Remove old clause operands from acc.kernels operation since
the new dataOperands is now in place.
private and firstprivate will receive some redesign but are
not part of the new dataOperands.
Depends on D150224
Reviewed By: vzakhari
Differential Revision: https://reviews.llvm.org/D150225
Remove old clause operands from acc.serial operation since
the new dataOperands is now in place.
private and firstprivate will receive some redesign but are
not part of the new dataOperands.
Depends on D150207
Reviewed By: vzakhari
Differential Revision: https://reviews.llvm.org/D150224
The order of operand in clauses that are decomposed was not
preserved. This patch change how operands are handled and preserve
the user ordering for the entry data operation on the acc.parallel
operation.
Reviewed By: vzakhari
Differential Revision: https://reviews.llvm.org/D150214
The order of operand in clauses that are decomposed was not
preserved. This patch change how operands are handled and preserve
the user ordering for the entry data operation on the acc.data
operation.
Reviewed By: vzakhari
Differential Revision: https://reviews.llvm.org/D150213
Remove old clause operands from acc.parallel operation since
the new dataOperands is now in place.
private, firstprivate and reductions will receive some redesign but are
not part of the new dataOperands.
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D150207
The lowerbound was not correctly normalized to 0 when the bound
are generated for an array without subscripts.
```
integer :: a(11:20)
!$acc enter data create(a)
```
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D150208
The implied-do index value has 'index' type, and it has to be
converted to the original ac-do-variable's data type.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D150150
Since the new data operand operations have been added in D148389 and
adopted on acc.data in D149673, the old clause operands are no longer
needed.
The LegalizeDataOpForLLVMTranslation will become obsolete when all
operations will be cleaned. For the time being only the appropriate
part are being removed.
processOperands will also receive some updates once all the operands
will be coming from an acc data operand operation.
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D150155
The lowerbound was not correctly normalized to 0 when the bound
are generated for an array without subscripts.
```
integer :: a(11:20)
!$acc enter data create(a)
```
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D150208
Since the new data operand operations have been added in D148389 and
adopted on acc.exit_data in D149601, the old clause operands are no longer
needed.
The LegalizeDataOpForLLVMTranslation will become obsolete when all
operations will be cleaned. For the time being only the appropriate
part are being removed.
processOperands will also receive some updates once all the operands
will be coming from an acc data operand operation.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D150145
Self clause is the same same as the host clause. Lower it
in a simmilar way.
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D150174
Follow up of https://reviews.llvm.org/D149979 for lowering.
In Fortran, it is possible to assign a pointer to another pointer
with an undefined association status.
When using the runtime do to none trivial pointer association, if the
descriptor are garbage, the runtime cannot safely detect that it has
a garbage descriptor, and it cannot safely know the descriptor size
leading to undefined behavior.
Another reason to initialize descriptor of pointers is to record any
non deferred length parameter value.
Hence, although this is not required by Fortran, f18 always initialize
pointers to NULL().
This was already done in lowering for whole pointer object, but not for
pointer components.
This patch uses the related semantics patch that updated
derivedTypeSpe::HasDefaultInitialization to ensure pointer components
of local and global object are always initialized.
It adds tests to ensure that allocation of such derived type uses the
runtime to ensure the storage is initialized, and that structure
constructors are setting the descriptor component to NULL() if no
initial target is given.
Differential Revision: https://reviews.llvm.org/D150180
Since the new data operand operations have been added in D148389 and
adopted on acc.enter_data in D148721, the old clause operands are no longer
needed.
The LegalizeDataOpForLLVMTranslation will become obsolete when all
operations will be cleaned. For the time being only the appropriate
part are being removed.
processOperands will also receive some updates once all the operands
will be coming from an acc data operand operation.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D150132
The `if_present` clause is modeled as an attribute on the
acc.update operation. The lowering was not adding correctly the attribute
when the clause was present. This patch update the lowering to add
the attribute when needed.
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D150171
This patch lowers assignments to vector subscripted designators into the
newly added hlfir.elemental_addr and hlfir.region_assign.
Note that the codegen of these operation to FIR is still TODO and will
still emit a TODO message when trying to compile programs end to end.
Differential Revision: https://reviews.llvm.org/D149962
Lower Forall to the previously added hlfir.forall, hlfir.forall_mask.
hlfir.forall_index, and hlfir.region_assign operations.
The HLFIR assignment code lowering is moved into genDataAssignment for
more readability and so that user defined assignment (still a TODO),
will be able to share most of the logic.
Differential Revision: https://reviews.llvm.org/D149878
This fixes LEN inquiry on array constructors where the length
expression includes references to the ac-do-variable.
Reviewed By: klausler
Differential Revision: https://reviews.llvm.org/D150149
Make __builtin_c_loc() into an intrinsic function and verify the
special semantic requirements on its actual arguments.
Differential Revision: https://reviews.llvm.org/D149988
Code in Evaluate/type.cpp was maintaining a very old assumption that
procedure pointers would need to be represented as descriptors. This
is not the case -- they are code or thunk addresses.
Differential Revision: https://reviews.llvm.org/D149984
Pointer components without default initialization pose some
difficult (or impossible) problems when they appear as right-hand
side targets in pointer assignment statements; they may contain
garbage or stale data that looks enough like a valid descriptor
to cause a crash. Solve the problem by avoiding it -- ensure
that pointers' descriptors are at least minimally established.
Differential Revision: https://reviews.llvm.org/D149979
Since the new data operand operations have been added in D148389 and
adopted on acc.update in D149909, the old clause operands are no longer
needed. This is a first patch to start cleaning the OpenACC operations
with data clause operands.
The `LegalizeDataOpForLLVMTranslation` will become obsolete when all
operations will be cleaned. For the time being only the appropriate
part are being removed.
`processOperands` will also receive some updates once all the operands
will be coming from an acc data operand operation.
Reviewed By: razvanlupusoru, jeanPerier
Differential Revision: https://reviews.llvm.org/D150053
Data operands associated with acc.update should comes
from acc data entry/exit operations or acc.getdeviceptr.
Reviewed By: razvanlupusoru
Differential Revision: https://reviews.llvm.org/D149990
Currently `CallOpInterface` has a method `getCallableForCallee` to have a consistent way to get the callee from an operation with `CallOpInterface`, but missing a consistent way to set a callee for an operation with `CallOpInterface`.
A set callee method is useful for transformations that operate on `CallOpInterface`, and change the callee, e.g., a pass that specialize function, which clone the callee, and change the `CallOpInterface`'s callee to the cloned version. Without such method, transformation would need to understand the implementation for every operations with `CallOpInterface`, and have a type switch to handle them.
This review adds a method to set callee for operation with `CallOpInterface`.
Reviewed By: gysit, zero9178o
Differential Revision: https://reviews.llvm.org/D149763
I removed the barriers a bit fast. Some mold is needed on
hlfir.elemental for polymorphic vector subscripted designator (and for
parenthesized polymorphic) so that a temporary can be later created.
The parenthesized array case may also just used asExpr and that
could later use AssignTemporary to deal with this. But the vector
subscripted designator case will need to use some new runtime
to get some mold allocation done for the temp.
Add TODOs in the meantime.
Differential Revision: https://reviews.llvm.org/D149970
Addresses the same issue as the following abandoned revision: D104391.
Rewrite leading declarative allocations so they are nested within their respective executable allocate directive
Original:
ExecutionPartConstruct -> OpenMPDeclarativeAllocate
ExecutionPartConstruct -> OpenMPDeclarativeAllocate
ExecutionPartConstruct -> OpenMPExecutableAllocate
After rewriting:
ExecutionPartConstruct -> OpenMPExecutableAllocate
| ExecutionPartConstruct -> OpenMPDeclarativeAllocate
| ExecutionPartConstruct -> OpenMPDeclarativeAllocate
Reviewed By: kiranchandramohan
Differential Revision: https://reviews.llvm.org/D148409
Co-authored-by: Isaac Perry <isaac.perry@arm.com>
Addresses the same issue as the following abandoned revision: D104391.
Rewrite leading declarative allocations so they are nested within their respective executable allocate directive
Original:
```
ExecutionPartConstruct -> OpenMPDeclarativeAllocate
ExecutionPartConstruct -> OpenMPDeclarativeAllocate
ExecutionPartConstruct -> OpenMPExecutableAllocate
```
After rewriting:
```
ExecutionPartConstruct -> OpenMPExecutableAllocate
| ExecutionPartConstruct -> OpenMPDeclarativeAllocate
| ExecutionPartConstruct -> OpenMPDeclarativeAllocate
```
Co-authored-by: Isaac Perry <isaac.perry@arm.com>
Reviewed By: kiranchandramohan
Differential Revision: https://reviews.llvm.org/D148409
This reverts commit 5faf45a3d24e603cbc8fe4eb45da386653dae5e5.
Once again arcanist stripped the co-author metadata. I'm going to add it to the revision description and try one last time.
Addresses the same issue as the following abandoned revision: D104391.
Rewrite leading declarative allocations so they are nested within their respective executable allocate directive
Original:
```
ExecutionPartConstruct -> OpenMPDeclarativeAllocate
ExecutionPartConstruct -> OpenMPDeclarativeAllocate
ExecutionPartConstruct -> OpenMPExecutableAllocate
```
After rewriting:
```
ExecutionPartConstruct -> OpenMPExecutableAllocate
| ExecutionPartConstruct -> OpenMPDeclarativeAllocate
| ExecutionPartConstruct -> OpenMPDeclarativeAllocate
```
Reviewed By: kiranchandramohan
Differential Revision: https://reviews.llvm.org/D148409
Update OpenACC update construct lowering to create
the data operand operations for host and device clauses.
Depends on D149909
Reviewed By: razvanlupusoru, jeanPerier
Differential Revision: https://reviews.llvm.org/D149910
Introduce acc.update_host and acc.update_device as data
operands operation to be used by the acc.update op.
Reviewed By: razvanlupusoru, jeanPerier
Differential Revision: https://reviews.llvm.org/D149909
Addresses the same issue as the following abandoned revision: D104391.
Rewrite leading declarative allocations so they are nested within their respective executable allocate directive
Original:
```
ExecutionPartConstruct -> OpenMPDeclarativeAllocate
ExecutionPartConstruct -> OpenMPDeclarativeAllocate
ExecutionPartConstruct -> OpenMPExecutableAllocate
```
After rewriting:
```
ExecutionPartConstruct -> OpenMPExecutableAllocate
| ExecutionPartConstruct -> OpenMPDeclarativeAllocate
| ExecutionPartConstruct -> OpenMPDeclarativeAllocate
```
Reviewed By: kiranchandramohan
Differential Revision: https://reviews.llvm.org/D148409
First issue is that the clean-ups were generated after the yield_element
operation that must be the terminator. Second issue is that codegen for
elemenal operation was not working properly with nested elemental ops.
Differential Revision: https://reviews.llvm.org/D149921
Instead of filling uninitialized global variables with "undef",
initialize them with 0. Only for Integer, Float or Logical type
variables. Complex, user defined data structures, arrays, etc
are not supported at this point.
This patch fixes the main problem of
https://github.com/llvm/llvm-project/issues/62432
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D149877
Currently complex division is lowered to a fir.divc operation and the
fir.divc is later converted to a sequence of llvm operations to perform
complex division, however this causes issues for extreme values when
the calculations overflow.
This patch changes the lowering of complex division to use the Intrinsic
Call functionality to lower into library calls (for single, double,
extended and quad precisions) or an MLIR complex dialect division operation
(for half and bfloat precisions).
Note 1: If the Complex To Standard conversion of division operation
matures then we can use it for all precisions. Currently it has the
same issues as the conversion of fir.divc.
Note 2: A previous patch (D145808) did the same but during conversion of
the fir.divc operation. But using function calls at that stage leads to
ABI issues since the conversion to LLVM is not aware of the complex target
rewrite.
Note 3: If the patch is accepted, fir.divc can be removed from FIR.
Reviewed By: vzakhari, PeteSteinfeld, DavidTruby
Differential Revision: https://reviews.llvm.org/D149546
Prior to this change, if you define a module as such with a declare target in it:
module test_0
implicit none
!$omp declare target(no_implicit_materialization_1)
end module test_0
The compiler will crash rather than give some form of reasonable
diagnostic. This patch attempts to fix that.
Reviewers: kiranchandramohan
Differential Revision: https://reviews.llvm.org/D149913
I plan to implement lowering from parse tree to HLFIR first for forall
and where to ease testing of the rewrite pass while writing it.
To avoid cryptic errors in ConvertToFir pass about unhandled operations,
this patch already defines the pass that will further lower these
operations and make it throw clear TODO messages.
Differential Revision: https://reviews.llvm.org/D149852
This is the last piece required to lower Forall (except pointer
assignments, where an operation may be needed to deal with bounds
remapping).
Lowering requires symbols to be mapped to memory SSA values produced
by a fir_FortranVariableOpInterface operation. This applies to
forall index-values, that are symbols.
fir.alloca/fir.store/hlfir.declare are not allowed inside the body of
an hlfir.forall that only accept operations with the
hlfir_OrderedAssignmentTreeOpInterface so that the forall structure is
well defined and easy to transform.
Allowing such operations in the forall body would open the doors to
generating ill-formed programs where such operation would be used for
non index-values.
Instead, add an hlfir.forall_index with both required interface to
produce a memory address for a forall index.
As a bonus, since forall index-value are by nature read-only, the
loads of hlfir.forall_index can be canonicalized, which will help
simplifying the hlfir.forall nested code (it is unclear we will be
able to tell MLIR enough about hlfir.forall and hlfir.where structure
so that it could safely do a generic mem-to-reg inside it, and getting
rid of read-effect operations will benefit the forall rewrite pass).
Differential Revision: https://reviews.llvm.org/D149836
This patch lowers the data clause on the OpenACC compute construct
to their corresponding acc data operand operation.
The decomposition is the same as in D149673.
Note that `private` and `firstprivate` are not lowered to data operand operation as they do not have one and will likely have dedicated design/process.
Depends on D149673
Reviewed By: razvanlupusoru, jeanPerier
Differential Revision: https://reviews.llvm.org/D149785
This patch lowers the data clause on the OpenACC data construct
to their corresponding acc data operand operation.
The copy clause is decomposed into acc.copyin before and acc.copyout after
the acc.data operation.
The copyout close is decomposed into acc.create before and acc.copyout after
the acc.data operation.
The attach clause is decomposed into acc.attach before and acc.detach after
the acc.data operation.
Depends on D149601
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D149673
Add support to lower data exit operations and use it for the
copyout, detach and delete clause on the `!$acc exit data` construct.
This patch refactor the data entry op creation so most part of the code
can be shared with the data exit op.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D149601
hlfir.assign, in general, ends up calling the Assign runtime that asserts
that the types of LHS and RHS match. In case of implicit logical<->integer
conversions (allowed as an extension) the operands of hlfir.assign
have non-matching types. This change makes sure that the lowering
produces explicit type cast (either as a scalar fir.convert or
as a hlfir.elemental producing array expression).
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D149765
The shared code for lowering the sum and product operations in
flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIRIntrinsics.cpp have
been moved into a new class HlfirReductionIntrinsicConverion.
Depends on: D148719
Differential Revision: https://reviews.llvm.org/D149644
