re PR middle-end/62178 ([AArch64] Performance regression on matrix matrix multiply...
authorBin Cheng <bin.cheng@arm.com>
Thu, 18 Dec 2014 02:53:42 +0000 (02:53 +0000)
committerBin Cheng <amker@gcc.gnu.org>
Thu, 18 Dec 2014 02:53:42 +0000 (02:53 +0000)
commit6326a5f515dfa3922d0fdfc20baf005b40503b61
treed21aa5196aa4af2765aa35beba33169adc7adccb
parent29ca9bfb905a864546e5b22ca3f5b3ebd5cbb0a8
re PR middle-end/62178 ([AArch64] Performance regression on matrix matrix multiply due to r211211)

PR tree-optimization/62178
* tree-ssa-loop-ivopts.c (cheaper_cost_with_cand): New function.
(iv_ca_replace): New function.
(try_improve_iv_set): New parameter try_replace_p.
Break local optimal fixed-point by calling iv_ca_replace.
(find_optimal_iv_set_1): Pass new argument to try_improve_iv_set.

gcc/testsuite/ChangeLog
PR tree-optimization/62178
* gcc.target/aarch64/pr62178.c: New test.

From-SVN: r218855
gcc/ChangeLog
gcc/testsuite/ChangeLog
gcc/testsuite/gcc.target/aarch64/pr62178.c [new file with mode: 0644]
gcc/tree-ssa-loop-ivopts.c