Summary:
There are few oddities that occur due to v1i1, v8i1, v16i1 being legal without v2i1 and v4i1 being legal when we don't have VLX. Particularly during legalization of v2i32/v4i32/v2i64/v4i64 masked gather/scatter/load/store. We end up promoting the mask argument to these during type legalization and then have to widen the promoted type to v8iX/v16iX and truncate it to get the element size back down to v8i1/v16i1 to use a 512-bit operation. Since need to fill the upper bits of the mask we have to fill with 0s at the promoted type.
It would be better if we could just have the v2i1/v4i1 types as legal so they don't undergo any promotion. Then we can just widen with 0s directly in a k register. There are no real v4i1/v2i1 instructions anyway. Everything is done on a larger register anyway.
This also fixes an issue that we couldn't implement a masked vextractf32x4 from zmm to xmm properly.
We now have to support widening more compares to 512-bit to get a mask result out so new tablegen patterns got added.
I had to hack the legalizer for widening the operand of a setcc a bit so it didn't try create a setcc returning v4i32, extract from it, then try to promote it using a sign extend to v2i1. Now we create the setcc with v4i1 if the original setcc's result type is v2i1. Then extract that and don't sign extend it at all.
There's definitely room for improvement with some follow up patches.
Reviewers: RKSimon, zvi, guyblank
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D41560
llvm-svn: 321967
This field is defined as kmp_int32, so we should use neither
pointers to kmp_int64 nor 64 bit atomic instructions.
(Found while testing on a Raspberry Pi, 32 bit ARM)
Differential Revision: https://reviews.llvm.org/D41656
llvm-svn: 321964
Summary:
After rL319736 for D28253 (which fixes PR28929), gcc cannot compile `<memory>` anymore in pre-C+11 modes, complaining:
```
In file included from /usr/include/c++/v1/memory:648:0,
from test.cpp:1:
/usr/include/c++/v1/memory: In static member function 'static std::__1::shared_ptr<_Tp> std::__1::shared_ptr<_Tp>::make_shared(_A0&, _A1&, _A2&)':
/usr/include/c++/v1/memory:4365:5: error: wrong number of template arguments (4, should be at least 1)
static_assert((is_constructible<_Tp, _A0, _A1, _A2>::value), "Can't construct object in make_shared" );
^
In file included from /usr/include/c++/v1/memory:649:0,
from test.cpp:1:
/usr/include/c++/v1/type_traits:3198:29: note: provided for 'template<class _Tp, class _A0, class _A1> struct std::__1::is_constructible'
struct _LIBCPP_TEMPLATE_VIS is_constructible
^~~~~~~~~~~~~~~~
In file included from /usr/include/c++/v1/memory:648:0,
from test.cpp:1:
/usr/include/c++/v1/memory:4365:5: error: template argument 1 is invalid
static_assert((is_constructible<_Tp, _A0, _A1, _A2>::value), "Can't construct object in make_shared" );
^
/usr/include/c++/v1/memory: In static member function 'static std::__1::shared_ptr<_Tp> std::__1::shared_ptr<_Tp>::allocate_shared(const _Alloc&, _A0&, _A1&, _A2&)':
/usr/include/c++/v1/memory:4444:5: error: wrong number of template arguments (4, should be at least 1)
static_assert((is_constructible<_Tp, _A0, _A1, _A2>::value), "Can't construct object in allocate_shared" );
^
In file included from /usr/include/c++/v1/memory:649:0,
from test.cpp:1:
/usr/include/c++/v1/type_traits:3198:29: note: provided for 'template<class _Tp, class _A0, class _A1> struct std::__1::is_constructible'
struct _LIBCPP_TEMPLATE_VIS is_constructible
^~~~~~~~~~~~~~~~
In file included from /usr/include/c++/v1/memory:648:0,
from test.cpp:1:
/usr/include/c++/v1/memory:4444:5: error: template argument 1 is invalid
static_assert((is_constructible<_Tp, _A0, _A1, _A2>::value), "Can't construct object in allocate_shared" );
^
```
This is also reported in https://bugs.freebsd.org/224946 (FreeBSD is apparently one of the very few projects that regularly builds programs against libc++ with gcc).
The reason is that the static assertions are invoking `is_constructible` with three arguments, while gcc does not have the built-in `is_constructible` feature, and the pre-C++11 `is_constructible` wrappers in `<type_traits>` only provide up to two arguments.
I have added additional wrappers for three arguments, modified the `is_constructible` entry point to take three arguments instead, and added a simple test to is_constructible.pass.cpp.
Reviewers: EricWF, mclow.lists
Reviewed By: EricWF
Subscribers: krytarowski, cfe-commits, emaste
Differential Revision: https://reviews.llvm.org/D41805
llvm-svn: 321963
Another small step forward to move VPlan stuff outside of LoopVectorize.cpp.
VPlanBuilder.h is renamed to LoopVectorizationPlanner.h
LoopVectorizationPlanner class is moved from LoopVectorize.cpp to
LoopVectorizationPlanner.h LoopVectorizationCostModel::VectorizationFactor
class is moved to LoopVectorizationPlanner.h (used by the planner class) ---
this needs further streamlining work in later patches and thus all I did was
take it out of the CostModel class and moved to the header file. The callback
function had to stay inside LoopVectorize.cpp since it calls an
InnerLoopVectorizer member function declared in it. Next Steps: Make
InnerLoopVectorizer, LoopVectorizationCostModel, and other classes more modular
and more aligned with VPlan direction, in small increments.
Previous step was: r320900 (https://reviews.llvm.org/D41045)
Patch by Hideki Saito, thanks!
Differential Revision: https://reviews.llvm.org/D41420
llvm-svn: 321962
In addition to target-dependent attributes, we can also preserve a
white-listed subset of target independent function attributes. The white-list
excludes problematic attributes, most prominently:
* attributes related to memory accesses, as alloca instructions
could be moved in/out of the extracted block
* control-flow dependent attributes, like no_return or thunk, as the
relerelevant instructions might or might not get extracted.
Thanks @efriedma and @aemerson for providing a set of attributes that cannot be
propagated.
Reviewers: efriedma, davidxl, davide, silvas
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D41334
llvm-svn: 321961
Summary:
I believe legalization is really expecting that ReplaceNodeResults will return something with the same type as the thing that's being legalized. Ultimately, it uses the output to replace the uses in the DAG so the type should match to make that work.
There are two relevant cases here. When crbits are enabled, then i1 is a legal type and getSetCCResultType should return i1. In this case, the truncate will be between i1 and i1 and should be removed (SelectionDAG::getNode does this). Otherwise, getSetCCResultType will be i32 and the legalizer will promote the truncate to be i32 -> i32 which will be similarly removed.
With this fixed we can remove some code from PromoteIntRes_SETCC that seemed to only exist to deal with the intrinsic being replaced with a larger type without changing the other operand. With the truncate being used for connectivity this doesn't happen anymore.
Reviewers: hfinkel
Reviewed By: hfinkel
Subscribers: nemanjai, llvm-commits, kbarton
Differential Revision: https://reviews.llvm.org/D41654
llvm-svn: 321959
The instructions that load 64-bits or an xmm register should be TB_NO_REVERSE to avoid the load being widened during unfold. The instructions that load 128-bits need to ensure 128-bit alignment.
llvm-svn: 321956
The memory form of the xmm->xmm version only writes 64-bits. If we use it in the folding tables and its get used for a stack spill, only half the slot will be written. Then a reload may read all 128-bits which will pull in garbage. But without the spill the upper bits of the register would have been zero. By not folding we would preserve the zeros.
llvm-svn: 321950
These just overloads for _Float128. They're supported by GCC 7 and used
by glibc. APFloat support is already there so just add the overloads.
__builtin_copysignf128
__builtin_fabsf128
__builtin_huge_valf128
__builtin_inff128
__builtin_nanf128
__builtin_nansf128
This is the same support that GCC has, according to the documentation,
but limited to _Float128.
llvm-svn: 321948
We don't do fine grained feature control like this on features prior to AVX512.
We do still have checks in place in the assembly parser itself that prevents %zmm references or %xmm16-31 from being parsed without at least -mattr=avx512f. Same for rounding control and mask operands. That will prevent the table matcher from matching for any instructions that need those features and that's probably good enough.
llvm-svn: 321947
If the varargs are not accessed by a function, we can inline the
function.
Reviewers: dblaikie, chandlerc, davide, efriedma, rnk, hfinkel
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D41335
llvm-svn: 321940
In the minimal case, this won't remove instructions, but it still improves
uses of existing values.
In the motivating example from PR35834, it does remove instructions, and
sets that case up to be optimized by something like D41603:
https://reviews.llvm.org/D41603
llvm-svn: 321936
This is the last step needed to fix PR33325:
https://bugs.llvm.org/show_bug.cgi?id=33325
We're trading branch and compares for loads and logic ops.
This makes the code smaller and hopefully faster in most cases.
The 24-byte test shows an interesting construct: we load the trailing scalar
elements into vector registers and generate the same pcmpeq+movmsk code that
we expected for a pair of full vector elements (see the 32- and 64-byte tests).
Differential Revision: https://reviews.llvm.org/D41714
llvm-svn: 321934
Use full PATH when looking up test tools rather than just llvm tools
directory. r320813 has added a lookup for 'lldb-test' which is part
of LLDB tools rather than LLVM, and therefore is not present
in llvm_tools_dir before LLDB is installed.
While technically we could introduce separate per-directory lookup
logic, there is no real reason not to use the PATH formed earlier here,
and this is what other tools are doing.
Differential Revision: https://reviews.llvm.org/D41726
llvm-svn: 321932
Any hashing for methods should be able to compile this test case without
emitting an error. Since the class and method come from the same header from
each module, there should be no messages about ODR violations.
llvm-svn: 321924
Reviewed as https://reviews.llvm.org/D41748
* These tests use function objects from functional, back_inserter from iterator, and equal from algorithm, so add those headers.
* The use of iota targeting vector<unsigned char> with an int parameter triggers warnings on MSVC++ assigning an into a unsigned char&; so change the parameter to unsigned char with a static_cast.
* Avoid naming unary_function in identity here as that is removed in '17. (This also fixes naming _VSTD, _NOEXCEPT_, and other libcxx-isms)
* Change the predicate in the transform tests to add_ten so that problems with multiple application are caught.
llvm-svn: 321922
Attempting to recompute it are doomed to fail because the IDNS of a declaration
is not necessarily preserved across serialization and deserialization (in turn
because whether a friend declaration is visible depends on whether some prior
non-friend declaration exists).
llvm-svn: 321921
Craig Topper