diff options
Diffstat (limited to 'gcc/tree-vectorizer.c')
| -rw-r--r-- | gcc/tree-vectorizer.c | 6 |
1 files changed, 3 insertions, 3 deletions
diff --git a/gcc/tree-vectorizer.c b/gcc/tree-vectorizer.c index ed37d5f..3e186a3 100644 --- a/gcc/tree-vectorizer.c +++ b/gcc/tree-vectorizer.c @@ -1762,7 +1762,7 @@ vect_is_simple_use (tree operand, loop_vec_info loop_vinfo, tree *def_stmt, vector form (i.e., when operating on arguments of type VECTYPE). The two kinds of widening operations we currently support are - NOP and WIDEN_MULT. This function checks if these oprations + NOP and WIDEN_MULT. This function checks if these operations are supported by the target platform either directly (via vector tree-codes), or via target builtins. @@ -1796,9 +1796,9 @@ supportable_widening_operation (enum tree_code code, tree stmt, tree vectype, vect1: [res1,res2,res3,res4], vect2: [res5,res6,res7,res8]. However, in the special case that the result of the widening operation is - used in a reduction copmutation only, the order doesn't matter (because + used in a reduction computation only, the order doesn't matter (because when vectorizing a reduction we change the order of the computation). - Some targets can take advatage of this and generate more efficient code. + Some targets can take advantage of this and generate more efficient code. For example, targets like Altivec, that support widen_mult using a sequence of {mult_even,mult_odd} generate the following vectors: vect1: [res1,res3,res5,res7], vect2: [res2,res4,res6,res8]. */ |