diff options
| author | Jakub Jelinek <jakub@redhat.com> | 2020-04-23 21:57:50 +0200 |
|---|---|---|
| committer | Jakub Jelinek <jakub@redhat.com> | 2020-04-23 21:59:01 +0200 |
| commit | f51be2fb8653f81092f8158a0f0527275f86603b (patch) | |
| tree | 138937756a7595d95dae0ac71e325a51598f5103 | |
| parent | 3dbc772128e944819b09e21021d4fcd5dc17f2d4 (diff) | |
| download | gcc-f51be2fb8653f81092f8158a0f0527275f86603b.zip gcc-f51be2fb8653f81092f8158a0f0527275f86603b.tar.gz gcc-f51be2fb8653f81092f8158a0f0527275f86603b.tar.bz2 | |
Shortcut identity VEC_PERM expansion [PR94710]
This PR is about the rs6000 backend emitting wrong assembly
for whole vector shift by 0, and while I think it is desirable
to fix the backend, I don't see a point why the expander should
try to emit that, whole vector shift by 0 is identity, we can just
return the operand.
2020-04-23 Jakub Jelinek <jakub@redhat.com>
PR target/94710
* optabs.c (expand_vec_perm_const): For shift_amt const0_rtx
just return v2.
| -rw-r--r-- | gcc/ChangeLog | 4 | ||||
| -rw-r--r-- | gcc/optabs.c | 2 |
2 files changed, 6 insertions, 0 deletions
diff --git a/gcc/ChangeLog b/gcc/ChangeLog index 5ddb433..54e207c 100644 --- a/gcc/ChangeLog +++ b/gcc/ChangeLog @@ -1,5 +1,9 @@ 2020-04-23 Jakub Jelinek <jakub@redhat.com> + PR target/94710 + * optabs.c (expand_vec_perm_const): For shift_amt const0_rtx + just return v2. + PR middle-end/94724 * tree.c (get_narrower): Instead of creating COMPOUND_EXPRs temporarily with non-final second operand and updating it later, diff --git a/gcc/optabs.c b/gcc/optabs.c index 1456e59..d85ce47 100644 --- a/gcc/optabs.c +++ b/gcc/optabs.c @@ -5627,6 +5627,8 @@ expand_vec_perm_const (machine_mode mode, rtx v0, rtx v1, if (shift_amt) { class expand_operand ops[3]; + if (shift_amt == const0_rtx) + return v2; if (shift_code != CODE_FOR_nothing) { create_output_operand (&ops[0], target, mode); |