Get rid of all fixmes and base heuristic on `num-clustered-dwords`. The main intuition behind this is as
follows. The existing heuristic roughly summarizes as below:
* Assume, all the mem ops instructions participating in the clustering process, loads/stores same num bytes
* If num bytes loaded by each mem op is 4 bytes, then cluster at max 5 mem ops, that is at max 20 bytes
* If num bytes loaded by each mem op is 8 bytes, then cluster at max 3 mem ops, that is at max 24 bytes
* If num bytes loaded by each mem op is 16 bytes, then cluster at max 2 mem ops, that is at max 32 bytes
So, we need to make sure that the new heuristic do not completey deviate away from the above one, and it
properly handles both the sub-word loads and the wide loads.
Reviewed By: arsenm, rampitec
Differential Revision: https://reviews.llvm.org/D84354
Not matching the (real) variadic declaration makes the interceptor take garbage inputs on Darwin/AArch64.
Differential Revision: https://reviews.llvm.org/D84570
Both tests define clang::tidy::test::TestCheck::registerMatchers().
This is UB and causes linker to sometimes choose the wrong overload.
Put classes into anonymous namespaces to avoid the problem.
Differential Revision: https://reviews.llvm.org/D84902
In cases where the alignment of the datatype is smaller than
expected by the instruction, the address is aligned. The aligned
address is used for the load, but wasn't used for the store
conditional, which resulted in a run-time alignment exception.
We parse .arch so that some `.arch i386; .code32` code can assemble. It seems
that X86AsmParser does not do a good job tracking what features are needed to
assemble instructions. GNU as's x86 port supports a very wide range of .arch
operands. Ignore the operand for now.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D84900
When we use the OpenMPIRBuilder for the parallel region we need to also
use it to get the thread ID (among other things) in the body. This is
because CGOpenMPRuntime::getThreadID() and
CGOpenMPRuntime::emitUpdateLocation implicitly assumes that if they are
called from within a parallel region there is a certain structure to the
code and certain members of the OMPRegionInfo are initialized. It might
make sense to initialize them even if we use the OpenMPIRBuilder but we
would preferably get rid of such state instead.
Bug reported by Anchu Rajendran Sudhakumari.
Depends on D82470.
Reviewed By: anchu-rajendran
Differential Revision: https://reviews.llvm.org/D82822
We need to keep track of the alloca insertion point (which we already
communicate via the callback to the user) as we place allocas as well.
Reviewed By: fghanim, SouraVX
Differential Revision: https://reviews.llvm.org/D82470
Need to map the base pointer for all directives, not only target
data-based ones.
The base pointer is mapped for array sections, array subscript, array
shaping and other array-like constructs with the base pointer. Also,
codegen for use_device_ptr clause was modified to correctly handle
mapping combination of array like constructs + use_device_ptr clause.
The data for use_device_ptr clause is emitted as the last records in the
data mapping array.
Reviewed By: ye-luo
Differential Revision: https://reviews.llvm.org/D84767
Temporarily disable IncrementalProcessingTest partially until the static
initialization implementation on AIX is recovered.
Differential Revision: https://reviews.llvm.org/D84880
The operand to these instructions is both input and output.
These are not yet emitted by the compiler and the assembler already
works fine, so can't test in this patch. But D75044 will use XPACI
and provide test coverage for this patch as well.
Differential Revision: https://reviews.llvm.org/D84298
To make it easier for lowering to identify which symbols from the host
are captured by internal subprograms, create HostAssocDetails for them.
In particular, if a symbol is referenced and it is contained in a
subprogram or main program that is not the same as the containing
program unit of the reference, a HostAssocDetails symbol is created
in the current scope.
Differential Revision: https://reviews.llvm.org/D84889
Need to map the base pointer for all directives, not only target
data-based ones.
The base pointer is mapped for array sections, array subscript, array
shaping and other array-like constructs with the base pointer. Also,
codegen for use_device_ptr clause was modified to correctly handle
mapping combination of array like constructs + use_device_ptr clause.
The data for use_device_ptr clause is emitted as the last records in the
data mapping array.
It applies only for global pointers.
Differential Revision: https://reviews.llvm.org/D84767
Summary:
This patch implements -ffunction-sections on AIX.
This patch focuses on assembly generation.
Follow-on patch needs to handle:
1. -ffunction-sections implication for jump table.
2. Object file generation path and associated testing.
Differential Revision: https://reviews.llvm.org/D83875
This removes some unneeded block masks when we don't have any
reductions. It should not have any effect on codegen as the values
created are dead anyway.
Differential Revision: https://reviews.llvm.org/D81415
Now that we can have a memref of index type, we no longer need to materialize shapes in i64 and then index_cast.
Differential Revision: https://reviews.llvm.org/D84938
-- Introduces a pass that normalizes the affine layout maps to the identity layout map both within and across functions by rewriting function arguments and call operands where necessary.
-- Memref normalization is now implemented entirely in the module pass '-normalize-memrefs' and the limited intra-procedural version has been removed from '-simplify-affine-structures'.
-- Run using -normalize-memrefs.
-- Return ops are not handled and would be handled in the subsequent revisions.
Signed-off-by: Abhishek Varma <abhishek.varma@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D84490
The intrinsic lowering facility is based on the generic intrinsic names to avoid
duplicating implementations. Specific intrinsics call are re-written to call to
the generic versions by the front-end but this cannot be done when specific intrinsics
are passed as arguments (the rewrite would give illegal/ambiguous unparsed Fortran).
Solve the issue by making the specific to generic name mapping accessible to lowering
and can be later used to generate the unrestricted intrinsic functions.
Reviewed By: schweitz
Differential Revision: https://reviews.llvm.org/D84842
When lowering to the standard dialect, we currently support only the extent
tensor variant of the shape.rank operation. This change lets the conversion
pattern fail in a well-defined manner.
Differential Revision: https://reviews.llvm.org/D84852
As far as I know, ipconstprop has not been used in years and ipsccp has
been used instead. This has the potential for confusion and sometimes
leads people to spend time finding & reporting bugs as well as
updating it to work with the latest API changes.
This patch moves the tests over to SCCP. There's one functional difference
I am aware of: ipconstprop propagates for each call-site individually, so
for functions that are called with different constant arguments it can sometimes
produce better results than ipsccp (at much higher compile-time cost).But
IPSCCP can be thought to do so as well for internal functions and as mentioned
earlier, the pass seems unused in practice (and there are no plans on working
towards enabling it anytime).
Also discussed on llvm-dev:
http://lists.llvm.org/pipermail/llvm-dev/2020-July/143773.html
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D84447
As long as we can extract the lowest 128-bit subvector from the pre-truncated source vector, then we don't care what size it is.
The next stage will be to support non-zero extraction indices, as long as its still coming from the lowest 128-bit subvector.
This is the last missing bit in the core remote index implementation. The only
remaining bits are some API refactorings (replacing Optional with Expected and
being better at reporting errors).
Reviewed By: kadircet
Differential Revision: https://reviews.llvm.org/D84894
Let's just return a std::string to make this safe. formatv seemed overkill for formatting
the return values as they all just append an integer value to a constant string.
Reviewed By: labath
Differential Revision: https://reviews.llvm.org/D84505
Without this patch the word occurrence search always returns the first token of the file.
Despite of that, `findNeardyIdentifier()` returns the correct result (but inefficently) until there are several matched tokens with the same value `floor(log2(<token line> - <word line>))` (e.g. several matched tokens on the same line).
Reviewed By: kadircet
Differential Revision: https://reviews.llvm.org/D84912
This patch makes the 'Length' field of the address range table optional.
Reviewed By: jhenderson
Differential Revision: https://reviews.llvm.org/D84911
This patch makes the 'AddressSize' and 'SegmentSelectorSize' fields of
address range table optional.
Reviewed By: jhenderson
Differential Revision: https://reviews.llvm.org/D84907
Ordering of options isn't important so an `llvm::StringMap` is a much better container for this purpose.
Reviewed By: gribozavr2
Differential Revision: https://reviews.llvm.org/D84868
Summary:
Currently the binaries are output directly into the bin subdirectory of the
build directory. This doesn't work correctly with multi-config generators which
should output the binaries into <CONFIG_NAME>/bin instead.
Reviewers: sscalpone, richard.barton.arm
Subscribers: mgorny, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D84022
As requested in the review, this patch removes the additional conditions in
the `COMPILER_RT_HAS_VERSION_SCRIPT` tests.
Tested on `amd64-pc-solaris2.11` and `x86_64-pc-linux-gnu`.
Differential Revision: https://reviews.llvm.org/D84559
This patch introduces new intrinsics in LLVM dialect:
- `llvm.intr.floor`
- `llvm.intr.maxnum`
- `llvm.intr.minnum`
- `llvm.intr.smax`
- `llvm.intr.smin`
These intrinsics correspond to SPIR-V ops from GLSL
extended instruction set (`spv.GLSL.Floor`, `spv.GLSL.FMax`,
`spv.GLSL.FMin`, `spv.GLSL.SMax` and `spv.GLSL.SMin`
respectively). Also conversion patterns for them were added.
Reviewed By: antiagainst
Differential Revision: https://reviews.llvm.org/D84661
Summary:
In the phi-node-elimination pass, we set the killed flag incorrectly.
When we eliminate the PHI node, we replace the PHI with a copy for the
incoming value.
Before this patch, we will set incoming value as killed(PHICopy). And
we will remove the killed flag from last using incoming value(OldKill).
This is correct, only if the new PHICopy is after the OldKill.
Reviewed By: bjope
Differential Revision: https://reviews.llvm.org/D80886
This is a second patch on conversion of GLSL ops to LLVM dialect.
It introduces patterns to convert `spv.InverseSqrt` and `spv.Tanh`.
Reviewed By: antiagainst
Differential Revision: https://reviews.llvm.org/D84633
The uniqueing decl in PathDiagnostic is the declaration with the
uniqueing loc, as stated by documentation comments.
It is enough to include the uniqueing loc in the profile. It is possible
to have objects with different uniqueing decl but same location, at
least with templates. These belong to the same class and should have
same profile.
Reviewed By: vsavchenko, NoQ
Differential Revision: https://reviews.llvm.org/D84843
hsmahesha