From 583b68f34fd9bdb6114fa48dadc8def2ec96edac Mon Sep 17 00:00:00 2001 From: David Greene Date: Thu, 17 Feb 2011 19:18:59 +0000 Subject: [AVX] Recorganize X86ShuffleDecode into its own library (LLVMX86Utils.a) to break cyclic library dependencies between LLVMX86CodeGen.a and LLVMX86AsmParser.a. Previously this code was in a header file and marked static but AVX requires some additional functionality here that won't be used by all clients. Since including unused static functions causes a gcc compiler warning, keeping it as a header would break builds that use -Werror. Putting this in its own library solves both problems at once. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125765 91177308-0d34-0410-b5e6-96231b3b80d8 --- lib/Target/X86/InstPrinter/X86InstComments.cpp | 2 +- lib/Target/X86/Makefile | 2 +- lib/Target/X86/Utils/CMakeLists.txt | 6 + lib/Target/X86/Utils/Makefile | 15 +++ lib/Target/X86/Utils/X86ShuffleDecode.cpp | 148 +++++++++++++++++++++++ lib/Target/X86/Utils/X86ShuffleDecode.h | 69 +++++++++++ lib/Target/X86/X86ISelLowering.cpp | 2 +- lib/Target/X86/X86ShuffleDecode.h | 155 ------------------------- 8 files changed, 241 insertions(+), 158 deletions(-) create mode 100644 lib/Target/X86/Utils/CMakeLists.txt create mode 100644 lib/Target/X86/Utils/Makefile create mode 100644 lib/Target/X86/Utils/X86ShuffleDecode.cpp create mode 100644 lib/Target/X86/Utils/X86ShuffleDecode.h delete mode 100644 lib/Target/X86/X86ShuffleDecode.h diff --git a/lib/Target/X86/InstPrinter/X86InstComments.cpp b/lib/Target/X86/InstPrinter/X86InstComments.cpp index da9d5a3579..12144e3f50 100644 --- a/lib/Target/X86/InstPrinter/X86InstComments.cpp +++ b/lib/Target/X86/InstPrinter/X86InstComments.cpp @@ -16,7 +16,7 @@ #include "X86GenInstrNames.inc" #include "llvm/MC/MCInst.h" #include "llvm/Support/raw_ostream.h" -#include "../X86ShuffleDecode.h" +#include "../Utils/X86ShuffleDecode.h" using namespace llvm; //===----------------------------------------------------------------------===// diff --git a/lib/Target/X86/Makefile b/lib/Target/X86/Makefile index 9c6415d6e8..12fb090d4d 100644 --- a/lib/Target/X86/Makefile +++ b/lib/Target/X86/Makefile @@ -20,6 +20,6 @@ BUILT_SOURCES = X86GenRegisterInfo.h.inc X86GenRegisterNames.inc \ X86GenCallingConv.inc X86GenSubtarget.inc \ X86GenEDInfo.inc -DIRS = InstPrinter AsmParser Disassembler TargetInfo +DIRS = InstPrinter AsmParser Disassembler TargetInfo Utils include $(LEVEL)/Makefile.common diff --git a/lib/Target/X86/Utils/CMakeLists.txt b/lib/Target/X86/Utils/CMakeLists.txt new file mode 100644 index 0000000000..3ad5f991c8 --- /dev/null +++ b/lib/Target/X86/Utils/CMakeLists.txt @@ -0,0 +1,6 @@ +include_directories( ${CMAKE_CURRENT_BINARY_DIR}/.. ${CMAKE_CURRENT_SOURCE_DIR}/.. ) + +add_llvm_library(LLVMX86Utils + X86ShuffleDecode.cpp + ) +add_dependencies(LLVMX86Utils X86CodeGenTable_gen) diff --git a/lib/Target/X86/Utils/Makefile b/lib/Target/X86/Utils/Makefile new file mode 100644 index 0000000000..1df6f0f561 --- /dev/null +++ b/lib/Target/X86/Utils/Makefile @@ -0,0 +1,15 @@ +##===- lib/Target/X86/Utils/Makefile -----------------------*- Makefile -*-===## +# +# The LLVM Compiler Infrastructure +# +# This file is distributed under the University of Illinois Open Source +# License. See LICENSE.TXT for details. +# +##===----------------------------------------------------------------------===## +LEVEL = ../../../.. +LIBRARYNAME = LLVMX86Utils + +# Hack: we need to include 'main' x86 target directory to grab private headers +CPP.Flags += -I$(PROJ_OBJ_DIR)/.. -I$(PROJ_SRC_DIR)/.. + +include $(LEVEL)/Makefile.common diff --git a/lib/Target/X86/Utils/X86ShuffleDecode.cpp b/lib/Target/X86/Utils/X86ShuffleDecode.cpp new file mode 100644 index 0000000000..12879774d7 --- /dev/null +++ b/lib/Target/X86/Utils/X86ShuffleDecode.cpp @@ -0,0 +1,148 @@ +//===-- X86ShuffleDecode.h - X86 shuffle decode logic ---------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// +// Define several functions to decode x86 specific shuffle semantics into a +// generic vector mask. +// +//===----------------------------------------------------------------------===// + +#include "X86ShuffleDecode.h" + +//===----------------------------------------------------------------------===// +// Vector Mask Decoding +//===----------------------------------------------------------------------===// + +namespace llvm { + +void DecodeINSERTPSMask(unsigned Imm, SmallVectorImpl &ShuffleMask) { + // Defaults the copying the dest value. + ShuffleMask.push_back(0); + ShuffleMask.push_back(1); + ShuffleMask.push_back(2); + ShuffleMask.push_back(3); + + // Decode the immediate. + unsigned ZMask = Imm & 15; + unsigned CountD = (Imm >> 4) & 3; + unsigned CountS = (Imm >> 6) & 3; + + // CountS selects which input element to use. + unsigned InVal = 4+CountS; + // CountD specifies which element of destination to update. + ShuffleMask[CountD] = InVal; + // ZMask zaps values, potentially overriding the CountD elt. + if (ZMask & 1) ShuffleMask[0] = SM_SentinelZero; + if (ZMask & 2) ShuffleMask[1] = SM_SentinelZero; + if (ZMask & 4) ShuffleMask[2] = SM_SentinelZero; + if (ZMask & 8) ShuffleMask[3] = SM_SentinelZero; +} + +// <3,1> or <6,7,2,3> +void DecodeMOVHLPSMask(unsigned NElts, + SmallVectorImpl &ShuffleMask) { + for (unsigned i = NElts/2; i != NElts; ++i) + ShuffleMask.push_back(NElts+i); + + for (unsigned i = NElts/2; i != NElts; ++i) + ShuffleMask.push_back(i); +} + +// <0,2> or <0,1,4,5> +void DecodeMOVLHPSMask(unsigned NElts, + SmallVectorImpl &ShuffleMask) { + for (unsigned i = 0; i != NElts/2; ++i) + ShuffleMask.push_back(i); + + for (unsigned i = 0; i != NElts/2; ++i) + ShuffleMask.push_back(NElts+i); +} + +void DecodePSHUFMask(unsigned NElts, unsigned Imm, + SmallVectorImpl &ShuffleMask) { + for (unsigned i = 0; i != NElts; ++i) { + ShuffleMask.push_back(Imm % NElts); + Imm /= NElts; + } +} + +void DecodePSHUFHWMask(unsigned Imm, + SmallVectorImpl &ShuffleMask) { + ShuffleMask.push_back(0); + ShuffleMask.push_back(1); + ShuffleMask.push_back(2); + ShuffleMask.push_back(3); + for (unsigned i = 0; i != 4; ++i) { + ShuffleMask.push_back(4+(Imm & 3)); + Imm >>= 2; + } +} + +void DecodePSHUFLWMask(unsigned Imm, + SmallVectorImpl &ShuffleMask) { + for (unsigned i = 0; i != 4; ++i) { + ShuffleMask.push_back((Imm & 3)); + Imm >>= 2; + } + ShuffleMask.push_back(4); + ShuffleMask.push_back(5); + ShuffleMask.push_back(6); + ShuffleMask.push_back(7); +} + +void DecodePUNPCKLMask(unsigned NElts, + SmallVectorImpl &ShuffleMask) { + for (unsigned i = 0; i != NElts/2; ++i) { + ShuffleMask.push_back(i); + ShuffleMask.push_back(i+NElts); + } +} + +void DecodePUNPCKHMask(unsigned NElts, + SmallVectorImpl &ShuffleMask) { + for (unsigned i = 0; i != NElts/2; ++i) { + ShuffleMask.push_back(i+NElts/2); + ShuffleMask.push_back(i+NElts+NElts/2); + } +} + +void DecodeSHUFPSMask(unsigned NElts, unsigned Imm, + SmallVectorImpl &ShuffleMask) { + // Part that reads from dest. + for (unsigned i = 0; i != NElts/2; ++i) { + ShuffleMask.push_back(Imm % NElts); + Imm /= NElts; + } + // Part that reads from src. + for (unsigned i = 0; i != NElts/2; ++i) { + ShuffleMask.push_back(Imm % NElts + NElts); + Imm /= NElts; + } +} + +void DecodeUNPCKHPMask(unsigned NElts, + SmallVectorImpl &ShuffleMask) { + for (unsigned i = 0; i != NElts/2; ++i) { + ShuffleMask.push_back(i+NElts/2); // Reads from dest + ShuffleMask.push_back(i+NElts+NElts/2); // Reads from src + } +} + + +/// DecodeUNPCKLPMask - This decodes the shuffle masks for unpcklps/unpcklpd +/// etc. NElts indicates the number of elements in the vector allowing it to +/// handle different datatypes and vector widths. +void DecodeUNPCKLPMask(unsigned NElts, + SmallVectorImpl &ShuffleMask) { + for (unsigned i = 0; i != NElts/2; ++i) { + ShuffleMask.push_back(i); // Reads from dest + ShuffleMask.push_back(i+NElts); // Reads from src + } +} + +} // llvm namespace diff --git a/lib/Target/X86/Utils/X86ShuffleDecode.h b/lib/Target/X86/Utils/X86ShuffleDecode.h new file mode 100644 index 0000000000..50d9ccbfa6 --- /dev/null +++ b/lib/Target/X86/Utils/X86ShuffleDecode.h @@ -0,0 +1,69 @@ +//===-- X86ShuffleDecode.h - X86 shuffle decode logic -----------*-C++-*---===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// +// +// Define several functions to decode x86 specific shuffle semantics into a +// generic vector mask. +// +//===----------------------------------------------------------------------===// + +#ifndef X86_SHUFFLE_DECODE_H +#define X86_SHUFFLE_DECODE_H + +#include "llvm/ADT/SmallVector.h" + +//===----------------------------------------------------------------------===// +// Vector Mask Decoding +//===----------------------------------------------------------------------===// + +namespace llvm { +enum { + SM_SentinelZero = ~0U +}; + +void DecodeINSERTPSMask(unsigned Imm, SmallVectorImpl &ShuffleMask); + +// <3,1> or <6,7,2,3> +void DecodeMOVHLPSMask(unsigned NElts, + SmallVectorImpl &ShuffleMask); + +// <0,2> or <0,1,4,5> +void DecodeMOVLHPSMask(unsigned NElts, + SmallVectorImpl &ShuffleMask); + +void DecodePSHUFMask(unsigned NElts, unsigned Imm, + SmallVectorImpl &ShuffleMask); + +void DecodePSHUFHWMask(unsigned Imm, + SmallVectorImpl &ShuffleMask); + +void DecodePSHUFLWMask(unsigned Imm, + SmallVectorImpl &ShuffleMask); + +void DecodePUNPCKLMask(unsigned NElts, + SmallVectorImpl &ShuffleMask); + +void DecodePUNPCKHMask(unsigned NElts, + SmallVectorImpl &ShuffleMask); + +void DecodeSHUFPSMask(unsigned NElts, unsigned Imm, + SmallVectorImpl &ShuffleMask); + +void DecodeUNPCKHPMask(unsigned NElts, + SmallVectorImpl &ShuffleMask); + + +/// DecodeUNPCKLPMask - This decodes the shuffle masks for unpcklps/unpcklpd +/// etc. NElts indicates the number of elements in the vector allowing it to +/// handle different datatypes and vector widths. +void DecodeUNPCKLPMask(unsigned NElts, + SmallVectorImpl &ShuffleMask); + +} // llvm namespace + +#endif diff --git a/lib/Target/X86/X86ISelLowering.cpp b/lib/Target/X86/X86ISelLowering.cpp index b7ed745485..e9aa5eda83 100644 --- a/lib/Target/X86/X86ISelLowering.cpp +++ b/lib/Target/X86/X86ISelLowering.cpp @@ -16,9 +16,9 @@ #include "X86.h" #include "X86InstrBuilder.h" #include "X86ISelLowering.h" -#include "X86ShuffleDecode.h" #include "X86TargetMachine.h" #include "X86TargetObjectFile.h" +#include "Utils/X86ShuffleDecode.h" #include "llvm/CallingConv.h" #include "llvm/Constants.h" #include "llvm/DerivedTypes.h" diff --git a/lib/Target/X86/X86ShuffleDecode.h b/lib/Target/X86/X86ShuffleDecode.h deleted file mode 100644 index df040520bc..0000000000 --- a/lib/Target/X86/X86ShuffleDecode.h +++ /dev/null @@ -1,155 +0,0 @@ -//===-- X86ShuffleDecode.h - X86 shuffle decode logic ---------------------===// -// -// The LLVM Compiler Infrastructure -// -// This file is distributed under the University of Illinois Open Source -// License. See LICENSE.TXT for details. -// -//===----------------------------------------------------------------------===// -// -// Define several functions to decode x86 specific shuffle semantics into a -// generic vector mask. -// -//===----------------------------------------------------------------------===// - -#ifndef X86_SHUFFLE_DECODE_H -#define X86_SHUFFLE_DECODE_H - -#include "llvm/ADT/SmallVector.h" -using namespace llvm; - -//===----------------------------------------------------------------------===// -// Vector Mask Decoding -//===----------------------------------------------------------------------===// - -enum { - SM_SentinelZero = ~0U -}; - -static inline -void DecodeINSERTPSMask(unsigned Imm, SmallVectorImpl &ShuffleMask) { - // Defaults the copying the dest value. - ShuffleMask.push_back(0); - ShuffleMask.push_back(1); - ShuffleMask.push_back(2); - ShuffleMask.push_back(3); - - // Decode the immediate. - unsigned ZMask = Imm & 15; - unsigned CountD = (Imm >> 4) & 3; - unsigned CountS = (Imm >> 6) & 3; - - // CountS selects which input element to use. - unsigned InVal = 4+CountS; - // CountD specifies which element of destination to update. - ShuffleMask[CountD] = InVal; - // ZMask zaps values, potentially overriding the CountD elt. - if (ZMask & 1) ShuffleMask[0] = SM_SentinelZero; - if (ZMask & 2) ShuffleMask[1] = SM_SentinelZero; - if (ZMask & 4) ShuffleMask[2] = SM_SentinelZero; - if (ZMask & 8) ShuffleMask[3] = SM_SentinelZero; -} - -// <3,1> or <6,7,2,3> -static void DecodeMOVHLPSMask(unsigned NElts, - SmallVectorImpl &ShuffleMask) { - for (unsigned i = NElts/2; i != NElts; ++i) - ShuffleMask.push_back(NElts+i); - - for (unsigned i = NElts/2; i != NElts; ++i) - ShuffleMask.push_back(i); -} - -// <0,2> or <0,1,4,5> -static void DecodeMOVLHPSMask(unsigned NElts, - SmallVectorImpl &ShuffleMask) { - for (unsigned i = 0; i != NElts/2; ++i) - ShuffleMask.push_back(i); - - for (unsigned i = 0; i != NElts/2; ++i) - ShuffleMask.push_back(NElts+i); -} - -static void DecodePSHUFMask(unsigned NElts, unsigned Imm, - SmallVectorImpl &ShuffleMask) { - for (unsigned i = 0; i != NElts; ++i) { - ShuffleMask.push_back(Imm % NElts); - Imm /= NElts; - } -} - -static void DecodePSHUFHWMask(unsigned Imm, - SmallVectorImpl &ShuffleMask) { - ShuffleMask.push_back(0); - ShuffleMask.push_back(1); - ShuffleMask.push_back(2); - ShuffleMask.push_back(3); - for (unsigned i = 0; i != 4; ++i) { - ShuffleMask.push_back(4+(Imm & 3)); - Imm >>= 2; - } -} - -static void DecodePSHUFLWMask(unsigned Imm, - SmallVectorImpl &ShuffleMask) { - for (unsigned i = 0; i != 4; ++i) { - ShuffleMask.push_back((Imm & 3)); - Imm >>= 2; - } - ShuffleMask.push_back(4); - ShuffleMask.push_back(5); - ShuffleMask.push_back(6); - ShuffleMask.push_back(7); -} - -static void DecodePUNPCKLMask(unsigned NElts, - SmallVectorImpl &ShuffleMask) { - for (unsigned i = 0; i != NElts/2; ++i) { - ShuffleMask.push_back(i); - ShuffleMask.push_back(i+NElts); - } -} - -static void DecodePUNPCKHMask(unsigned NElts, - SmallVectorImpl &ShuffleMask) { - for (unsigned i = 0; i != NElts/2; ++i) { - ShuffleMask.push_back(i+NElts/2); - ShuffleMask.push_back(i+NElts+NElts/2); - } -} - -static void DecodeSHUFPSMask(unsigned NElts, unsigned Imm, - SmallVectorImpl &ShuffleMask) { - // Part that reads from dest. - for (unsigned i = 0; i != NElts/2; ++i) { - ShuffleMask.push_back(Imm % NElts); - Imm /= NElts; - } - // Part that reads from src. - for (unsigned i = 0; i != NElts/2; ++i) { - ShuffleMask.push_back(Imm % NElts + NElts); - Imm /= NElts; - } -} - -static void DecodeUNPCKHPMask(unsigned NElts, - SmallVectorImpl &ShuffleMask) { - for (unsigned i = 0; i != NElts/2; ++i) { - ShuffleMask.push_back(i+NElts/2); // Reads from dest - ShuffleMask.push_back(i+NElts+NElts/2); // Reads from src - } -} - - -/// DecodeUNPCKLPMask - This decodes the shuffle masks for unpcklps/unpcklpd -/// etc. NElts indicates the number of elements in the vector allowing it to -/// handle different datatypes and vector widths. -static void DecodeUNPCKLPMask(unsigned NElts, - SmallVectorImpl &ShuffleMask) { - for (unsigned i = 0; i != NElts/2; ++i) { - ShuffleMask.push_back(i); // Reads from dest - ShuffleMask.push_back(i+NElts); // Reads from src - } -} - -#endif -- cgit v1.2.3