mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-11-22 02:33:06 +01:00
c112624b0b
As discussed in D101191, this patch adds a poison-safe folding of overflow bit check:
```
%Op0 = icmp ne i4 %X, 0
%Agg = call { i4, i1 } @llvm.[us]mul.with.overflow.i4(i4 %X, i4 %Y)
%Op1 = extractvalue { i4, i1 } %Agg, 1
%ret = select i1 %Op0, i1 %Op1, i1 false
=>
%Y.fr = freeze %Y
%Agg = call { i4, i1 } @llvm.[us]mul.with.overflow.i4(i4 %X, i4 %Y.fr)
%Op1 = extractvalue { i4, i1 } %Agg, 1
%ret = %Op1
```
https://alive2.llvm.org/ce/z/zgPUGT
https://alive2.llvm.org/ce/z/h2gZ_6
Note that there are cases where inserting freeze is not necessary: e.g. %Y is `noundef`.
In this case, LLVM is already good because `%ret` is already successfully folded into `and`,
triggering the pre-existing optimization in InstSimplify: https://godbolt.org/z/v6qena15K
Differential Revision: https://reviews.llvm.org/D101423
46 lines
1.7 KiB
C++
46 lines
1.7 KiB
C++
//===-- OverflowInstAnalysis.h - Utils to fold overflow insts ----*- C++ -*-==//
|
|
//
|
|
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
|
|
// See https://llvm.org/LICENSE.txt for license information.
|
|
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
//
|
|
// This file holds routines to help analyse overflow instructions
|
|
// and fold them into constants or other overflow instructions
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
#ifndef LLVM_ANALYSIS_OVERFLOWINSTANALYSIS_H
|
|
#define LLVM_ANALYSIS_OVERFLOWINSTANALYSIS_H
|
|
|
|
#include "llvm/IR/InstrTypes.h"
|
|
|
|
namespace llvm {
|
|
class Value;
|
|
class Use;
|
|
|
|
/// Match one of the patterns up to the select/logic op:
|
|
/// %Op0 = icmp ne i4 %X, 0
|
|
/// %Agg = call { i4, i1 } @llvm.[us]mul.with.overflow.i4(i4 %X, i4 %Y)
|
|
/// %Op1 = extractvalue { i4, i1 } %Agg, 1
|
|
/// %ret = select i1 %Op0, i1 %Op1, i1 false / %ret = and i1 %Op0, %Op1
|
|
///
|
|
/// %Op0 = icmp eq i4 %X, 0
|
|
/// %Agg = call { i4, i1 } @llvm.[us]mul.with.overflow.i4(i4 %X, i4 %Y)
|
|
/// %NotOp1 = extractvalue { i4, i1 } %Agg, 1
|
|
/// %Op1 = xor i1 %NotOp1, true
|
|
/// %ret = select i1 %Op0, i1 true, i1 %Op1 / %ret = or i1 %Op0, %Op1
|
|
///
|
|
/// Callers are expected to align that with the operands of the select/logic.
|
|
/// IsAnd is set to true if the Op0 and Op1 are used as the first pattern.
|
|
/// If Op0 and Op1 match one of the patterns above, return true and fill Y's
|
|
/// use.
|
|
|
|
bool isCheckForZeroAndMulWithOverflow(Value *Op0, Value *Op1, bool IsAnd,
|
|
Use *&Y);
|
|
bool isCheckForZeroAndMulWithOverflow(Value *Op0, Value *Op1, bool IsAnd);
|
|
} // end namespace llvm
|
|
|
|
#endif
|