diff options
author | Bill Schmidt <wschmidt@linux.vnet.ibm.com> | 2012-02-10 16:38:37 +0000 |
---|---|---|
committer | William Schmidt <wschmidt@gcc.gnu.org> | 2012-02-10 16:38:37 +0000 |
commit | 8bd373026eaf4fe6d6ce7c5047f10dd85e29b942 (patch) | |
tree | 1455fc4d37f458d4ad3a9c635dedf65878af17e6 /gcc/targhooks.c | |
parent | 19f326e8ec1917f33e54e165e036f2c6ff091ce8 (diff) | |
download | gcc-8bd373026eaf4fe6d6ce7c5047f10dd85e29b942.zip gcc-8bd373026eaf4fe6d6ce7c5047f10dd85e29b942.tar.gz gcc-8bd373026eaf4fe6d6ce7c5047f10dd85e29b942.tar.bz2 |
re PR tree-optimization/50031 (Sphinx3 has a 10% regression going from GCC 4.5 to GCC 4.6 on powerpc)
2012-02-10 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
Ira Rosen <irar@il.ibm.com>
PR tree-optimization/50031
* targhooks.c (default_builtin_vectorization_cost): Handle
vec_promote_demote.
* target.h (enum vect_cost_for_stmt): Add vec_promote_demote.
* tree-vect-loop.c (vect_get_single_scalar_iteraion_cost): Handle
all types of reduction and pattern statements.
(vect_estimate_min_profitable_iters): Likewise.
* tree-vect-stmts.c (vect_model_promotion_demotion_cost): New function.
(vect_get_load_cost): Use vec_perm for permutations; add dump logic
for explicit realigns.
(vectorizable_conversion): Call vect_model_promotion_demotion_cost.
* config/spu/spu.c (spu_builtin_vectorization_cost): Handle
vec_promote_demote.
* config/i386/i386.c (ix86_builtin_vectorization_cost): Likewise.
* config/rs6000/rs6000.c (rs6000_builtin_vectorization_cost): Update
vec_perm for VSX and handle vec_promote_demote.
Co-Authored-By: Ira Rosen <irar@il.ibm.com>
From-SVN: r184102
Diffstat (limited to 'gcc/targhooks.c')
-rw-r--r-- | gcc/targhooks.c | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/gcc/targhooks.c b/gcc/targhooks.c index 2b4fd27..8e3d74e 100644 --- a/gcc/targhooks.c +++ b/gcc/targhooks.c @@ -514,6 +514,7 @@ default_builtin_vectorization_cost (enum vect_cost_for_stmt type_of_cost, case scalar_to_vec: case cond_branch_not_taken: case vec_perm: + case vec_promote_demote: return 1; case unaligned_load: |