| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
|
|
|
|
|
|
| |
Summary: Add StreamExecutor kernel types.
Reviewers: jlebar, tra
Subscribers: parallel_libs-commits
Differential Revision: https://reviews.llvm.org/D23138
llvm-svn: 277827
|
|
|
|
|
|
|
|
| |
MSVC doesn't have an exact equivalent for __builtin_frame_address, but
_AddressOfReturnAddress() + sizeof(void*) should be equivalent for all
frames build with -fno-omit-frame-pointer.
llvm-svn: 277826
|
|
|
|
|
|
|
|
| |
Mask for value read from ebx register returned by CPUID expanded to 0xFFFF.
Differential Revision: https://reviews.llvm.org/D23203
llvm-svn: 277825
|
|
|
|
|
|
|
|
| |
wavefront result.
Differential Revision: http://reviews.llvm.org/D22934
llvm-svn: 277824
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There were two locations where fast-isel would generate a LFD instruction
with a target register class VSFRC instead of F8RC when VSX was enabled.
This can ccause invalid registers to be used in certain cases, like:
lfd 36, ...
instead of using a VSX load instruction. The wrong register number gets
silently truncated, causing invalid code to be generated.
The first place is PPCFastISel::PPCEmitLoad, which had multiple problems:
1.) The IsVSSRC and IsVSFRC flags are not initialized correctly, since they
are computed from resultReg, which is still zero at this point in many cases.
Fixed by changing the helper routines to operate on a register class instead
of a register and passing in UseRC.
2.) Even with this fixed, Is64VSXLoad is still wrong due to a typo:
bool Is32VSXLoad = IsVSSRC && Opc == PPC::LFS;
bool Is64VSXLoad = IsVSSRC && Opc == PPC::LFD;
The second line needs to use isVSFRC (like PPCEmitStore does).
3.) Once both the above are fixed, we're now generating a VSX instruction --
but an incorrect one, since generation of an indexed instruction with null
index is wrong. Fixed by copying the code handling the same issue in
PPCEmitStore.
The second place is PPCFastISel::PPCMaterializeFP, where we would emit an
LFD to load a constant from the literal pool, and use the wrong result
register class. Fixed by hardcoding a F8RC class even on systems
supporting VSX.
Fixes: https://llvm.org/bugs/show_bug.cgi?id=28630
Differential Revision: https://reviews.llvm.org/D22632
llvm-svn: 277823
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
Add instruction formats E, RSI, SSd, SSE, and SSF.
Added BRXH, BRXLE, PR, MVCK, STRAG, and ECTG instructions to test out
those formats.
Reviewers: uweigand
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D23179
llvm-svn: 277822
|
|
|
|
|
|
| |
nondeterminism.
llvm-svn: 277821
|
|
|
|
|
|
|
|
|
|
|
| |
This generated IR based on the order of evaluation, which is different
between GCC and Clang. With that in mind you get bootstrap miscompares
if you compare a Clang built with GCC-built Clang vs. Clang built with
Clang-built Clang. Diagnosing that made my head hurt.
This also reverts commit r277337, which "fixed" the test case.
llvm-svn: 277820
|
|
|
|
| |
llvm-svn: 277819
|
|
|
|
|
|
|
|
| |
This patch is to implement sema and parsing for 'teams distribute' pragma.
Differential Revision: https://reviews.llvm.org/D23189
llvm-svn: 277818
|
|
|
|
|
|
|
|
| |
value type directly instead of via the input value type.
Preparation for adding 2 input support so we want to avoid unnecessary references to the input value type.
llvm-svn: 277817
|
|
|
|
| |
llvm-svn: 277816
|
|
|
|
|
|
|
|
| |
size calculations. NFCI.
Preparation for adding 2 input support so we want to avoid unnecessary references to the input value type.
llvm-svn: 277814
|
|
|
|
|
|
|
|
|
| |
0x7FF4000000000000".
We should not use double (or float) in the LLVM, unless it is really needed. x87 FP register doesn't preserve SNaN to move the value.
FIXME: APFloat() may have the constructor by raw bit.
llvm-svn: 277813
|
|
|
|
| |
llvm-svn: 277812
|
|
|
|
|
|
|
|
|
|
| |
Reviewers: bkramer
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D23199
llvm-svn: 277811
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When entering the dependence computation and the max_operations is set, the
operations counter may have already exceeded the counter, thus aborting any ISL
computation from the start. The counter is reset at the end of the dependence
calculation such that a follow-up recomputation might succeed, ie. the success
of the first dependence calculation depends on unrelated ISL operations that
happened before, giving it a disadvantage to the following calculations.
This patch resets the operations counter at the beginning of the dependence
recalculation to not depend on previous actions. Otherwise additional
preprocessing of the Scop that aims to improve its schedulability (eg. DeLICM)
do have the effect that DependenceInfo and hence the scheduling fail more
likely, contraproductive to the goal of said preprocessing.
llvm-svn: 277810
|
|
|
|
| |
llvm-svn: 277809
|
|
|
|
|
|
| |
Added matchBinaryPermuteVectorShuffle and moved the blend+zero and insertps matching code into it.
llvm-svn: 277808
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
a registry"
This differs from the previous version by being more careful about template
instantiation/specialization in order to prevent errors when building with
clang -Werror. Specifically:
* begin is not defined in the template and is instead instantiated when Head
is. I think the warning when we don't do that is wrong (PR28815) but for now
at least do it this way to avoid the warning.
* Instead of performing template specializations in LLVM_INSTANTIATE_REGISTRY
instead provide a template definition then do explicit instantiation. No
compiler I've tried has problems with doing it the other way, but strictly
speaking it's not permitted by the C++ standard so better safe than sorry.
Original commit message:
Currently the Registry class contains the vestiges of a previous attempt to
allow plugins to be used on Windows without using BUILD_SHARED_LIBS, where a
plugin would have its own copy of a registry and export it to be imported by
the tool that's loading the plugin. This only works if the plugin is entirely
self-contained with the only interface between the plugin and tool being the
registry, and in particular this conflicts with how IR pass plugins work.
This patch changes things so that instead the add_node function of the registry
is exported by the tool and then imported by the plugin, which solves this
problem and also means that instead of every plugin having to export every
registry they use instead LLVM only has to export the add_node functions. This
allows plugins that use a registry to work on Windows if
LLVM_EXPORT_SYMBOLS_FOR_PLUGINS is used.
llvm-svn: 277806
|
|
|
|
|
|
|
|
|
|
| |
This patch fixes passing long double type arguments to function in
soft float mode. If there is less than 4 argument registers free
(long double type is mapped in 4 gpr registers in soft float mode)
long double type argument must be passed through stack.
Differential Revision: https://reviews.llvm.org/D20114.
llvm-svn: 277804
|
|
|
|
|
|
|
|
|
|
| |
Before this commit we generated the array type in reverse order and we also
added the outermost dimension size to the new array declaration, which is
incorrect as Polly additionally assumed an additional unsized outermost
dimension, such that we had an off-by-one error in the linearization of access
expressions.
llvm-svn: 277802
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
Turn (select C, (sext A), B) into (sext (select C, A, B')) when A is i1 and
B is a compatible constant, also for zext instead of sext. This will then be
further folded into logical operations.
The transformation would be valid for non-i1 types as well, but other parts of
InstCombine prefer to have sext from non-i1 as an operand of select.
Motivated by the shader compiler frontend in Mesa for AMDGPU, which emits i32
for boolean operations. With this change, the boolean logic is fully
recovered.
Reviewers: majnemer, spatel, tstellarAMD
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D22747
llvm-svn: 277801
|
|
|
|
| |
llvm-svn: 277800
|
|
|
|
|
|
|
|
| |
These annotations ensure that the NVIDIA PTX assembler limits the number of
registers used such that we can be certain the resulting kernel can be executed
for the number of threads in a thread block that we are planning to use.
llvm-svn: 277799
|
|
|
|
|
|
|
|
| |
Reviewers: kcc
Differential Revision: https://reviews.llvm.org/D23190
llvm-svn: 277798
|
|
|
|
|
|
|
|
|
| |
When the type being diffed is a type alias, and the orginal type is not a
templated type, then there will be no unsugared TemplateSpecializationType.
When this happens, exit early from the constructor. Also add assertions to
the other iterator accessor to prevent the iterator from being used.
llvm-svn: 277797
|
|
|
|
|
|
|
|
| |
Silence the -Wbitfield-constant-conversion warning for when -1 or other
negative values are assigned to unsigned bitfields, provided that the bitfield
is wider than the minimum number of bits needed to encode the negative value.
llvm-svn: 277796
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
Often, a code will call multiple virtual methods of a given object.
If they go in a linear block, it should be possible to check vtable
before the first call, then store vtable pointer and reuse it for
the second vcall without any additional checks.
This is expected to have a positive performance impact on a hot
path in Blink, see https://crbug.com/634139.
Reviewers: kcc
Differential Revision: https://reviews.llvm.org/D23151
llvm-svn: 277795
|
|
|
|
| |
llvm-svn: 277794
|
|
|
|
| |
llvm-svn: 277793
|
|
|
|
| |
llvm-svn: 277792
|
|
|
|
| |
llvm-svn: 277791
|
|
|
|
|
|
| |
Because this function depends only on its arguments.
llvm-svn: 277790
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also re-write how most of the directory indexing is done - as it has
grown over the years, it has become a bit of a mess and was overdue
for a cleanup.
Most importantly, this allows you to specify a directory with the
platform.plugin.darwin-kernel.kext-directories setting and now lldb
will search for kexts and kernels in those directories recursively.
<rdar://problem/20754467>
llvm-svn: 277789
|
|
|
|
|
|
|
|
|
|
|
|
| |
Followup from r277778, after Mehdi's comments.
Expand %ld64 to perform the necessary preload instead, that way new
tests do not need to worry about setting up DYLD_INSERT_LIBRARIES
themselves.
rdar://problem/24300926
llvm-svn: 277788
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
For memset (and others) we can get diagnostics like:
struct stat { int x; };
void foo(struct stat *stamps) {
bzero(stamps, sizeof(stamps));
memset(stamps, 0, sizeof(stamps));
}
t.c:7:28: warning: 'memset' call operates on objects of type 'struct stat' while the size is based on a different type 'struct stat *' [-Wsizeof-pointer-memaccess]
memset(stamps, 0, sizeof(stamps));
~~~~~~ ^~~~~~
t.c:7:28: note: did you mean to dereference the argument to 'sizeof' (and multiply it by the number of elements)?
memset(stamps, 0, sizeof(stamps));
^~~~~~
This patch implements the same class of warnings for bzero.
Differential Revision: https://reviews.llvm.org/D22525
rdar://problem/18963514
llvm-svn: 277787
|
|
|
|
| |
llvm-svn: 277786
|
|
|
|
|
|
|
|
|
|
| |
The patch splits a complex && if condition into easier to read and understand
logic. That wrong early exit condition was letting some instructions with not
all operands available pass through when HoistingGeps was true.
Differential Revision: https://reviews.llvm.org/D23174
llvm-svn: 277785
|
|
|
|
|
|
|
|
|
|
| |
Add a generalized IRBuilderCallbackInserter, which is just given a
callback to execute after insertion. This can be used to get rid of
the custom inserter in InstCombine, which will in turn allow me to add
target specific InstCombineCalls API for intrinsics without horrible
layering violations.
llvm-svn: 277784
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary: The corresponding LLVM change is D23146.
Reviewers: dblaikie, chandlerc
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D23147
llvm-svn: 277783
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Shifts with a uniform but non-constant count were considered very expensive to
vectorize, because the splat of the uniform count and the shift would tend to
appear in different blocks. That made the splat invisible to ISel, and we'd
scalarize the shift at codegen time.
Since r201655, CodeGenPrepare sinks those splats to be next to their use, and we
are able to select the appropriate vector shifts. This updates the cost model to
to take this into account by making shifts by a uniform cheap again.
Differential Revision: https://reviews.llvm.org/D23049
llvm-svn: 277782
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
The comparator function to compare input sections as instructed by
SORT command was a bit too complicated because it needed to handle
four different cases. This patch split it into two function calls.
This patch also simplifies the parser.
Reviewers: grimar
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D23140
llvm-svn: 277780
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
constant vectors
This concludes the splat vector enhancements for foldICmpEqualityWithConstant().
Other commits in this series:
https://reviews.llvm.org/rL277762
https://reviews.llvm.org/rL277752
https://reviews.llvm.org/rL277738
https://reviews.llvm.org/rL277731
https://reviews.llvm.org/rL277659
https://reviews.llvm.org/rL277638
https://reviews.llvm.org/rL277629
llvm-svn: 277779
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Green Dragon's darwin stage2 asan bot fails on some checks:
http://lab.llvm.org:8080/green/job/clang-stage2-cmake-RgSan_check
test/tools/lto/hide-linkonce-odr.ll
test/tools/lto/opt-level.ll
ERROR: Interceptors are not working. This may be because
AddressSanitizer is loaded too late (e.g. via dlopen)
To fix this, %ld64 needs to load 'libclang_rt.asan_osx_dynamic.dylib'
before libLTO.dylib, via DYLD_INSERT_LIBRARIES. This won't work by
updating config.environment, since some shim binary in the way scrubs
the env vars. Instead, provide the path to this lib through %asanrtlib,
which can then be used by tests directly with DYLD_INSERT_LIBRARIES.
rdar://problem/24300926
llvm-svn: 277778
|
|
|
|
|
|
|
|
| |
These are meant to only be included on certain targets. This only disables it
for Windows ARM for now. Ideally these would be conditionally included as
appropriate.
llvm-svn: 277777
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
is not a nullptr
when we are pointed at real data.
David Blaikie pointed out some odd logic in the case the Err value was a nullptr and
Lang Hames suggested it could be cleaned it up with an assert to know that Err is
not a nullptr when we are pointed at real data. As only in the case of constructing
the sentinel value by pointing it at null data is Err is permitted to be a nullptr,
since no error could occur in that case.
With this change the testing for “if (Err)” is removed from the constructor’s logic
and *Err is used directly without any check after the assert().
llvm-svn: 277776
|
|
|
|
|
|
|
| |
These are the operations that are trivially identical. Division is omitted for
now because you need to use the correct sign/zero extension.
llvm-svn: 277775
|
|
|
|
| |
llvm-svn: 277774
|
|
|
|
|
|
|
|
|
| |
PR28848 had a very nice reduction of the underlying cause of the bug.
Our ValueMap had, in an entry for an Instruction, a ConstantInt.
This is not at all unexpected but should be handled properly.
llvm-svn: 277773
|