>From 5071322839ac6200b713bebbc4152d9659a82c9c Mon Sep 17 00:00:00 2001 From: Maxim Kuvyrkov Date: Fri, 3 Feb 2017 11:26:24 +0000 Subject: [PATCH 4/4] tFrom: Maxim Kuvyrkov Prefetch tuning for ThunderX. --- gcc/config/aarch64/aarch64-cores.def | 4 +-- gcc/config/aarch64/aarch64.c | 52 +++++++++++++++++++++++++++++++++--- 2 files changed, 50 insertions(+), 6 deletions(-) diff --git a/gcc/config/aarch64/aarch64-cores.def b/gcc/config/aarch64/aarch64-cores.def index 92b57cf..b4f6d94 100644 --- a/gcc/config/aarch64/aarch64-cores.def +++ b/gcc/config/aarch64/aarch64-cores.def @@ -63,8 +63,8 @@ AARCH64_CORE("qdf24xx", qdf24xx, cortexa57, 8A, AARCH64_FL_FOR_ARCH8 | AA AARCH64_CORE("thunderx", thunderx, thunderx, 8A, AARCH64_FL_FOR_ARCH8 | AARCH64_FL_CRC | AARCH64_FL_CRYPTO, thunderx, 0x43, 0x0a0, -1) /* Do not swap around "thunderxt88p1" and "thunderxt88", this order is required to handle variant correctly. */ -AARCH64_CORE("thunderxt88p1", thunderxt88p1, thunderx, 8A, AARCH64_FL_FOR_ARCH8 | AARCH64_FL_CRC | AARCH64_FL_CRYPTO, thunderx, 0x43, 0x0a1, 0) -AARCH64_CORE("thunderxt88", thunderxt88, thunderx, 8A, AARCH64_FL_FOR_ARCH8 | AARCH64_FL_CRC | AARCH64_FL_CRYPTO, thunderx, 0x43, 0x0a1, -1) +AARCH64_CORE("thunderxt88p1", thunderxt88p1, thunderx, 8A, AARCH64_FL_FOR_ARCH8 | AARCH64_FL_CRC | AARCH64_FL_CRYPTO, thunderxt88, 0x43, 0x0a1, 0) +AARCH64_CORE("thunderxt88", thunderxt88, thunderx, 8A, AARCH64_FL_FOR_ARCH8 | AARCH64_FL_CRC | AARCH64_FL_CRYPTO, thunderxt88, 0x43, 0x0a1, -1) AARCH64_CORE("thunderxt81", thunderxt81, thunderx, 8A, AARCH64_FL_FOR_ARCH8 | AARCH64_FL_CRC | AARCH64_FL_CRYPTO, thunderx, 0x43, 0x0a2, -1) AARCH64_CORE("thunderxt83", thunderxt83, thunderx, 8A, AARCH64_FL_FOR_ARCH8 | AARCH64_FL_CRC | AARCH64_FL_CRYPTO, thunderx, 0x43, 0x0a3, -1) AARCH64_CORE("thunderx2t99", thunderx2t99, thunderx2t99, 8_1A, AARCH64_FL_FOR_ARCH8_1 | AARCH64_FL_CRYPTO, thunderx2t99, 0x43, 0x0af, -1) diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c index bdcdf02..00165e2 100644 --- a/gcc/config/aarch64/aarch64.c +++ b/gcc/config/aarch64/aarch64.c @@ -554,12 +554,30 @@ static const cpu_prefetch_tune qdf24xx_prefetch_tune = 3 /* default_opt_level */ }; +static const cpu_prefetch_tune thunderx_prefetch_tune = +{ + 8, /* num_slots */ + 32, /* l1_cache_size */ + 128, /* l1_cache_line_size */ + 0, /* l2_cache_size */ + -1 /* default_opt_level */ +}; + +static const cpu_prefetch_tune thunderxt88_prefetch_tune = +{ + 8, /* num_slots */ + 32, /* l1_cache_size */ + 128, /* l1_cache_line_size */ + 16*1024, /* l2_cache_size */ + 3 /* default_opt_level */ +}; + static const cpu_prefetch_tune thunderx2t99_prefetch_tune = { - 0, /* num_slots */ - -1, /* l1_cache_size */ + 8, /* num_slots */ + 32, /* l1_cache_size */ 64, /* l1_cache_line_size */ - -1, /* l2_cache_size */ + 256, /* l2_cache_size */ -1 /* default_opt_level */ }; @@ -767,7 +785,33 @@ static const struct tune_params thunderx_tunings = 0, /* max_case_values. */ tune_params::AUTOPREFETCHER_OFF, /* autoprefetcher_model. */ (AARCH64_EXTRA_TUNE_SLOW_UNALIGNED_LDPW), /* tune_flags. */ - &generic_prefetch_tune + &thunderx_prefetch_tune +}; + +/* Tunings for ThunderX CN88xx */ +static const struct tune_params thunderxt88_tunings = +{ + &thunderx_extra_costs, + &generic_addrcost_table, + &thunderx_regmove_cost, + &thunderx_vector_cost, + &generic_branch_cost, + &generic_approx_modes, + 6, /* memmov_cost */ + 2, /* issue_rate */ + AARCH64_FUSE_CMP_BRANCH, /* fusible_ops */ + 8, /* function_align. */ + 8, /* jump_align. */ + 8, /* loop_align. */ + 2, /* int_reassoc_width. */ + 4, /* fp_reassoc_width. */ + 1, /* vec_reassoc_width. */ + 2, /* min_div_recip_mul_sf. */ + 2, /* min_div_recip_mul_df. */ + 0, /* max_case_values. */ + tune_params::AUTOPREFETCHER_OFF, /* autoprefetcher_model. */ + (AARCH64_EXTRA_TUNE_SLOW_UNALIGNED_LDPW), /* tune_flags. */ + &thunderxt88_prefetch_tune }; static const struct tune_params xgene1_tunings = -- 2.7.4