Hi,
I noticed that commit r14-6192 can't help PR112606 #c3 as
it only takes care of SF/DF but TF/KF can still suffer the
issue. Similar to commit r14-6192, this patch is to take
care of copysign<mode>3 with IEEE128 as well.
Bootstrapped and regtested on powerpc64-linux-gnu P8/P9
and powerpc64le-linux-gnu P9 and P10.
I'm going to push this soon.
BR,
Kewen
-----
PR target/112606
gcc/ChangeLog:
* config/rs6000/rs6000.md (copysign<mode>3 IEEE128): Change predicate
of the last argument from altivec_register_operand to any_operand. If
operands[2] is CONST_DOUBLE, emit abs or neg abs depending on its sign
otherwise if it doesn't satisfy altivec_register_operand, force it to
REG using copy_to_mode_reg.
---
gcc/config/rs6000/rs6000.md | 20 +++++++++++++++++++-
1 file changed, 19 insertions(+), 1 deletion(-)
diff --git a/gcc/config/rs6000/rs6000.md b/gcc/config/rs6000/rs6000.md
index c880cec33a2..bc8bc6ab060 100644
--- a/gcc/config/rs6000/rs6000.md
+++ b/gcc/config/rs6000/rs6000.md
@@ -15020,9 +15020,27 @@ (define_insn "sqrt<mode>2"
(define_expand "copysign<mode>3"
[(use (match_operand:IEEE128 0 "altivec_register_operand"))
(use (match_operand:IEEE128 1 "altivec_register_operand"))
- (use (match_operand:IEEE128 2 "altivec_register_operand"))]
+ (use (match_operand:IEEE128 2 "any_operand"))]
"FLOAT128_IEEE_P (<MODE>mode)"
{
+ /* Middle-end canonicalizes -fabs (x) to copysign (x, -1),
+ but PowerPC prefers -fabs (x). */
+ if (CONST_DOUBLE_AS_FLOAT_P (operands[2]))
+ {
+ if (real_isneg (CONST_DOUBLE_REAL_VALUE (operands[2])))
+ {
+ rtx abs_res = gen_reg_rtx (<MODE>mode);
+ emit_insn (gen_abs<mode>2 (abs_res, operands[1]));
+ emit_insn (gen_neg<mode>2 (operands[0], abs_res));
+ }
+ else
+ emit_insn (gen_abs<mode>2 (operands[0], operands[1]));
+ DONE;
+ }
+
+ if (!altivec_register_operand (operands[2], <MODE>mode))
+ operands[2] = copy_to_mode_reg (<MODE>mode, operands[2]);
+
if (TARGET_FLOAT128_HW)
emit_insn (gen_copysign<mode>3_hard (operands[0], operands[1],
operands[2]));
--
2.42.0
