- Add a parameter to the `Assign` function to be able to use a different
`memmove` function. This is preparatory work to be able to use the
`Assign` function between host and device data.
- Expose the `Assign` function so it can be used from different files.
- The new `memmoveFct` is not used in `BlankPadCharacterAssignment` yet
since it is not clear if there is a need. It will be updated in case it
is needed.
Passing a descriptor as a `const Descriptor &` or a `const Descriptor *`
generates a FIR signature where the box is passed by value.
This is an issue, as it requires a load of the box to be passed. But
since, ultimately, all boxes are passed by reference a temporary is
generated in LLVM and the reference to the temporary is passed.
The boxes addresses are registered with the CUDA runtime but the
temporaries are not, thus preventing the runtime to properly map a host
side address to its device side counterpart.
To address this issue, this PR changes the signatures to the transfer
functions to pass a descriptor as a `Descriptor *`, which will in turn
generate a FIR signature with that takes a box reference as an argument.
Issue deprecation warning for these directives.
Lowering currently supports parallel master, for all other combined or
composite directives involving master, issue TODO errors.
Note: The first commit changes the formatting and generalizes the
deprecation message emission for reuse in the second commit. I can pull
it out into a separate commit if required.
Parse the locator list in OmpDependClause as an OmpObjectList (instead
of a list of Designators). When a common block appears in the locator
list, show an informative message.
Implement resolving symbols in DependSinkVec in a dedicated visitor
instead of having a visitor for OmpDependClause.
Resolve unresolved names common blocks in OmpObjectList.
Minor changes to the code organization:
- rename OmpDependenceType to OmpTaskDependenceType (to follow 5.2
terminology),
- rename Depend::WithLocators to Depend::DepType,
- add comments with more detailed spec references to parse-tree.h.
---------
Co-authored-by: Kiran Chandramohan <kiran.chandramohan@arm.com>
nsw is now added to do-variable increment when -fno-wrapv is enabled as
GFortran seems to do.
That means the option introduced by #91579 isn't necessary any more.
Note that the feature of -flang-experimental-integer-overflow is enabled
by default.
Implement parsing of the AFFINITY clause on TASK construct, conversion
from the parser class to omp::Clause.
Lowering to HLFIR is unsupported, a TODO message is displayed.
Define `OmpIteratorSpecifier` and `OmpIteratorModifier` parser classes,
and add parsing for them. Those are reusable between any clauses that
use iterator modifiers.
Add support for iterator modifiers to the MAP clause up to lowering,
where a TODO message is emitted.
See https://reviews.llvm.org/D157626 for the rational of declare having
side effects.
The write effect is to scary for passes that look for read/write effects
without caring about the resource affected. I know Slava asked for it,
but I think the creation of the `DebuggingResource` was enough and that
a write is too much. The alloca effect is sufficient to prevent DCE to
remove it, which is all we care about currently.
This currently is flag as a reason for creating LHS temporary in
assignment to vector subscripted entity with array constructor.
There is a lot of read/write side effect analysis in the
"lower-hlfir-ordered-assignments" pass, and I feel like we will just
keep adding weird "debug ressource" bypassing here and there with these
side effects.
hlfir.assign currently has the `MemoryEffects<[MemWrite]` which makes it
look like it can write to anything. This is good for some cases where
the assign effect cannot be precisely described through the MLIR side
effect API (e.g., when the LHS is a descriptor and it is not possible to
get an OpOperand describing the data address, or when derived type are
involved and finalization could be called, or user defined assignment
for some components). For the most common case of hlfir.assign on
intrinsic types without whole allocatable LHS, this is pessimistic.
This patch implements a finer description of the side effects when
possible, and also adds the proper read/allocate/free effects when
relevant.
The ultimate goal is to suppress the generation of temporary for the LHS
address when dealing with an assignment to a vector subscripted LHS
where the vector subscript is an array constructor that does not refer
to the LHS (as in `x([a,b]) = y`).
Two more patches will follow to enable this.
The convention is to use enum names that match the source spelling (up
to upper/lower case), including names with underscores.
Remove the special case from unparser, update tests.
For consistency with other dialects and other CUF passes and files, this
patch renames passes CufOpConversion to CUFOpConversion,
CufImplicitDeviceGlobal to CUFDeviceGlobal.
It also renames the file.
Fix missing extern templates for llvm::Registry use in other projects of
llvm
Windows doesn't implicitly import and merge exported symbols across
shared libraries
like Linux does so we need to explicitly export/import each
instantiation of llvm::Registry.
Updated LLVM_INSTANTIATE_REGISTRY to just be a full explicit template
instantiation.
This is part of the work to enable LLVM_BUILD_LLVM_DYLIB and LLVM
plugins on window.
Flang generates many globals to handle derived types. There was a check
in debug info to filter them based on the information that their names
start with a period. This changed since PR#104859 where 'X' is being
used instead of '.'.
This PR fixes this issue by also adding 'X' in that list. As user
variables gets lower cased by the NameUniquer, there is no risk that
those will be filtered out. I added a test for that to be sure.
This PR adds an OpenMP dialect related pass for FIR/HLFIR which creates
`MapInfoOp` instances for certain privatized symbols. For example, if an
allocatable variable is used in a private clause attached to a
`omp.target` op, then the allocatable variable's descriptor will be
needed on the device (e.g. GPU). This descriptor needs to be separately
mapped onto the device. This pass creates the necessary `omp.map.info`
ops for this.
getElementType() was missing from Sequence and Vector types. Did a
replace of the obvious places getEleTy() was used for these two types
and updated to use this name instead.
Co-authored-by: Scott Manley <scmanley@nvidia.com>
Remove 'if std::string' that is covered by another branch of the
if-statement.
Add printing of 'bool' and 'int' values, since they have corresponding
`GetNodeName` definitions.
I just introduced a dependency from the Evaluate library to the
Semantics library, which is circular in a shared library build.
Rearrange the code a little to ensure that the dependence is only on a
header.
Move the ErfcScaled template function from the runtime into a new header
file in flang/include/Common, then use it in constant folding to
implement folding for the erfc_scaled() intrinsic function.
It is possible for the compiler to emit an impossible error message
about dummy argument character length incompatibility in the case of a
MODULE SUBROUTINE or FUNCTION defined later in a submodule with MODULE
PROCEDURE, when the character length is defined by USE association in
its interface. The checking for separate module procedure interface
compatibility needs to use a more flexible check than just operator== on
a semantics::ParamValue.
Nearly every Fortran compiler supports "PRINT namelistname" as a synonym
for "WRITE (*, NML=namelistname)". Implement this extension via parse
tree rewriting.
Fixes https://github.com/llvm/llvm-project/issues/111738.
Type interoperability warnings current issue for intrinsic types when
their type, kind, or length do not meet the requirements for C
interoperability. This turns out to be too noisy for the case of
one-byte characters with lengths other than one when creating C pointers
from C_LOC or C_F_POINTER -- it is not uncommon for programs to use
pointers to longer character objects.
So split the interoperability warning so that the case of a known bad
character length for an otherwise interoperable type is controlled by
its own UsageWarning enumerator, and leave that usage warning off by
default. This will better fit expectations in the default case while
still showing a warning under -pedantic.
The operation will be used in the CUF constructor to register the kernel
functions. This allow to delay this until codegen when the gpu.binary
will be available.
Reland of #112268 with correct shared library build support.
This PR applies the changes discussed in [[RFC] Rationale for Flang
AliasAnalysis pointer component
logic](https://discourse.llvm.org/t/rfc-rationale-for-flang-aliasanalysis-pointer-component-logic/79252).
In summary, this PR replaces the existing pointer component logic in
Flang's AliasAnalysis implementation. That logic focuses on aliasing
between pointers and non-pointer, non-target composites that have
pointer components. However, it is more conservative than necessary, and
some existing tests expect its current results when less conservative
results seem reasonable.
This PR splits the logic into two cases:
1. Source values are the same: Return MayAlias when one value is the
address of a composite, and the other value is statically the address of
a pointer component of that composite.
2. Source values are different: Return MayAlias when one value is the
address of a composite (actual argument), and the other value is the
address of a pointer (dummy arg) that might dynamically be a component
of that composite.
In both cases, the actual implementation is still more conservative than
described above, but it can be improved further later. Details appear in
the comments.
Additionally, this PR revises the logic that reports MayAlias for a
pointer/target vs. another pointer/target. It constrains the existing
logic to handle only isData cases, and it adds less conservative
handling of !isData cases elsewhere. First, it extends case 2 listed
above to cover the case where the actual argument is the address of a
pointer rather than a composite. Second, it adds a third case: where
target attributes enable aliasing with a dummy argument.
