>
> Note I think Andrews comment#7 in the PR is spot-on then, the issue
> isn't the bitfield inserts but the compare where combine elides
> the sign_extend in favor of a subreg. That's likely some wrongdoing
> in simplify-rtx in the context of WORD_REGISTER_OPERATIONS.
>
Yes. There are 2 problems here. Any one of them can make this problem.
1) jump_insn eats sign_extend (and truncate) in
/* Simplify X, an IF_THEN_ELSE expression. Return the new expression. */
static rtx simplify_if_then_else (rtx x)
2) MIPS claims sign_extend to be deleted.
(define_insn_and_split "extendsidi2"
[(set (match_operand:DI 0 "register_operand" "=d,l,d")
(sign_extend:DI (match_operand:SI 1 "nonimmediate_operand" "0,0,m")))]
"TARGET_64BIT"
"@
#
#
lw\t%0,%1"
"&& reload_completed && register_operand (operands[1], VOIDmode)"
[(const_int 0)]
{
emit_note (NOTE_INSN_DELETED);
DONE;
}
[(set_attr "move_type" "move,move,load")
(set_attr "mode" "DI")])
--
YunQiang Su
