diff options
| author | Kewen Lin <linkw@linux.ibm.com> | 2022-09-25 22:01:50 -0500 |
|---|---|---|
| committer | Kewen Lin <linkw@linux.ibm.com> | 2022-09-25 22:23:44 -0500 |
| commit | bfad7069b74c97000b698191c1945f07a6192db5 (patch) | |
| tree | 2e084e4b94b727b05798e0ccc5c698e48ed2eb14 /gcc | |
| parent | 3db8e9c2422d924a958336fd0871b24cce3e65d1 (diff) | |
| download | gcc-bfad7069b74c97000b698191c1945f07a6192db5.zip gcc-bfad7069b74c97000b698191c1945f07a6192db5.tar.gz gcc-bfad7069b74c97000b698191c1945f07a6192db5.tar.bz2 | |
rs6000: Fix condition of define_expand vec_shr_<mode> [PR100645]
PR100645 exposes one latent bug in define_expand vec_shr_<mode>
that the current condition TARGET_ALTIVEC is too loose. The
mode iterator VEC_L contains a few modes, they are not always
supported as vector mode, VECTOR_UNIT_ALTIVEC_OR_VSX_P should
be used like some other VEC_L usages.
PR target/100645
gcc/ChangeLog:
* config/rs6000/vector.md (vec_shr_<mode>): Replace condition
TARGET_ALTIVEC with VECTOR_UNIT_ALTIVEC_OR_VSX_P.
gcc/testsuite/ChangeLog:
* gcc.target/powerpc/pr100645.c: New test.
Diffstat (limited to 'gcc')
| -rw-r--r-- | gcc/config/rs6000/vector.md | 2 | ||||
| -rw-r--r-- | gcc/testsuite/gcc.target/powerpc/pr100645.c | 13 |
2 files changed, 14 insertions, 1 deletions
diff --git a/gcc/config/rs6000/vector.md b/gcc/config/rs6000/vector.md index a0d33d2..0171705 100644 --- a/gcc/config/rs6000/vector.md +++ b/gcc/config/rs6000/vector.md @@ -1475,7 +1475,7 @@ [(match_operand:VEC_L 0 "vlogical_operand") (match_operand:VEC_L 1 "vlogical_operand") (match_operand:QI 2 "reg_or_short_operand")] - "TARGET_ALTIVEC" + "VECTOR_UNIT_ALTIVEC_OR_VSX_P (<MODE>mode)" { rtx bitshift = operands[2]; rtx shift; diff --git a/gcc/testsuite/gcc.target/powerpc/pr100645.c b/gcc/testsuite/gcc.target/powerpc/pr100645.c new file mode 100644 index 0000000..c4e92cc --- /dev/null +++ b/gcc/testsuite/gcc.target/powerpc/pr100645.c @@ -0,0 +1,13 @@ +/* { dg-require-effective-target powerpc_altivec_ok } */ +/* { dg-options "-mdejagnu-cpu=power6 -maltivec" } */ + +/* This used to ICE. */ + +typedef long long v2di __attribute__ ((vector_size (16))); + +v2di +foo_v2di_l (v2di x) +{ + return __builtin_shuffle ((v2di){0, 0}, x, (v2di){3, 0}); +} + |