From: Jan Hubicka Date: Tue, 10 Jan 2017 09:14:54 +0000 (+0100) Subject: re PR middle-end/77484 (Static branch predictor causes ~6-8% regression of SPEC2000... X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=ad730facfbf77c30042b711802b1aff317e2b8e2;p=gcc.git re PR middle-end/77484 (Static branch predictor causes ~6-8% regression of SPEC2000 GAP) PR middle-end/77484 * predict.def (PRED_CALL): Set to 67. From-SVN: r244260 --- diff --git a/gcc/ChangeLog b/gcc/ChangeLog index 5f728841fe8..48782726272 100644 --- a/gcc/ChangeLog +++ b/gcc/ChangeLog @@ -1,3 +1,8 @@ +2017-01-01 Jan Hubicka + + PR middle-end/77484 + * predict.def (PRED_CALL): Set to 67. + 2017-01-09 Eric Botcazou * expr.c (store_field): In the bitfield case, if the value comes from diff --git a/gcc/predict.def b/gcc/predict.def index f3b17c7ecdd..e96be12be4a 100644 --- a/gcc/predict.def +++ b/gcc/predict.def @@ -116,7 +116,7 @@ DEF_PREDICTOR (PRED_TREE_OPCODE_NONEQUAL, "opcode values nonequal (on trees)", DEF_PREDICTOR (PRED_TREE_FPOPCODE, "fp_opcode (on trees)", HITRATE (90), 0) /* Branch guarding call is probably taken. */ -DEF_PREDICTOR (PRED_CALL, "call", HITRATE (55), 0) +DEF_PREDICTOR (PRED_CALL, "call", HITRATE (67), 0) /* PRED_CALL is not very reliable predictor and it turns out to be even less reliable for indirect calls and polymorphic calls. For spec2k6