Add support for opencl_unroll_hint attribute from OpenCL v2.0 s6.11.5.
Reusing most of metadata generation from CGLoopInfo helper class.
The code is based on Khronos OpenCL compiler:
https://github.com/KhronosGroup/SPIR/tree/spirv-1.0
Patch by Liu Yaxun (Sam)!
Differential Revision: http://reviews.llvm.org/D16686
llvm-svn: 261350
Now that we don't always add an element to AllocatedStackSlots if we
don't find a pre-existing unallocated stack slot, bumping
StatepointMaxSlotsRequired to `NumSlots + 1` is not correct. Instead
bump the statistic near the push_back, to
Builder.FuncInfo.StatepointStackSlots.size().
llvm-svn: 261348
The check on MFI->getObjectSize() has to be on the FrameIndex, not on
the index of the FrameIndex in AllocatedStackSlots. Weirdly, the tests
I added in rL261336 didn't catch this.
llvm-svn: 261347
This patch enables the vectorization of first-order recurrences. A first-order
recurrence is a non-reduction recurrence relation in which the value of the
recurrence in the current loop iteration equals a value defined in the previous
iteration. The load PRE of the GVN pass often creates these recurrences by
hoisting loads from within loops.
In this patch, we add a new recurrence kind for first-order phi nodes and
attempt to vectorize them if possible. Vectorization is performed by shuffling
the values for the current and previous iterations. The vectorization cost
estimate is updated to account for the added shuffle instruction.
Contributed-by: Matthew Simpson and Chad Rosier <mcrosier@codeaurora.org>
Differential Revision: http://reviews.llvm.org/D16197
llvm-svn: 261346
This patches does the following:
+ fix return type: ClangExpressionParser::Parse returns unsigned, but was actually returning a signed value, num_errors.
+ use helper clang::TextDiagnosticBuffer::getNumErrors() instead of counting the errors ourself.
+ limit scoping of block-level automatic variables as much as practical.
+ remove reused multipurpose TextDiagnosticBuffer::const_iterator in favour of loop-scoped err, warn, and note variables in the diagnostic printing code.
+ refactor diagnostic printing loops to use a proper loop invariant.
Author: Luke Drummond <luke.drummond@codeplay.com>
Differential Revision: http://reviews.llvm.org/D17273
llvm-svn: 261345
There is a report in the PR from several months ago that it failed
intermittently, but it is passing consistently for me on FreeBSD 10
and 11. We can re-add a decorator if further testing shows it is
still flakey.
llvm.org/pr17214
llvm-svn: 261340
I ran the test suite yesterday and when I came back this morning the
queue_user_work_item.cc test was hung. This could be why the
sanitizer-windows buildbot keeps randomly timing out. I updated all the
usages of WaitForSingleObject involving threading events. I'm assuming
the API can reliably wait for subprocesses, which is what the majority
of call sites use it for.
While I'm at it, we can simplify some EH tests now that clang can
compile C++ EH.
llvm-svn: 261338
NFCI. They key motivation here is that I'd like to use
SmallBitVector::all() in a later change. Also, using a bit vector here
seemed better in general.
The only interesting change here is that in the failure case of
allocateStackSlot, we no longer (the equivalent of) push_back(true) to
AllocatedStackSlots. As far as I can tell, this is fine, since we'd
never re-use those slots in the same StatepointLoweringState instance.
Technically there was no need to change the operator[] type accesses to
set() and test(), but I thought it'd be nice to make it obvious that
we're using something other than a std::vector like thing.
llvm-svn: 261337
allocateStackSlot did not consider the size of the value to be spilled
before deciding to re-use a spill slot. This was originally okay (since
originally we'd only ever spill pointers), but it became not okay when
we changed our scheme to directly spill vectors of pointers.
While this change fixes the bug pointed out, it has two performance
caveats:
- It matches spill slot and spillee size exactly, while in theory we
can spill, e.g., an 8 byte pointer into a 16 byte slot. This is
slightly complicated to fix since in the stackmaps section, we report
the size of the spill slot as the size of the "indirect value"; and
if they're no longer equivalent, we'll have to keep track of the
(indirect) value size separately from the stack slot size.
- It will "spuriously run out" of reusable slots, since we now have an
second check in the search loop in addition to the availablity
check (e.g. you had two free scalar slots, and you first ask for a
vector slot followed by a scalar slot). I'll fix this in a later
commit.
llvm-svn: 261336
This removes the unusual loop structure in allocateStackSlot in favor of
something more straightforward. I've also removed the cautionary
comment in the function, which I suspect is historical cruft now, and
confuses more than it enlightens.
llvm-svn: 261335
Both Linux and FreeBSD had a comment "This needs to be root-caused."
It looks like the failure has been fixed on both, and the Linux XFAIL
decorator was removed in r233716 (Mar 2015).
llvm-svn: 261333
Summary:
If we don't have the first and last access of an interleaved load group,
the first and last wide load in the loop can do an out of bounds
access. Even though we discard results from speculative loads,
this can cause problems, since it can technically generate page faults
(or worse).
We now discard interleaved load groups that don't have the first and
load in the group.
Reviewers: hfinkel, rengolin
Subscribers: rengolin, llvm-commits, mzolotukhin, anemet
Differential Revision: http://reviews.llvm.org/D17332
llvm-svn: 261331
Summary:
This was broken in r260694 which swapped the address and data operands
for flat store instructions. The code in SIInsertWaits assumes
that the data operand always comes before the address operand, so
we need to add a special case for flat.
Reviewers: arsenm
Subscribers: arsenm, llvm-commits
Differential Revision: http://reviews.llvm.org/D17366
llvm-svn: 261330
As discussed on PR24580, this patch adds some (more to come) initial fast-isel codegen tests to match the IR generated in clang/test/CodeGen/avx-builtins.c
llvm-svn: 261329
[git 65dafa83] introduced the GetBuiltinIncludePath function copied from cfe/lib/Driver/CC1Options.cpp
This function is no longer used in lldb's expression parser and I believe it is safe to remove it.
Author: Luke Drummond <luke.drummond@codeplay.com>
Differential Revision: http://reviews.llvm.org/D17266
llvm-svn: 261328
This reduces the .rodata of scyladb from 4501932 to 4334639 bytes (1.038
times smaller).
I don't think it is critical to support tail merging, just exact
duplicates, but given the code organization it was actually a bit easier
to support both.
llvm-svn: 261327
According to the SystemZ ABI, 128-bit integer types should be
passed and returned via implicit reference. However, this is
not currently implemented at the LLVM IR level for the i128
type. This does not matter when compiling C/C++ code, since
clang will implement the implicit reference itself.
However, it turns out that when calling libgcc helper routines
operating on 128-bit integers, LLVM will use i128 argument and
return value types; the resulting code is not compatible with
the ABI used in libgcc, leading to crashes (see PR26559).
This should be simple to fix, except that i128 currently is not
even a legal type for the SystemZ back end. Therefore, common
code will already split arguments and return values into multiple
parts. The bulk of this patch therefore consists of detecting
such parts, and correctly handling passing via implicit reference
of a value split into multiple parts. If at some time in the
future, i128 becomes a legal type, this code can be removed again.
This fixes PR26559.
llvm-svn: 261325
"aligned", by Vladimir Yakovlev
Fix clang/gcc incompatibility of bitfields layout in the presence of
pragma packed and attributes aligned and packed.
Differential Revision: http://reviews.llvm.org/D17023
llvm-svn: 261321
We now always print the reason why the code did not pass the LLVM verifier and
we also allow to disable verfication with -polly-codegen-verify=false. Before
this change the first assertion had generally no information why or what might
have gone wrong and it was also impossible to -view-cfg without recompile. This
change makes debugging bugs that result in incorrect IR a lot easier.
llvm-svn: 261320
This change is improving the instruction emulation based unwinding to
handle when the frame pointer is adjusted (increment/decrement) after
it has been initialized. The situation can occur in the prologue of
some function where FP is adjusted before it is copied back to SP.
Example code (thumb, generated by gcc 4.8):
< +0>: push {r4, r7, lr}
< +2>: sub sp, #0x14
< +4>: add r7, sp, #0x0
...
<+50>: adds r7, #0x14 ; The CL fixes the handling of this instruction
<+52>: mov sp, r7 ; Previously unwinding from here was broken
<+54>: pop {r4, r7, pc}
Differential revision: http://reviews.llvm.org/D17295
llvm-svn: 261318
* Else-ifs in ScriptParser::run() replaced with std::function + map
* Reordered members of ScriptParser
Differential revision: http://reviews.llvm.org/D17256
llvm-svn: 261317
routine.
We were getting this wrong in small ways and generally being very
inconsistent about it across loop passes. Instead, let's have a common
place where we do this. One minor downside is that this will require
some analyses like SCEV in more places than they are strictly needed.
However, this seems benign as these analyses are complete no-ops, and
without this consistency we can in many cases end up with the legacy
pass manager scheduling deciding to split up a loop pass pipeline in
order to run the function analysis half-way through. It is very, very
annoying to fix these without just being very pedantic across the board.
The only loop passes I've not updated here are ones that use
AU.setPreservesAll() such as IVUsers (an analysis) and the pass printer.
They seemed less relevant.
With this patch, almost all of the problems in PR24804 around loop pass
pipelines are fixed. The one remaining issue is that we run simplify-cfg
and instcombine in the middle of the loop pass pipeline. We've recently
added some loop variants of these passes that would seem substantially
cleaner to use, but this at least gets us much closer to the previous
state. Notably, the seven loop pass managers is down to three.
I've not updated the loop passes using LoopAccessAnalysis because that
analysis hasn't been fully wired into LoopSimplify/LCSSA, and it isn't
clear that those transforms want to support those forms anyways. They
all run late anyways, so this is harmless. Similarly, LSR is left alone
because it already carefully manages its forms and doesn't need to get
fused into a single loop pass manager with a bunch of other loop passes.
LoopReroll didn't use loop simplified form previously, and I've updated
the test case to match the trivially different output.
Finally, I've also factored all the pass initialization for the passes
that use this technique as well, so that should be done regularly and
reliably.
Thanks to James for the help reviewing and thinking about this stuff,
and Ben for help thinking about it as well!
Differential Revision: http://reviews.llvm.org/D17435
llvm-svn: 261316
especially the *structure* of it with respect to various pass managers.
This uncovers an absolute horror show of problems. This test shows just
how bad PR24804 is: we have a totaly of *seven* loop pass managers in
the main optimization pipeline.
I've tried to comment the various bits to the best of my knowledge, but
more enhancements here would be great.
Also great would be folks adding various test for other pipelines, I'm
focused on trying to fix the O2 pipeline. I just wanted a test to show
what I'm changing.
llvm-svn: 261305
Certain optimization passes (like globaldce) can prune function
declaration that SjLjEHPrepare assumed would exit when it'd
runOnFunction.
This fixes PR26669.
llvm-svn: 261303
more places to prevent gratuitous re-"runs" of these passes.
The passes themselves don't do any work when run, but we keep spending
time scheduling and running these needlessly when we really don't need
to do so.
This is the first patch towards fixing the really horrible loop pass
pipeline fragmentation pointed out by Sanjoy in PR24804.
llvm-svn: 261302
For templates, fields can have incomplete types:
template <class T>
struct A2 {
struct B;
B b;
};
Don't try to touch the DefinitionData of those fields.
llvm-svn: 261301
Dan Gohman