procyberian/gcc
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

aarch64: Fix mismatched SVE predicate modes [PR94606]

Browse source 
For this testcase we ended up generating the invalid rtl:

(insn 10 9 11 2 (set (reg:VNx16BI 105)
        (and:VNx16BI (xor:VNx16BI (reg:VNx8BI 103)
                (reg:VNx16BI 104))
            (reg:VNx16BI 104))) "/tmp/bar.c":9:12 -1
     (nil))

Fixed by taking the VNx16BI lowpart.  It's safe to do that here because
the gp (r104) masks out the extra odd-indexed bits.

2020-04-16  Richard Sandiford  <richard.sandiford@arm.com>

gcc/
	PR target/94606
	* config/aarch64/aarch64.c (aarch64_expand_sve_const_pred_eor): Take
	the VNx16BI lowpart of the recursively-generated constant.

gcc/testsuite/
	PR target/94606
	* gcc.dg/vect/pr94606.c: New test.
This commit is contained in:
Richard Sandiford 2020-04-15 13:52:20 +01:00
parent d7a65edb62
commit 26bebf576d
4 changed files with 25 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
gcc/ChangeLog
View file

@ -1,3 +1,9 @@
2020-04-16 Richard Sandiford <richard.sandiford@arm.com>
PR target/94606
* config/aarch64/aarch64.c (aarch64_expand_sve_const_pred_eor): Take
the VNx16BI lowpart of the recursively-generated constant.
2020-04-16 Martin Liska <mliska@suse.cz>
Jakub Jelinek <jakub@redhat.com>

1
gcc/config/aarch64/aarch64.c
View file

@ -4742,6 +4742,7 @@ aarch64_expand_sve_const_pred_eor (rtx target, rtx_vector_builder &builder,
/* EOR the result with an ELT_SIZE PTRUE. */
rtx mask = aarch64_ptrue_all (elt_size);
mask = force_reg (VNx16BImode, mask);
inv = gen_lowpart (VNx16BImode, inv);
target = aarch64_target_reg (target, VNx16BImode);
emit_insn (gen_aarch64_pred_z (XOR, VNx16BImode, target, mask, inv, mask));
return target;

5
gcc/testsuite/ChangeLog
View file

@ -1,3 +1,8 @@
2020-04-16 Richard Sandiford <richard.sandiford@arm.com>
PR target/94606
* gcc.dg/vect/pr94606.c: New test.
2020-04-16 Martin Liska <mliska@suse.cz>
Jakub Jelinek <jakub@redhat.com>

13
gcc/testsuite/gcc.dg/vect/pr94606.c Normal file
View file

@ -0,0 +1,13 @@
/* { dg-do compile } */
/* { dg-additional-options "-march=armv8.2-a+sve -msve-vector-bits=256" { target aarch64*-*-* } } */
const short mask[] = { 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 1, 1, 1, 1, 1 };
int
foo (short *restrict x, short *restrict y)
{
for (int i = 0; i < 16; ++i)
if (mask[i])
x[i] += y[i];
}