llvm-6502/lib/Target/ARM/ARMSubtarget.cpp
Evan Cheng 48575f6ea7 Making use of VFP / NEON floating point multiply-accumulate / subtraction is
difficult on current ARM implementations for a few reasons.
1. Even though a single vmla has latency that is one cycle shorter than a pair
   of vmul + vadd, a RAW hazard during the first (4? on Cortex-a8) can cause
   additional pipeline stall. So it's frequently better to single codegen
   vmul + vadd.
2. A vmla folowed by a vmul, vmadd, or vsub causes the second fp instruction to
   stall for 4 cycles. We need to schedule them apart.
3. A vmla followed vmla is a special case. Obvious issuing back to back RAW
   vmla + vmla is very bad. But this isn't ideal either:
     vmul
     vadd
     vmla
   Instead, we want to expand the second vmla:
     vmla
     vmul
     vadd
   Even with the 4 cycle vmul stall, the second sequence is still 2 cycles
   faster.

Up to now, isel simply avoid codegen'ing fp vmla / vmls. This works well enough
but it isn't the optimial solution. This patch attempts to make it possible to
use vmla / vmls in cases where it is profitable.

A. Add missing isel predicates which cause vmla to be codegen'ed.
B. Make sure the fmul in (fadd (fmul)) has a single use. We don't want to
   compute a fmul and a fmla.
C. Add additional isel checks for vmla, avoid cases where vmla is feeding into
   fp instructions (except for the #3 exceptional case).
D. Add ARM hazard recognizer to model the vmla / vmls hazards.
E. Add a special pre-regalloc case to expand vmla / vmls when it's likely the
   vmla / vmls will trigger one of the special hazards.

Work in progress, only A+B are enabled.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@120960 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-05 22:04:16 +00:00

236 lines
7.1 KiB
C++

//===-- ARMSubtarget.cpp - ARM Subtarget Information ------------*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file implements the ARM specific subclass of TargetSubtarget.
//
//===----------------------------------------------------------------------===//
#include "ARMSubtarget.h"
#include "ARMGenSubtarget.inc"
#include "llvm/GlobalValue.h"
#include "llvm/Target/TargetOptions.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/ADT/SmallVector.h"
using namespace llvm;
static cl::opt<bool>
ReserveR9("arm-reserve-r9", cl::Hidden,
cl::desc("Reserve R9, making it unavailable as GPR"));
static cl::opt<bool>
UseMOVT("arm-use-movt",
cl::init(true), cl::Hidden);
static cl::opt<bool>
StrictAlign("arm-strict-align", cl::Hidden,
cl::desc("Disallow all unaligned memory accesses"));
ARMSubtarget::ARMSubtarget(const std::string &TT, const std::string &FS,
bool isT)
: ARMArchVersion(V4)
, ARMProcFamily(Others)
, ARMFPUType(None)
, UseNEONForSinglePrecisionFP(false)
, SlowFPVMLx(false)
, SlowFPBrcc(false)
, IsThumb(isT)
, ThumbMode(Thumb1)
, NoARM(false)
, PostRAScheduler(false)
, IsR9Reserved(ReserveR9)
, UseMovt(UseMOVT)
, HasFP16(false)
, HasD16(false)
, HasHardwareDivide(false)
, HasT2ExtractPack(false)
, HasDataBarrier(false)
, Pref32BitThumb(false)
, HasMPExtension(false)
, FPOnlySP(false)
, AllowsUnalignedMem(false)
, stackAlignment(4)
, CPUString("generic")
, TargetType(isELF) // Default to ELF unless otherwise specified.
, TargetABI(ARM_ABI_APCS) {
// Default to soft float ABI
if (FloatABIType == FloatABI::Default)
FloatABIType = FloatABI::Soft;
// Determine default and user specified characteristics
// When no arch is specified either by CPU or by attributes, make the default
// ARMv4T.
const char *ARMArchFeature = "";
if (CPUString == "generic" && (FS.empty() || FS == "generic")) {
ARMArchVersion = V4T;
ARMArchFeature = ",+v4t";
}
// Set the boolean corresponding to the current target triple, or the default
// if one cannot be determined, to true.
unsigned Len = TT.length();
unsigned Idx = 0;
if (Len >= 5 && TT.substr(0, 4) == "armv")
Idx = 4;
else if (Len >= 6 && TT.substr(0, 5) == "thumb") {
IsThumb = true;
if (Len >= 7 && TT[5] == 'v')
Idx = 6;
}
if (Idx) {
unsigned SubVer = TT[Idx];
if (SubVer >= '7' && SubVer <= '9') {
ARMArchVersion = V7A;
ARMArchFeature = ",+v7a";
if (Len >= Idx+2 && TT[Idx+1] == 'm') {
ARMArchVersion = V7M;
ARMArchFeature = ",+v7m";
}
} else if (SubVer == '6') {
ARMArchVersion = V6;
ARMArchFeature = ",+v6";
if (Len >= Idx+3 && TT[Idx+1] == 't' && TT[Idx+2] == '2') {
ARMArchVersion = V6T2;
ARMArchFeature = ",+v6t2";
}
} else if (SubVer == '5') {
ARMArchVersion = V5T;
ARMArchFeature = ",+v5t";
if (Len >= Idx+3 && TT[Idx+1] == 't' && TT[Idx+2] == 'e') {
ARMArchVersion = V5TE;
ARMArchFeature = ",+v5te";
}
} else if (SubVer == '4') {
if (Len >= Idx+2 && TT[Idx+1] == 't') {
ARMArchVersion = V4T;
ARMArchFeature = ",+v4t";
} else {
ARMArchVersion = V4;
ARMArchFeature = "";
}
}
}
if (Len >= 10) {
if (TT.find("-darwin") != std::string::npos)
// arm-darwin
TargetType = isDarwin;
}
if (TT.find("eabi") != std::string::npos)
TargetABI = ARM_ABI_AAPCS;
// Parse features string. If the first entry in FS (the CPU) is missing,
// insert the architecture feature derived from the target triple. This is
// important for setting features that are implied based on the architecture
// version.
std::string FSWithArch;
if (FS.empty())
FSWithArch = std::string(ARMArchFeature);
else if (FS.find(',') == 0)
FSWithArch = std::string(ARMArchFeature) + FS;
else
FSWithArch = FS;
CPUString = ParseSubtargetFeatures(FSWithArch, CPUString);
// Thumb2 implies at least V6T2.
if (ARMArchVersion >= V6T2)
ThumbMode = Thumb2;
else if (ThumbMode >= Thumb2)
ARMArchVersion = V6T2;
if (isAAPCS_ABI())
stackAlignment = 8;
if (isTargetDarwin())
IsR9Reserved = ReserveR9 | (ARMArchVersion < V6);
if (!isThumb() || hasThumb2())
PostRAScheduler = true;
// v6+ may or may not support unaligned mem access depending on the system
// configuration.
if (!StrictAlign && hasV6Ops() && isTargetDarwin())
AllowsUnalignedMem = true;
}
/// GVIsIndirectSymbol - true if the GV will be accessed via an indirect symbol.
bool
ARMSubtarget::GVIsIndirectSymbol(const GlobalValue *GV,
Reloc::Model RelocM) const {
if (RelocM == Reloc::Static)
return false;
// Materializable GVs (in JIT lazy compilation mode) do not require an extra
// load from stub.
bool isDecl = GV->isDeclaration() && !GV->isMaterializable();
if (!isTargetDarwin()) {
// Extra load is needed for all externally visible.
if (GV->hasLocalLinkage() || GV->hasHiddenVisibility())
return false;
return true;
} else {
if (RelocM == Reloc::PIC_) {
// If this is a strong reference to a definition, it is definitely not
// through a stub.
if (!isDecl && !GV->isWeakForLinker())
return false;
// Unless we have a symbol with hidden visibility, we have to go through a
// normal $non_lazy_ptr stub because this symbol might be resolved late.
if (!GV->hasHiddenVisibility()) // Non-hidden $non_lazy_ptr reference.
return true;
// If symbol visibility is hidden, we have a stub for common symbol
// references and external declarations.
if (isDecl || GV->hasCommonLinkage())
// Hidden $non_lazy_ptr reference.
return true;
return false;
} else {
// If this is a strong reference to a definition, it is definitely not
// through a stub.
if (!isDecl && !GV->isWeakForLinker())
return false;
// Unless we have a symbol with hidden visibility, we have to go through a
// normal $non_lazy_ptr stub because this symbol might be resolved late.
if (!GV->hasHiddenVisibility()) // Non-hidden $non_lazy_ptr reference.
return true;
}
}
return false;
}
unsigned ARMSubtarget::getMispredictionPenalty() const {
// If we have a reasonable estimate of the pipeline depth, then we can
// estimate the penalty of a misprediction based on that.
if (isCortexA8())
return 13;
else if (isCortexA9())
return 8;
// Otherwise, just return a sensible default.
return 10;
}
bool ARMSubtarget::enablePostRAScheduler(
CodeGenOpt::Level OptLevel,
TargetSubtarget::AntiDepBreakMode& Mode,
RegClassVector& CriticalPathRCs) const {
Mode = TargetSubtarget::ANTIDEP_CRITICAL;
CriticalPathRCs.clear();
CriticalPathRCs.push_back(&ARM::GPRRegClass);
return PostRAScheduler && OptLevel >= CodeGenOpt::Default;
}