This is the autoconf analog of r251201. I realize autoconf is
deprecated, but while it's in tree, it should at least be kept working.

Also add the deprecation message to configure.ac such that AutoRegen
actually picks ip up.

llvm-svn: 254215

llvm-svn: 254214

llvm-svn: 254213

This adds two thresholds to the sample profiler to affect inlining
decisions: the concept of global hotness and coldness.

Functions that have accumulated more than a certain fraction of samples at
runtime, are annotated with the InlineHint attribute. Conversely,
functions that accumulate less than a certain fraction of samples, are
annotated with the Cold attribute.

This is very similar to the hints emitted by Clang when using
instrumentation profiles.

Notice that this is a very blunt instrument. A function may have
globally collected a significant fraction of samples, but that does not
necessarily mean that every callsite for that function is hot.

Ideally, we would annotate each callsite with the samples collected at
that callsite. This way, the inliner can incorporate all these weights
into its cost model.

Once the inliner offers this functionality, we can change the hints
emitted here to a more precise per-callsite annotation. For now, this is
providing some measure of speedups with our internal benchmarks. I've
observed speedups of up to 23% (though the geo mean is about 3%). I expect
these numbers to improve as the inliner gets better annotations.

llvm-svn: 254212

Based on testing of internal benchmarks, I'm lowering this threshold to
a value of 0.1%.  This means that SamplePGO will respect 99.9% of the
original inline decisions when following a profile.

The performance difference is noticeable in some tests. With the
previous threshold, the speedups over baseline -O2 was about 0.63%. With
the new default, the speedups are around 3% on average.

The point of this threshold is not to do more aggressive inlining. When
an inlined callsite crosses this threshold, SamplePGO will redo the
inline decision so that it can better apply the input profile.

By respecting most original inline decisions, we can apply more of the
input profile because the shape of the code follows the profile more
closely.

In the next series, I'll be looking at adding some inline hints for the
cold callsites and for toplevel functions that are hot/cold as well.

llvm-svn: 254211

Remove out of date comment.
Pass .ll files to llvm-link.

llvm-svn: 254210

Now the ValueMapper has two callbacks. The first one maps the
declaration. The ValueMapper records the mapping and then materializes
the body/initializer.

llvm-svn: 254209

llvm-svn: 254208

http://reviews.llvm.org/D15029

llvm-svn: 254207

Changing the return type to void.

llvm-svn: 254206

Summary: This patch adds support for the interrupt attribute for mips32r2+.

Patch by Simon Dardis.

Reviewers: dsanders, aaron.ballman

Subscribers: aaron.ballman, cfe-commits

Differential Revision: http://reviews.llvm.org/D10802

llvm-svn: 254205

I forgot to credit the author.

llvm-svn: 254204

Summary: This patch adds support for the interrupt attribute for mips32r2+.

Reviewers: dsanders, aaron.ballman

Subscribers: aaron.ballman, cfe-commits

Differential Revision: http://reviews.llvm.org/D10802

llvm-svn: 254203

llvm-svn: 254202

Summary:
Since this build attribute corresponds to a whole module, and
different functions in a module may differ in the optimizations
enabled for them, this attribute is emitted after all functions,
and only in the case that the optimization goals for all
functions match.

Reviewers: logan, hans

Subscribers: aemerson, rengolin, llvm-commits

Differential Revision: http://reviews.llvm.org/D14934

llvm-svn: 254201

Summary:
This makes sure we do not attempt to send output over the gdb-remote protocol when the client is
not expecting it (i.e., after sending the stop-reply packet). Normally, this should not happen
(the process cannot generate output when it is stopped), but due to the fact that pty
communication is asynchronous in the linux kernel (llvm.org/pr25652), we may sometimes get this
output too late. Instead, we just hold the output, and send it next time we resume. This is not
ideal, but at least it makes sure we do not violate the remote protocol. Given that this happens
extremely rarely it's not worth trying to work around it with sleeps or something like that.

I also remove the m_stdio_communication_mutex, as all of LLGS is now single-threaded anyway.

Reviewers: tberghammer, ovyalov

Subscribers: lldb-commits

Differential Revision: http://reviews.llvm.org/D15019

llvm-svn: 254200

handling is corrected, the primary reason for forcing IAS as default is
gone and the remaining issues are still somewhat problematic in common
situations.

llvm-svn: 254199

ARMv8.2-A adds 16-bit floating point versions of all existing VFP
floating-point instructions. This is an optional extension, so all of
these instructions require the FeatureFullFP16 subtarget feature.

Most of these instructions are the same as the 32- and 64-bit versions,
but with the type field (bits 23-22) set to 0b11. Previously the top bit
of the size field was always 0, so the instruction classes only provided
a 1-bit size field, which I have widened to 2 bits.

Differential Revision: http://reviews.llvm.org/D15014

llvm-svn: 254198

    
This patch reorganize the platform specific mapping information to
export the application mask on a external variable. This exported
variable will be used by intrumentation phase to create code to be
used on architecture with multiple VMA range.
    
The patch creates a new header, dfsan_platform.h, and move all the
mapping information and also create function accessors to the
mapping value.

Also for aarch64 it initialize application exported mask to the 
value based on runtime VMA detection.

llvm-svn: 254197

This patch changes the DFSan instrumentation for aarch64 to instead
of using fixes application mask defined by SANITIZER_AARCH64_VMA
to read the application shadow mask value from compiler-rt. The value
is initialized based on runtime VAM detection.

Along with this patch a compiler-rt one will also be added to export
the shadow mask variable.

llvm-svn: 254196

Exclusion of /usr/include and /usr/local/include headers paths for MCU target.

Differential Revision: http://reviews.llvm.org/D14954

llvm-svn: 254195

The test is flakey but it fails too often with gcc 4.9.2 on x86_64 to
be marked only as expected flakey.

llvm-svn: 254194

llvm-svn: 254193

[TableGen] Sort pattern predicates before concatenating into a string so that different orders of the same set will produce the same string. This can reduce the number of unique predicates in the isel tables. NFC

llvm-svn: 254192

[X86] Pair a NoVLX with HasAVX512 to match the others and remove a unique predicate check in the isel tables. NFC

llvm-svn: 254191

tool_path will be None for llvm-go if Go cannot be found

llvm-svn: 254190

llvm-svn: 254189

Summary:
Force exporting __morestack from llgoi, so that
the symbol is available to the execution engine
when linking with libLLVM.so. The engine does
not reference __morestack explicitly, so must
be provided by the host program.

Reviewers: pcc

Subscribers: axw, llvm-commits

Differential Revision: http://reviews.llvm.org/D12625

llvm-svn: 254188

Summary:
When running tests, pass the GO_EXECUTABLE CMake
cache variable to llvm-go. The "go" binary may
not be in $PATH, or may be different to the one
passed to CMake.

Reviewers: pcc

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D14041

llvm-svn: 254187

llvm-svn: 254186

llvm-svn: 254185

It now covers a regular function replacing an available_externally one.

llvm-svn: 254184

The COFF object writer was previously adding unnecessary symbols to its
temporary data structures and cleaning them up later. This made the code
harder to understand and caused a bug (aliases classed as temporary symbols
would cause an assertion failure). A much simpler way of handling such
symbols is to ask the layout for their section-relative position when needed.

Tested with a bootstrap on Windows and by building Chrome.

Differential Revision: http://reviews.llvm.org/D14975

llvm-svn: 254183

Summary:
Inside a range-based for-loop over an array, the compiler
generates pointer arithmetic (end = array + size). Don't flag this.

Reviewers: alexfh, sbenza, bkramer, aaron.ballman

Subscribers: cfe-commits

Differential Revision: http://reviews.llvm.org/D14582

llvm-svn: 254182

Remove tabs.

llvm-svn: 254181

As discussed on D14909

llvm-svn: 254180

The order in which instructions are truncated in truncateToMinimalBitwidths
effects code generation. Switch to a map with a determinisic order, since the
iteration order over a DenseMap is not defined.

This code is not hot, so the difference in container performance isn't
interesting.

Many thanks to David Blaikie for making me aware of MapVector!

Fixes PR25490.

Differential Revision: http://reviews.llvm.org/D14981

llvm-svn: 254179

If an argument of the INPUT directive is a regular path, linker should
lookup it in the current folder first.

The fix does not contain any test cases because I think it is not a good
idea to pollute a current folder (which in general might be arbitrary)
by test files.

Differential Revision: http://reviews.llvm.org/D15027

llvm-svn: 254178

[X86] Now that X86VPermt2 is used in all the avx512_perm_t_sizes just hardcode it into the patterns instead of passing as an argument. NFC

llvm-svn: 254177

[X86] Merge X86VPermt2Fp and X86VPermt2Int back together by weakening them just enough. The SDTCisSameSizeAs introduced in r254138 helps here.

llvm-svn: 254176