[GlobalISel] Enable legalizing non-power-of-2 sized types.

This changes the interface of how targets describe how to legalize, see
the below description.

1. Interface for targets to describe how to legalize.

In GlobalISel, the API in the LegalizerInfo class is the main interface
for targets to specify which types are legal for which operations, and
what to do to turn illegal type/operation combinations into legal ones.

For each operation the type sizes that can be legalized without having
to change the size of the type are specified with a call to setAction.
This isn't different to how GlobalISel worked before. For example, for a
target that supports 32 and 64 bit adds natively:

  for (auto Ty : {s32, s64})
    setAction({G_ADD, 0, s32}, Legal);

or for a target that needs a library call for a 32 bit division:

  setAction({G_SDIV, s32}, Libcall);

The main conceptual change to the LegalizerInfo API, is in specifying
how to legalize the type sizes for which a change of size is needed. For
example, in the above example, how to specify how all types from i1 to
i8388607 (apart from s32 and s64 which are legal) need to be legalized
and expressed in terms of operations on the available legal sizes
(again, i32 and i64 in this case). Before, the implementation only
allowed specifying power-of-2-sized types (e.g. setAction({G_ADD, 0,
s128}, NarrowScalar).  A worse limitation was that if you'd wanted to
specify how to legalize all the sized types as allowed by the LLVM-IR
LangRef, i1 to i8388607, you'd have to call setAction 8388607-3 times
and probably would need a lot of memory to store all of these
specifications.

Instead, the legalization actions that need to change the size of the
type are specified now using a "SizeChangeStrategy".  For example:

   setLegalizeScalarToDifferentSizeStrategy(
       G_ADD, 0, widenToLargerAndNarrowToLargest);

This example indicates that for type sizes for which there is a larger
size that can be legalized towards, do it by Widening the size.
For example, G_ADD on s17 will be legalized by first doing WidenScalar
to make it s32, after which it's legal.
The "NarrowToLargest" indicates what to do if there is no larger size
that can be legalized towards. E.g. G_ADD on s92 will be legalized by
doing NarrowScalar to s64.

Another example, taken from the ARM backend is:
   for (unsigned Op : {G_SDIV, G_UDIV}) {
     setLegalizeScalarToDifferentSizeStrategy(Op, 0,
         widenToLargerTypesUnsupportedOtherwise);
     if (ST.hasDivideInARMMode())
       setAction({Op, s32}, Legal);
     else
       setAction({Op, s32}, Libcall);
   }

For this example, G_SDIV on s8, on a target without a divide
instruction, would be legalized by first doing action (WidenScalar,
s32), followed by (Libcall, s32).

The same principle is also followed for when the number of vector lanes
on vector data types need to be changed, e.g.:

   setAction({G_ADD, LLT::vector(8, 8)}, LegalizerInfo::Legal);
   setAction({G_ADD, LLT::vector(16, 8)}, LegalizerInfo::Legal);
   setAction({G_ADD, LLT::vector(4, 16)}, LegalizerInfo::Legal);
   setAction({G_ADD, LLT::vector(8, 16)}, LegalizerInfo::Legal);
   setAction({G_ADD, LLT::vector(2, 32)}, LegalizerInfo::Legal);
   setAction({G_ADD, LLT::vector(4, 32)}, LegalizerInfo::Legal);
   setLegalizeVectorElementToDifferentSizeStrategy(
       G_ADD, 0, widenToLargerTypesUnsupportedOtherwise);

As currently implemented here, vector types are legalized by first
making the vector element size legal, followed by then making the number
of lanes legal. The strategy to follow in the first step is set by a
call to setLegalizeVectorElementToDifferentSizeStrategy, see example
above.  The strategy followed in the second step
"moreToWiderTypesAndLessToWidest" (see code for its definition),
indicating that vectors are widened to more elements so they map to
natively supported vector widths, or when there isn't a legal wider
vector, split the vector to map it to the widest vector supported.

Therefore, for the above specification, some example legalizations are:
  * getAction({G_ADD, LLT::vector(3, 3)})
    returns {WidenScalar, LLT::vector(3, 8)}
  * getAction({G_ADD, LLT::vector(3, 8)})
    then returns {MoreElements, LLT::vector(8, 8)}
  * getAction({G_ADD, LLT::vector(20, 8)})
    returns {FewerElements, LLT::vector(16, 8)}


2. Key implementation aspects.

How to legalize a specific (operation, type index, size) tuple is
represented by mapping intervals of integers representing a range of
size types to an action to take, e.g.:

       setScalarAction({G_ADD, LLT:scalar(1)},
                       {{1, WidenScalar},  // bit sizes [ 1, 31[
                        {32, Legal},       // bit sizes [32, 33[
                        {33, WidenScalar}, // bit sizes [33, 64[
                        {64, Legal},       // bit sizes [64, 65[
                        {65, NarrowScalar} // bit sizes [65, +inf[
                       });

Please note that most of the code to do the actual lowering of
non-power-of-2 sized types is currently missing, this is just trying to
make it possible for targets to specify what is legal, and how non-legal
types should be legalized.  Probably quite a bit of further work is
needed in the actual legalizing and the other passes in GlobalISel to
support non-power-of-2 sized types.

I hope the documentation in LegalizerInfo.h and the examples provided in the
various {Target}LegalizerInfo.cpp and LegalizerInfoTest.cpp explains well
enough how this is meant to be used.

This drops the need for LLT::{half,double}...Size().


Differential Revision: https://reviews.llvm.org/D30529

llvm-svn: 317560

2 files changed, 323 insertions, 132 deletions

diff --git a/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp b/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp
index 99f605abe06..a8cfe0b89a0 100644
--- a/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp
+++ b/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp
@@ -173,12 +173,18 @@ LegalizerHelper::LegalizeResult LegalizerHelper::narrowScalar(MachineInstr &MI,
 
   MIRBuilder.setInstr(MI);
 
+  int64_t SizeOp0 = MRI.getType(MI.getOperand(0).getReg()).getSizeInBits();
+  int64_t NarrowSize = NarrowTy.getSizeInBits();
+
   switch (MI.getOpcode()) {
   default:
     return UnableToLegalize;
   case TargetOpcode::G_IMPLICIT_DEF: {
-    int NumParts = MRI.getType(MI.getOperand(0).getReg()).getSizeInBits() /
-                   NarrowTy.getSizeInBits();
+    // FIXME: add support for when SizeOp0 isn't an exact multiple of
+    // NarrowSize.
+    if (SizeOp0 % NarrowSize != 0)
+      return UnableToLegalize;
+    int NumParts = SizeOp0 / NarrowSize;
 
     SmallVector<unsigned, 2> DstRegs;
     for (int i = 0; i < NumParts; ++i) {
@@ -191,9 +197,12 @@ LegalizerHelper::LegalizeResult LegalizerHelper::narrowScalar(MachineInstr &MI,
     return Legalized;
   }
   case TargetOpcode::G_ADD: {
+    // FIXME: add support for when SizeOp0 isn't an exact multiple of
+    // NarrowSize.
+    if (SizeOp0 % NarrowSize != 0)
+      return UnableToLegalize;
     // Expand in terms of carry-setting/consuming G_ADDE instructions.
-    int NumParts = MRI.getType(MI.getOperand(0).getReg()).getSizeInBits() /
-                   NarrowTy.getSizeInBits();
+    int NumParts = SizeOp0 / NarrowTy.getSizeInBits();
 
     SmallVector<unsigned, 2> Src1Regs, Src2Regs, DstRegs;
     extractParts(MI.getOperand(1).getReg(), NarrowTy, NumParts, Src1Regs);
@@ -221,9 +230,12 @@ LegalizerHelper::LegalizeResult LegalizerHelper::narrowScalar(MachineInstr &MI,
     if (TypeIdx != 1)
       return UnableToLegalize;
 
-    int64_t NarrowSize = NarrowTy.getSizeInBits();
-    int NumParts =
-        MRI.getType(MI.getOperand(1).getReg()).getSizeInBits() / NarrowSize;
+    int64_t SizeOp1 = MRI.getType(MI.getOperand(1).getReg()).getSizeInBits();
+    // FIXME: add support for when SizeOp1 isn't an exact multiple of
+    // NarrowSize.
+    if (SizeOp1 % NarrowSize != 0)
+      return UnableToLegalize;
+    int NumParts = SizeOp1 / NarrowSize;
 
     SmallVector<unsigned, 2> SrcRegs, DstRegs;
     SmallVector<uint64_t, 2> Indexes;
@@ -270,12 +282,12 @@ LegalizerHelper::LegalizeResult LegalizerHelper::narrowScalar(MachineInstr &MI,
     return Legalized;
   }
   case TargetOpcode::G_INSERT: {
-    if (TypeIdx != 0)
+    // FIXME: add support for when SizeOp0 isn't an exact multiple of
+    // NarrowSize.
+    if (SizeOp0 % NarrowSize != 0)
       return UnableToLegalize;
 
-    int64_t NarrowSize = NarrowTy.getSizeInBits();
-    int NumParts =
-        MRI.getType(MI.getOperand(0).getReg()).getSizeInBits() / NarrowSize;
+    int NumParts = SizeOp0 / NarrowSize;
 
     SmallVector<unsigned, 2> SrcRegs, DstRegs;
     SmallVector<uint64_t, 2> Indexes;
@@ -330,9 +342,11 @@ LegalizerHelper::LegalizeResult LegalizerHelper::narrowScalar(MachineInstr &MI,
     return Legalized;
   }
   case TargetOpcode::G_LOAD: {
-    unsigned NarrowSize = NarrowTy.getSizeInBits();
-    int NumParts =
-        MRI.getType(MI.getOperand(0).getReg()).getSizeInBits() / NarrowSize;
+    // FIXME: add support for when SizeOp0 isn't an exact multiple of
+    // NarrowSize.
+    if (SizeOp0 % NarrowSize != 0)
+      return UnableToLegalize;
+    int NumParts = SizeOp0 / NarrowSize;
     LLT OffsetTy = LLT::scalar(
         MRI.getType(MI.getOperand(1).getReg()).getScalarSizeInBits());
 
@@ -357,9 +371,11 @@ LegalizerHelper::LegalizeResult LegalizerHelper::narrowScalar(MachineInstr &MI,
     return Legalized;
   }
   case TargetOpcode::G_STORE: {
-    unsigned NarrowSize = NarrowTy.getSizeInBits();
-    int NumParts =
-        MRI.getType(MI.getOperand(0).getReg()).getSizeInBits() / NarrowSize;
+    // FIXME: add support for when SizeOp0 isn't an exact multiple of
+    // NarrowSize.
+    if (SizeOp0 % NarrowSize != 0)
+      return UnableToLegalize;
+    int NumParts = SizeOp0 / NarrowSize;
     LLT OffsetTy = LLT::scalar(
         MRI.getType(MI.getOperand(1).getReg()).getScalarSizeInBits());
 
@@ -381,9 +397,11 @@ LegalizerHelper::LegalizeResult LegalizerHelper::narrowScalar(MachineInstr &MI,
     return Legalized;
   }
   case TargetOpcode::G_CONSTANT: {
-    unsigned NarrowSize = NarrowTy.getSizeInBits();
-    int NumParts =
-        MRI.getType(MI.getOperand(0).getReg()).getSizeInBits() / NarrowSize;
+    // FIXME: add support for when SizeOp0 isn't an exact multiple of
+    // NarrowSize.
+    if (SizeOp0 % NarrowSize != 0)
+      return UnableToLegalize;
+    int NumParts = SizeOp0 / NarrowSize;
     const APInt &Cst = MI.getOperand(1).getCImm()->getValue();
     LLVMContext &Ctx = MIRBuilder.getMF().getFunction()->getContext();
 
@@ -410,9 +428,12 @@ LegalizerHelper::LegalizeResult LegalizerHelper::narrowScalar(MachineInstr &MI,
     // ...
     // AN = BinOp<Ty/N> BN, CN
     // A = G_MERGE_VALUES A1, ..., AN
-    unsigned NarrowSize = NarrowTy.getSizeInBits();
-    int NumParts =
-        MRI.getType(MI.getOperand(0).getReg()).getSizeInBits() / NarrowSize;
+
+    // FIXME: add support for when SizeOp0 isn't an exact multiple of
+    // NarrowSize.
+    if (SizeOp0 % NarrowSize != 0)
+      return UnableToLegalize;
+    int NumParts = SizeOp0 / NarrowSize;
 
     // List the registers where the destination will be scattered.
     SmallVector<unsigned, 2> DstRegs;
@@ -854,7 +875,12 @@ LegalizerHelper::fewerElementsVector(MachineInstr &MI, unsigned TypeIdx,
   case TargetOpcode::G_ADD: {
     unsigned NarrowSize = NarrowTy.getSizeInBits();
     unsigned DstReg = MI.getOperand(0).getReg();
-    int NumParts = MRI.getType(DstReg).getSizeInBits() / NarrowSize;
+    unsigned Size = MRI.getType(DstReg).getSizeInBits();
+    int NumParts = Size / NarrowSize;
+    // FIXME: Don't know how to handle the situation where the small vectors
+    // aren't all the same size yet.
+    if (Size % NarrowSize != 0)
+      return UnableToLegalize;
 
     MIRBuilder.setInstr(MI);
 
diff --git a/llvm/lib/CodeGen/GlobalISel/LegalizerInfo.cpp b/llvm/lib/CodeGen/GlobalISel/LegalizerInfo.cpp
index e7a46eadb44..2bd8da3d83b 100644
--- a/llvm/lib/CodeGen/GlobalISel/LegalizerInfo.cpp
+++ b/llvm/lib/CodeGen/GlobalISel/LegalizerInfo.cpp
@@ -28,46 +28,130 @@
 #include "llvm/Support/MathExtras.h"
 #include "llvm/Target/TargetOpcodes.h"
 #include <algorithm>
-#include <cassert>
-#include <tuple>
-#include <utility>
-
+#include <map>
 using namespace llvm;
 
-LegalizerInfo::LegalizerInfo() {
-  DefaultActions[TargetOpcode::G_IMPLICIT_DEF] = NarrowScalar;
-
-  // FIXME: these two can be legalized to the fundamental load/store Jakob
-  // proposed. Once loads & stores are supported.
-  DefaultActions[TargetOpcode::G_ANYEXT] = Legal;
-  DefaultActions[TargetOpcode::G_TRUNC] = Legal;
+LegalizerInfo::LegalizerInfo() : TablesInitialized(false) {
+  // Set defaults.
+  // FIXME: these two (G_ANYEXT and G_TRUNC?) can be legalized to the
+  // fundamental load/store Jakob proposed. Once loads & stores are supported.
+  setScalarAction(TargetOpcode::G_ANYEXT, 1, {{1, Legal}});
+  setScalarAction(TargetOpcode::G_ZEXT, 1, {{1, Legal}});
+  setScalarAction(TargetOpcode::G_SEXT, 1, {{1, Legal}});
+  setScalarAction(TargetOpcode::G_TRUNC, 0, {{1, Legal}});
+  setScalarAction(TargetOpcode::G_TRUNC, 1, {{1, Legal}});
 
-  DefaultActions[TargetOpcode::G_INTRINSIC] = Legal;
-  DefaultActions[TargetOpcode::G_INTRINSIC_W_SIDE_EFFECTS] = Legal;
+  setScalarAction(TargetOpcode::G_INTRINSIC, 0, {{1, Legal}});
+  setScalarAction(TargetOpcode::G_INTRINSIC_W_SIDE_EFFECTS, 0, {{1, Legal}});
 
-  DefaultActions[TargetOpcode::G_ADD] = NarrowScalar;
-  DefaultActions[TargetOpcode::G_LOAD] = NarrowScalar;
-  DefaultActions[TargetOpcode::G_STORE] = NarrowScalar;
-  DefaultActions[TargetOpcode::G_OR] = NarrowScalar;
+  setLegalizeScalarToDifferentSizeStrategy(
+      TargetOpcode::G_IMPLICIT_DEF, 0, narrowToSmallerAndUnsupportedIfTooSmall);
+  setLegalizeScalarToDifferentSizeStrategy(
+      TargetOpcode::G_ADD, 0, widenToLargerTypesAndNarrowToLargest);
+  setLegalizeScalarToDifferentSizeStrategy(
+      TargetOpcode::G_OR, 0, widenToLargerTypesAndNarrowToLargest);
+  setLegalizeScalarToDifferentSizeStrategy(
+      TargetOpcode::G_LOAD, 0, narrowToSmallerAndUnsupportedIfTooSmall);
+  setLegalizeScalarToDifferentSizeStrategy(
+      TargetOpcode::G_STORE, 0, narrowToSmallerAndUnsupportedIfTooSmall);
 
-  DefaultActions[TargetOpcode::G_BRCOND] = WidenScalar;
-  DefaultActions[TargetOpcode::G_INSERT] = NarrowScalar;
-  DefaultActions[TargetOpcode::G_EXTRACT] = NarrowScalar;
-  DefaultActions[TargetOpcode::G_FNEG] = Lower;
+  setLegalizeScalarToDifferentSizeStrategy(
+      TargetOpcode::G_BRCOND, 0, widenToLargerTypesUnsupportedOtherwise);
+  setLegalizeScalarToDifferentSizeStrategy(
+      TargetOpcode::G_INSERT, 0, narrowToSmallerAndUnsupportedIfTooSmall);
+  setLegalizeScalarToDifferentSizeStrategy(
+      TargetOpcode::G_EXTRACT, 0, narrowToSmallerAndUnsupportedIfTooSmall);
+  setLegalizeScalarToDifferentSizeStrategy(
+      TargetOpcode::G_EXTRACT, 1, narrowToSmallerAndUnsupportedIfTooSmall);
+  setScalarAction(TargetOpcode::G_FNEG, 0, {{1, Lower}});
 }
 
 void LegalizerInfo::computeTables() {
-  for (unsigned Opcode = 0; Opcode <= LastOp - FirstOp; ++Opcode) {
-    for (unsigned Idx = 0, End = Actions[Opcode].size(); Idx != End; ++Idx) {
-      for (auto &Action : Actions[Opcode][Idx]) {
-        LLT Ty = Action.first;
-        if (!Ty.isVector())
-          continue;
-
-        auto &Entry = MaxLegalVectorElts[std::make_pair(Opcode + FirstOp,
-                                                        Ty.getElementType())];
-        Entry = std::max(Entry, Ty.getNumElements());
+  assert(TablesInitialized == false);
+
+  for (unsigned OpcodeIdx = 0; OpcodeIdx <= LastOp - FirstOp; ++OpcodeIdx) {
+    const unsigned Opcode = FirstOp + OpcodeIdx;
+    for (unsigned TypeIdx = 0; TypeIdx != SpecifiedActions[OpcodeIdx].size();
+         ++TypeIdx) {
+      // 0. Collect information specified through the setAction API, i.e.
+      // for specific bit sizes.
+      // For scalar types:
+      SizeAndActionsVec ScalarSpecifiedActions;
+      // For pointer types:
+      std::map<uint16_t, SizeAndActionsVec> AddressSpace2SpecifiedActions;
+      // For vector types:
+      std::map<uint16_t, SizeAndActionsVec> ElemSize2SpecifiedActions;
+      for (auto LLT2Action : SpecifiedActions[OpcodeIdx][TypeIdx]) {
+        const LLT Type = LLT2Action.first;
+        const LegalizeAction Action = LLT2Action.second;
+
+        auto SizeAction = std::make_pair(Type.getSizeInBits(), Action);
+        if (Type.isPointer())
+          AddressSpace2SpecifiedActions[Type.getAddressSpace()].push_back(
+              SizeAction);
+        else if (Type.isVector())
+          ElemSize2SpecifiedActions[Type.getElementType().getSizeInBits()]
+              .push_back(SizeAction);
+        else
+          ScalarSpecifiedActions.push_back(SizeAction);
+      }
+
+      // 1. Handle scalar types
+      {
+        // Decide how to handle bit sizes for which no explicit specification
+        // was given.
+        SizeChangeStrategy S = &unsupportedForDifferentSizes;
+        if (TypeIdx < ScalarSizeChangeStrategies[OpcodeIdx].size() &&
+            ScalarSizeChangeStrategies[OpcodeIdx][TypeIdx] != nullptr)
+          S = ScalarSizeChangeStrategies[OpcodeIdx][TypeIdx];
+        std::sort(ScalarSpecifiedActions.begin(), ScalarSpecifiedActions.end());
+        checkPartialSizeAndActionsVector(ScalarSpecifiedActions);
+        setScalarAction(Opcode, TypeIdx, S(ScalarSpecifiedActions));
       }
+
+      // 2. Handle pointer types
+      for (auto PointerSpecifiedActions : AddressSpace2SpecifiedActions) {
+        std::sort(PointerSpecifiedActions.second.begin(),
+                  PointerSpecifiedActions.second.end());
+        checkPartialSizeAndActionsVector(PointerSpecifiedActions.second);
+        // For pointer types, we assume that there isn't a meaningfull way
+        // to change the number of bits used in the pointer.
+        setPointerAction(
+            Opcode, TypeIdx, PointerSpecifiedActions.first,
+            unsupportedForDifferentSizes(PointerSpecifiedActions.second));
+      }
+
+      // 3. Handle vector types
+      SizeAndActionsVec ElementSizesSeen;
+      for (auto VectorSpecifiedActions : ElemSize2SpecifiedActions) {
+        std::sort(VectorSpecifiedActions.second.begin(),
+                  VectorSpecifiedActions.second.end());
+        const uint16_t ElementSize = VectorSpecifiedActions.first;
+        ElementSizesSeen.push_back({ElementSize, Legal});
+        checkPartialSizeAndActionsVector(VectorSpecifiedActions.second);
+        // For vector types, we assume that the best way to adapt the number
+        // of elements is to the next larger number of elements type for which
+        // the vector type is legal, unless there is no such type. In that case,
+        // legalize towards a vector type with a smaller number of elements.
+        SizeAndActionsVec NumElementsActions;
+        for (SizeAndAction BitsizeAndAction : VectorSpecifiedActions.second) {
+          assert(BitsizeAndAction.first % ElementSize == 0);
+          const uint16_t NumElements = BitsizeAndAction.first / ElementSize;
+          NumElementsActions.push_back({NumElements, BitsizeAndAction.second});
+        }
+        setVectorNumElementAction(
+            Opcode, TypeIdx, ElementSize,
+            moreToWiderTypesAndLessToWidest(NumElementsActions));
+      }
+      std::sort(ElementSizesSeen.begin(), ElementSizesSeen.end());
+      SizeChangeStrategy VectorElementSizeChangeStrategy =
+          &unsupportedForDifferentSizes;
+      if (TypeIdx < VectorElementSizeChangeStrategies[OpcodeIdx].size() &&
+          VectorElementSizeChangeStrategies[OpcodeIdx][TypeIdx] != nullptr)
+        VectorElementSizeChangeStrategy =
+            VectorElementSizeChangeStrategies[OpcodeIdx][TypeIdx];
+      setScalarInVectorAction(
+          Opcode, TypeIdx, VectorElementSizeChangeStrategy(ElementSizesSeen));
     }
   }
 
@@ -90,69 +174,24 @@ LegalizerInfo::getAction(const InstrAspect &Aspect) const {
       Aspect.Opcode == TargetOpcode::G_UNMERGE_VALUES)
     return std::make_pair(Legal, Aspect.Type);
 
-  LLT Ty = Aspect.Type;
-  LegalizeAction Action = findInActions(Aspect);
-  // LegalizerHelper is not able to handle non-power-of-2 types right now, so do
-  // not try to legalize them unless they are marked as Legal or Custom.
-  // FIXME: This is a temporary hack until the general non-power-of-2
-  // legalization works.
-  if (!isPowerOf2_64(Ty.getSizeInBits()) &&
-      !(Action == Legal || Action == Custom))
-    return std::make_pair(Unsupported, LLT());
-
-  if (Action != NotFound)
-    return findLegalAction(Aspect, Action);
-
-  unsigned Opcode = Aspect.Opcode;
-  if (!Ty.isVector()) {
-    auto DefaultAction = DefaultActions.find(Aspect.Opcode);
-    if (DefaultAction != DefaultActions.end() && DefaultAction->second == Legal)
-      return std::make_pair(Legal, Ty);
-
-    if (DefaultAction != DefaultActions.end() && DefaultAction->second == Lower)
-      return std::make_pair(Lower, Ty);
-
-    if (DefaultAction == DefaultActions.end() ||
-        DefaultAction->second != NarrowScalar)
-      return std::make_pair(Unsupported, LLT());
-    return findLegalAction(Aspect, NarrowScalar);
-  }
-
-  LLT EltTy = Ty.getElementType();
-  int NumElts = Ty.getNumElements();
-
-  auto ScalarAction = ScalarInVectorActions.find(std::make_pair(Opcode, EltTy));
-  if (ScalarAction != ScalarInVectorActions.end() &&
-      ScalarAction->second != Legal)
-    return findLegalAction(Aspect, ScalarAction->second);
-
-  // The element type is legal in principle, but the number of elements is
-  // wrong.
-  auto MaxLegalElts = MaxLegalVectorElts.lookup(std::make_pair(Opcode, EltTy));
-  if (MaxLegalElts > NumElts)
-    return findLegalAction(Aspect, MoreElements);
-
-  if (MaxLegalElts == 0) {
-    // Scalarize if there's no legal vector type, which is just a special case
-    // of FewerElements.
-    return std::make_pair(FewerElements, EltTy);
-  }
-
-  return findLegalAction(Aspect, FewerElements);
+  if (Aspect.Type.isScalar() || Aspect.Type.isPointer())
+    return findScalarLegalAction(Aspect);
+  assert(Aspect.Type.isVector());
+  return findVectorLegalAction(Aspect);
 }
 
 std::tuple<LegalizerInfo::LegalizeAction, unsigned, LLT>
 LegalizerInfo::getAction(const MachineInstr &MI,
                          const MachineRegisterInfo &MRI) const {
   SmallBitVector SeenTypes(8);
-  const MCInstrDesc &MCID = MI.getDesc();
-  const MCOperandInfo *OpInfo = MCID.OpInfo;
-  for (unsigned i = 0, e = MCID.getNumOperands(); i != e; ++i) {
+  const MCOperandInfo *OpInfo = MI.getDesc().OpInfo;
+  // FIXME: probably we'll need to cache the results here somehow?
+  for (unsigned i = 0; i < MI.getDesc().getNumOperands(); ++i) {
     if (!OpInfo[i].isGenericType())
       continue;
 
-    // We don't want to repeatedly check the same operand index, that
-    // could get expensive.
+    // We must only record actions once for each TypeIdx; otherwise we'd
+    // try to legalize operands multiple times down the line.
     unsigned TypeIdx = OpInfo[i].getGenericTypeIndex();
     if (SeenTypes[TypeIdx])
       continue;
@@ -172,38 +211,164 @@ bool LegalizerInfo::isLegal(const MachineInstr &MI,
   return std::get<0>(getAction(MI, MRI)) == Legal;
 }
 
-Optional<LLT> LegalizerInfo::findLegalType(const InstrAspect &Aspect,
-                                 LegalizeAction Action) const {
-  switch(Action) {
-  default:
-    llvm_unreachable("Cannot find legal type");
+bool LegalizerInfo::legalizeCustom(MachineInstr &MI, MachineRegisterInfo &MRI,
+                                   MachineIRBuilder &MIRBuilder) const {
+  return false;
+}
+
+LegalizerInfo::SizeAndActionsVec
+LegalizerInfo::increaseToLargerTypesAndDecreaseToLargest(
+    const SizeAndActionsVec &v, LegalizeAction IncreaseAction,
+    LegalizeAction DecreaseAction) {
+  SizeAndActionsVec result;
+  unsigned LargestSizeSoFar = 0;
+  if (v.size() >= 1 && v[0].first != 1)
+    result.push_back({1, IncreaseAction});
+  for (size_t i = 0; i < v.size(); ++i) {
+    result.push_back(v[i]);
+    LargestSizeSoFar = v[i].first;
+    if (i + 1 < v.size() && v[i + 1].first != v[i].first + 1) {
+      result.push_back({LargestSizeSoFar + 1, IncreaseAction});
+      LargestSizeSoFar = v[i].first + 1;
+    }
+  }
+  result.push_back({LargestSizeSoFar + 1, DecreaseAction});
+  return result;
+}
+
+LegalizerInfo::SizeAndActionsVec
+LegalizerInfo::decreaseToSmallerTypesAndIncreaseToSmallest(
+    const SizeAndActionsVec &v, LegalizeAction DecreaseAction,
+    LegalizeAction IncreaseAction) {
+  SizeAndActionsVec result;
+  if (v.size() == 0 || v[0].first != 1)
+    result.push_back({1, IncreaseAction});
+  for (size_t i = 0; i < v.size(); ++i) {
+    result.push_back(v[i]);
+    if (i + 1 == v.size() || v[i + 1].first != v[i].first + 1) {
+      result.push_back({v[i].first + 1, DecreaseAction});
+    }
+  }
+  return result;
+}
+
+LegalizerInfo::SizeAndAction
+LegalizerInfo::findAction(const SizeAndActionsVec &Vec, const uint32_t Size) {
+  assert(Size >= 1);
+  // Find the last element in Vec that has a bitsize equal to or smaller than
+  // the requested bit size.
+  // That is the element just before the first element that is bigger than Size.
+  auto VecIt = std::upper_bound(
+      Vec.begin(), Vec.end(), Size,
+      [](const uint32_t Size, const SizeAndAction lhs) -> bool {
+        return Size < lhs.first;
+      });
+  assert(VecIt != Vec.begin() && "Does Vec not start with size 1?");
+  --VecIt;
+  int VecIdx = VecIt - Vec.begin();
+
+  LegalizeAction Action = Vec[VecIdx].second;
+  switch (Action) {
   case Legal:
   case Lower:
   case Libcall:
   case Custom:
-    return Aspect.Type;
+    return {Size, Action};
+  case FewerElements:
+    // FIXME: is this special case still needed and correct?
+    // Special case for scalarization:
+    if (Vec == SizeAndActionsVec({{1, FewerElements}}))
+      return {1, FewerElements};
   case NarrowScalar: {
-    return findLegalizableSize(
-        Aspect, [&](LLT Ty) -> LLT { return Ty.halfScalarSize(); });
-  }
-  case WidenScalar: {
-    return findLegalizableSize(Aspect, [&](LLT Ty) -> LLT {
-      return Ty.getSizeInBits() < 8 ? LLT::scalar(8) : Ty.doubleScalarSize();
-    });
-  }
-  case FewerElements: {
-    return findLegalizableSize(
-        Aspect, [&](LLT Ty) -> LLT { return Ty.halfElements(); });
+    // The following needs to be a loop, as for now, we do allow needing to
+    // go over "Unsupported" bit sizes before finding a legalizable bit size.
+    // e.g. (s8, WidenScalar), (s9, Unsupported), (s32, Legal). if Size==8,
+    // we need to iterate over s9, and then to s32 to return (s32, Legal).
+    // If we want to get rid of the below loop, we should have stronger asserts
+    // when building the SizeAndActionsVecs, probably not allowing
+    // "Unsupported" unless at the ends of the vector.
+    for (int i = VecIdx - 1; i >= 0; --i)
+      if (!needsLegalizingToDifferentSize(Vec[i].second) &&
+          Vec[i].second != Unsupported)
+        return {Vec[i].first, Action};
+    llvm_unreachable("");
   }
+  case WidenScalar:
   case MoreElements: {
-    return findLegalizableSize(
-        Aspect, [&](LLT Ty) -> LLT { return Ty.doubleElements(); });
+    // See above, the following needs to be a loop, at least for now.
+    for (std::size_t i = VecIdx + 1; i < Vec.size(); ++i)
+      if (!needsLegalizingToDifferentSize(Vec[i].second) &&
+          Vec[i].second != Unsupported)
+        return {Vec[i].first, Action};
+    llvm_unreachable("");
   }
+  case Unsupported:
+    return {Size, Unsupported};
+  case NotFound:
+    llvm_unreachable("NotFound");
   }
 }
 
-bool LegalizerInfo::legalizeCustom(MachineInstr &MI,
-                                   MachineRegisterInfo &MRI,
-                                   MachineIRBuilder &MIRBuilder) const {
-  return false;
+std::pair<LegalizerInfo::LegalizeAction, LLT>
+LegalizerInfo::findScalarLegalAction(const InstrAspect &Aspect) const {
+  assert(Aspect.Type.isScalar() || Aspect.Type.isPointer());
+  if (Aspect.Opcode < FirstOp || Aspect.Opcode > LastOp)
+    return {NotFound, LLT()};
+  const unsigned OpcodeIdx = Aspect.Opcode - FirstOp;
+  if (Aspect.Type.isPointer() &&
+      AddrSpace2PointerActions[OpcodeIdx].find(Aspect.Type.getAddressSpace()) ==
+          AddrSpace2PointerActions[OpcodeIdx].end()) {
+    return {NotFound, LLT()};
+  }
+  const SmallVector<SizeAndActionsVec, 1> &Actions =
+      Aspect.Type.isPointer()
+          ? AddrSpace2PointerActions[OpcodeIdx]
+                .find(Aspect.Type.getAddressSpace())
+                ->second
+          : ScalarActions[OpcodeIdx];
+  if (Aspect.Idx >= Actions.size())
+    return {NotFound, LLT()};
+  const SizeAndActionsVec &Vec = Actions[Aspect.Idx];
+  // FIXME: speed up this search, e.g. by using a results cache for repeated
+  // queries?
+  auto SizeAndAction = findAction(Vec, Aspect.Type.getSizeInBits());
+  return {SizeAndAction.second,
+          Aspect.Type.isScalar() ? LLT::scalar(SizeAndAction.first)
+                                 : LLT::pointer(Aspect.Type.getAddressSpace(),
+                                                SizeAndAction.first)};
+}
+
+std::pair<LegalizerInfo::LegalizeAction, LLT>
+LegalizerInfo::findVectorLegalAction(const InstrAspect &Aspect) const {
+  assert(Aspect.Type.isVector());
+  // First legalize the vector element size, then legalize the number of
+  // lanes in the vector.
+  if (Aspect.Opcode < FirstOp || Aspect.Opcode > LastOp)
+    return {NotFound, Aspect.Type};
+  const unsigned OpcodeIdx = Aspect.Opcode - FirstOp;
+  const unsigned TypeIdx = Aspect.Idx;
+  if (TypeIdx >= ScalarInVectorActions[OpcodeIdx].size())
+    return {NotFound, Aspect.Type};
+  const SizeAndActionsVec &ElemSizeVec =
+      ScalarInVectorActions[OpcodeIdx][TypeIdx];
+
+  LLT IntermediateType;
+  auto ElementSizeAndAction =
+      findAction(ElemSizeVec, Aspect.Type.getScalarSizeInBits());
+  IntermediateType =
+      LLT::vector(Aspect.Type.getNumElements(), ElementSizeAndAction.first);
+  if (ElementSizeAndAction.second != Legal)
+    return {ElementSizeAndAction.second, IntermediateType};
+
+  auto i = NumElements2Actions[OpcodeIdx].find(
+      IntermediateType.getScalarSizeInBits());
+  if (i == NumElements2Actions[OpcodeIdx].end()) {
+    return {NotFound, IntermediateType};
+  }
+  const SizeAndActionsVec &NumElementsVec = (*i).second[TypeIdx];
+  auto NumElementsAndAction =
+      findAction(NumElementsVec, IntermediateType.getNumElements());
+  return {NumElementsAndAction.second,
+          LLT::vector(NumElementsAndAction.first,
+                      IntermediateType.getScalarSizeInBits())};
 }