Remove the second_result field from instructions.

Now that result values are stored in value lists in the DFG, these
head-of-list pointers are no longer needed.
This commit is contained in:
Jakob Stoklund Olesen
2017-04-12 09:08:29 -07:00
parent 71338bb31f
commit 78a036fa51
5 changed files with 13 additions and 83 deletions

View File

@@ -1442,7 +1442,6 @@ impl<'a> Parser<'a> {
InstructionData::UnarySplit {
opcode: opcode,
ty: VOID,
second_result: None.into(),
arg: self.match_value("expected SSA value operand")?,
}
}
@@ -1474,7 +1473,6 @@ impl<'a> Parser<'a> {
InstructionData::BinaryOverflow {
opcode: opcode,
ty: VOID,
second_result: None.into(),
args: [lhs, rhs],
}
}
@@ -1497,7 +1495,6 @@ impl<'a> Parser<'a> {
InstructionData::MultiAry {
opcode: opcode,
ty: VOID,
second_result: None.into(),
args: args.into_value_list(&[], &mut ctx.function.dfg.value_lists),
}
}
@@ -1610,7 +1607,6 @@ impl<'a> Parser<'a> {
InstructionData::Call {
opcode: opcode,
ty: VOID,
second_result: None.into(),
func_ref: func_ref,
args: args.into_value_list(&[], &mut ctx.function.dfg.value_lists),
}
@@ -1626,7 +1622,6 @@ impl<'a> Parser<'a> {
InstructionData::IndirectCall {
opcode: opcode,
ty: VOID,
second_result: None.into(),
sig_ref: sig_ref,
args: args.into_value_list(&[callee], &mut ctx.function.dfg.value_lists),
}