by Raul Penacoba.
The size of kmp_depend_info and the number of dependencies are computed multiplying the iterator sizes, which not right.
Now size is computed as:
itersize1*numclausedeps1 + itersize2*numclausedeps2 + ... + itersizeN*numclausedepsN
where itersizeX is the size of the iterator and numclausedepsX the number of dependencies in that depend clause.
Reviewed By: ABataev
Differential Revision: https://reviews.llvm.org/D111045
This patch adds OpenMP assumption attributes to call sites in applicable
regions. Currently this applies the caller's assumption attributes to
any calls contained within it. So, if a call occurs inside an OpenMP
assumes region to a function outside that region, we will assume that
call respects the assumptions. This is primarily useful for inline
assembly calls used heavily in the OpenMP GPU device runtime, which
allows us to then make judgements about what the ASM will do.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D110655
This patch adds a new RTL function for worksharing. Currently we use
`__kmpc_for_static_init` for both the `distribute` and `parallel`
portion of the loop clause. This patch replaces the `distribute` portion
with a new runtime call `__kmpc_distribute_static_init`. Currently this
will be used exactly the same way, but will make it easier in the future
to fine-tune the distribute and parallel portion of the loop.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D110429
This is a follow-up of D110029, which uses bitset to indicate execution mode. This patches makes the changes in the function call.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D110279
The execution mode of a kernel is stored in a global variable, whose value means:
- 0 - SPMD mode
- 1 - indicates generic mode
- 2 - SPMD mode execution with generic mode semantics
We are going to add support for SIMD execution mode. It will be come with another
execution mode, such as SIMD-generic mode. As a result, this value-based indicator
is not flexible.
This patch changes to bitset based solution to encode execution mode. Each
position is:
[0] - generic mode
[1] - SPMD mode
[2] - SIMD mode (will be added later)
In this way, `0x1` is generic mode, `0x2` is SPMD mode, and `0x3` is SPMD mode
execution with generic mode semantics. In the future after we add the support for
SIMD mode, `0b1xx` will be in SIMD mode.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D110029
This improves diagnostic (& important to me, DWARF) accuracy - otherwise
there could be ambiguities between "std::nullptr_t" and some user-defined
type that's /actually/ "nullptr_t" defined in the global namespace.
Differential Revision: https://reviews.llvm.org/D110044
Parallel regions are outlined as functions with capture variables explicitly generated as distinct parameters in the function's argument list. That complicates the fork_call interface in the OpenMP runtime: (1) the fork_call is variadic since there is a variable number of arguments to forward to the outlined function, (2) wrapping/unwrapping arguments happens in the OpenMP runtime, which is sub-optimal, has been a source of ABI bugs, and has a hardcoded limit (16) in the number of arguments, (3) forwarded arguments must cast to pointer types, which complicates debugging. This patch avoids those issues by aggregating captured arguments in a struct to pass to the fork_call.
Reviewed By: jdoerfert, jhuber6
Differential Revision: https://reviews.llvm.org/D102107
This patch supports OpenMP 5.0 metadirective features.
It is implemented keeping the OpenMP 5.1 features like dynamic user condition in mind.
A new function, getBestWhenMatchForContext, is defined in llvm/Frontend/OpenMP/OMPContext.h
Currently this function return the index of the when clause with the highest score from the ones applicable in the Context.
But this function is declared with an array which can be used in OpenMP 5.1 implementation to select all the valid when clauses which can be resolved in runtime. Currently this array is set to null by default and its implementation is left for future.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D91944
The new device runtime uses an internal variable to set debugging. This
variable was originally privately linked because every module will have
a copy of it. This caused problems with merging the device bitcode
library because it would get renamed and there was not a way to refer to
an external, private symbol. This changes the symbol to weak_odr so it
can be defined multiply, but will not be renamed.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D109997
This patch supports OpenMP 5.0 metadirective features.
It is implemented keeping the OpenMP 5.1 features like dynamic user condition in mind.
A new function, getBestWhenMatchForContext, is defined in llvm/Frontend/OpenMP/OMPContext.h
Currently this function return the index of the when clause with the highest score from the ones applicable in the Context.
But this function is declared with an array which can be used in OpenMP 5.1 implementation to select all the valid when clauses which can be resolved in runtime. Currently this array is set to null by default and its implementation is left for future.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D91944
This patch supports OpenMP 5.0 metadirective features.
It is implemented keeping the OpenMP 5.1 features like dynamic user condition in mind.
A new function, getBestWhenMatchForContext, is defined in llvm/Frontend/OpenMP/OMPContext.h
Currently this function return the index of the when clause with the highest score from the ones applicable in the Context.
But this function is declared with an array which can be used in OpenMP 5.1 implementation to select all the valid when clauses which can be resolved in runtime. Currently this array is set to null by default and its implementation is left for future.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D91944
Use irbuilder as default and remove redundant Clang codegen for masked construct and master construct.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D100874
fae0dfa changed code to check 128-bit float availability, since it
introduced a new 128-bit double type on PowerPC. However, there're other
long float types besides IEEE float128 and PPC double-double requiring
this feature.
Reviewed By: ronlieb
Differential Revision: https://reviews.llvm.org/D109943
This patch supports construct trait set selector by using the existed
declare variant infrastructure inside `OMPContext` and simd selector is
currently not supported. The goal of this patch is to pass the declare variant
test inside sollve test suite.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D109635
The patch adds missing diagnostics for cases like:
float F3 = ((__float128)F1 * (__float128)F2) / 2.0f;
Sema::checkDeviceDecl (renamed to checkTypeSupport) is changed to work
with a type without the corresponding ValueDecl. It is also refactored
so that host diagnostics for unsupported types can be added here as
well.
Differential Revision: https://reviews.llvm.org/D109315
This patch introduces the flags `-fopenmp-target-debug` and
`-fopenmp-target-debug=` to set the value of a global in the device.
This will be used to enable or disable debugging features statically in
the device runtime library.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D109544
Since these assumptions are coming from OpenMP it makes sense to mark
them as such in the generic IR encoding. Standardized assumptions will
be named
omp_ASSUMPTION_NAME
and extensions will be named
ompx_ASSUMPTION_NAME
which is the OpenMP 5.2 syntax for "extensions" of any kind.
This also matches what the OpenMP-Opt pass expects.
Summarized,
#pragma omp [...] assume[s] no_parallelism
now generates the same IR assumption annotation as
__attribute__((assume("omp_no_parallelism")))
Reviewed By: jhuber6
Differential Revision: https://reviews.llvm.org/D105937
Recommit of 707ce34b06. Don't introduce a
dependency to the LLVMPasses component, instead register the required
passes individually.
Add methods for loop unrolling to the OpenMPIRBuilder class and use them in Clang if `-fopenmp-enable-irbuilder` is enabled. The unrolling methods are:
* `unrollLoopFull`
* `unrollLoopPartial`
* `unrollLoopHeuristic`
`unrollLoopPartial` and `unrollLoopHeuristic` can use compiler heuristics to automatically determine the unroll factor. If possible, that is if no CanonicalLoopInfo is required to pass to another method, metadata for LLVM's LoopUnrollPass is added. Otherwise the unroll factor is determined using the same heurstics as user by LoopUnrollPass. Not requiring a CanonicalLoopInfo, especially with `unrollLoopHeuristic` allows greater flexibility.
With full unrolling and partial unrolling with known unroll factor, instead of duplicating instructions by the OpenMPIRBuilder, the full unroll is still delegated to the LoopUnrollPass. In case of partial unrolling the loop is first tiled using the existing `tileLoops` methods, then the inner loop fully unrolled using the same mechanism.
Reviewed By: jdoerfert, kiranchandramohan
Differential Revision: https://reviews.llvm.org/D107764
The test driver-fopenmp-extensions.c is failing on platforms that does
not use integrated-as. It can be reproduced using -fno-integrated-as on
Linux too.
bin/clang -c -Xclang -verify=omp -fopenmp -fopenmp-extensions
-fno-openmp-extensions
../llvm-project/clang/test/OpenMP/driver-fopenmp-extensions.c
-fno-integrated-as
Assembler messages:
Error: can't open /tmp/driver-fopenmp-extensions-8fafe8.s for reading:
No such file or directory
clang-14: error: assembler command failed with exit code 1 (use -v to
see invocation)
The goal of this test is to verify syntax diags only,
so we should use clang_cc1 to test.
Reviewed By: jdenny, ABataev
Differential Revision: https://reviews.llvm.org/D109255
Add support for ordered directive in the OpenMPIRBuilder.
This patch also modidies clang to use the ordered directive when the
option -fopenmp-enable-irbuilder is enabled.
Also fix one ICE when parsing one canonical for loop with the relational
operator LE or GE in openmp region by replacing unary increment
operation of the expression of the variable "Expr A" minus the variable
"Expr B" (++(Expr A - Expr B)) with binary addition operation of the
experssion of the variable "Expr A" minus the variable "Expr B" and the
expression with constant value "1" (Expr A - Expr B + "1").
Reviewed By: Meinersbur, kiranchandramohan
Differential Revision: https://reviews.llvm.org/D107430
Breaks build with -DBUILD_SHARED_LIBS=ON
```
CMake Error: The inter-target dependency graph contains the following strongly connected component (cycle):
"LLVMFrontendOpenMP" of type SHARED_LIBRARY
depends on "LLVMPasses" (weak)
"LLVMipo" of type SHARED_LIBRARY
depends on "LLVMFrontendOpenMP" (weak)
"LLVMCoroutines" of type SHARED_LIBRARY
depends on "LLVMipo" (weak)
"LLVMPasses" of type SHARED_LIBRARY
depends on "LLVMCoroutines" (weak)
depends on "LLVMipo" (weak)
At least one of these targets is not a STATIC_LIBRARY. Cyclic dependencies are allowed only among static libraries.
CMake Generate step failed. Build files cannot be regenerated correctly.
```
This reverts commit 707ce34b06.
Add methods for loop unrolling to the OpenMPIRBuilder class and use them in Clang if `-fopenmp-enable-irbuilder` is enabled. The unrolling methods are:
* `unrollLoopFull`
* `unrollLoopPartial`
* `unrollLoopHeuristic`
`unrollLoopPartial` and `unrollLoopHeuristic` can use compiler heuristics to automatically determine the unroll factor. If possible, that is if no CanonicalLoopInfo is required to pass to another method, metadata for LLVM's LoopUnrollPass is added. Otherwise the unroll factor is determined using the same heurstics as user by LoopUnrollPass. Not requiring a CanonicalLoopInfo, especially with `unrollLoopHeuristic` allows greater flexibility.
With full unrolling and partial unrolling with known unroll factor, instead of duplicating instructions by the OpenMPIRBuilder, the full unroll is still delegated to the LoopUnrollPass. In case of partial unrolling the loop is first tiled using the existing `tileLoops` methods, then the inner loop fully unrolled using the same mechanism.
Reviewed By: jdoerfert, kiranchandramohan
Differential Revision: https://reviews.llvm.org/D107764
This patch implements Clang support for an original OpenMP extension
we have developed to support OpenACC: the `ompx_hold` map type
modifier. The next patch in this series, D106510, implements OpenMP
runtime support.
Consider the following example:
```
#pragma omp target data map(ompx_hold, tofrom: x) // holds onto mapping of x
{
foo(); // might have map(delete: x)
#pragma omp target map(present, alloc: x) // x is guaranteed to be present
printf("%d\n", x);
}
```
The `ompx_hold` map type modifier above specifies that the `target
data` directive holds onto the mapping for `x` throughout the
associated region regardless of any `target exit data` directives
executed during the call to `foo`. Thus, the presence assertion for
`x` at the enclosed `target` construct cannot fail. (As usual, the
standard OpenMP reference count for `x` must also reach zero before
the data is unmapped.)
Justification for inclusion in Clang and LLVM's OpenMP runtime:
* The `ompx_hold` modifier supports OpenACC functionality (structured
reference count) that cannot be achieved in standard OpenMP, as of
5.1.
* The runtime implementation for `ompx_hold` (next patch) will thus be
used by Flang's OpenACC support.
* The Clang implementation for `ompx_hold` (this patch) as well as the
runtime implementation are required for the Clang OpenACC support
being developed as part of the ECP Clacc project, which translates
OpenACC to OpenMP at the directive AST level. These patches are the
first step in upstreaming OpenACC functionality from Clacc.
* The Clang implementation for `ompx_hold` is also used by the tests
in the runtime implementation. That syntactic support makes the
tests more readable than low-level runtime calls can. Moreover,
upstream Flang and Clang do not yet support OpenACC syntax
sufficiently for writing the tests.
* More generally, the Clang implementation enables a clean separation
of concerns between OpenACC and OpenMP development in LLVM. That
is, LLVM's OpenMP developers can discuss, modify, and debug LLVM's
extended OpenMP implementation and test suite without directly
considering OpenACC's language and execution model, which can be
handled by LLVM's OpenACC developers.
* OpenMP users might find the `ompx_hold` modifier useful, as in the
above example.
See new documentation introduced by this patch in `openmp/docs` for
more detail on the functionality of this extension and its
relationship with OpenACC. For example, it explains how the runtime
must support two reference counts, as specified by OpenACC.
Clang recognizes `ompx_hold` unless `-fno-openmp-extensions`, a new
command-line option introduced by this patch, is specified.
Reviewed By: ABataev, jdoerfert, protze.joachim, grokos
Differential Revision: https://reviews.llvm.org/D106509
Previously when emitting a C++ guarded initializer, we tried to work out what
the enclosing function would be used for and added it to the COMDAT containing
the variable if we thought that doing so would be correct. But this was done
from a context in which we didn't -- and realistically couldn't -- correctly
infer how the enclosing function would be used.
Instead, add the initialization function to a COMDAT from the code that
creates it, in the case where it makes sense to do so: when we know that
the one and only reference to the initialization function is in
@llvm.global.ctors and that reference is in the same COMDAT.
Reviewed By: rjmccall
Differential Revision: https://reviews.llvm.org/D108680
Mapping expressions that have `this` as their base expression aren't
considered a valid base variable and the rest of the runtime expects
this. However, if we have an expression with no value declaration we can
try to extract it manually to provide more helpful debuggin information.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D108483
A new rule is added in 5.0:
If a list item appears in a reduction, lastprivate or linear clause
on a combined target construct then it is treated as if it also appears
in a map clause with a map-type of tofrom.
Currently map clauses for all capture variables are added implicitly.
But missing for list item of expression for array elements or array
sections.
The change is to add implicit map clause for array of elements used in
reduction clause. Skip adding map clause if the expression is not
mappable.
Noted: For linear and lastprivate, since only variable name is
accepted, the map has been added though capture variables.
To do so:
During the mappable checking, if error, ignore diagnose and skip
adding implicit map clause.
The changes:
1> Add code to generate implicit map in ActOnOpenMPExecutableDirective,
for omp 5.0 and up.
2> Add extra default parameter NoDiagnose in ActOnOpenMPMapClause:
Use that to skip error as well as skip adding implicit map during the
mappable checking.
Note: there are only tow places need to be check for NoDiagnose. Rest
of them either the check is for < omp 5.0 or the error already generated for
reduction clause.
Differential Revision: https://reviews.llvm.org/D108132
Use uint64_t for lanemask on all GPU architectures at the interface
with clang. Updates tests. The deviceRTL is always linked as IR so the zext
and trunc introduced for wave32 architectures will fold after inlining.
Simplification partly motivated by amdgpu gfx10 which will be wave32 and
is awkward to express in the current arch-dependant typedef interface.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D108317
We were using an OpaqueValueExpr allocated on the stack to store
the size of a VLA. Because the VLASizeMap in CodegenFunction
uses the address of the expression to avoid recomputing VLAs,
we were accidentally reusing an earlier llvm::Value. This led to
invalid LLVM IR.
This is a temporary solution until VLASizeMap can be pushed and popped
based on the context.
Differential Revision: https://reviews.llvm.org/D107666
After D94315 we add the `NoInline` attribute to the outlined function to handle
data environments in the OpenMP if clause. This conflicted with the `AlwaysInline`
attribute added to the outlined function. for better performance in D106799.
The data environments should ideally not require NoInline, but for now this
fixes PR51349.
Reviewed By: mikerice
Differential Revision: https://reviews.llvm.org/D107649
The root problem is a null pointer is accessed during the call to
checkOpenMPLoop, because loop up bound expr is an error expression
due to error diagnostic was emit early.
To fix this, in setLCDeclAndLB, setUB and setStep instead return false,
return true when LB, UB or Step contains Error, so that the checking is
stopped in checkOpenMPLoop.
Differential Revision: https://reviews.llvm.org/D107385
Clang diagnostics should not start with a capital letter or use
trailing punctuation (https://clang.llvm.org/docs/InternalsManual.html#the-format-string),
but quite a few driver diagnostics were not following this advice. This
corrects the grammar and punctuation to improve consistency, but does
not change the circumstances under which the diagnostics are produced.
where should not.
Currently we are using QTy->isIncompleteType(&ND) to check incomplete
type. But before doing that, need to instantiate for a class template
specialization or a class member of a class template specialization,
or an array with known size of such..., so that we know it is really
incomplete type.
To fix this using RequireCompleteType instead.
The new test is added into "test/OpenMP/target_update_messages.cpp"
The different of using RequireCompleteType is when emit incomplete type,
an additional note is also emitted to point to where incomplete type
is declared. Because this change, many tests are needed to be fixed
by adding additional note.
This is to fix https://bugs.llvm.org/show_bug.cgi?id=50508
Differential Revision: https://reviews.llvm.org/D107200
Target-dependent constant folding will fold these down to simple
constants (or at least, expressions that don't involve a GEP). We don't
need heroics to try to optimize the form of the expression before that
happens.
Fixes https://bugs.llvm.org/show_bug.cgi?id=51232 .
Differential Revision: https://reviews.llvm.org/D107116
The device runtime contains several calls to __kmpc_get_hardware_num_threads_in_block
and __kmpc_get_hardware_num_blocks. If the thread_limit and the num_teams are constant,
these calls can be folded to the constant value.
In commit D106033 we have the optimization phase. This commit adds the attributes to
the outlined function for the grid size. the two attributes are `omp_target_num_teams` and
`omp_target_thread_limit`. These values are added as long as they are constant.
Two functions are created `getNumThreadsExprForTargetDirective` and
`getNumTeamsExprForTargetDirective`. The original functions `emitNumTeamsForTargetDirective`
and `emitNumThreadsForTargetDirective` identify the expresion and emit the code.
However, for the Device version of the outlined function, we cannot emit anything.
Therefore, this is a first attempt to separate emision of code from deduction of the
values.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D106298
This patch adds the always inline attribute to the outlined functions generated
by OpenMP regions. Because there is only a single instance of this function and
it always has internal linkage it is safe to inline in every instance it is
created. This could potentially lead to performance degredation due to
inflated register counts in the parallel region.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D106799
In OpenMP 5.1:
> If the `write` or `update` clause is specifieded, the atomic operation is not an atomic conditional update for which the comparison fails, and the effective memory ordering is `release`, `acq_rel`, or `seq_cst`, the strong flush on entry to the atomic operation is also a release flush. If the `read` or `update` clause is specified and the effective memory ordering is `acquire`, `acq_rel`, or `seq_cst` then the strong flush on exit from the atomic operation is also an acquire flush.
In OpenMP 5.0:
> If the `write`, `update`, or **`capture`** clause is specified and the `release`, `acq_rel`, or `seq_cst` clause is specified then the strong flush on entry to the atomic operation is also a release flush. If the `read` or `capture` clause is specified and the `acquire`, `acq_rel`, or `seq_cst` clause is specified then the strong flush on exit from the atomic operation is also an acquire flush.
From my understanding, in OpenMP 5.1, `capture` is removed from the requirement for flush, therefore we don't have to enforce it.
Reviewed By: ABataev
Differential Revision: https://reviews.llvm.org/D100768
This patch changes `__kmpc_free_shared` to take an additional argument
corresponding to the associated allocation's size. This makes it easier to
implement the allocator in the runtime.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D106496
Parallel regions are outlined as functions with capture variables explicitly generated as distinct parameters in the function's argument list. That complicates the fork_call interface in the OpenMP runtime: (1) the fork_call is variadic since there is a variable number of arguments to forward to the outlined function, (2) wrapping/unwrapping arguments happens in the OpenMP runtime, which is sub-optimal, has been a source of ABI bugs, and has a hardcoded limit (16) in the number of arguments, (3) forwarded arguments must cast to pointer types, which complicates debugging. This patch avoids those issues by aggregating captured arguments in a struct to pass to the fork_call.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D102107
Alexey Bataev