Summary:
When performing cmp for EQ/NE and the operand is sign extended, we can
avoid the truncaton if the bits to be tested are no less than origianl
bits.
Reviewers: eli.friedman
Subscribers: eli.friedman, aemerson, nemanjai, t.p.northover, llvm-commits
Differential Revision: https://reviews.llvm.org/D22933
llvm-svn: 277252
This should fix UB warnings from the sanitizer bots: LLD performs bit
manipulations on enums of this type, and these are UB if the underlying
storage type isn't specified.
llvm-svn: 277251
These come in two variants for now: G_INTRINSIC and G_INTRINSIC_W_SIDE_EFFECTS.
We may decide to split the latter up with finer-grained restrictions later, if
necessary.
llvm-svn: 277224
in r277177 and added back this test which was deleted in r277196 while
I tracked down these problems.
Changed from constructing Twine's to std::string's as Twine's don't work
across statements. Also removed a few unneeded Twine() constructions.
Fix the write_escaped() calls to not pass the unintended second argument
fixing the warning on the ld-x86_64-win7 bot.
llvm-svn: 277223
Up until now, we only had code to match PSADBW patterns that look like what
comes out of the loop vectorizer - a partial reduction inside the loop body
that gets fed into a horizontal operation in a different basic block.
This adds support for straight-line patterns, like those generated by the
SLP vectorizer.
Differential Revision: https://reviews.llvm.org/D22889
llvm-svn: 277219
Support for lowering to VBROADCASTF128 etc. in D22460 was not correctly ensuring that the only users of the 128-bit vector load were the insertions of the vector into the lower/upper subvectors.
llvm-svn: 277214
Previously this change was submitted from a Windows machine, so
changes made to the case of filenames and directory names did
not survive the commit, and as a result the CMake source file
names and the on-disk file names did not match on case-sensitive
file systems.
I'm resubmitting this patch from a Linux system, which hopefully
allows the case changes to make it through unfettered.
llvm-svn: 277213
This patch enables balanced affinity on machines that do not have
hardware threads and have cores clustered into packages. In facts,
balacing algorithm could be generalized for any arrangement with
at least two levels of hierarchy (depth > 1).
Differential Revision: https://reviews.llvm.org/D22365
llvm-svn: 277212
Summary:
Error handling in StreamExecutor is based on llvm::Error and
llvm::Expected. This CL sets up the StreamExecutor wrapper classes in
the streamexecutor namespace.
All the other StreamExecutor code makes use of this error handling code,
so this is the first CL for checking in StreamExecutor.
Reviewers: jlebar, tra
Subscribers: parallel_libs-commits
Differential Revision: https://reviews.llvm.org/D22687
llvm-svn: 277210
The MachO debug support code (committed in r276935) occasionally needs to
allocate string copies, and was doing so by creating std::strings on a
BumpPtrAllocator. The strings were untracked, so the destructors weren't being
run and we were leaking the memory when the allocator was thrown away. Since
it's easier than tracking the strings, this patch switches the copies to char
buffers allocated directly in the bump-ptr allocator.
llvm-svn: 277208
The recent change implementing __final forgot to initialize a variable.
This was caught by the Memory Sanitizer.
Properly initialize the value to nullptr to ensure proper memory reads.
Patch by Erich Keane!
Differential Revision: https://reviews.llvm.org/D22970
llvm-svn: 277206
LoopUnroll is a loop pass, so the analysis of OptimizationRemarkEmitter
is added to the common function analysis passes that loop passes
depend on.
The BFI and indirectly BPI used in this pass is computed lazily so no
overhead should be observed unless -pass-remarks-with-hotness is used.
This is how the patch affects the O3 pipeline:
Dominator Tree Construction
Natural Loop Information
Canonicalize natural loops
Loop-Closed SSA Form Pass
Basic Alias Analysis (stateless AA impl)
Function Alias Analysis Results
Scalar Evolution Analysis
+ Lazy Branch Probability Analysis
+ Lazy Block Frequency Analysis
+ Optimization Remark Emitter
Loop Pass Manager
Rotate Loops
Loop Invariant Code Motion
Unswitch loops
Simplify the CFG
Dominator Tree Construction
Basic Alias Analysis (stateless AA impl)
Function Alias Analysis Results
Combine redundant instructions
Natural Loop Information
Canonicalize natural loops
Loop-Closed SSA Form Pass
Scalar Evolution Analysis
+ Lazy Branch Probability Analysis
+ Lazy Block Frequency Analysis
+ Optimization Remark Emitter
Loop Pass Manager
Induction Variable Simplification
Recognize loop idioms
Delete dead loops
Unroll loops
...
llvm-svn: 277203
The previous change was insufficient to mark the content as read-write as the
structure itself was marked constant. Adjust this and add tests to ensure that
the section is marked appropriately as being read-write.
llvm-svn: 277200
1. Add description of the arguments to these attributes.
2. Add missing declarations to some of the MPI code examples.
3. Made clarifications where possible.
Based on the write-up by: Craig Flores.
Differential Revision: https://reviews.llvm.org/D22717
llvm-svn: 277192
In a previous patch, it was suggested to use all caps instead of
rolling caps for initialisms, so this patch changes everything
to do this.
llvm-svn: 277190
Patch by Sunita Marathe
Third try, now following fixes to MSan to handle mempcy in such a way that this commit won't break the MSan buildbots. (Thanks, Evegenii!)
llvm-svn: 277189
The following pattern was being layed out poorly:
A
/ \
B C
/ \ / \
D E ? (Doesn't matter)
Where A->B is far more likely than A->C, and prob(B->D) = prob(B->E)
The current algorithm gives:
A,B,C,E (D goes on worklist)
It does this even if C has a frequency count of 0. This patch
adjusts the layout calculation so that if freq(B->E) >> freq(C->E)
then we go ahead and layout E rather than C. Fallthrough half the time
is better than fallthrough never, or fallthrough very rarely. The
resulting layout is:
A,B,E, (C and D are in a worklist)
llvm-svn: 277187
frame to another was triggering an early stop when stepping back out to a
real frame. Check that we're doing this only for inlined frames.
<rdar://problem/26482931>
llvm-svn: 277185
The DAG combiner will try to merge consecutive stores into a bigger
store, unless the resulting store is not fast. Misaligned vector stores
are allowed on Hexagon, but are not fast. Add a testcase to make sure
this type of merging does not occur.
Patch by Pranav Bhandarkar.
llvm-svn: 277182
Adding extension cl_khr_mipmap_image to clang's OpenCL Extensions and initiated inside AMDGPU Target.
Patch by Aaron En Ye Shi.
Differential Revision: https://reviews.llvm.org/D22637
llvm-svn: 277181