[PATCH 1/3] AArch64: Add baseline tune

Wilco Dijkstra Thu, 14 Nov 2024 10:01:22 -0800

Cleanup the extra tune defines by introducing AARCH64_EXTRA_TUNE_BASE as a
common base supported by all modern cores.  Initially set it to
AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND.  No change in generated code.


Passes regress & bootstrap, OK for commit?

gcc/ChangeLog:

        * config/aarch64/aarch64-tuning-flags.def (AARCH64_EXTRA_TUNE_BASE): 
New define.
        * config/aarch64/tuning_models/ampere1b.h: Use AARCH64_EXTRA_TUNE_BASE.
        * config/aarch64/tuning_models/cortexx925.h: Likewise.
        * config/aarch64/tuning_models/fujitsu_monaka.h: Likewise.
        * config/aarch64/tuning_models/generic_armv8_a.h: Likewise.
        * config/aarch64/tuning_models/generic_armv9_a.h: Likewise.
        * config/aarch64/tuning_models/neoversen1.h: Likewise.
        * config/aarch64/tuning_models/neoversen2.h: Likewise.
        * config/aarch64/tuning_models/neoversen3.h: Likewise.
        * config/aarch64/tuning_models/neoversev1.h: Likewise.
        * config/aarch64/tuning_models/neoversev2.h: Likewise.
        * config/aarch64/tuning_models/neoversev3.h: Likewise.
        * config/aarch64/tuning_models/neoversev3ae.h: Likewise.

---

diff --git a/gcc/config/aarch64/aarch64-tuning-flags.def 
b/gcc/config/aarch64/aarch64-tuning-flags.def
index 
5939602576bdff18329d7bd1f8d944637db96462..ffbff20e29c78c00fc211adbba962c20827370aa
 100644
--- a/gcc/config/aarch64/aarch64-tuning-flags.def
+++ b/gcc/config/aarch64/aarch64-tuning-flags.def
@@ -50,4 +50,7 @@ AARCH64_EXTRA_TUNING_OPTION ("fully_pipelined_fma", 
FULLY_PIPELINED_FMA)
    rather than re-use an input predicate register.  */
 AARCH64_EXTRA_TUNING_OPTION ("avoid_pred_rmw", AVOID_PRED_RMW)
 
+/* Baseline tuning settings suitable for all modern cores.  */
+#define AARCH64_EXTRA_TUNE_BASE (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND)
+
 #undef AARCH64_EXTRA_TUNING_OPTION
diff --git a/gcc/config/aarch64/tuning_models/ampere1b.h 
b/gcc/config/aarch64/tuning_models/ampere1b.h
index 
9f0ee807da29ffba3525cc5542d37da5d5abcb6f..936fe7ad390edbf70f670d50843bc5caa4fa55e5
 100644
--- a/gcc/config/aarch64/tuning_models/ampere1b.h
+++ b/gcc/config/aarch64/tuning_models/ampere1b.h
@@ -102,7 +102,7 @@ static const struct tune_params ampere1b_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_STRONG,  /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
+  (AARCH64_EXTRA_TUNE_BASE
    | AARCH64_EXTRA_TUNE_AVOID_CROSS_LOOP_FMA
    | AARCH64_EXTRA_TUNE_FULLY_PIPELINED_FMA), /* tune_flags.  */
   &ampere1b_prefetch_tune,
diff --git a/gcc/config/aarch64/tuning_models/cortexx925.h 
b/gcc/config/aarch64/tuning_models/cortexx925.h
index 
eb9b89984b0472858bc08dba924c962ec4ba53bd..b2ff716157a452f4ff0260c5be8ddc0355e1a9e1
 100644
--- a/gcc/config/aarch64/tuning_models/cortexx925.h
+++ b/gcc/config/aarch64/tuning_models/cortexx925.h
@@ -217,7 +217,7 @@ static const struct tune_params cortexx925_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
+  (AARCH64_EXTRA_TUNE_BASE
    | AARCH64_EXTRA_TUNE_CSE_SVE_VL_CONSTANTS
    | AARCH64_EXTRA_TUNE_USE_NEW_VECTOR_COSTS
    | AARCH64_EXTRA_TUNE_MATCHED_VECTOR_THROUGHPUT
diff --git a/gcc/config/aarch64/tuning_models/fujitsu_monaka.h 
b/gcc/config/aarch64/tuning_models/fujitsu_monaka.h
index 
6a0984977591ff1e96f97fd5110ed51611bc29ed..2d704ecd1100b5ed04a81c297f4d1508089fa78b
 100644
--- a/gcc/config/aarch64/tuning_models/fujitsu_monaka.h
+++ b/gcc/config/aarch64/tuning_models/fujitsu_monaka.h
@@ -54,7 +54,7 @@ static const struct tune_params fujitsu_monaka_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
+  (AARCH64_EXTRA_TUNE_BASE
    | AARCH64_EXTRA_TUNE_USE_NEW_VECTOR_COSTS
    | AARCH64_EXTRA_TUNE_MATCHED_VECTOR_THROUGHPUT),    /* tune_flags.  */
   &generic_prefetch_tune,
diff --git a/gcc/config/aarch64/tuning_models/generic_armv8_a.h 
b/gcc/config/aarch64/tuning_models/generic_armv8_a.h
index 
9b1cbfc5bd2a828959833fb80e0e7b2fce71b3a4..bdd309ab03d7737a38c2b12b16db669424d43b3a
 100644
--- a/gcc/config/aarch64/tuning_models/generic_armv8_a.h
+++ b/gcc/config/aarch64/tuning_models/generic_armv8_a.h
@@ -181,7 +181,7 @@ static const struct tune_params generic_armv8_a_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
+  (AARCH64_EXTRA_TUNE_BASE
    | AARCH64_EXTRA_TUNE_CSE_SVE_VL_CONSTANTS
    | AARCH64_EXTRA_TUNE_USE_NEW_VECTOR_COSTS
    | AARCH64_EXTRA_TUNE_MATCHED_VECTOR_THROUGHPUT),    /* tune_flags.  */
diff --git a/gcc/config/aarch64/tuning_models/generic_armv9_a.h 
b/gcc/config/aarch64/tuning_models/generic_armv9_a.h
index 
48353a59939d84647c6981d6d0551af7ce9df751..a05a9ab92a27e8f24949aa2ffa5b5512c1487518
 100644
--- a/gcc/config/aarch64/tuning_models/generic_armv9_a.h
+++ b/gcc/config/aarch64/tuning_models/generic_armv9_a.h
@@ -248,7 +248,7 @@ static const struct tune_params generic_armv9_a_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
+  (AARCH64_EXTRA_TUNE_BASE
    | AARCH64_EXTRA_TUNE_USE_NEW_VECTOR_COSTS
    | AARCH64_EXTRA_TUNE_MATCHED_VECTOR_THROUGHPUT),    /* tune_flags.  */
   &generic_armv9a_prefetch_tune,
diff --git a/gcc/config/aarch64/tuning_models/neoversen1.h 
b/gcc/config/aarch64/tuning_models/neoversen1.h
index 
82def6b2736df8162d9b606440d260c951f3ef99..3079eb2d9ec37d77b7c851eb3eb24d8e39323b54
 100644
--- a/gcc/config/aarch64/tuning_models/neoversen1.h
+++ b/gcc/config/aarch64/tuning_models/neoversen1.h
@@ -51,7 +51,7 @@ static const struct tune_params neoversen1_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND),     /* tune_flags.  */
+  (AARCH64_EXTRA_TUNE_BASE),   /* tune_flags.  */
   &generic_prefetch_tune,
   AARCH64_LDP_STP_POLICY_ALWAYS,   /* ldp_policy_model.  */
   AARCH64_LDP_STP_POLICY_ALWAYS    /* stp_policy_model.  */
diff --git a/gcc/config/aarch64/tuning_models/neoversen2.h 
b/gcc/config/aarch64/tuning_models/neoversen2.h
index 
18199ac206c6cbfcef8695b497401b78a8f77f38..fd5f8f3737054e037428527206101f9bc726116d
 100644
--- a/gcc/config/aarch64/tuning_models/neoversen2.h
+++ b/gcc/config/aarch64/tuning_models/neoversen2.h
@@ -217,7 +217,7 @@ static const struct tune_params neoversen2_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
+  (AARCH64_EXTRA_TUNE_BASE
    | AARCH64_EXTRA_TUNE_CSE_SVE_VL_CONSTANTS
    | AARCH64_EXTRA_TUNE_USE_NEW_VECTOR_COSTS
    | AARCH64_EXTRA_TUNE_MATCHED_VECTOR_THROUGHPUT
diff --git a/gcc/config/aarch64/tuning_models/neoversen3.h 
b/gcc/config/aarch64/tuning_models/neoversen3.h
index 
4da85cfac0d185a5d59439f6d19d90ace0354e8f..8b156c2fe4d27c5df70446c4eb7d9153e8082268
 100644
--- a/gcc/config/aarch64/tuning_models/neoversen3.h
+++ b/gcc/config/aarch64/tuning_models/neoversen3.h
@@ -217,7 +217,7 @@ static const struct tune_params neoversen3_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
+  (AARCH64_EXTRA_TUNE_BASE
    | AARCH64_EXTRA_TUNE_CSE_SVE_VL_CONSTANTS
    | AARCH64_EXTRA_TUNE_USE_NEW_VECTOR_COSTS
    | AARCH64_EXTRA_TUNE_MATCHED_VECTOR_THROUGHPUT),    /* tune_flags.  */
diff --git a/gcc/config/aarch64/tuning_models/neoversev1.h 
b/gcc/config/aarch64/tuning_models/neoversev1.h
index 
dd9120eee48ae4fa30d589173788389c2413cdfe..23c121d865248c38c2bfef5e34c314207014649b
 100644
--- a/gcc/config/aarch64/tuning_models/neoversev1.h
+++ b/gcc/config/aarch64/tuning_models/neoversev1.h
@@ -226,10 +226,10 @@ static const struct tune_params neoversev1_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CSE_SVE_VL_CONSTANTS
+  (AARCH64_EXTRA_TUNE_BASE
+   | AARCH64_EXTRA_TUNE_CSE_SVE_VL_CONSTANTS
    | AARCH64_EXTRA_TUNE_USE_NEW_VECTOR_COSTS
    | AARCH64_EXTRA_TUNE_MATCHED_VECTOR_THROUGHPUT
-   | AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
    | AARCH64_EXTRA_TUNE_AVOID_PRED_RMW),       /* tune_flags.  */
   &generic_prefetch_tune,
   AARCH64_LDP_STP_POLICY_ALWAYS,   /* ldp_policy_model.  */
diff --git a/gcc/config/aarch64/tuning_models/neoversev2.h 
b/gcc/config/aarch64/tuning_models/neoversev2.h
index 
1369de73991cb9413ea3ea8fb292b7589420bab2..40af5f47f4f62757e8e374abbb29cec5d1a8f7f3
 100644
--- a/gcc/config/aarch64/tuning_models/neoversev2.h
+++ b/gcc/config/aarch64/tuning_models/neoversev2.h
@@ -230,7 +230,7 @@ static const struct tune_params neoversev2_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
+  (AARCH64_EXTRA_TUNE_BASE
    | AARCH64_EXTRA_TUNE_CSE_SVE_VL_CONSTANTS
    | AARCH64_EXTRA_TUNE_USE_NEW_VECTOR_COSTS
    | AARCH64_EXTRA_TUNE_MATCHED_VECTOR_THROUGHPUT
diff --git a/gcc/config/aarch64/tuning_models/neoversev3.h 
b/gcc/config/aarch64/tuning_models/neoversev3.h
index 
d8c82255378c81d4d2437412d41f3cbbd456f9c3..d65d74bfecfb69eac6c523870a372fc9d32e8687
 100644
--- a/gcc/config/aarch64/tuning_models/neoversev3.h
+++ b/gcc/config/aarch64/tuning_models/neoversev3.h
@@ -217,7 +217,7 @@ static const struct tune_params neoversev3_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
+  (AARCH64_EXTRA_TUNE_BASE
    | AARCH64_EXTRA_TUNE_CSE_SVE_VL_CONSTANTS
    | AARCH64_EXTRA_TUNE_USE_NEW_VECTOR_COSTS
    | AARCH64_EXTRA_TUNE_MATCHED_VECTOR_THROUGHPUT
diff --git a/gcc/config/aarch64/tuning_models/neoversev3ae.h 
b/gcc/config/aarch64/tuning_models/neoversev3ae.h
index 
7f050501ede7108e1109ea0b5efae355380d07dc..7b7fa0b4b081bb3bd17c8418b8ce9fc598cb19ce
 100644
--- a/gcc/config/aarch64/tuning_models/neoversev3ae.h
+++ b/gcc/config/aarch64/tuning_models/neoversev3ae.h
@@ -217,7 +217,7 @@ static const struct tune_params neoversev3ae_tunings =
   2,   /* min_div_recip_mul_df.  */
   0,   /* max_case_values.  */
   tune_params::AUTOPREFETCHER_WEAK,    /* autoprefetcher_model.  */
-  (AARCH64_EXTRA_TUNE_CHEAP_SHIFT_EXTEND
+  (AARCH64_EXTRA_TUNE_BASE
    | AARCH64_EXTRA_TUNE_CSE_SVE_VL_CONSTANTS
    | AARCH64_EXTRA_TUNE_USE_NEW_VECTOR_COSTS
    | AARCH64_EXTRA_TUNE_MATCHED_VECTOR_THROUGHPUT

[PATCH 1/3] AArch64: Add baseline tune

Reply via email to