diff options
| author | Uros Bizjak <ubizjak@gmail.com> | 2021-05-26 20:44:49 +0200 |
|---|---|---|
| committer | Uros Bizjak <ubizjak@gmail.com> | 2021-05-26 20:46:26 +0200 |
| commit | 28484d00c45b7bf094a22a4fddf9ffdc7482c7e1 (patch) | |
| tree | 7004265233600a5c10b51e27431d7a5e03b21c00 /gcc/doc | |
| parent | a49a045b92f982f5617c3bbde97a33157237e25b (diff) | |
| download | gcc-28484d00c45b7bf094a22a4fddf9ffdc7482c7e1.zip gcc-28484d00c45b7bf094a22a4fddf9ffdc7482c7e1.tar.gz gcc-28484d00c45b7bf094a22a4fddf9ffdc7482c7e1.tar.bz2 | |
i386: Autovectorize 4-byte vectors
2021-05-26 Uroš Bizjak <ubizjak@gmail.com>
gcc/
* config/i386/i386.c (ix86_autovectorize_vector_modes):
Add V4QImode and V16QImode for TARGET_SSE2.
* doc/sourcebuild.texi (Vector-specific attributes):
Add vect64 and vect32 description.
gcc/testsuite/
* lib/target-supports.exp (check_effective_target_vect32): New.
(available_vector_sizes): Append 32 for x86 targets.
* gcc.dg/vect/pr71264.c (dg-final): Xfail scan dump for vect32 targets.
* gcc.dg/vect/slp-28.c (dg-final): Adjust dumps for vect32 targets.
* gcc.dg/vect/slp-3.c (dg-final): Ditto.
* gcc.target/i386/pr100637-3b.c: New test.
* gcc.target/i386/pr100637-3w.c: Ditto.
* gcc.target/i386/pr100637-4b.c: Ditto.
* gcc.target/i386/pr100637-4w.c: Ditto.
Diffstat (limited to 'gcc/doc')
| -rw-r--r-- | gcc/doc/sourcebuild.texi | 6 |
1 files changed, 6 insertions, 0 deletions
diff --git a/gcc/doc/sourcebuild.texi b/gcc/doc/sourcebuild.texi index cf309874..16c6a3b 100644 --- a/gcc/doc/sourcebuild.texi +++ b/gcc/doc/sourcebuild.texi @@ -1740,6 +1740,12 @@ circumstances. @item vect_variable_length Target has variable-length vectors. +@item vect64 +Target supports vectors of 64 bits. + +@item vect32 +Target supports vectors of 32 bits. + @item vect_widen_sum_hi_to_si Target supports a vector widening summation of @code{short} operands into @code{int} results, or can promote (unpack) from @code{short} |