i386.md (*popcounthi2_1): New insn_and_split pattern.

* config/i386/i386.md (*popcounthi2_1): New insn_and_split pattern.

testsuite/ChangeLog:

	* gcc.target/i386/pr59874-3.c: New test.

From-SVN: r243811
This commit is contained in:
Uros Bizjak 2016-12-19 21:23:08 +01:00
parent f56013e541
commit 6418b07737
4 changed files with 50 additions and 16 deletions

View File

@ -1,6 +1,10 @@
2016-12-19 Uros Bizjak <ubizjak@gmail.com>
* config/i386/i386.md (*popcounthi2_1): New insn_and_split pattern.
2016-12-19 Sandra Loosemore <sandra@codesourcery.com>
* doc/cpp.texi: Clean up anachronistic C99 references and remove
* doc/cpp.texi: Clean up anachronistic C99 references and remove
discussion of very old GCC versions.
(Differences from previous versions): Delete entire section.
@ -183,8 +187,7 @@
* config/arc/arc.md (mulsidi_600): Change to insn_and_split,
generate new mul64 insn for core multiplication work.
(umulsidi_600): Likewise, but use mulu64 insn.
(mul64): New pattern, content taken from old mulsidi_600 insn
pattern.
(mul64): New pattern, content taken from old mulsidi_600 insn pattern.
(mulu64): Likewise, but using umulsidi_600.
(mulsidi3): Remove move to destination, this is now handled by
mulsidi_600 insn_and_split.

View File

@ -13221,6 +13221,24 @@
(set_attr "type" "bitmanip")
(set_attr "mode" "<MODE>")])
(define_insn_and_split "*popcounthi2_1"
[(set (match_operand:SI 0 "register_operand")
(popcount:SI
(zero_extend:SI (match_operand:HI 1 "nonimmediate_operand"))))
(clobber (reg:CC FLAGS_REG))]
"TARGET_POPCNT
&& can_create_pseudo_p ()"
"#"
"&& 1"
[(const_int 0)]
{
rtx tmp = gen_reg_rtx (HImode);
emit_insn (gen_popcounthi2 (tmp, operands[1]));
emit_insn (gen_zero_extendhisi2 (operands[0], tmp));
DONE;
})
(define_insn "popcounthi2"
[(set (match_operand:HI 0 "register_operand" "=r")
(popcount:HI

View File

@ -1,18 +1,21 @@
2016-12-19 Will Schmidt <will_schmidt@vnet.ibm.com>
2016-12-19 Uros Bizjak <ubizjak@gmail.com>
* gcc.dg/vmx/mult-even-odd-be-order.c: Mark
variables as volatile.
* gcc.target/powerpc/fold-vec-mult-char.c: New.
* gcc.target/powerpc/fold-vec-mult-float.c: New.
* gcc.target/powerpc/fold-vec-mult-floatdouble.c: New.
* gcc.target/powerpc/fold-vec-mult-int.c: New.
* gcc.target/powerpc/fold-vec-mult-int128-p8.c: New.
* gcc.target/powerpc/fold-vec-mult-int128-p9.c: New.
* gcc.target/powerpc/fold-vec-mult-longlong.c: New.
* gcc.target/powerpc/fold-vec-mult-short.c: New.
* gcc.target/i386/pr59874-3.c: New test.
2016-12-19 Will Schmidt <will_schmidt@vnet.ibm.com>
* gcc.dg/vmx/mult-even-odd-be-order.c: Mark variables as volatile.
* gcc.target/powerpc/fold-vec-mult-char.c: New.
* gcc.target/powerpc/fold-vec-mult-float.c: New.
* gcc.target/powerpc/fold-vec-mult-floatdouble.c: New.
* gcc.target/powerpc/fold-vec-mult-int.c: New.
* gcc.target/powerpc/fold-vec-mult-int128-p8.c: New.
* gcc.target/powerpc/fold-vec-mult-int128-p9.c: New.
* gcc.target/powerpc/fold-vec-mult-longlong.c: New.
* gcc.target/powerpc/fold-vec-mult-short.c: New.
2016-12-19 Will Schmidt <will_schmidt@vnet.ibm.com>
* gcc.target/powerpc/fold-vec-sub-char.c: New.
* gcc.target/powerpc/fold-vec-sub-float.c: New.
* gcc.target/powerpc/fold-vec-sub-floatdouble.c: New.
@ -39,7 +42,7 @@
2016-12-17 Steven G. Kargl <kargl@gcc.gnu.org>
PR fortran/78746
* charlen_03.f90: Remove test.
* charlen_03.f90: Remove test.
* charlen_10.f90: Ditto.
2016-12-17 Jakub Jelinek <jakub@redhat.com>

View File

@ -0,0 +1,10 @@
/* PR target/59874 */
/* { dg-do compile } */
/* { dg-options "-O2 -mpopcnt -masm=att" } */
/* { dg-final { scan-assembler "popcntw" } } */
unsigned int
foo (unsigned short x)
{
return __builtin_popcount (x);
}