diff options
| author | Richard Biener <rguenther@suse.de> | 2020-12-03 10:25:14 +0100 |
|---|---|---|
| committer | Richard Biener <rguenther@suse.de> | 2020-12-07 12:07:12 +0100 |
| commit | ebdfd1606da6b5aa586b0cd156b69b659235c9c2 (patch) | |
| tree | 5a5649a6f9dd5d39848c76e182968f356e2d32c0 /gcc/testsuite/gcc.dg | |
| parent | cdcbef3c3310a14f2994982b44cb1f8e14c77232 (diff) | |
| download | gcc-ebdfd1606da6b5aa586b0cd156b69b659235c9c2.zip gcc-ebdfd1606da6b5aa586b0cd156b69b659235c9c2.tar.gz gcc-ebdfd1606da6b5aa586b0cd156b69b659235c9c2.tar.bz2 | |
tree-optimization/98113 - vectorize a sequence of BIT_INSERT_EXPRs
This adds the capability to handle a sequence of vector BIT_INSERT_EXPRs
to be vectorized similar as to how we vectorize vector constructors.
2020-12-03 Richard Biener <rguenther@suse.de>
PR tree-optimization/98113
* tree-vectorizer.h (struct slp_root): New.
(_bb_vec_info::roots): New member.
* tree-vect-slp.c (vect_analyze_slp): Also walk BB info
roots.
(_bb_vec_info::_bb_vec_info): Adjust.
(_bb_vec_info::~_bb_vec_info): Likewise.
(vld_cmp): New.
(vect_slp_is_lane_insert): Likewise.
(vect_slp_check_for_constructors): Match a series of
BIT_INSERT_EXPRs as vector constructor.
(vect_slp_analyze_bb_1): Continue if BB info roots is
not empty.
(vect_slp_analyze_bb_1): Mark the whole BIT_INSERT_EXPR root
sequence as pure_slp.
* gcc.dg/vect/bb-slp-70.c: New testcase.
Diffstat (limited to 'gcc/testsuite/gcc.dg')
| -rw-r--r-- | gcc/testsuite/gcc.dg/vect/bb-slp-70.c | 17 |
1 files changed, 17 insertions, 0 deletions
diff --git a/gcc/testsuite/gcc.dg/vect/bb-slp-70.c b/gcc/testsuite/gcc.dg/vect/bb-slp-70.c new file mode 100644 index 0000000..0eb7011 --- /dev/null +++ b/gcc/testsuite/gcc.dg/vect/bb-slp-70.c @@ -0,0 +1,17 @@ +/* { dg-do compile } */ +/* { dg-additional-options "-mavx512vl -mavx512vpopcntdq" { target avx512vpopcntdq } } */ + +typedef unsigned uv4si __attribute__((vector_size(16))); + +uv4si __attribute__((noinline)) +vpopctf (uv4si a) +{ + uv4si r; + r[2] = __builtin_popcount (a[2]); + r[1] = __builtin_popcount (a[1]); + r[0] = __builtin_popcount (a[0]); + r[3] = __builtin_popcount (a[3]); + return r; +} + +/* { dg-final { scan-tree-dump "optimized: basic block" "slp2" { target avx512vpopcntdq } } } */ |