nicolas/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

LoadStoreVectorizer: Remove TargetBaseAlign. Keep alignment for stack adjustments.

Browse Source 
Summary:
TargetBaseAlign is no longer required since LSV checks if target allows misaligned accesses.
A constant defining a base alignment is still needed for stack accesses where alignment can be adjusted.

Previous patch (D22936) was reverted because tests were failing. This patch also fixes the cause of those failures:
- x86 failing tests either did not have the right target, or the right alignment.
- NVPTX failing tests did not have the right alignment.
- AMDGPU failing test (merge-stores) should allow vectorization with the given alignment but the target info
  considers <3xi32> a non-standard type and gives up early. This patch removes the condition and only checks
  for a maximum size allowed and relies on the next condition checking for %4 for correctness.
  This should be revisited to include 3xi32 as a MVT type (on arsenm's non-immediate todo list).

Note that checking the sizeInBits for a MVT is undefined (leads to an assertion failure),
so we need to create an EVT, hence the interface change in allowsMisaligned to include the Context.

Reviewers: arsenm, jlebar, tstellarAMD

Subscribers: jholewinski, arsenm, mzolotukhin, llvm-commits

Differential Revision: https://reviews.llvm.org/D23068

llvm-svn: 277735
This commit is contained in:
Alina Sbirlea
2016-08-04 16:38:44 +00:00
parent 98d78405b0
commit 6f937b1144
12 changed files with 65 additions and 53 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
llvm/lib/Transforms/Vectorize/LoadStoreVectorizer.cpp
View File

@@ -40,9 +40,8 @@ STATISTIC(NumScalarsVectorized, "Number of scalar accesses vectorized");
namespace {
// TODO: Remove this
static const unsigned TargetBaseAlign = 4;
// FIXME: Assuming stack alignment of 4 is always good enough
static const unsigned StackAdjustedAlignment = 4;
typedef SmallVector<Instruction *, 8> InstrList;
typedef MapVector<Value *, InstrList> InstrListMap;
@@ -798,8 +797,8 @@ bool Vectorizer::vectorizeStoreChain(
// so we can cheat and change it!
Value *V = GetUnderlyingObject(S0->getPointerOperand(), DL);
if (AllocaInst *AI = dyn_cast_or_null<AllocaInst>(V)) {
AI->setAlignment(TargetBaseAlign);
Alignment = TargetBaseAlign;
AI->setAlignment(StackAdjustedAlignment);
Alignment = StackAdjustedAlignment;
} else {
return false;
}
@@ -948,8 +947,8 @@ bool Vectorizer::vectorizeLoadChain(
// so we can cheat and change it!
Value *V = GetUnderlyingObject(L0->getPointerOperand(), DL);
if (AllocaInst *AI = dyn_cast_or_null<AllocaInst>(V)) {
AI->setAlignment(TargetBaseAlign);
Alignment = TargetBaseAlign;
AI->setAlignment(StackAdjustedAlignment);
Alignment = StackAdjustedAlignment;
} else {
return false;
}
@@ -1029,10 +1028,13 @@ bool Vectorizer::vectorizeLoadChain(
bool Vectorizer::accessIsMisaligned(unsigned SzInBytes, unsigned AddressSpace,
unsigned Alignment) {
if (Alignment % SzInBytes == 0)
return false;
bool Fast = false;
bool Allows = TTI.allowsMisalignedMemoryAccesses(SzInBytes * 8, AddressSpace,
bool Allows = TTI.allowsMisalignedMemoryAccesses(F.getParent()->getContext(),
SzInBytes * 8, AddressSpace,
Alignment, &Fast);
// TODO: Remove TargetBaseAlign
return !(Allows && Fast) && (Alignment % SzInBytes) != 0 &&
(Alignment % TargetBaseAlign) != 0;
DEBUG(dbgs() << "LSV: Target said misaligned is allowed? " << Allows
<< " and fast? " << Fast << "\n";);
return !Allows || !Fast;
}