct,znver5-fpu1|znver5-fpu2,znver5-fp-store") + +(define_insn_reservation "znver5_sse_log1_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sselog1") + (and (eq_attr "mode" "V4SF,V8SF,V2DF,V4DF,QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "both")))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2,znver5-fp-store") + +(define_insn_reservation "znver5_sse_comi" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecomi") + (eq_attr "memory" "store"))) + "znver5-double,znver5-fpu2|znver5-fpu3,znver5-fp-store") + +(define_insn_reservation "znver5_sse_comi_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecomi") + (eq_attr "memory" "both"))) + "znver5-double,znver5-load,znver5-fpu2|znver5-fpu3,znver5-fp-store") + +(define_insn_reservation "znver5_sse_test" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "prefix_extra" "1") + (and (eq_attr "type" "ssecomi") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_test_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "prefix_extra" "1") + (and (eq_attr "type" "ssecomi") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_imul" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseimul") + (and (eq_attr "mode" "QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_imul_load" 8 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseimul") + (and (eq_attr "mode" "QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_mov" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemov") + (and (eq_attr "mode" "QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_mov_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemov") + (and (eq_attr "mode" "QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_mov_store" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemov") + (and (eq_attr "mode" "QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "store")))) + "znver5-direct,znver5-fpu1|znver5-fpu2,znver5-fp-store") + +(define_insn_reservation "znver5_sse_mov_fp" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemov") + (and (eq_attr "mode" "V16SF,V8DF,V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu") + +(define_insn_reservation "znver5_sse_mov_fp_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemov") + (and (eq_attr "mode" "V16SF,V8DF,V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu") + +(define_insn_reservation "znver5_sse_mov_fp_store" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemov") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "store")))) + "znver5-direct,znver5-fp-store") + +(define_insn_reservation "znver5_sse_mov_fp_store_512" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemov") + (and (eq_attr "mode" "V16SF,V8DF") + (eq_attr "memory" "store")))) + "znver5-direct,znver5-fp-store-512") + +(define_insn_reservation "znver5_sse_add" 2 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseadd") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_add_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseadd") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_add1" 4 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseadd1") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "none")))) + "znver5-vector,znver5-fvector*2") + +(define_insn_reservation "znver5_sse_add1_load" 9 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseadd1") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "load")))) + "znver5-vector,znver5-load,znver5-fvector*2") + +(define_insn_reservation "znver5_sse_iadd" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseiadd") + (and (eq_attr "mode" "QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu") + +(define_insn_reservation "znver5_sse_iadd_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseiadd") + (and (eq_attr "mode" "QI,HI,SI,DI,TI,OI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu") + +(define_insn_reservation "znver5_sse_mul" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemul") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_mul_load" 8 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemul") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_div_pd" 13 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssediv") + (and (eq_attr "mode" "V4DF,V2DF,V1DF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fdiv*5") + +(define_insn_reservation "znver5_sse_div_ps" 10 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssediv") + (and (eq_attr "mode" "V8SF,V4SF,V2SF,SF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fdiv*3") + +(define_insn_reservation "znver5_sse_div_pd_load" 18 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssediv") + (and (eq_attr "mode" "V4DF,V2DF,V1DF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fdiv*5") + +(define_insn_reservation "znver5_sse_div_ps_load" 15 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssediv") + (and (eq_attr "mode" "V8SF,V4SF,V2SF,SF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fdiv*3") + +(define_insn_reservation "znver5_sse_cmp_avx" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecmp") + (and (eq_attr "prefix" "vex") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_cmp_avx_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecmp") + (and (eq_attr "prefix" "vex") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_comi_avx" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecomi") + (eq_attr "memory" "store"))) + "znver5-direct,znver5-fpu2+znver5-fpu3,znver5-fp-store") + +(define_insn_reservation "znver5_sse_comi_avx_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecomi") + (eq_attr "memory" "both"))) + "znver5-direct,znver5-load,znver5-fpu2+znver5-fpu3,znver5-fp-store") + +(define_insn_reservation "znver5_sse_cvt" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecvt") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_cvt_load" 8 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecvt") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_icvt" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecvt") + (and (eq_attr "mode" "SI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_icvt_store" 4 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecvt") + (and (eq_attr "mode" "SI") + (eq_attr "memory" "store")))) + "znver5-double,znver5-fpu2|znver5-fpu3,znver5-fp-store") + +(define_insn_reservation "znver5_sse_shuf" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseshuf") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu1|znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_shuf_load" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseshuf") + (and (eq_attr "mode" "V8SF,V4DF,V4SF,V2DF,V2SF,V1DF,SF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu") + +(define_insn_reservation "znver5_sse_ishuf" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseshuf") + (and (eq_attr "mode" "OI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_ishuf_load" 8 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseshuf") + (and (eq_attr "mode" "OI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2") + +;; AVX512 instructions +(define_insn_reservation "znver5_sse_log_evex" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sselog") + (and (eq_attr "mode" "V16SF,V8DF,XI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu0|znver5-fpu1|znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_log_evex_load" 7 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sselog") + (and (eq_attr "mode" "V16SF,V8DF,XI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu1|znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_log1_evex" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sselog1") + (and (eq_attr "mode" "V16SF,V8DF,XI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu1|znver5-fpu2,znver5-fp-store-512") + +(define_insn_reservation "znver5_sse_log1_evex_load" 7 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sselog1") + (and (eq_attr "mode" "V16SF,V8DF,XI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2,znver5-fp-store-512") + +(define_insn_reservation "znver5_sse_mul_evex" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemul") + (and (eq_attr "mode" "V16SF,V8DF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_mul_evex_load" 9 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemul") + (and (eq_attr "mode" "V16SF,V8DF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_imul_evex" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseimul") + (and (eq_attr "mode" "XI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu0|znver5-fpu3") + +(define_insn_reservation "znver5_sse_imul_evex_load" 9 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseimul") + (and (eq_attr "mode" "XI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_mov_evex" 2 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemov") + (and (eq_attr "mode" "XI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_mov_evex_load" 8 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemov") + (and (eq_attr "mode" "XI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_mov_evex_store" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemov") + (and (eq_attr "mode" "XI") + (eq_attr "memory" "store")))) + "znver5-direct,znver5-fpu1|znver5-fpu2,znver5-fp-store-512") + +(define_insn_reservation "znver5_sse_add_evex" 2 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseadd") + (and (eq_attr "mode" "V16SF,V8DF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_add_evex_load" 8 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseadd") + (and (eq_attr "mode" "V16SF,V8DF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_iadd_evex" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseiadd") + (and (eq_attr "mode" "XI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu0|znver5-fpu1|znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_iadd_evex_load" 7 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseiadd") + (and (eq_attr "mode" "XI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu1|znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_div_pd_evex" 13 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssediv") + (and (eq_attr "mode" "V8DF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fdiv*9") + +(define_insn_reservation "znver5_sse_div_ps_evex" 10 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssediv") + (and (eq_attr "mode" "V16SF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fdiv*6") + +(define_insn_reservation "znver5_sse_div_pd_evex_load" 19 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssediv") + (and (eq_attr "mode" "V8DF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fdiv*9") + +(define_insn_reservation "znver5_sse_div_ps_evex_load" 16 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssediv") + (and (eq_attr "mode" "V16SF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fdiv*6") + +(define_insn_reservation "znver5_sse_cmp_avx128" 3 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecmp") + (and (eq_attr "mode" "V4SF,V2DF,V2SF,V1DF,SF") + (and (eq_attr "prefix" "evex") + (eq_attr "memory" "none"))))) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_cmp_avx128_load" 9 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecmp") + (and (eq_attr "mode" "V4SF,V2DF,V2SF,V1DF,SF") + (and (eq_attr "prefix" "evex") + (eq_attr "memory" "load"))))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_cmp_avx256" 4 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecmp") + (and (eq_attr "mode" "V8SF,V4DF") + (and (eq_attr "prefix" "evex") + (eq_attr "memory" "none"))))) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_cmp_avx256_load" 10 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecmp") + (and (eq_attr "mode" "V8SF,V4DF") + (and (eq_attr "prefix" "evex") + (eq_attr "memory" "load"))))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_cmp_avx512" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecmp") + (and (eq_attr "mode" "V16SF,V8DF") + (and (eq_attr "prefix" "evex") + (eq_attr "memory" "none"))))) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_cmp_avx512_load" 11 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecmp") + (and (eq_attr "mode" "V16SF,V8DF") + (and (eq_attr "prefix" "evex") + (eq_attr "memory" "load"))))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_cvt_evex" 6 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecvt") + (and (eq_attr "mode" "V16SF,V8DF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu1|znver5-fpu2,znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_cvt_evex_load" 12 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssecvt") + (and (eq_attr "mode" "V16SF,V8DF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2,znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_shuf_evex" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseshuf") + (and (eq_attr "mode" "V16SF,V8DF") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu0|znver5-fpu1|znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_shuf_evex_load" 7 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseshuf") + (and (eq_attr "mode" "V16SF,V8DF") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu0|znver5-fpu1|znver5-fpu2|znver5-fpu3") + +(define_insn_reservation "znver5_sse_ishuf_evex" 5 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseshuf") + (and (eq_attr "mode" "XI") + (eq_attr "memory" "none")))) + "znver5-direct,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_ishuf_evex_load" 10 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseshuf") + (and (eq_attr "mode" "XI") + (eq_attr "memory" "load")))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2") + +(define_insn_reservation "znver5_sse_muladd" 4 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "ssemuladd") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_muladd_load" 10 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "sseshuf") + (eq_attr "memory" "load"))) + "znver5-direct,znver5-load,znver5-fpu1|znver5-fpu2") + +;; AVX512 mask instructions + +(define_insn_reservation "znver5_sse_mskmov" 2 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "mskmov") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-fpu0|znver5-fpu1") + +(define_insn_reservation "znver5_sse_msklog" 1 + (and (eq_attr "cpu" "znver5") + (and (eq_attr "type" "msklog") + (eq_attr "memory" "none"))) + "znver5-direct,znver5-fpu0|znver5-fpu3") diff --git a/gcc/doc/extend.texi b/gcc/doc/extend.texi index 2b8ba1949bf..22b4aceb217 100644 --- a/gcc/doc/extend.texi +++ b/gcc/doc/extend.texi @@ -26174,6 +26174,9 @@ AMD Family 19h Zen version 3. @item znver4 AMD Family 19h Zen version 4. + +@item znver5 +AMD Family 1ah Zen version 5. @end table Here is an example: diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.texi index 71339b8b30f..96b666fc9de 100644 --- a/gcc/doc/invoke.texi +++ b/gcc/doc/invoke.texi @@ -34418,6 +34418,16 @@ WBNOINVD, PKU, VPCLMULQDQ, VAES, AVX512F, AVX512DQ, AVX512IFMA, AVX512CD, AVX512BW, AVX512VL, AVX512BF16, AVX512VBMI, AVX512VBMI2, AVX512VNNI, AVX512BITALG, AVX512VPOPCNTDQ, GFNI and 64-bit instruction set extensions.) +@item znver5 +AMD Family 1ah core based CPUs with x86-64 instruction set support. (This +supersets BMI, BMI2, CLWB, F16C, FMA, FSGSBASE, AVX, AVX2, ADCX, RDSEED, +MWAITX, SHA, CLZERO, AES, PCLMUL, CX16, MOVBE, MMX, SSE, SSE2, SSE3, SSE4A, +SSSE3, SSE4.1, SSE4.2, ABM, XSAVEC, XSAVES, CLFLUSHOPT, POPCNT, RDPID, +WBNOINVD, PKU, VPCLMULQDQ, VAES, AVX512F, AVX512DQ, AVX512IFMA, AVX512CD, +AVX512BW, AVX512VL, AVX512BF16, AVX512VBMI, AVX512VBMI2, AVX512VNNI, +AVX512BITALG, AVX512VPOPCNTDQ, GFNI, AVXVNNI, MOVDIRI, MOVDIR64B, +AVX512VP2INTERSECT, PREFETCHI and 64-bit instruction set extensions.) + @item btver1 CPUs based on AMD Family 14h cores with x86-64 instruction set support. (This supersets MMX, SSE, SSE2, SSE3, SSSE3, SSE4A, CX16, ABM and 64-bit diff --git a/gcc/testsuite/g++.target/i386/mv29.C b/gcc/testsuite/g++.target/i386/mv29.C index a8dd8ac4803..ab229534edd 100644 --- a/gcc/testsuite/g++.target/i386/mv29.C +++ b/gcc/testsuite/g++.target/i386/mv29.C @@ -53,6 +53,10 @@ int __attribute__ ((target("arch=znver4"))) foo () { return 10; } +int __attribute__ ((target("arch=znver5"))) foo () { + return 11; +} + int main () { int val = foo (); @@ -77,6 +81,8 @@ int main () assert (val == 9); else if (__builtin_cpu_is ("znver4")) assert (val == 10); + else if (__builtin_cpu_is ("znver5")) + assert (val == 11); else assert (val == 0); diff --git a/gcc/testsuite/gcc.target/i386/funcspec-56.inc b/gcc/testsuite/gcc.target/i386/funcspec-56.inc index e910e1f9211..2a50f5bf67c 100644 --- a/gcc/testsuite/gcc.target/i386/funcspec-56.inc +++ b/gcc/testsuite/gcc.target/i386/funcspec-56.inc @@ -224,6 +224,7 @@ extern void test_arch_znver1 (void) __attribute__((__target__("arch= extern void test_arch_znver2 (void) __attribute__((__target__("arch=znver2"))); extern void test_arch_znver3 (void) __attribute__((__target__("arch=znver3"))); extern void test_arch_znver4 (void) __attribute__((__target__("arch=znver4"))); +extern void test_arch_znver5 (void) __attribute__((__target__("arch=znver5"))); extern void test_tune_nocona (void) __attribute__((__target__("tune=nocona"))); extern void test_tune_core2 (void) __attribute__((__target__("tune=core2"))); @@ -249,6 +250,7 @@ extern void test_tune_znver1 (void) __attribute__((__target__("tune= extern void test_tune_znver2 (void) __attribute__((__target__("tune=znver2"))); extern void test_tune_znver3 (void) __attribute__((__target__("tune=znver3"))); extern void test_tune_znver4 (void) __attribute__((__target__("tune=znver4"))); +extern void test_tune_znver5 (void) __attribute__((__target__("tune=znver5"))); extern void test_fpmath_sse (void) __attribute__((__target__("sse2,fpmath=sse"))); extern void test_fpmath_387 (void) __attribute__((__target__("sse2,fpmath=387"))); -- 2.34.1