| Commit message (Collapse) | Author | Age | Files | Lines |
| ... | |
| |
|
|
| |
llvm-svn: 218725
|
| |
|
|
|
|
|
|
|
|
| |
VPBROADCAST.
This has the somewhat expected pervasive impact. I don't know why
I forgot about this. Everything seems good with lots of significant
improvements in the tests.
llvm-svn: 218724
|
| |
|
|
|
|
|
|
| |
is unavailable.
Note, mingw uses its own printf instead of msvcrt.
llvm-svn: 218723
|
| |
|
|
|
|
|
|
| |
used with -mtriple.
Also, remove object-emission. test/DebugInfo/X86 doesn't require it.
llvm-svn: 218722
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In special cases select instructions can be eliminated by
replacing them with a cheaper bitwise operation even when the
select result is used outside its home block. The instances implemented
are patterns like
%x=icmp.eq
%y=select %x,%r, null
%z=icmp.eq|neq %y, null
br %z,true, false
==> %x=icmp.ne
%y=icmp.eq %r,null
%z=or %x,%y
br %z,true,false
The optimization is integrated into the instruction
combiner and performed only when all uses of the select result can
be replaced by the select operand proper. For this dominator information
is used and dominance is now a required analysis pass in the combiner.
The optimization itself is iterative. The critical step is to replace the
select result with the non-constant select operand. So the select becomes
local and the combiner iteratively works out simpler code pattern and
eventually eliminates the select.
rdar://17853760
llvm-svn: 218721
|
| |
|
|
|
|
| |
same coverage as the global checker.
llvm-svn: 218720
|
| |
|
|
| |
llvm-svn: 218719
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
The darwin linker has the -demangle option which directs it to demangle C++
(and soon Swift) mangled symbol names. Long term we need some Diagnostics object
for formatting errors and warnings. But for now we have the Core linker just
writing messages to llvm::errs(). So, to enable demangling, I changed the
Resolver to call a LinkingContext method on the symbol name.
To make this more interesting, the demangling code is done via __cxa_demangle()
which is part of the C++ ABI, which is only supported on some platforms, so I
had to conditionalize the code with the config generated HAVE_CXXABI_H.
llvm-svn: 218718
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
being on by default. -fno-cxx-modules can still be used to enable C modules but
not C++ modules, but C++ modules is not significantly less stable than C
modules any more.
Also remove some of the scare words from the modules documentation. We're
certainly not going to remove modules support (though we might change the
interface), and it works well enough to bootstrap and build lots of
non-trivial code.
Note that this does not represent a commitment to the current interface nor
implementation, and we still intend to follow whatever direction the C and C++
committees take regarding modules support.
llvm-svn: 218717
|
| |
|
|
|
|
|
| |
The optimization for -gmlt/-gline-tables-only introduced in r218129 happened to break on Darwin and produce no line number information due to
an incompatibility with dsymutil. ASan tests have been failing because of that and we disabled the use of -gmlt for the tests in r218545. This patch re-enables the use of -gmlt, because we have conditionally disabled the incompatible optimization in LLVM, so -gmlt now works on Darwin. Once Darwin's dsymutil is modified to allow this optimization, we can re-enable the optimization in LLVM.
llvm-svn: 218716
|
| |
|
|
| |
llvm-svn: 218715
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Two related things:
1. Fixes a bug when calculating the offset in GetLinearExpression. The code
previously used zext to extend the offset, so negative offsets were converted
to large positive ones.
2. Enhance aliasGEP to deduce that, if the difference between two GEP
allocations is positive and all the variables that govern the offset are also
positive (i.e. the offset is strictly after the higher base pointer), then
locations that fit in the gap between the two base pointers are NoAlias.
Patch by Nick White!
llvm-svn: 218714
|
| |
|
|
|
|
|
|
|
|
|
| |
common spot.
No functional change. Pre-emptive refactoring before I start pushing
some of this subprogram creation down into DWARFCompileUnit so I can
build different subprograms in the skeleton unit from the dwo unit for
adding -gmlt-like data to the skeleton.
llvm-svn: 218713
|
| |
|
|
|
|
|
| |
It would previously not continue the platforms loop
unless it could find the latest toolset directory.
llvm-svn: 218712
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
This patch adds a threshold that controls the number of bonus instructions
allowed for folding branches with common destination. The original code allows
at most one bonus instruction. With this patch, users can customize the
threshold to allow multiple bonus instructions. The default threshold is still
1, so that the code behaves the same as before when users do not specify this
threshold.
The motivation of this change is that tuning this threshold significantly (up
to 25%) improves the performance of some CUDA programs in our internal code
base. In general, branch instructions are very expensive for GPU programs.
Therefore, it is sometimes worth trading more arithmetic computation for a more
straightened control flow. Here's a reduced example:
__global__ void foo(int a, int b, int c, int d, int e, int n,
const int *input, int *output) {
int sum = 0;
for (int i = 0; i < n; ++i)
sum += (((i ^ a) > b) && (((i | c ) ^ d) > e)) ? 0 : input[i];
*output = sum;
}
The select statement in the loop body translates to two branch instructions "if
((i ^ a) > b)" and "if (((i | c) ^ d) > e)" which share a common destination.
With the default threshold, SimplifyCFG is unable to fold them, because
computing the condition of the second branch "(i | c) ^ d > e" requires two
bonus instructions. With the threshold increased, SimplifyCFG can fold the two
branches so that the loop body contains only one branch, making the code
conceptually look like:
sum += (((i ^ a) > b) & (((i | c ) ^ d) > e)) ? 0 : input[i];
Increasing the threshold significantly improves the performance of this
particular example. In the configuration where both conditions are guaranteed
to be true, increasing the threshold from 1 to 2 improves the performance by
18.24%. Even in the configuration where the first condition is false and the
second condition is true, which favors shortcuts, increasing the threshold from
1 to 2 still improves the performance by 4.35%.
We are still looking for a good threshold and maybe a better cost model than
just counting the number of bonus instructions. However, according to the above
numbers, we think it is at least worth adding a threshold to enable more
experiments and tuning. Let me know what you think. Thanks!
Test Plan: Added one test case to check the threshold is in effect
Reviewers: nadav, eliben, meheff, resistor, hfinkel
Reviewed By: hfinkel
Subscribers: hfinkel, llvm-commits
Differential Revision: http://reviews.llvm.org/D5529
llvm-svn: 218711
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
cases.
While clearly we don't need the AVX vector width, these ISA extensions
often cause us to select different instructions and we should cover them
even with the narrow vector width.
Also, while here, nuke the stress_test2 contents. There is no reason to
try to FileCheck this entire body when it is mostly a test for
successfully surviving the code generator.
llvm-svn: 218710
|
| |
|
|
|
|
|
|
|
|
|
| |
shuffle tests to match that used in the script I posted and now used
consistently in 128-bit tests.
Nothing interesting changing here, just using the label name as the
FileCheck label and a slightly more general comment marker consumption
strategy.
llvm-svn: 218709
|
| |
|
|
|
|
| |
isn't built.
llvm-svn: 218708
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
updating script so that they are more thorough and consistent.
Specific fixes here include:
- Actually test VEX-encoded AVX mnemonics.
- Actually use an SSE 4.1 run to test SSE 4.1 features!
- Correctly check instructions sequences from the start of the function.
- Elide the shuffle operands and comment designator in a consistent way.
- Test all of the architectures instead of just the ones I was motivated
to manually author.
I've gone back through and fixed up any egregious issues I spotted. Let
me know if I missed something you really dislike.
One downside to this is that we're now not as diligently using FileCheck
variables for registers. I would be much more concerned with this if we
had larger register usage, but there just aren't that interesting of
register choices here and most of the registers are constrained by the
ABI. Ultimately, I don't think this is likely to be the maintenance
burden for these tests and updating them again should be staright
forward.
llvm-svn: 218707
|
| |
|
|
|
|
|
|
| |
This is yet another edge case of ambiguous name resolution.
When a symbol is specified with /entry:SYM, SYM may be resolved
to the C++ mangled function name (?SYM@@YAXXZ).
llvm-svn: 218706
|
| |
|
|
|
|
| |
No functionality change intended.
llvm-svn: 218705
|
| |
|
|
| |
llvm-svn: 218704
|
| |
|
|
|
|
|
|
|
|
|
| |
This is a minimally useful pass to construct the __unwind_info section in a
final object from the various __compact_unwind inputs. Currently it doesn't
produce any compressed pages, only works for x86_64 and will fail if any
function ends up without __compact_unwind.
rdar://problem/18208653
llvm-svn: 218703
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
issues with dsymutil.
r218129 omits DW_TAG_subprograms which have no inlined subroutines when
emitting -gmlt data. This makes -gmlt very low cost for -O0 builds.
Darwin's dsymutil reasonably considers a CU empty if it has no
subprograms (which occurs with the above optimization in -O0 programs
without any force_inline function calls) and drops the line table, CU,
and everything in this situation, making backtraces impossible.
Until dsymutil is modified to account for this, disable this
optimization on Darwin to preserve the desired functionality.
(see r218545, which should be reverted after this patch, for other
discussion/details)
Footnote:
In the long term, it doesn't look like this scheme (of simplified debug
info to describe inlining to enable backtracing) is tenable, it is far
too size inefficient for optimized code (the DW_TAG_inlined_subprograms,
even once compressed, are nearly twice as large as the line table
itself (also compressed)) and we'll be considering things like Cary's
two level line table proposal to encode all this information directly in
the line table.
llvm-svn: 218702
|
| |
|
|
|
|
| |
refinement of an estimate. NFC.
llvm-svn: 218700
|
| |
|
|
|
|
|
|
| |
before using it.
<rdar://problem/18491391>
llvm-svn: 218699
|
| |
|
|
|
|
|
|
|
|
|
|
| |
It was hacky to use an opcode as a switch because it won't always match
(rsqrte != sqrte), and it looks like we'll need to add more special casing
per arch than I had hoped for. Eg, x86 will prefer a different NR estimate
implementation. ARM will want to use it's 'step' instructions. There also
don't appear to be any new estimate instructions in any arch in a long,
long time. Altivec vloge and vexpte may have been the first and last in
that field...
llvm-svn: 218698
|
| |
|
|
| |
llvm-svn: 218697
|
| |
|
|
|
|
| |
PRIVATE option is also an undocumented feature.
llvm-svn: 218696
|
| |
|
|
|
|
|
|
|
|
|
| |
MSDN doesn't say about /export:foo=bar style option, but
it turned out MSVC link.exe actually accepts that. So we need that
too.
It also means that the export directive in the module definition
file and /export command line option are functionally equivalent.
llvm-svn: 218695
|
| |
|
|
|
|
|
|
|
|
|
| |
In code-completion, don't assume there is a MacroInfo for everything,
since we aren't serializing the def corresponding to a later #undef in
the same module. Also setup the HadMacro bit correctly for undefs to
avoid an assertion failure.
rdar://18416901
llvm-svn: 218694
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Note: This version fixed an issue with the TBZ/TBNZ instructions that were
generated in FastISel. The issue was that the 64bit version of TBZ (TBZX)
automagically sets the upper bit of the immediate field that is used to specify
the bit we want to test. To test for any of the lower 32bits we have to first
extract the subregister and use the 32bit version of the TBZ instruction (TBZW).
Original commit message:
Teach selectBranch to fold bit test and branch into a single instruction (TBZ or
TBNZ).
llvm-svn: 218693
|
| |
|
|
|
|
|
|
| |
No tests for omod since nothing uses it yet, but
this should get rid of the remaining annoying trailing
zeros after some instructions.
llvm-svn: 218692
|
| |
|
|
|
|
| |
abs / neg are now part of the srcN_modifiers operands
llvm-svn: 218691
|
| |
|
|
|
|
|
|
| |
"__imp_" prefix always starts with double underscores.
When I was writing the original code I misunderstood
that it's "_imp_" on x64.
llvm-svn: 218690
|
| |
|
|
|
|
|
|
|
| |
Specifically, this also counts for stuff like (with style "inline"):
var x = function() {
return 1;
};
llvm-svn: 218689
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
r218624 implemented target inference for implicit special members. However,
other entities can be implicit - for example intrinsics. These can not have
inference running on them, so they should be marked host device as before. This
is the safest and most flexible setting, since by construction these functions
don't invoke anything, and we'd like them to be invokable from both host and
device code. LLVM's intrinsics definitions (where these intrinsics come from in
the case of CUDA/NVPTX) have no notion of target, so both host and device
intrinsics can be supported this way.
llvm-svn: 218688
|
| |
|
|
| |
llvm-svn: 218687
|
| |
|
|
|
|
| |
check-in.
llvm-svn: 218686
|
| |
|
|
| |
llvm-svn: 218685
|
| |
|
|
| |
llvm-svn: 218684
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
presence of deferred stop notification still pending.
There is a state transition that seems potentially buggy that I am capturing and
logging here, and including an explicit test to demonstrate expected behavior. See new test
for detailed description. Added logging around this area since, if we hit it, we
may have a usage bug, or a new state transition we really need to investigate.
This is around this scenario:
Thread C deferred stop notification awaiting thread A and thread B to stop.
Thread A stops.
Thread A requests resume.
Thread B stops.
Here we will explicitly signal the deferred stop notification after thread B
stops even though thread A is now resumed. Copious logging happens here.
llvm-svn: 218683
|
| |
|
|
| |
llvm-svn: 218682
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
Add numeric extend, trunctate to mips fast-isel
Reactivates D4827
Test Plan:
fpext.ll
loadstoreconv.ll
Reviewers: dsanders
Subscribers: mcrosier
Differential Revision: http://reviews.llvm.org/D5251
llvm-svn: 218681
|
| |
|
|
| |
llvm-svn: 218680
|
| |
|
|
| |
llvm-svn: 218679
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
to convert it into a shuffle.
Currently, the DAG Combiner only tries to convert type-legal build_vector nodes
into shuffles. This patch simply moves the logic that checks if a
build_vector has a legal value type up before we even start analyzing the
operands. This allows to early exit immediately from method
'visitBUILD_VECTOR' if the node type is known to be illegal.
No functional change intended.
llvm-svn: 218677
|
| |
|
|
|
|
| |
Test causes buildbot failures.
llvm-svn: 218676
|
| |
|
|
|
|
| |
does not work yet.
llvm-svn: 218675
|
| |
|
|
|
|
| |
And add a missing return in main, just in case.
llvm-svn: 218674
|