This reverts commit https://llvm.org/svn/llvm-project/polly/trunk@260853
We unfortunately still have two bugs left which show only up with
-polly-process-unprofitable and which I forgot to test before committing.
llvm-svn: 260854
First support for this feature was committed in r259784. Support for
loop invariant load hoisting with different types was added by Johannes
Doerfert in r260045. This fixed the last known bug.
llvm-svn: 260853
C++ programs compiled for profiling (using `-pg`) should be linked with
`-lc++_p` (or `-lstdc++_p`, depending on the `-stdlib=` setting), not
with the regular C++ libraries.
Add a `FreeBSD::AddCXXStdlibLibArgs()` override to handle this, and add
a test case for it. While here, extend the test case for the proper
passing of -lm and -lm_p.
Reviewers: compnerd, davide, dws, emaste
Reviewed By: compnerd
Differential Revision: http://reviews.llvm.org/D16264
llvm-svn: 260851
FreeBSD also needs to have sanitizer_linux_libcdep.cc included,
otherwise linking will fail with "undefined reference to
`__sanitizer::GetRSS()'".
While here, tabify the FreeBSD part, similar to the other parts.
llvm-svn: 260839
This ensures that all of the various pieces are working. The next patch
will wire up commandline-driven alias analysis chain building and allow
BasicAA to work with the AAManager.
llvm-svn: 260838
Summary:
Unlike other outlined regions in OpenMP, offloading entry points have to have be visible (external linkage) for the device side. Using dots in the names of the entries can be therefore problematic for some toolchains, e.g. NVPTX.
Also the patch drops the column information in the unique name of the entry points. The parsing of directives ignore unknown tokens, preventing several target regions to be implemented in the same line. Therefore, the line information is sufficient for the name to be unique. Also, the preprocessor printer does not preserve the column information, causing offloading-entry detection issues if the host uses an integrated preprocessor and the target doesn't (or vice versa).
Reviewers: hfinkel, arpith-jacob, carlo.bertolli, kkwli0, ABataev
Subscribers: cfe-commits, fraggamuffin, caomhin
Differential Revision: http://reviews.llvm.org/D17179
llvm-svn: 260837
into the new pass manager and fix the latent bugs there.
This lets everything live together nicely, but it isn't really useful
yet. I never finished wiring the AA layer up for the new pass manager,
and so subsequent patches will change this to do that wiring and get AA
stuff more fully integrated into the new pass manager. Turns out this is
necessary even to get functionattrs ported over. =]
llvm-svn: 260836
r180893 added an indirect include of llvm/Config/Targets.def to
llvm/Support/CodeGen.h, which in turn is included by things like
llvm/IR/Module.h. After a full build of LLVM and Clang, ninja had to
rebuild 1274 files after reconfiguring.
This commit strips CodeGen.h back down to just a pile of enums and moves
the expensive includes over to CodeGenCWrappers.h (which is only
included in two places). This gets ninja down to 88 files if you
reconfigure with, e.g., -DLLVM_TARGETS_TO_BUILD=X86.
llvm-svn: 260835
This patch attempts to represent a shuffle as a repeating shuffle (recognisable by is128BitLaneRepeatedShuffleMask) with the source input(s) in their original lanes, followed by a single permutation of the 128-bit lanes to their final destinations.
On AVX2 we can additionally attempt to match using 64-bit sub-lane permutation. AVX2 can also now match a similar 'broadcasted' repeating shuffle.
This patch has several benefits:
* Avoids prematurely matching with lowerVectorShuffleByMerging128BitLanes which can require both inputs to have their input lanes permuted before shuffling.
* Can replace PERMPS/PERMD instructions - although these are useful for cross-lane unary shuffling, they require their shuffle mask to be pre-loaded (and increase register pressure).
* Matching the repeating shuffle makes use of a lot of existing shuffle lowering.
There is an outstanding minor AVX1 regression (combine_unneeded_subvector1 in vector-shuffle-combining.ll) of a previously 128-bit shuffle + subvector splat being converted to a subvector splat + (2 instruction) 256-bit shuffle, I intend to fix this in a followup patch for review.
Differential Revision: http://reviews.llvm.org/D16537
llvm-svn: 260834
Storing std::strings in attributes simply doesn't work, we never call
the destructor. Use an array of StringRefs instead of std::strings and
copy the data into memory taken from the ASTContext.
llvm-svn: 260831
As shown in:
https://llvm.org/bugs/show_bug.cgi?id=23203
...we currently die because lowering believes that mfence is allowed without SSE2 on x86-64,
but the instruction def doesn't know that.
I don't know if allowing mfence without SSE is right, but if not, at least now it's consistently wrong. :)
Differential Revision: http://reviews.llvm.org/D17219
llvm-svn: 260828
Since the origin AccFuncMap in ScopInfo is used by the underlying Scop
only, and it must stay alive until we delete the Scop. It will be better
if we simply move the origin AccFuncMap in ScopInfo into the Scop class.
llvm-svn: 260820
Make Scop become more portable such that we can use it in a CallGraphSCC pass.
The first step is to drop the analyses that are only used during Scop construction.
This patch drop LoopInfo from Scop.
llvm-svn: 260819
Make Scop become more portable such that we can use it in a CallGraphSCC pass.
The first step is to drop the analyses that are only used during Scop construction.
This patch drop DominatorTree from Scop.
llvm-svn: 260818
Make Scop become more portable such that we can use it in a CallGraphSCC pass.
The first step is to drop the analyses that are only used during Scop construction.
This patch drop ScopDecection from Scop.
llvm-svn: 260817
than the SCC object, and have it scan the instruction stream directly
rather than relying on call records.
This makes the behavior of this routine consistent between libc routines
and LLVM intrinsics for libc routines. We can go and start teaching it
about those being norecurse, but we should behave the same for the
intrinsic and the libc routine rather than differently. I chatted with
James Molloy and the inconsistency doesn't seem intentional and likely
is due to intrinsic calls not being modelled in the call graph analyses.
This also fixes a bug where we would deduce norecurse on optnone
functions, when generally we try to handle optnone functions as-if they
were replaceable and thus unanalyzable.
llvm-svn: 260813