During instruction selection, the AArch64 backend can recognise the
following pattern and generate an [U|S]MADDL instruction, i.e. a
multiply of two 32-bit operands with a 64-bit result:
(mul (sext i32), (sext i32))
However, when one of the operands is constant, the sign extension
gets folded into the constant in SelectionDAG::getNode(). This means
that the instruction selection sees this:
(mul (sext i32), i64)
...which doesn't match the pattern. Sign-extension and 64-bit
multiply instructions are generated, which are slower than one 32-bit
multiply.
Add a pattern to match this and generate the correct instruction, for
both signed and unsigned multiplies.
Patch by Chris Diamand!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259800 91177308-0d34-0410-b5e6-96231b3b80d8
Fix the lit bug that enabled this "feature" (empty triple is substring
of all possible target triples) and change the two outliers to use the
documented * syntax.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259799 91177308-0d34-0410-b5e6-96231b3b80d8
This patch corresponds to review:
http://reviews.llvm.org/D16847
There are some files in glibc that use the output operand modifier even though
it was deprecated in GCC. This patch just adds support for it to prevent issues
with such files.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259798 91177308-0d34-0410-b5e6-96231b3b80d8
This patch allows the mixing of scaled and unscaled load/stores to form
load/store pairs.
PR24465
http://reviews.llvm.org/D12116
Many thanks to Ahmed and Michael for fixes and code review.
This is a reapplication of r246769, which was reverted in r246782 due to a
test-suite failure. I'm unable to reproduce the issue at this time.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259790 91177308-0d34-0410-b5e6-96231b3b80d8
This adds -DLLVM_ENABLE_LTO, rather than forcing people to manually
add -flto to the various _FLAGS variables.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259766 91177308-0d34-0410-b5e6-96231b3b80d8
Current SCEV expansion will expand SCEV as a sequence of operations
and doesn't utilize the value already existed. This will introduce
redundent computation which may not be cleaned up throughly by
following optimizations.
This patch introduces an ExprValueMap which is a map from SCEV to the
set of equal values with the same SCEV. When a SCEV is expanded, the
set of values is checked and reused whenever possible before generating
a sequence of operations.
The original commit triggered regressions in Polly tests. The regressions
exposed two problems which have been fixed in current version.
1. Polly will generate a new function based on the old one. To generate an
instruction for the new function, it builds SCEV for the old instruction,
applies some tranformation on the SCEV generated, then expands the transformed
SCEV and insert the expanded value into new function. Because SCEV expansion
may reuse value cached in ExprValueMap, the value in old function may be
inserted into new function, which is wrong.
In SCEVExpander::expand, there is a logic to check the cached value to
be used should dominate the insertion point. However, for the above
case, the check always passes. That is because the insertion point is
in a new function, which is unreachable from the old function. However
for unreachable node, DominatorTreeBase::dominates thinks it will be
dominated by any other node.
The fix is to simply add a check that the cached value to be used in
expansion should be in the same function as the insertion point instruction.
2. When the SCEV is of scConstant type, expanding it directly is cheaper than
reusing a normal value cached. Although in the cached value set in ExprValueMap,
there is a Constant type value, but it is not easy to find it out -- the cached
Value set is not sorted according to the potential cost. Existing reuse logic
in SCEVExpander::expand simply chooses the first legal element from the cached
value set.
The fix is that when the SCEV is of scConstant type, don't try the reuse
logic. simply expand it.
Differential Revision: http://reviews.llvm.org/D12090
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259736 91177308-0d34-0410-b5e6-96231b3b80d8
enabled): ensure that we do not invoke the sized deallocator for MemoryBuffer
subclasses that have tail-allocated data.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259735 91177308-0d34-0410-b5e6-96231b3b80d8
D16251)
Summary:
This is a simpler fix to the problem than the dominator approach in
http://reviews.llvm.org/D16251. It adds only values into the gather() while loop
that have been seen before.
The actual endless loop is in the constant compare gather() routine in
Utils/SimplifyCFG.cpp. The same value ret.0.off0.i is pushed back into the
queue:
%.ret.0.off0.i = or i1 %.ret.0.off0.i, %cmp10.i
Here is what happens at the IR level:
for.cond.i: ; preds = %if.end6.i,
%if.end.i54
%ix.0.i = phi i32 [ 0, %if.end.i54 ], [ %inc.i55, %if.end6.i ]
%ret.0.off0.i = phi i1 [false, %if.end.i54], [%.ret.0.off0.i, %if.end6.i] <<<
%cmp2.i = icmp ult i32 %ix.0.i, %11
br i1 %cmp2.i, label %for.body.i, label %LBJ_TmpSimpleNeedExt.exit
if.end6.i: ; preds = %for.body.i
%cmp10.i = icmp ugt i32 %conv.i, %add9.i
%.ret.0.off0.i = or i1 %ret.0.off0.i, %cmp10.i <<<
When if.end.i54 gets eliminated which removes the definition of ret.0.off0.i.
The result is the expression %.ret.0.off0.i = or i1 %.ret.0.off0.i, %cmp10.i
(Note the first ‘or’ operand is now %.ret.0.off0.i, and *NOT* %ret.0.off0.i).
And
now there is use of .ret.0.off0.i before a definition which triggers the
“endless” loop in gather():
while(!DFT.empty()) {
V = DFT.pop_back_val(); // V is .ret.0.off0.i
if (Instruction *I = dyn_cast<Instruction>(V)) {
// If it is a || (or && depending on isEQ), process the operands.
if (I->getOpcode() == (isEQ ? Instruction::Or : Instruction::And)) {
DFT.push_back(I->getOperand(1)); // This is now .ret.0.off0.i also
DFT.push_back(I->getOperand(0));
continue; // “endless loop” for .ret.0.off0.i
}
Reviewers: reames, ahatanak
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D16839
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259730 91177308-0d34-0410-b5e6-96231b3b80d8
Unfortunately, ProgramInfo::ProcessId is signed on Unix and unsigned on
Windows, breaking the standard fix of using '0U' in the gtest
expectation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259704 91177308-0d34-0410-b5e6-96231b3b80d8
Bail out if we have a PHI on an EHPad that gets a value from a
CatchSwitchInst. Because the CatchSwitchInst cannot be split, there is
no good place to stick any instructions.
This fixes PR26373.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259702 91177308-0d34-0410-b5e6-96231b3b80d8
This is mostly about having shorter lines and standardizing on one
interface, but it also avoids some needless indirection.
No functional change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259697 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
In r257979, I added code to ensure that we wouldn't merge DebugLocEntries if
the pieces they describe overlap. Unfortunately, I failed to cover the case,
where there may have multiple active Expressions in the entry, in which case we
need to make sure that no two values overlap before we can perform the merge.
This fixed PR26148.
Reviewers: aprantl
Differential Revision: http://reviews.llvm.org/D16742
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259696 91177308-0d34-0410-b5e6-96231b3b80d8
The IR/Value class had a linkage issue present when LLVM was built
as a library, and the LLVM library build time had different settings
for NDEBUG than the client of the LLVM library. Clients could get
into a state where the LLVM lib expected
Value::assertModuleIsMaterialized() to be inline-defined in the header
but clients expected that method to be defined in the LLVM library.
See this llvm-commits thread for more details:
http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160201/329667.html
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259695 91177308-0d34-0410-b5e6-96231b3b80d8
This patch consists of two parts: a performance fix in DAGCombiner.cpp
and a correctness fix in SelectionDAG.cpp.
The test case tests the bug that's uncovered by the performance fix, and
fixed by the correctness fix.
The performance fix keeps the containers required by the
hasPredecessorHelper (which is a lazy DFS) and reuse them. Since
hasPredecessorHelper is called in a loop, the overall efficiency reduced
from O(n^2) to O(n), where n is the number of SDNodes.
The correctness fix keeps iterating the neighbor list even if it's time
to early return. It will return after finishing adding all neighbors to
Worklist, so that no neighbors are discarded due to the original early
return.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259691 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
This patch adds a reserve call to an expensive function
(`llvm::LoadIntrinsics`), and may fix a few other low hanging
performance fruit (I've put them in comments for now, so we can
discuss).
**Motivation:**
As I'm sure other developers do, when I build LLVM, I build the entire
project with the same config (`Debug`, `MinSizeRel`, `Release`, or
`RelWithDebInfo`). However, the `Debug` config also builds llvm-tblgen
in `Debug` mode. Later build steps that run llvm-tblgen then can
actually be the slowest steps in the entire build. Nobody likes slow
builds.
Reviewers: rnk, dblaikie
Differential Revision: http://reviews.llvm.org/D16832
Patch by Alexander G. Riccio
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259683 91177308-0d34-0410-b5e6-96231b3b80d8
Add support for TLS access for Windows on ARM. This generates a similar access
to MSVC for ARM.
The changes to the tablegen data is needed to support loading an external symbol
global that is not for a call. The adjustments to the DAG to DAG transforms are
needed to preserve the 32-bit move.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@259676 91177308-0d34-0410-b5e6-96231b3b80d8
