From the standard: The taskloop construct specifies that the iterations of one
or more associated loops will be executed in parallel using OpenMP tasks. The
iterations are distributed across tasks created by the construct and scheduled
to be executed.
This initial implementation uses a simple linear tasks distribution algorithm.
Later we can add other algorithms to speedup generation of huge number of tasks
(i.e., tree-like tasks generation should be faster).
This needs to be put into the OpenMP runtime library in order for the
compiler team to develop the compiler side of the implementation.
Differential Revision: http://reviews.llvm.org/D17404
llvm-svn: 262535
Turns "foo" into 'foo' (or vice versa, depending on configuration).
This makes it more convenient to follow the Google JavaScript style
guide:
https://google.github.io/styleguide/javascriptguide.xml?showone=Strings#Strings
This functionality is behind the option "JavaScriptQuotes", which can be:
* "leave" (no re-quoting)
* "single" (change to single quotes)
* "double" (change to double quotes)
This also changes single quoted JavaScript string literals to be treated
as tok::string_literal, not tok::char_literal, which fixes two unrelated
tests.
Patch by Martin Probst. Thank you.
llvm-svn: 262534
From the standard: A doacross loop nest is a loop nest that has cross-iteration
dependence. An iteration is dependent on one or more lexicographically earlier
iterations. The ordered clause parameter on a loop directive identifies the
loop(s) associated with the doacross loop nest.
The init/fini routines allocate/free doacross buffer(s) for each loop for each
thread. The wait routine waits for a flag designated by the dependence vector.
The post routine sets the flag designated by current iteration vector. We use
a similar technique of shared buffer indices that covers up to 7 nowait loops
executed simultaneously by different threads (number 7 has no real meaning,
just heuristic value). Also, the size of structures are kept intact via
reducing dummy arrays.
This needs to be put into the OpenMP runtime library in order for the compiler
team to develop the compiler side of the implementation.
Differential Revision: http://reviews.llvm.org/D17399
llvm-svn: 262532
Summary: This is the last step toward supporting aggregate memory access in instcombine. This explodes stores of arrays into a serie of stores for each element, allowing them to be optimized.
Reviewers: joker.eph, reames, hfinkel, majnemer, mgrang
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D17828
llvm-svn: 262530
PDB is Microsoft's debug information format, and although we
cannot yet generate it, we still must be able to consume it.
Reason for this is that debug information for system libraries
(e.g. kernel32, C Runtime Library, etc) only have debug info
in PDB format, so in order to be able to support debugging
of system code, we must support it.
Currently this code should compile on every platform, but on
non-Windows platforms the PDB plugin will return 0 capabilities,
meaning that for now PDB is only supported on Windows. This
may change in the future, but the API is designed in such a way
that this will require few (if any) changes on the LLDB side.
In the future we can just flip a switch and everything will
work.
This patch only adds support for line tables. It does not return
information about functions, types, global variables, or anything
else. This functionality will be added in a followup patch.
Differential Revision: http://reviews.llvm.org/D17363
Reviewed by: Greg Clayton
llvm-svn: 262528
Summary:
Hi David, SCE folks,
What is implemented in this patch is enough for the upstream libprofile to
work for PGO with the PS4 game codebase I tested ("game7" for you SCE
folks; this is with a standalone build of compiler-rt).
The first change, which is simple, is to stub out gethostname. PS4
doesn't have a simple analog for this that doesn't bring in extra
OS libraries, so for now we do not support `%h` expansion.
This is consistent with internal B#136272.
The second change implies future work, but is a simple change at present.
PS4 does not have `getenv`, so for now we will introduce a shim.
This obviously makes it impossible for many of the tests to be run since
they require setting `LLVM_PROFILE_FILE=`.
I see two paths forward:
1. In the tests we are already wrapping execution with `%run` and so by
setting a PS4-specific expansion for `%run` we can pass the information
in another way We can adapt the getenv shim as appropriate.
We will need to experiment with this internally.
Maggie, Phillip, Filipe? Any ideas? Maybe ping me internally since we
may need to get into some PS4 vagaries. I'm thinking a fake getenv
library that uses some side channel for communication.
2. Another possibility which is more verbose is to use a separate clang
invocation with `-profile-generate=<filename>` to set the filename in
each test.
This might require redundant clang invocations though which may be
undesirable for upstream. David, thoughts?
Also, this is a fairly libprofile-specific workaround, so it e.g.
doesn't help Filipe's ASan work.
Overall, this approach sounds like a bit of a hack to me.
Small detail:
InstrProfilingPort.h seems like the natural place for the getenv shim,
but GCDAProfiling.c needs it as well. InstrProfilingUtil.h is currently
the only header common between InstrProfilingFile.c and GCDAProfiling.c.
I can move the shim to InstrProfilingPort.h and add an include to
GCDAProfiling.c as per your preference David.
Reviewers: davidxl, MaggieYi, phillip.power, filcab
Subscribers: simon.f.whittaker, slingn, probinson, llvm-commits
Differential Revision: http://reviews.llvm.org/D17676
llvm-svn: 262527
Summary: This is another step toward improving fca support. This unpack load of array in a series of load to array's elements.
Reviewers: chandlerc, joker.eph, majnemer, reames, hfinkel
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D15890
llvm-svn: 262521
* Run cc with -triple x86_64-linux-gnu to make symbol mangling predictable.
* Use temporary file as a fake GPU input so its content
does not interfere with pattern matching.
llvm-svn: 262516
This patch changes cc1 option for PGO profile use from
-fprofile-instr-use=<path> to -fprofile-instrument-use-path=<path>.
-fprofile-instr-use=<path> is now a driver only option.
In addition to decouple the cc1 option from the driver level option, this patch
also enables IR level profile use. cc1 option handling now reads the profile
header and sets CodeGenOpt ProfileUse (valid values are {None, Clang, LLVM}
-- this is a common enum for -fprofile-instrument={}, for the profile
instrumentation), and invoke the pipeline to enable the respective PGO use pass.
Reviewers: silvas, davidxl
Differential Revision: http://reviews.llvm.org/D17737
llvm-svn: 262515
Summary:
This makes cloning (and therefore the whole build) faster.
The checkout step goes from ~4m to ~30s on my host.
Reviewers: tfiala
Subscribers: lldb-commits
Differential Revision: http://reviews.llvm.org/D17425
llvm-svn: 262513
This is a test that Akira Hatanaka wrote to test GlobalOpt's handling of
aliases with GEP operands. David Majnemer independently made the same
change to GlobalOpt in r212079. Akira's test is a useful addition, so I'm
pulling it over from the llvm repo for Swift on GitHub.
llvm-svn: 262510
When div+rem calls on the same arguments are found, the ARM back-end merges the
two calls into one __aeabi_divmod call for up to 32-bits values. However,
for 64-bit values, which also have a lib call (__aeabi_ldivmod), it wasn't
merging the calls, and thus calling ldivmod twice and spilling the temporary
results, which generated pretty bad code.
This patch legalises 64-bit lib calls for divmod, so that now all the spilling
and the second call are gone. It also relaxes the DivRem combiner a bit on the
legal type check, since it was already checking for isLegalOrCustom on every
value, so the extra check for isTypeLegal was redundant.
This patch fixes PR17193 (and a long time FIXME in the tests).
llvm-svn: 262507
This is like r262493, but for pragma detect_mismatch instead of pragma comment.
The two pragmas have similar behavior, so use the same approach for both.
llvm-svn: 262506
This reverts commit r262370.
It turns out there is code out there that does sequences of allocas
greater than 4K: http://crbug.com/591404
The goal of this change was to improve the code size of inalloca call
sequences, but we got tangled up in the mess of dynamic allocas.
Instead, we should come back later with a separate MI pass that uses
dominance to optimize the full sequence. This should also be able to
remove the often unneeded stacksave/stackrestore pairs around the call.
llvm-svn: 262505
Most of the time ARM has the CCR.UNALIGN_TRP bit set to false which
means that unaligned loads/stores do not trap and even extensive testing
will not catch these bugs. However the multi/double variants are not
affected by this bit and will still trap. In effect a more aggressive
load/store optimization will break existing (bad) code.
These bugs do not necessarily manifest in the broken code where the
misaligned pointer is formed but often later in perfectly legal code
where it is accessed. This means recompiling system libraries (which
have no alignment bugs) with a newer compiler will break existing
applications (with alignment bugs) that worked before.
So (under protest) I implemented this safe mode which limits the
formation of multi/double operations to cases that are not affected by
user code (stack operations like spills/reloads) or cases where the
normal operations trap anyway (floating point load/stores). It is
disabled by default.
Differential Revision: http://reviews.llvm.org/D17015
llvm-svn: 262504
Now that DarwinLdDriver is the only derived class of Driver.
This patch merges them and actually removed the class because
they can now just be non-member functions. This change simplifies
a common header, Driver.h.
http://reviews.llvm.org/D17788
llvm-svn: 262502
The placement new calls here were all calling the allocation function
in RecyclingAllocator/Recycler for SDNode, instead of the function for
the specific subclass we were constructing.
Since this particular allocator always overallocates it more or less
worked, but would hide what we're actually doing from any memory
tools. Also, if you tried to change this allocator so something like a
BumpPtrAllocator or MallocAllocator, the compiler would crash horribly
all the time.
Part of llvm.org/PR26808.
llvm-svn: 262500
... and register them with CUDA runtime.
This is needed for commonly used cudaMemcpy*() APIs that use address of
host-side shadow to access their counterparts on device side.
Fixes PR26340
Differential Revision: http://reviews.llvm.org/D17779
llvm-svn: 262498
Summary:
This change enables frame pointer elimination in non-leaf functions.
The -fomit-frame-pointer option still needs to be used when compiling
via clang (or an equivalent method of not setting the
'no-frame-pointer-elim*' function attributes if generating llvm IR via
some other method) to take advantage of this optimization.
This change should be NFC when compiling via clang without
-fomit-frame-pointer.
Reviewers: t.p.northover
Subscribers: aemerson, rengolin, tberghammer, qcolombet, llvm-commits, danalbert, mcrosier, srhines
Differential Revision: http://reviews.llvm.org/D17730
llvm-svn: 262495
`#pragma comment` was handled by Sema calling a function on ASTConsumer, and
CodeGen then implementing this function and writing things to its output.
Instead, introduce a PragmaCommentDecl AST node and hang one off the
TranslationUnitDecl for every `#pragma comment` line, and then use the regular
serialization machinery. (Since PragmaCommentDecl has codegen relevance, it's
eagerly deserialized.)
http://reviews.llvm.org/D17799
llvm-svn: 262493
Otherwise users get messages from CheckAtomic about missing libatomic
instead of a sensible message that says "use GCC 4.7 or newer".
I structured the change along the lines of HandleLLVMStdlib.cmake, so
that the standalone build of Clang still gets the compiler version
check.
Reviewers: beanz
Differential Revision: http://reviews.llvm.org/D17789
llvm-svn: 262491
parts of the AA interface out of the base class of every single AA
result object.
Because this logic reformulates the query in terms of some other aspect
of the API, it would easily cause O(n^2) query patterns in alias
analysis. These could in turn be magnified further based on the number
of call arguments, and then further based on the number of AA queries
made for a particular call. This ended up causing problems for Rust that
were actually noticable enough to get a bug (PR26564) and probably other
places as well.
When originally re-working the AA infrastructure, the desire was to
regularize the pattern of refinement without losing any generality.
While I think it was successful, that is clearly proving to be too
costly. And the cost is needless: we gain no actual improvement for this
generality of making a direct query to tbaa actually be able to
re-use some other alias analysis's refinement logic for one of the other
APIs, or some such. In short, this is entirely wasted work.
To the extent possible, delegation to other API surfaces should be done
at the aggregation layer so that we can avoid re-walking the
aggregation. In fact, this significantly simplifies the logic as we no
longer need to smuggle the aggregation layer into each alias analysis
(or the TargetLibraryInfo into each alias analysis just so we can form
argument memory locations!).
However, we also have some delegation logic inside of BasicAA and some
of it even makes sense. When the delegation logic is baking in specific
knowledge of aliasing properties of the LLVM IR, as opposed to simply
reformulating the query to utilize a different alias analysis interface
entry point, it makes a lot of sense to restrict that logic to
a different layer such as BasicAA. So one aspect of the delegation that
was in every AA base class is that when we don't have operand bundles,
we re-use function AA results as a fallback for callsite alias results.
This relies on the IR properties of calls and functions w.r.t. aliasing,
and so seems a better fit to BasicAA. I've lifted the logic up to that
point where it seems to be a natural fit. This still does a bit of
redundant work (we query function attributes twice, once via the
callsite and once via the function AA query) but it is *exactly* twice
here, no more.
The end result is that all of the delegation logic is hoisted out of the
base class and into either the aggregation layer when it is a pure
retargeting to a different API surface, or into BasicAA when it relies
on the IR's aliasing properties. This should fix the quadratic query
pattern reported in PR26564, although I don't have a stand-alone test
case to reproduce it.
It also seems general goodness. Now the numerous AAs that don't need
target library info don't carry it around and depend on it. I think
I can even rip out the general access to the aggregation layer and only
expose that in BasicAA as it is the only place where we re-query in that
manner.
However, this is a non-trivial change to the AA infrastructure so I want
to get some additional eyes on this before it lands. Sadly, it can't
wait long because we should really cherry pick this into 3.8 if we're
going to go this route.
Differential Revision: http://reviews.llvm.org/D17329
llvm-svn: 262490
The LNT test suite with -polly-process-unprofitable
-polly-position=before-vectorizer currenty fails 59 tests. With this
barrier added, only 16 keep failing. This is probably because Polly's
code generation currently does not correctly preserve all analyses it
promised to preserve. Temporarily add this barrier until further
investigation.
llvm-svn: 262488
Matt Arsenault