Try 2 to merge 4fbd1d6c872e8228f23a6e13914222af40ca6461.
Flang algebraic simplification pass will run algebraic simplification
rewrite patterns for Math/Complex/etc. dialects. It is enabled
under opt-for-speed optimization levels (i.e. for O1/O2/O3; Os/Oz will not
enable it).
With this change the FIR/MLIR optimization pipeline becomes affected
by the -O* optimization level switches. Until now these switches
only affected the middle-end and back-end.
Differential Revision: https://reviews.llvm.org/D130035
Tests that use `--mlir-pass-statistics-display=` from MLIR require the
following condition to hold: (extracted from LLVM's Statistics.h):
```
#define LLVM_ENABLE_STATS 1
```
This is normally enforced with `REQUIRES: asserts`. This patch updates
relevant Flang tests accordingly.
For "Release" builds (with assertions disabled), the affected tests will
be failing without this change.
Differential Revision: https://reviews.llvm.org/D130185
Flang algebraic simplification pass will run algebraic simplification
rewrite patterns for Math/Complex/etc. dialects. It is enabled
under opt-for-speed optimization levels (i.e. for O1/O2/O3; Os/Oz will not
enable it).
With this change the FIR/MLIR optimization pipeline becomes affected
by the -O* optimization level switches. Until now these switches
only affected the middle-end and back-end.
Differential Revision: https://reviews.llvm.org/D130035
This update makes sure that `flang-new -O2 -O0` will run at `-O0` rather
than `-O2`.
This bug was identified and originally fixed by Vyacheslav Zakharin in
https://reviews.llvm.org/D130035. I've extracted the fix into a separate
patch.
co-authored by: Vyacheslav Zakharin <vyacheslav.p.zakharin@intel.com>
Differential Revision: https://reviews.llvm.org/D130104
mathOperations should now support all intrinsics that are handled
by the llvmIntrinsics table + `tan` lowered as Math dialect operation +
f128 flavor of abs.
I am going to flip the default to late math lowering after this change,
but still keep the fallback via pgmath. This will allow getting rid
of the llvmIntrinsics table and continue populating
only the mathOperations table, otherwise, updating both tables
seems to be inconvenient.
Differential Revision: https://reviews.llvm.org/D130048
This patch implements lowering for the F08 bitwise comparison intrinsics
(BGE, BGT, BLE and BLT).
This does not create any runtime functions since the functionality is
simple enough to carry out in IR.
The existing semantic check has been changed because it unconditionally
converted the arguments to the largest possible integer type. This
resulted in the argument with the smaller bit-size being sign-extended.
However, the standard requires the argument with the smaller bit-size to
be zero-extended.
Reviewed By: klausler, jeanPerier
Differential Revision: https://reviews.llvm.org/D127805
Evaluations for the Where and Forall constructs previously did
not have their constructExit field fixed up. This could lead to
falling through to subsequent case blocks in select case
statements if either a Where or Forall construct was the final part
of one case block. Setting the constructExit field results in the
proper branching behavior.
Fixes issue: https://github.com/llvm/llvm-project/issues/56500
Differential Revision: https://reviews.llvm.org/D129879
Change-Id: Ia868df12084520a935f087524e118bcdf47f6d7a
Add the collective subroutine, co_sum, to the list of intrinsics.
In accordance with 16.9.50 and 16.9.137, add a check for and an
error if coindexed objects are being passed to certain arguments
in co_sum and in move_alloc. Add a semantics test to check that
this error is successfully caught in calls to move_alloc. Remove
the XFAIL directive, update the ERROR directives and add both
standard-conforming and non-standard conforming calls to the
semantics test for co_sum.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D114134
This patch improves semantic checks for hint clause.
It checks "hint-expression is a constant expression
that evaluates to a scalar value with kind
`omp_sync_hint_kind` and a value that is a valid
synchronization hint."
Reviewed By: peixin
Differential Revision: https://reviews.llvm.org/D127615
This patch adds lowering support for atomic update construct. A region
is associated with every `omp.atomic.update` operation wherein resides:
(1) the evaluation of the expression on the RHS of the atomic assignment
statement, and (2) a `omp.yield` operation that yields the extended value
of expression evaluated in (1).
Reviewed By: peixin
Differential Revision: https://reviews.llvm.org/D125668
Fortran permits forward references to derived types in contexts that don't
require knowledge of the derived type definition for semantic analysis,
such as in the declaration of a pointer or allocatable variable or component.
But when the forward-referenced derived type is used later for a component
reference, it is possible for the DerivedTypeSpec in he base variable or component
declaration to still have a null scope pointer even if the type has been defined,
since DerivedTypeSpec and TypeSpec objects are created in scopes of use
rather than in scopes of definition. The fix is to call
DerivedTypeSpec::Instantiate() in the name resolution of each component
name so that the scope gets filled in if it is still null.
Differential Revision: https://reviews.llvm.org/D129681
Most modern Fortran programs declare procedure pointers with a
procedure-declaration-stmt, but it's also possible to declare one
with a type-declaration-stmt with a POINTER attribute. In this
case, e.g. "real, external, pointer :: p => null()" the initializer
is required to be a null-init. The parse tree traversal in name
resolution would visit the null-init if the symbol were an object
pointer only, leading to a crash in the case of a procedure pointer.
That explanation of the bug is longer than the fix. In short,
ensure that a null-init in an entity-decl is visited for both
species of pointers.
Differential Revision: https://reviews.llvm.org/D129676
Some procedure pointers and EXTERNAL procedures have neither
explicit interfaces nor result types; these procedures are obviously
not known to be functions, but they could be, so semantics must not
assume that they are necessarily subroutines. Refine the procedure
pointer / dummy procedure compatibility check to handle these more
ambiguous cases and not elicit inappropriate error messages.
Differential Revision: https://reviews.llvm.org/D129674
Calling runtime TRANSPOSE requires a temporary array for the result,
and, sometimes, a temporary array for the argument. Lowering it inline
should provide faster code.
I added -opt-transpose control just for debugging purposes temporary.
I am going to make driver changes that will disable inline lowering
for -O0. For the time being I would like to enable it by default
to expose the code to more tests.
Differential Revision: https://reviews.llvm.org/D129497
When computing the base addresses of an array slice to make a
descriptor, codegen generated two LLVM GEPs. The first to compute
the address of the base character element, and a second one to
compute the substring base inside that element.
The previous code did not care about getting the result of the first
GEP right: it used the base array LLVM type as the result type.
This used to work when opaque pointer were not enabled (the actual GEP
result type was probably applied in some later pass). But with opaque
pointers, the second GEP ends-up computing an offset of len*<LLVM array
type> instead of len*<character width>. A previous attempt to fix the
issue was done in D129079, but it does not cover the cases where the
array slice contains subcomponents before the substring
(e.g: array(:)%char_field(5:10)).
This patch fix the issue by computing the actual GEP result type in
codegen. There is also enough knowledge now so that a single GEP can be
generated instead of two.
Differential Revision: https://reviews.llvm.org/D129481
Add a semantics test for the intrinsic function image_status. Add
a check and restriction on the image argument in image_status,
ensuring that it is a positive value. Add same check on the
size argument of the intrinsic ishftc. Add another check on
the shift argument of ishftc, ensuring that it is less than or
equal to the size argument. Add a short semantics test checking
these restrictions in ishftc function calls.
Reviewed By: klausler
Differential Revision: https://reviews.llvm.org/D128009
In case where the bound(s) of a workshare loop use(s) firstprivate var(s), currently, that use is not updated with the created clone. It still uses the shared variable. This patch fixes that.
Reviewed By: peixin
Differential Revision: https://reviews.llvm.org/D127137
Array-value-copy fails to generate a temporary array for case like this:
subroutine bug(b)
real, allocatable :: b(:)
b = b(2:1:-1)
end subroutine
Since LHS may need to be reallocated, lowering produces the following FIR:
%rhs_load = fir.array_load %b %slice
%lhs_mem = fir.if %b_is_allocated_with_right_shape {
fir.result %b
} else {
%new_storage = fir.allocmem %rhs_shape
fir.result %new_storage
}
%lhs = fir.array_load %lhs_mem
%loop = fir.do_loop {
....
}
fir.array_merge_store %lhs, %loop to %lhs_mem
// deallocate old storage if reallocation occured,
// and update b descriptor if needed.
Since %b in array_load and %lhs_mem in array_merge_store are not the same SSA
values, array-value-copy does not detect the conflict and does not produce
a temporary array. This causes incorrect result in runtime.
The suggested change in lowering is to generate this:
%rhs_load = fir.array_load %b %slice
%lhs_mem = fir.if %b_is_allocated_with_right_shape {
%lhs = fir.array_load %b
%loop = fir.do_loop {
....
}
fir.array_merge_store %lhs, %loop to %b
fir.result %b
} else {
%new_storage = fir.allocmem %rhs_shape
%lhs = fir.array_load %new_storage
%loop = fir.do_loop {
....
}
fir.array_merge_store %lhs, %loop to %new_storage
fir.result %new_storage
}
// deallocate old storage if reallocation occured,
// and update b descriptor if needed.
Note that there are actually 3 branches in FIR, so the assignment loops
are currently produced in three copies, which is a code-size issue.
It is possible to generate just two branches with two copies of the loops,
but it is not addressed in this change-set.
Differential Revision: https://reviews.llvm.org/D129314
In folding and in the runtime library for real MOD/MODULO(A,P),
detect overflow from the division A/P and return a properly signed
zero result. (When A/P overflows and both A and P are finite numbers
with nonzero P, the quotient would be a large integer when rounded to
the precision of the floating-point representation.)
Differential Revision: https://reviews.llvm.org/D129020
Create a TargetCharacteristics class to centralize the few items of
target specific information that are relevant to semantics. Use the
new class for all target queries, including derived type component layout
modeling.
Future work will initialize this class with target information
provided or forwarded by the drivers, and use it to fold layout-dependent
intrinsic functions like TRANSFER().
Differential Revision: https://reviews.llvm.org/D129018
Updates: Attempts to work around build issues on Windows.
When addressing a substring of a character array, codegen emits two
GEPs: one for to compute the address of the base element, and a second
one to address the first characters from that element.
The first GEP still returns the LLVM array type (if the FIR array type could be
translated to an array type. Therefore) so zero
indexes must be added to the second GEP in this case to cover for the
Fortran array dimensions before inserting the susbtring offset index.
Surprisingly, the previous code worked ok when MLIR emits none opaque
pointers. But with opaque pointers, the two GEPs are folded in an
invalid GEP where the substring offset becomes an offset for the outer
array dimension.
Note that I tried to fix the issue by modifying the first GEP to return the
element type, but this still gave bad results (here something might be
wrong with opaque pointer in MLIR or LLVM).
Differential Revision: https://reviews.llvm.org/D129079
The previous code made the assumption that the defining
operation is a fir::ConvertOp without checking. This results in
segmentation fault in code like the added test.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D129077
This test is added to check for multidimensional descriptor of array
substring/derived type component array.
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D128990
Co-authored-by: Jean Perier <jperier@nvidia.com>
Add source coordinates to BeginWait and BeginWaitAll calls
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D128970
Co-authored-by: V Donaldson <vdonaldson@nvidia.com>
Add non-standard conforming calls that violate the intent(inout)
of errmsg argument for co_sum, co_max, co_min, and co_broadcast.
Add non-standard conforming calls that violate the argument
typing of errmsg argument for co_max, co_min, and co_broadcast.
Add standard conforming calls that reorder keyword arguments
for co_sum and co_reduce.
Reviewed By: ktras
Differential Revision: https://reviews.llvm.org/D128468
Add new semantics test for team_number and rename existing
team_number semantics test.
Reviewed By: ktras
Differential Revision: https://reviews.llvm.org/D128309
As Fortran 2018 16.9.163, the reshape is the only intrinsic which
requires the shape argument to be rank-one integer array and the SIZE
of it to be one constant expression. The current expression lowering
converts the shape expression with slice in intrinsic into one box value
with the box element type of unknown extent. However, the genReshape
requires the box element type to be constant size. So, convert the box
value into one with box element type of sequence of 1 x constant. This
corner case is found in cam4 in SPEC 2017
https://github.com/llvm/llvm-project/issues/56140.
Reviewed By: Jean Perier
Differential Revision: https://reviews.llvm.org/D128597
Even though the array is declared with '*' upper bounds, it has an
initial value that has a statically known shape. Use the shape from
the type of the initializer when the declared size is '*'.
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D128889
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D128888
Co-authored-by: V Donaldson <vdonaldson@nvidia.com>
Co-authored-by: Jean Perier <jperier@nvidia.com>
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
The names of CHARACTER strings were being truncated leading to invalid
collisions and other failures. This change makes sure to use the entire
string as the seed for the unique name.
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D128884
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
Here is a character SELECT CASE construct that requires a temp to hold the
result of the TRIM intrinsic call:
```
module m
character(len=6) :: s
contains
subroutine sc
n = 0
if (lge(s,'00')) then
select case(trim(s))
case('11')
n = 1
case default
continue
case('22')
n = 2
case('33')
n = 3
case('44':'55','66':'77','88':)
n = 4
end select
end if
print*, n
end subroutine
end module m
```
This SELECT CASE construct is implemented as an IF/ELSE-IF/ELSE comparison
sequence. The temp must be retained until some comparison is successful.
At that point the temp may be freed. Generalize statement context processing
to allow multiple finalize calls to do this, such that the program always
executes exactly one freemem call.
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: klausler, vdonaldson
Differential Revision: https://reviews.llvm.org/D128852
Co-authored-by: V Donaldson <vdonaldson@nvidia.com>
- Add verifiers that determine if an Op requires type parameters or
not and checks that the correct number of parameters is specified.
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: PeteSteinfeld
Differential Revision: https://reviews.llvm.org/D128828
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
LEN(f(...)), where "f" is a non-intrinsic function, should not be folded
to anything else unless the result is a known constant value. While there
are conceivable cases in which we could do better (e.g., an internal function
whose length is a host-associated INTENT(IN) dummy argument), there are
other cases that we're getting wrong.
Differential Revision: https://reviews.llvm.org/D128759
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D128788
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
Uses of these markers are contrary to all other llvm-project subprojects
and carry no information for people who know the basics of a lit test.
I understand that there is an argument "this makes beginners get started
quickly" but I am unsure whether this is strong enough to deviate from
all other projects. https://llvm.org/docs/TestingGuide.html covers the
basics. Actually, some contributors were confused by the markers.
Differential Revision: https://reviews.llvm.org/D128763
