From 8a595b1d2edf3d5d987a2a498e7f696596d0eaa5 Mon Sep 17 00:00:00 2001 From: Adhemerval Zanella Date: Mon, 18 Mar 2019 18:23:23 +0000 Subject: [PATCH] [AArch64] Refactor floating point materialization. NFC It splits the login of actual instruction emission away from the logic that figures out the appropriate sequence on AArch64ExpandPseudo::expandMOVImm. The new function AArch64_IMM::expandMOVImm, which return the list of the instructions to materialize the immediate constant, is implemented on a separated unit because it will be used in a subsequent patch to optimize floating point materialization. Reviewers: efriedma Differential Revision: https://reviews.llvm.org/D58915 llvm-svn: 356387 --- llvm/lib/Target/AArch64/AArch64ExpandImm.cpp | 408 ++++++++++++++ llvm/lib/Target/AArch64/AArch64ExpandImm.h | 35 ++ .../AArch64/AArch64ExpandPseudoInsts.cpp | 504 ++---------------- llvm/lib/Target/AArch64/CMakeLists.txt | 1 + 4 files changed, 485 insertions(+), 463 deletions(-) create mode 100644 llvm/lib/Target/AArch64/AArch64ExpandImm.cpp create mode 100644 llvm/lib/Target/AArch64/AArch64ExpandImm.h diff --git a/llvm/lib/Target/AArch64/AArch64ExpandImm.cpp b/llvm/lib/Target/AArch64/AArch64ExpandImm.cpp new file mode 100644 index 000000000000..c8602dabcd54 --- /dev/null +++ b/llvm/lib/Target/AArch64/AArch64ExpandImm.cpp @@ -0,0 +1,408 @@ +//===- AArch64ExpandImm.h - AArch64 Immediate Expansion -------------------===// +// +// 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 implements the AArch64ExpandImm stuff. +// +//===----------------------------------------------------------------------===// + +#include "AArch64.h" +#include "AArch64ExpandImm.h" +#include "MCTargetDesc/AArch64AddressingModes.h" + +namespace llvm { + +namespace AArch64_IMM { + +/// Helper function which extracts the specified 16-bit chunk from a +/// 64-bit value. +static uint64_t getChunk(uint64_t Imm, unsigned ChunkIdx) { + assert(ChunkIdx < 4 && "Out of range chunk index specified!"); + + return (Imm >> (ChunkIdx * 16)) & 0xFFFF; +} + +/// Check whether the given 16-bit chunk replicated to full 64-bit width +/// can be materialized with an ORR instruction. +static bool canUseOrr(uint64_t Chunk, uint64_t &Encoding) { + Chunk = (Chunk << 48) | (Chunk << 32) | (Chunk << 16) | Chunk; + + return AArch64_AM::processLogicalImmediate(Chunk, 64, Encoding); +} + +/// Check for identical 16-bit chunks within the constant and if so +/// materialize them with a single ORR instruction. The remaining one or two +/// 16-bit chunks will be materialized with MOVK instructions. +/// +/// This allows us to materialize constants like |A|B|A|A| or |A|B|C|A| (order +/// of the chunks doesn't matter), assuming |A|A|A|A| can be materialized with +/// an ORR instruction. +static bool tryToreplicateChunks(uint64_t UImm, + SmallVectorImpl &Insn) { + using CountMap = DenseMap; + + CountMap Counts; + + // Scan the constant and count how often every chunk occurs. + for (unsigned Idx = 0; Idx < 4; ++Idx) + ++Counts[getChunk(UImm, Idx)]; + + // Traverse the chunks to find one which occurs more than once. + for (CountMap::const_iterator Chunk = Counts.begin(), End = Counts.end(); + Chunk != End; ++Chunk) { + const uint64_t ChunkVal = Chunk->first; + const unsigned Count = Chunk->second; + + uint64_t Encoding = 0; + + // We are looking for chunks which have two or three instances and can be + // materialized with an ORR instruction. + if ((Count != 2 && Count != 3) || !canUseOrr(ChunkVal, Encoding)) + continue; + + const bool CountThree = Count == 3; + + Insn.push_back({ AArch64::ORRXri, 0, Encoding }); + + unsigned ShiftAmt = 0; + uint64_t Imm16 = 0; + // Find the first chunk not materialized with the ORR instruction. + for (; ShiftAmt < 64; ShiftAmt += 16) { + Imm16 = (UImm >> ShiftAmt) & 0xFFFF; + + if (Imm16 != ChunkVal) + break; + } + + // Create the first MOVK instruction. + Insn.push_back({ AArch64::MOVKXi, Imm16, + AArch64_AM::getShifterImm(AArch64_AM::LSL, ShiftAmt) }); + + // In case we have three instances the whole constant is now materialized + // and we can exit. + if (CountThree) + return true; + + // Find the remaining chunk which needs to be materialized. + for (ShiftAmt += 16; ShiftAmt < 64; ShiftAmt += 16) { + Imm16 = (UImm >> ShiftAmt) & 0xFFFF; + + if (Imm16 != ChunkVal) + break; + } + Insn.push_back({ AArch64::MOVKXi, Imm16, + AArch64_AM::getShifterImm(AArch64_AM::LSL, ShiftAmt) }); + return true; + } + + return false; +} + +/// Check whether this chunk matches the pattern '1...0...'. This pattern +/// starts a contiguous sequence of ones if we look at the bits from the LSB +/// towards the MSB. +static bool isStartChunk(uint64_t Chunk) { + if (Chunk == 0 || Chunk == std::numeric_limits::max()) + return false; + + return isMask_64(~Chunk); +} + +/// Check whether this chunk matches the pattern '0...1...' This pattern +/// ends a contiguous sequence of ones if we look at the bits from the LSB +/// towards the MSB. +static bool isEndChunk(uint64_t Chunk) { + if (Chunk == 0 || Chunk == std::numeric_limits::max()) + return false; + + return isMask_64(Chunk); +} + +/// Clear or set all bits in the chunk at the given index. +static uint64_t updateImm(uint64_t Imm, unsigned Idx, bool Clear) { + const uint64_t Mask = 0xFFFF; + + if (Clear) + // Clear chunk in the immediate. + Imm &= ~(Mask << (Idx * 16)); + else + // Set all bits in the immediate for the particular chunk. + Imm |= Mask << (Idx * 16); + + return Imm; +} + +/// Check whether the constant contains a sequence of contiguous ones, +/// which might be interrupted by one or two chunks. If so, materialize the +/// sequence of contiguous ones with an ORR instruction. +/// Materialize the chunks which are either interrupting the sequence or outside +/// of the sequence with a MOVK instruction. +/// +/// Assuming S is a chunk which starts the sequence (1...0...), E is a chunk +/// which ends the sequence (0...1...). Then we are looking for constants which +/// contain at least one S and E chunk. +/// E.g. |E|A|B|S|, |A|E|B|S| or |A|B|E|S|. +/// +/// We are also looking for constants like |S|A|B|E| where the contiguous +/// sequence of ones wraps around the MSB into the LSB. +static bool trySequenceOfOnes(uint64_t UImm, + SmallVectorImpl &Insn) { + const int NotSet = -1; + const uint64_t Mask = 0xFFFF; + + int StartIdx = NotSet; + int EndIdx = NotSet; + // Try to find the chunks which start/end a contiguous sequence of ones. + for (int Idx = 0; Idx < 4; ++Idx) { + int64_t Chunk = getChunk(UImm, Idx); + // Sign extend the 16-bit chunk to 64-bit. + Chunk = (Chunk << 48) >> 48; + + if (isStartChunk(Chunk)) + StartIdx = Idx; + else if (isEndChunk(Chunk)) + EndIdx = Idx; + } + + // Early exit in case we can't find a start/end chunk. + if (StartIdx == NotSet || EndIdx == NotSet) + return false; + + // Outside of the contiguous sequence of ones everything needs to be zero. + uint64_t Outside = 0; + // Chunks between the start and end chunk need to have all their bits set. + uint64_t Inside = Mask; + + // If our contiguous sequence of ones wraps around from the MSB into the LSB, + // just swap indices and pretend we are materializing a contiguous sequence + // of zeros surrounded by a contiguous sequence of ones. + if (StartIdx > EndIdx) { + std::swap(StartIdx, EndIdx); + std::swap(Outside, Inside); + } + + uint64_t OrrImm = UImm; + int FirstMovkIdx = NotSet; + int SecondMovkIdx = NotSet; + + // Find out which chunks we need to patch up to obtain a contiguous sequence + // of ones. + for (int Idx = 0; Idx < 4; ++Idx) { + const uint64_t Chunk = getChunk(UImm, Idx); + + // Check whether we are looking at a chunk which is not part of the + // contiguous sequence of ones. + if ((Idx < StartIdx || EndIdx < Idx) && Chunk != Outside) { + OrrImm = updateImm(OrrImm, Idx, Outside == 0); + + // Remember the index we need to patch. + if (FirstMovkIdx == NotSet) + FirstMovkIdx = Idx; + else + SecondMovkIdx = Idx; + + // Check whether we are looking a chunk which is part of the contiguous + // sequence of ones. + } else if (Idx > StartIdx && Idx < EndIdx && Chunk != Inside) { + OrrImm = updateImm(OrrImm, Idx, Inside != Mask); + + // Remember the index we need to patch. + if (FirstMovkIdx == NotSet) + FirstMovkIdx = Idx; + else + SecondMovkIdx = Idx; + } + } + assert(FirstMovkIdx != NotSet && "Constant materializable with single ORR!"); + + // Create the ORR-immediate instruction. + uint64_t Encoding = 0; + AArch64_AM::processLogicalImmediate(OrrImm, 64, Encoding); + Insn.push_back({ AArch64::ORRXri, 0, Encoding }); + + const bool SingleMovk = SecondMovkIdx == NotSet; + Insn.push_back({ AArch64::MOVKXi, getChunk(UImm, FirstMovkIdx), + AArch64_AM::getShifterImm(AArch64_AM::LSL, + FirstMovkIdx * 16) }); + + // Early exit in case we only need to emit a single MOVK instruction. + if (SingleMovk) + return true; + + // Create the second MOVK instruction. + Insn.push_back({ AArch64::MOVKXi, getChunk(UImm, SecondMovkIdx), + AArch64_AM::getShifterImm(AArch64_AM::LSL, + SecondMovkIdx * 16) }); + + return true; +} + +/// \brief Expand a MOVi32imm or MOVi64imm pseudo instruction to a +/// MOVZ or MOVN of width BitSize followed by up to 3 MOVK instructions. +static inline void expandMOVImmSimple(uint64_t Imm, unsigned BitSize, + unsigned OneChunks, unsigned ZeroChunks, + SmallVectorImpl &Insn) { + const unsigned Mask = 0xFFFF; + + // Use a MOVZ or MOVN instruction to set the high bits, followed by one or + // more MOVK instructions to insert additional 16-bit portions into the + // lower bits. + bool isNeg = false; + + // Use MOVN to materialize the high bits if we have more all one chunks + // than all zero chunks. + if (OneChunks > ZeroChunks) { + isNeg = true; + Imm = ~Imm; + } + + unsigned FirstOpc; + if (BitSize == 32) { + Imm &= (1LL << 32) - 1; + FirstOpc = (isNeg ? AArch64::MOVNWi : AArch64::MOVZWi); + } else { + FirstOpc = (isNeg ? AArch64::MOVNXi : AArch64::MOVZXi); + } + unsigned Shift = 0; // LSL amount for high bits with MOVZ/MOVN + unsigned LastShift = 0; // LSL amount for last MOVK + if (Imm != 0) { + unsigned LZ = countLeadingZeros(Imm); + unsigned TZ = countTrailingZeros(Imm); + Shift = (TZ / 16) * 16; + LastShift = ((63 - LZ) / 16) * 16; + } + unsigned Imm16 = (Imm >> Shift) & Mask; + + Insn.push_back({ FirstOpc, Imm16, + AArch64_AM::getShifterImm(AArch64_AM::LSL, Shift) }); + + if (Shift == LastShift) + return; + + // If a MOVN was used for the high bits of a negative value, flip the rest + // of the bits back for use with MOVK. + if (isNeg) + Imm = ~Imm; + + unsigned Opc = (BitSize == 32 ? AArch64::MOVKWi : AArch64::MOVKXi); + while (Shift < LastShift) { + Shift += 16; + Imm16 = (Imm >> Shift) & Mask; + if (Imm16 == (isNeg ? Mask : 0)) + continue; // This 16-bit portion is already set correctly. + + Insn.push_back({ Opc, Imm16, + AArch64_AM::getShifterImm(AArch64_AM::LSL, Shift) }); + } +} + +/// Expand a MOVi32imm or MOVi64imm pseudo instruction to one or more +/// real move-immediate instructions to synthesize the immediate. +void expandMOVImm(uint64_t Imm, unsigned BitSize, + SmallVectorImpl &Insn) { + const unsigned Mask = 0xFFFF; + + // Scan the immediate and count the number of 16-bit chunks which are either + // all ones or all zeros. + unsigned OneChunks = 0; + unsigned ZeroChunks = 0; + for (unsigned Shift = 0; Shift < BitSize; Shift += 16) { + const unsigned Chunk = (Imm >> Shift) & Mask; + if (Chunk == Mask) + OneChunks++; + else if (Chunk == 0) + ZeroChunks++; + } + + // FIXME: Prefer MOVZ/MOVN over ORR because of the rules for the "mov" + // alias. + + // Try a single ORR. + uint64_t UImm = Imm << (64 - BitSize) >> (64 - BitSize); + uint64_t Encoding; + if (AArch64_AM::processLogicalImmediate(UImm, BitSize, Encoding)) { + unsigned Opc = (BitSize == 32 ? AArch64::ORRWri : AArch64::ORRXri); + Insn.push_back({ Opc, 0, Encoding }); + return; + } + + // One to up three instruction sequences. + // + // Prefer MOVZ/MOVN followed by MOVK; it's more readable, and possibly the + // fastest sequence with fast literal generation. + if (OneChunks >= (BitSize / 16) - 2 || ZeroChunks >= (BitSize / 16) - 2) { + expandMOVImmSimple(Imm, BitSize, OneChunks, ZeroChunks, Insn); + return; + } + + assert(BitSize == 64 && "All 32-bit immediates can be expanded with a" + "MOVZ/MOVK pair"); + + // Try other two-instruction sequences. + + // 64-bit ORR followed by MOVK. + // We try to construct the ORR immediate in three different ways: either we + // zero out the chunk which will be replaced, we fill the chunk which will + // be replaced with ones, or we take the bit pattern from the other half of + // the 64-bit immediate. This is comprehensive because of the way ORR + // immediates are constructed. + for (unsigned Shift = 0; Shift < BitSize; Shift += 16) { + uint64_t ShiftedMask = (0xFFFFULL << Shift); + uint64_t ZeroChunk = UImm & ~ShiftedMask; + uint64_t OneChunk = UImm | ShiftedMask; + uint64_t RotatedImm = (UImm << 32) | (UImm >> 32); + uint64_t ReplicateChunk = ZeroChunk | (RotatedImm & ShiftedMask); + if (AArch64_AM::processLogicalImmediate(ZeroChunk, BitSize, Encoding) || + AArch64_AM::processLogicalImmediate(OneChunk, BitSize, Encoding) || + AArch64_AM::processLogicalImmediate(ReplicateChunk, BitSize, + Encoding)) { + // Create the ORR-immediate instruction. + Insn.push_back({ AArch64::ORRXri, 0, Encoding }); + + // Create the MOVK instruction. + const unsigned Imm16 = getChunk(UImm, Shift / 16); + Insn.push_back({ AArch64::MOVKXi, Imm16, + AArch64_AM::getShifterImm(AArch64_AM::LSL, Shift) }); + return; + } + } + + // FIXME: Add more two-instruction sequences. + + // Three instruction sequences. + // + // Prefer MOVZ/MOVN followed by two MOVK; it's more readable, and possibly + // the fastest sequence with fast literal generation. (If neither MOVK is + // part of a fast literal generation pair, it could be slower than the + // four-instruction sequence, but we won't worry about that for now.) + if (OneChunks || ZeroChunks) { + expandMOVImmSimple(Imm, BitSize, OneChunks, ZeroChunks, Insn); + return; + } + + // Check for identical 16-bit chunks within the constant and if so materialize + // them with a single ORR instruction. The remaining one or two 16-bit chunks + // will be materialized with MOVK instructions. + if (BitSize == 64 && tryToreplicateChunks(UImm, Insn)) + return; + + // Check whether the constant contains a sequence of contiguous ones, which + // might be interrupted by one or two chunks. If so, materialize the sequence + // of contiguous ones with an ORR instruction. Materialize the chunks which + // are either interrupting the sequence or outside of the sequence with a + // MOVK instruction. + if (BitSize == 64 && trySequenceOfOnes(UImm, Insn)) + return; + + // We found no possible two or three instruction sequence; use the general + // four-instruction sequence. + expandMOVImmSimple(Imm, BitSize, OneChunks, ZeroChunks, Insn); +} + +} // end namespace AArch64_AM + +} // end namespace llvm diff --git a/llvm/lib/Target/AArch64/AArch64ExpandImm.h b/llvm/lib/Target/AArch64/AArch64ExpandImm.h new file mode 100644 index 000000000000..42c97d2c3e9b --- /dev/null +++ b/llvm/lib/Target/AArch64/AArch64ExpandImm.h @@ -0,0 +1,35 @@ +//===- AArch64ExpandImm.h - AArch64 Immediate Expansion ---------*- 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 contains the AArch64 immediate expansion stuff. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_LIB_TARGET_AARCH64_MCTARGETDESC_AARCH64EXPANDIMM_H +#define LLVM_LIB_TARGET_AARCH64_MCTARGETDESC_AARCH64EXPANDIMM_H + +#include "llvm/ADT/SmallVector.h" + +namespace llvm { + +namespace AArch64_IMM { + +struct ImmInsnModel { + unsigned Opcode; + uint64_t Op1; + uint64_t Op2; +}; + +void expandMOVImm(uint64_t Imm, unsigned BitSize, + SmallVectorImpl &Insn); + +} // end namespace AArch64_IMM + +} // end namespace llvm + +#endif diff --git a/llvm/lib/Target/AArch64/AArch64ExpandPseudoInsts.cpp b/llvm/lib/Target/AArch64/AArch64ExpandPseudoInsts.cpp index 2d0b52b67e57..2b73f3b9c10d 100644 --- a/llvm/lib/Target/AArch64/AArch64ExpandPseudoInsts.cpp +++ b/llvm/lib/Target/AArch64/AArch64ExpandPseudoInsts.cpp @@ -13,6 +13,7 @@ // //===----------------------------------------------------------------------===// +#include "AArch64ExpandImm.h" #include "AArch64InstrInfo.h" #include "AArch64Subtarget.h" #include "MCTargetDesc/AArch64AddressingModes.h" @@ -65,11 +66,6 @@ private: MachineBasicBlock::iterator &NextMBBI); bool expandMOVImm(MachineBasicBlock &MBB, MachineBasicBlock::iterator MBBI, unsigned BitSize); - bool expandMOVImmSimple(MachineBasicBlock &MBB, - MachineBasicBlock::iterator MBBI, - unsigned BitSize, - unsigned OneChunks, - unsigned ZeroChunks); bool expandCMP_SWAP(MachineBasicBlock &MBB, MachineBasicBlock::iterator MBBI, unsigned LdarOp, unsigned StlrOp, unsigned CmpOp, @@ -103,279 +99,6 @@ static void transferImpOps(MachineInstr &OldMI, MachineInstrBuilder &UseMI, } } -/// Helper function which extracts the specified 16-bit chunk from a -/// 64-bit value. -static uint64_t getChunk(uint64_t Imm, unsigned ChunkIdx) { - assert(ChunkIdx < 4 && "Out of range chunk index specified!"); - - return (Imm >> (ChunkIdx * 16)) & 0xFFFF; -} - -/// Check whether the given 16-bit chunk replicated to full 64-bit width -/// can be materialized with an ORR instruction. -static bool canUseOrr(uint64_t Chunk, uint64_t &Encoding) { - Chunk = (Chunk << 48) | (Chunk << 32) | (Chunk << 16) | Chunk; - - return AArch64_AM::processLogicalImmediate(Chunk, 64, Encoding); -} - -/// Check for identical 16-bit chunks within the constant and if so -/// materialize them with a single ORR instruction. The remaining one or two -/// 16-bit chunks will be materialized with MOVK instructions. -/// -/// This allows us to materialize constants like |A|B|A|A| or |A|B|C|A| (order -/// of the chunks doesn't matter), assuming |A|A|A|A| can be materialized with -/// an ORR instruction. -static bool tryToreplicateChunks(uint64_t UImm, MachineInstr &MI, - MachineBasicBlock &MBB, - MachineBasicBlock::iterator &MBBI, - const AArch64InstrInfo *TII) { - using CountMap = DenseMap; - - CountMap Counts; - - // Scan the constant and count how often every chunk occurs. - for (unsigned Idx = 0; Idx < 4; ++Idx) - ++Counts[getChunk(UImm, Idx)]; - - // Traverse the chunks to find one which occurs more than once. - for (CountMap::const_iterator Chunk = Counts.begin(), End = Counts.end(); - Chunk != End; ++Chunk) { - const uint64_t ChunkVal = Chunk->first; - const unsigned Count = Chunk->second; - - uint64_t Encoding = 0; - - // We are looking for chunks which have two or three instances and can be - // materialized with an ORR instruction. - if ((Count != 2 && Count != 3) || !canUseOrr(ChunkVal, Encoding)) - continue; - - const bool CountThree = Count == 3; - // Create the ORR-immediate instruction. - MachineInstrBuilder MIB = - BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(AArch64::ORRXri)) - .add(MI.getOperand(0)) - .addReg(AArch64::XZR) - .addImm(Encoding); - - const unsigned DstReg = MI.getOperand(0).getReg(); - const bool DstIsDead = MI.getOperand(0).isDead(); - - unsigned ShiftAmt = 0; - uint64_t Imm16 = 0; - // Find the first chunk not materialized with the ORR instruction. - for (; ShiftAmt < 64; ShiftAmt += 16) { - Imm16 = (UImm >> ShiftAmt) & 0xFFFF; - - if (Imm16 != ChunkVal) - break; - } - - // Create the first MOVK instruction. - MachineInstrBuilder MIB1 = - BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(AArch64::MOVKXi)) - .addReg(DstReg, - RegState::Define | getDeadRegState(DstIsDead && CountThree)) - .addReg(DstReg) - .addImm(Imm16) - .addImm(AArch64_AM::getShifterImm(AArch64_AM::LSL, ShiftAmt)); - - // In case we have three instances the whole constant is now materialized - // and we can exit. - if (CountThree) { - transferImpOps(MI, MIB, MIB1); - MI.eraseFromParent(); - return true; - } - - // Find the remaining chunk which needs to be materialized. - for (ShiftAmt += 16; ShiftAmt < 64; ShiftAmt += 16) { - Imm16 = (UImm >> ShiftAmt) & 0xFFFF; - - if (Imm16 != ChunkVal) - break; - } - - // Create the second MOVK instruction. - MachineInstrBuilder MIB2 = - BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(AArch64::MOVKXi)) - .addReg(DstReg, RegState::Define | getDeadRegState(DstIsDead)) - .addReg(DstReg) - .addImm(Imm16) - .addImm(AArch64_AM::getShifterImm(AArch64_AM::LSL, ShiftAmt)); - - transferImpOps(MI, MIB, MIB2); - MI.eraseFromParent(); - return true; - } - - return false; -} - -/// Check whether this chunk matches the pattern '1...0...'. This pattern -/// starts a contiguous sequence of ones if we look at the bits from the LSB -/// towards the MSB. -static bool isStartChunk(uint64_t Chunk) { - if (Chunk == 0 || Chunk == std::numeric_limits::max()) - return false; - - return isMask_64(~Chunk); -} - -/// Check whether this chunk matches the pattern '0...1...' This pattern -/// ends a contiguous sequence of ones if we look at the bits from the LSB -/// towards the MSB. -static bool isEndChunk(uint64_t Chunk) { - if (Chunk == 0 || Chunk == std::numeric_limits::max()) - return false; - - return isMask_64(Chunk); -} - -/// Clear or set all bits in the chunk at the given index. -static uint64_t updateImm(uint64_t Imm, unsigned Idx, bool Clear) { - const uint64_t Mask = 0xFFFF; - - if (Clear) - // Clear chunk in the immediate. - Imm &= ~(Mask << (Idx * 16)); - else - // Set all bits in the immediate for the particular chunk. - Imm |= Mask << (Idx * 16); - - return Imm; -} - -/// Check whether the constant contains a sequence of contiguous ones, -/// which might be interrupted by one or two chunks. If so, materialize the -/// sequence of contiguous ones with an ORR instruction. -/// Materialize the chunks which are either interrupting the sequence or outside -/// of the sequence with a MOVK instruction. -/// -/// Assuming S is a chunk which starts the sequence (1...0...), E is a chunk -/// which ends the sequence (0...1...). Then we are looking for constants which -/// contain at least one S and E chunk. -/// E.g. |E|A|B|S|, |A|E|B|S| or |A|B|E|S|. -/// -/// We are also looking for constants like |S|A|B|E| where the contiguous -/// sequence of ones wraps around the MSB into the LSB. -static bool trySequenceOfOnes(uint64_t UImm, MachineInstr &MI, - MachineBasicBlock &MBB, - MachineBasicBlock::iterator &MBBI, - const AArch64InstrInfo *TII) { - const int NotSet = -1; - const uint64_t Mask = 0xFFFF; - - int StartIdx = NotSet; - int EndIdx = NotSet; - // Try to find the chunks which start/end a contiguous sequence of ones. - for (int Idx = 0; Idx < 4; ++Idx) { - int64_t Chunk = getChunk(UImm, Idx); - // Sign extend the 16-bit chunk to 64-bit. - Chunk = (Chunk << 48) >> 48; - - if (isStartChunk(Chunk)) - StartIdx = Idx; - else if (isEndChunk(Chunk)) - EndIdx = Idx; - } - - // Early exit in case we can't find a start/end chunk. - if (StartIdx == NotSet || EndIdx == NotSet) - return false; - - // Outside of the contiguous sequence of ones everything needs to be zero. - uint64_t Outside = 0; - // Chunks between the start and end chunk need to have all their bits set. - uint64_t Inside = Mask; - - // If our contiguous sequence of ones wraps around from the MSB into the LSB, - // just swap indices and pretend we are materializing a contiguous sequence - // of zeros surrounded by a contiguous sequence of ones. - if (StartIdx > EndIdx) { - std::swap(StartIdx, EndIdx); - std::swap(Outside, Inside); - } - - uint64_t OrrImm = UImm; - int FirstMovkIdx = NotSet; - int SecondMovkIdx = NotSet; - - // Find out which chunks we need to patch up to obtain a contiguous sequence - // of ones. - for (int Idx = 0; Idx < 4; ++Idx) { - const uint64_t Chunk = getChunk(UImm, Idx); - - // Check whether we are looking at a chunk which is not part of the - // contiguous sequence of ones. - if ((Idx < StartIdx || EndIdx < Idx) && Chunk != Outside) { - OrrImm = updateImm(OrrImm, Idx, Outside == 0); - - // Remember the index we need to patch. - if (FirstMovkIdx == NotSet) - FirstMovkIdx = Idx; - else - SecondMovkIdx = Idx; - - // Check whether we are looking a chunk which is part of the contiguous - // sequence of ones. - } else if (Idx > StartIdx && Idx < EndIdx && Chunk != Inside) { - OrrImm = updateImm(OrrImm, Idx, Inside != Mask); - - // Remember the index we need to patch. - if (FirstMovkIdx == NotSet) - FirstMovkIdx = Idx; - else - SecondMovkIdx = Idx; - } - } - assert(FirstMovkIdx != NotSet && "Constant materializable with single ORR!"); - - // Create the ORR-immediate instruction. - uint64_t Encoding = 0; - AArch64_AM::processLogicalImmediate(OrrImm, 64, Encoding); - MachineInstrBuilder MIB = - BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(AArch64::ORRXri)) - .add(MI.getOperand(0)) - .addReg(AArch64::XZR) - .addImm(Encoding); - - const unsigned DstReg = MI.getOperand(0).getReg(); - const bool DstIsDead = MI.getOperand(0).isDead(); - - const bool SingleMovk = SecondMovkIdx == NotSet; - // Create the first MOVK instruction. - MachineInstrBuilder MIB1 = - BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(AArch64::MOVKXi)) - .addReg(DstReg, - RegState::Define | getDeadRegState(DstIsDead && SingleMovk)) - .addReg(DstReg) - .addImm(getChunk(UImm, FirstMovkIdx)) - .addImm( - AArch64_AM::getShifterImm(AArch64_AM::LSL, FirstMovkIdx * 16)); - - // Early exit in case we only need to emit a single MOVK instruction. - if (SingleMovk) { - transferImpOps(MI, MIB, MIB1); - MI.eraseFromParent(); - return true; - } - - // Create the second MOVK instruction. - MachineInstrBuilder MIB2 = - BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(AArch64::MOVKXi)) - .addReg(DstReg, RegState::Define | getDeadRegState(DstIsDead)) - .addReg(DstReg) - .addImm(getChunk(UImm, SecondMovkIdx)) - .addImm( - AArch64_AM::getShifterImm(AArch64_AM::LSL, SecondMovkIdx * 16)); - - transferImpOps(MI, MIB, MIB2); - MI.eraseFromParent(); - return true; -} - /// Expand a MOVi32imm or MOVi64imm pseudo instruction to one or more /// real move-immediate instructions to synthesize the immediate. bool AArch64ExpandPseudo::expandMOVImm(MachineBasicBlock &MBB, @@ -384,7 +107,6 @@ bool AArch64ExpandPseudo::expandMOVImm(MachineBasicBlock &MBB, MachineInstr &MI = *MBBI; unsigned DstReg = MI.getOperand(0).getReg(); uint64_t Imm = MI.getOperand(1).getImm(); - const unsigned Mask = 0xFFFF; if (DstReg == AArch64::XZR || DstReg == AArch64::WZR) { // Useless def, and we don't want to risk creating an invalid ORR (which @@ -393,194 +115,50 @@ bool AArch64ExpandPseudo::expandMOVImm(MachineBasicBlock &MBB, return true; } - // Scan the immediate and count the number of 16-bit chunks which are either - // all ones or all zeros. - unsigned OneChunks = 0; - unsigned ZeroChunks = 0; - for (unsigned Shift = 0; Shift < BitSize; Shift += 16) { - const unsigned Chunk = (Imm >> Shift) & Mask; - if (Chunk == Mask) - OneChunks++; - else if (Chunk == 0) - ZeroChunks++; - } + SmallVector Insn; + AArch64_IMM::expandMOVImm(Imm, BitSize, Insn); + assert(Insn.size() != 0); - // FIXME: Prefer MOVZ/MOVN over ORR because of the rules for the "mov" - // alias. + SmallVector MIBS; + for (auto I = Insn.begin(), E = Insn.end(); I != E; ++I) { + bool LastItem = std::next(I) == E; + switch (I->Opcode) + { + default: llvm_unreachable("unhandled!"); break; - // Try a single ORR. - uint64_t UImm = Imm << (64 - BitSize) >> (64 - BitSize); - uint64_t Encoding; - if (AArch64_AM::processLogicalImmediate(UImm, BitSize, Encoding)) { - unsigned Opc = (BitSize == 32 ? AArch64::ORRWri : AArch64::ORRXri); - MachineInstrBuilder MIB = - BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(Opc)) - .add(MI.getOperand(0)) - .addReg(BitSize == 32 ? AArch64::WZR : AArch64::XZR) - .addImm(Encoding); - transferImpOps(MI, MIB, MIB); - MI.eraseFromParent(); - return true; - } - - // Two instruction sequences. - // - // Prefer MOVZ/MOVN followed by MOVK; it's more readable, and possibly the - // fastest sequence with fast literal generation. - if (OneChunks >= (BitSize / 16) - 2 || ZeroChunks >= (BitSize / 16) - 2) - return expandMOVImmSimple(MBB, MBBI, BitSize, OneChunks, ZeroChunks); - - assert(BitSize == 64 && "All 32-bit immediates can be expanded with a" - "MOVZ/MOVK pair"); - - // Try other two-instruction sequences. - - // 64-bit ORR followed by MOVK. - // We try to construct the ORR immediate in three different ways: either we - // zero out the chunk which will be replaced, we fill the chunk which will - // be replaced with ones, or we take the bit pattern from the other half of - // the 64-bit immediate. This is comprehensive because of the way ORR - // immediates are constructed. - for (unsigned Shift = 0; Shift < BitSize; Shift += 16) { - uint64_t ShiftedMask = (0xFFFFULL << Shift); - uint64_t ZeroChunk = UImm & ~ShiftedMask; - uint64_t OneChunk = UImm | ShiftedMask; - uint64_t RotatedImm = (UImm << 32) | (UImm >> 32); - uint64_t ReplicateChunk = ZeroChunk | (RotatedImm & ShiftedMask); - if (AArch64_AM::processLogicalImmediate(ZeroChunk, BitSize, Encoding) || - AArch64_AM::processLogicalImmediate(OneChunk, BitSize, Encoding) || - AArch64_AM::processLogicalImmediate(ReplicateChunk, - BitSize, Encoding)) { - // Create the ORR-immediate instruction. - MachineInstrBuilder MIB = - BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(AArch64::ORRXri)) - .add(MI.getOperand(0)) - .addReg(AArch64::XZR) - .addImm(Encoding); - - // Create the MOVK instruction. - const unsigned Imm16 = getChunk(UImm, Shift / 16); - const unsigned DstReg = MI.getOperand(0).getReg(); - const bool DstIsDead = MI.getOperand(0).isDead(); - MachineInstrBuilder MIB1 = - BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(AArch64::MOVKXi)) - .addReg(DstReg, RegState::Define | getDeadRegState(DstIsDead)) - .addReg(DstReg) - .addImm(Imm16) - .addImm(AArch64_AM::getShifterImm(AArch64_AM::LSL, Shift)); - - transferImpOps(MI, MIB, MIB1); - MI.eraseFromParent(); - return true; + case AArch64::ORRWri: + case AArch64::ORRXri: + MIBS.push_back(BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(I->Opcode)) + .add(MI.getOperand(0)) + .addReg(BitSize == 32 ? AArch64::WZR : AArch64::XZR) + .addImm(I->Op2)); + break; + case AArch64::MOVNWi: + case AArch64::MOVNXi: + case AArch64::MOVZWi: + case AArch64::MOVZXi: { + bool DstIsDead = MI.getOperand(0).isDead(); + MIBS.push_back(BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(I->Opcode)) + .addReg(DstReg, RegState::Define | + getDeadRegState(DstIsDead && LastItem)) + .addImm(I->Op1) + .addImm(I->Op2)); + } break; + case AArch64::MOVKWi: + case AArch64::MOVKXi: { + unsigned DstReg = MI.getOperand(0).getReg(); + bool DstIsDead = MI.getOperand(0).isDead(); + MIBS.push_back(BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(I->Opcode)) + .addReg(DstReg, + RegState::Define | + getDeadRegState(DstIsDead && LastItem)) + .addReg(DstReg) + .addImm(I->Op1) + .addImm(I->Op2)); + } break; } } - - // FIXME: Add more two-instruction sequences. - - // Three instruction sequences. - // - // Prefer MOVZ/MOVN followed by two MOVK; it's more readable, and possibly - // the fastest sequence with fast literal generation. (If neither MOVK is - // part of a fast literal generation pair, it could be slower than the - // four-instruction sequence, but we won't worry about that for now.) - if (OneChunks || ZeroChunks) - return expandMOVImmSimple(MBB, MBBI, BitSize, OneChunks, ZeroChunks); - - // Check for identical 16-bit chunks within the constant and if so materialize - // them with a single ORR instruction. The remaining one or two 16-bit chunks - // will be materialized with MOVK instructions. - if (BitSize == 64 && tryToreplicateChunks(UImm, MI, MBB, MBBI, TII)) - return true; - - // Check whether the constant contains a sequence of contiguous ones, which - // might be interrupted by one or two chunks. If so, materialize the sequence - // of contiguous ones with an ORR instruction. Materialize the chunks which - // are either interrupting the sequence or outside of the sequence with a - // MOVK instruction. - if (BitSize == 64 && trySequenceOfOnes(UImm, MI, MBB, MBBI, TII)) - return true; - - // We found no possible two or three instruction sequence; use the general - // four-instruction sequence. - return expandMOVImmSimple(MBB, MBBI, BitSize, OneChunks, ZeroChunks); -} - -/// \brief Expand a MOVi32imm or MOVi64imm pseudo instruction to a -/// MOVZ or MOVN of width BitSize followed by up to 3 MOVK instructions. -bool AArch64ExpandPseudo::expandMOVImmSimple(MachineBasicBlock &MBB, - MachineBasicBlock::iterator MBBI, - unsigned BitSize, - unsigned OneChunks, - unsigned ZeroChunks) { - MachineInstr &MI = *MBBI; - unsigned DstReg = MI.getOperand(0).getReg(); - uint64_t Imm = MI.getOperand(1).getImm(); - const unsigned Mask = 0xFFFF; - - // Use a MOVZ or MOVN instruction to set the high bits, followed by one or - // more MOVK instructions to insert additional 16-bit portions into the - // lower bits. - bool isNeg = false; - - // Use MOVN to materialize the high bits if we have more all one chunks - // than all zero chunks. - if (OneChunks > ZeroChunks) { - isNeg = true; - Imm = ~Imm; - } - - unsigned FirstOpc; - if (BitSize == 32) { - Imm &= (1LL << 32) - 1; - FirstOpc = (isNeg ? AArch64::MOVNWi : AArch64::MOVZWi); - } else { - FirstOpc = (isNeg ? AArch64::MOVNXi : AArch64::MOVZXi); - } - unsigned Shift = 0; // LSL amount for high bits with MOVZ/MOVN - unsigned LastShift = 0; // LSL amount for last MOVK - if (Imm != 0) { - unsigned LZ = countLeadingZeros(Imm); - unsigned TZ = countTrailingZeros(Imm); - Shift = (TZ / 16) * 16; - LastShift = ((63 - LZ) / 16) * 16; - } - unsigned Imm16 = (Imm >> Shift) & Mask; - bool DstIsDead = MI.getOperand(0).isDead(); - MachineInstrBuilder MIB1 = - BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(FirstOpc)) - .addReg(DstReg, RegState::Define | - getDeadRegState(DstIsDead && Shift == LastShift)) - .addImm(Imm16) - .addImm(AArch64_AM::getShifterImm(AArch64_AM::LSL, Shift)); - - // If a MOVN was used for the high bits of a negative value, flip the rest - // of the bits back for use with MOVK. - if (isNeg) - Imm = ~Imm; - - if (Shift == LastShift) { - transferImpOps(MI, MIB1, MIB1); - MI.eraseFromParent(); - return true; - } - - MachineInstrBuilder MIB2; - unsigned Opc = (BitSize == 32 ? AArch64::MOVKWi : AArch64::MOVKXi); - while (Shift < LastShift) { - Shift += 16; - Imm16 = (Imm >> Shift) & Mask; - if (Imm16 == (isNeg ? Mask : 0)) - continue; // This 16-bit portion is already set correctly. - MIB2 = BuildMI(MBB, MBBI, MI.getDebugLoc(), TII->get(Opc)) - .addReg(DstReg, - RegState::Define | - getDeadRegState(DstIsDead && Shift == LastShift)) - .addReg(DstReg) - .addImm(Imm16) - .addImm(AArch64_AM::getShifterImm(AArch64_AM::LSL, Shift)); - } - - transferImpOps(MI, MIB1, MIB2); + transferImpOps(MI, MIBS.front(), MIBS.back()); MI.eraseFromParent(); return true; } diff --git a/llvm/lib/Target/AArch64/CMakeLists.txt b/llvm/lib/Target/AArch64/CMakeLists.txt index e219f5f3b67d..743196877a1d 100644 --- a/llvm/lib/Target/AArch64/CMakeLists.txt +++ b/llvm/lib/Target/AArch64/CMakeLists.txt @@ -31,6 +31,7 @@ add_llvm_target(AArch64CodeGen AArch64CondBrTuning.cpp AArch64ConditionalCompares.cpp AArch64DeadRegisterDefinitionsPass.cpp + AArch64ExpandImm.cpp AArch64ExpandPseudoInsts.cpp AArch64FalkorHWPFFix.cpp AArch64FastISel.cpp