This patch adds a function attribute to the kernel function generated in
OpenMP offloading. We already create a `nvvm.annotations` metadata node
indicating the kernels present in the program. However, this created
some indirection when trying to identify if a specific function was an
entry. We add a single function attribute for each function now to
simplify this.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D118708
Turning on `enable_noundef_analysis` flag allows better codegen by removing freeze instructions.
I modified clang by renaming `enable_noundef_analysis` flag to `disable-noundef-analysis` and turning it off by default.
Test updates are made as a separate patch: D108453
Reviewed By: eugenis
Differential Revision: https://reviews.llvm.org/D105169
This patch changes the default aligntment from 8 to 16, and encodes this
information in the `__kmpc_alloc_shared` runtime call to communicate it
to the HeapToStack pass. The previous alignment of 8 was not sufficient
for the maximum size of primitive types on 64-bit systems, and needs to
be increaesd. This reduces the amount of space availible in the data
sharing stack, so this implementation will need to be improved later to
include the alignment requirements in the allocation call, and use it
properly in the data sharing stack in the runtime.
Depends on D115888
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D115971
Turning on `enable_noundef_analysis` flag allows better codegen by removing freeze instructions.
I modified clang by renaming `enable_noundef_analysis` flag to `disable-noundef-analysis` and turning it off by default.
Test updates are made as a separate patch: D108453
Reviewed By: eugenis
Differential Revision: https://reviews.llvm.org/D105169
[Clang/Test]: Rename enable_noundef_analysis to disable-noundef-analysis and turn it off by default (2)
This patch updates test files after D105169.
Autogenerated test codes are changed by `utils/update_cc_test_checks.py,` and non-autogenerated test codes are changed as follows:
(1) I wrote a python script that (partially) updates the tests using regex: {F18594904} The script is not perfect, but I believe it gives hints about which patterns are updated to have `noundef` attached.
(2) The remaining tests are updated manually.
Reviewed By: eugenis
Differential Revision: https://reviews.llvm.org/D108453
Resolve lit failures in clang after 8ca4b3e's land
Fix lit test failures in clang-ppc* and clang-x64-windows-msvc
Fix missing failures in clang-ppc64be* and retry fixing clang-x64-windows-msvc
Fix internal_clone(aarch64) inline assembly
Turning on `enable_noundef_analysis` flag allows better codegen by removing freeze instructions.
I modified clang by renaming `enable_noundef_analysis` flag to `disable-noundef-analysis` and turning it off by default.
Test updates are made as a separate patch: D108453
Reviewed By: eugenis
Differential Revision: https://reviews.llvm.org/D105169
This patch updates test files after D105169.
Autogenerated test codes are changed by `utils/update_cc_test_checks.py,` and non-autogenerated test codes are changed as follows:
(1) I wrote a python script that (partially) updates the tests using regex: {F18594904} The script is not perfect, but I believe it gives hints about which patterns are updated to have `noundef` attached.
(2) The remaining tests are updated manually.
Reviewed By: eugenis
Differential Revision: https://reviews.llvm.org/D108453
Sequel patch to https://reviews.llvm.org/D111293.
Remove call to CodeGenFunction::InitTempAlloca() from OpenMP related
codegen part.
Also remove the metadata `!llvm.access.group` from the updated lit
tests.
Reviewed By: rjmccall
Differential Revision: https://reviews.llvm.org/D111316
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
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 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
In the spirit of TRegions [0], this patch provides a simpler and uniform
interface for a kernel to set up the device runtime. The OMPIRBuilder is
used for reuse in Flang. A custom state machine will be generated in the
follow up patch.
The "surplus" threads of the "master warp" will not exit early anymore
so we need to use non-aligned barriers. The new runtime will not have an
extra warp but also require these non-aligned barriers.
[0] https://link.springer.com/chapter/10.1007/978-3-030-28596-8_11
This was in parts extracted from D59319.
Reviewed By: ABataev, JonChesterfield
Differential Revision: https://reviews.llvm.org/D101976
Summary:
Memory globalization is required to maintain OpenMP standard semantics for data sharing between
worker and master threads. The GPU cannot share data between its threads so must allocate global or
shared memory to store the data in. Currently this is implemented fully in the frontend using the
`__kmpc_data_sharing_push_stack` and __kmpc_data_sharing_pop_stack` functions to emulate standard
CPU stack sharing. The front-end scans the target region for variables that escape the region and
must be shared between the threads. Each variable then has a field created for it in a global record
type.
This patch replaces this functinality with a single allocation command, effectively mimicing an
alloca instruction for the variables that must be shared between the threads. This will be much
slower than the current solution, but makes it much easier to optimize as we can analyze each
variable independently and determine if it is not captured. In the future, we can replace these
calls with an `alloca` and small allocations can be pushed to shared memory.
Reviewed By: tianshilei1992
Differential Revision: https://reviews.llvm.org/D97680
This patch refactors a subset of Clang OpenMP tests, generating checklines using the update_cc_test_checks script. This refactoring facilitates updating the Clang OpenMP code generation codebase by automating test generation.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D101849
This patch renames the replace-function-regex to replace-value-regex to indicate that the existing regex replacement functionality can replace any IR value besides functions.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D101934
This patch refactors a subset of Clang OpenMP tests, generating checklines using the update_cc_test_checks script. This refactoring facilitates updating the Clang OpenMP code generation codebase by automating test generation.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D101849
This revision simplifies Clang codegen for parallel regions in OpenMP GPU target offloading and corresponding changes in libomptarget: SPMD/non-SPMD parallel calls are unified under a single `kmpc_parallel_51` runtime entry point for parallel regions (which will be commonized between target, host-side parallel regions), data sharing is internalized to the runtime. Tests have been auto-generated using `update_cc_test_checks.py`. Also, the revision contains changes to OpenMPOpt for remark creation on target offloading regions.
Reviewed By: jdoerfert, Meinersbur
Differential Revision: https://reviews.llvm.org/D95976
[libomptarget][nvptx] Undef, weak shared variables
Shared variables on nvptx, and LDS on amdgcn, are uninitialized at
the start of kernel execution. Therefore create the variables with
undef instead of zeros, motivated in part by the amdgcn back end
rejecting LDS+initializer.
Common is zero initialized, which seems incompatible with shared. Thus
change them to weak, following the direction of
https://reviews.llvm.org/rG7b3eabdcd215
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D90248
There are various runtime calls in the device runtime with unused, or
always fixed, arguments. This is bad for all sorts of reasons. Clean up
two before as we match them in OpenMPOpt now.
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D83268
Summary:
Added support for dynamic memory allocation for globalized variables in
case if execution of target regions in parallel is required.
Reviewers: jdoerfert
Subscribers: jholewinski, yaxunl, guansong, sstefan1, cfe-commits, caomhin
Tags: #clang
Differential Revision: https://reviews.llvm.org/D82324
If we're going to assume references are dereferenceable, we should also
assume they're aligned: otherwise, we can't actually dereference them.
See also D80072.
Differential Revision: https://reviews.llvm.org/D80166
performance.
Internally generated functions must be marked as always_inlines in most
cases. Patch marks some extra reduction function + outlined parallel
functions as always_inline for better performance, but only if the
optimization is requested.
llvm-svn: 361269
nvvm_barrier0.
Use runtime functions instead of the direct call to the nvvm intrinsics.
It allows to prevent some dangerous LLVM optimizations, that breaks the
code for the NVPTX target.
llvm-svn: 350328
buffer.
Seems to me, nvlink has a bug with the proper support of the weakly
linked symbols. It does not allow to define several shared memory buffer
with the different sizes even with the weak linkage. Instead we always
use 128 bytes buffer to prevent nvlink from the error message emission.
llvm-svn: 349540
reductions.
Fixed previously committed code for the reduction support in
teams/parallel constructs taking into account new design of the NVPTX
support in the compiler. Teams reduction are not fully functional yet,
it is going to be fixed in the following patches.
llvm-svn: 347081
target|teams|distribute variables.
If the total size of the variables, declared in target|teams|distribute
regions, is less than the maximal size of shared memory available, the
buffer is allocated in the shared memory.
llvm-svn: 346507
target/teams/distribute regions.
Target/teams/distribute regions exist for all the time the kernel is
executed. Thus, if the variable is declared in their context and then
escape it, we can allocate global memory statically instead of
allocating it dynamically.
Patch captures all the globalized variables in target/teams/distribute
contexts, merges them into the records, one per each target region.
Those records are then joined into the union, one per compilation unit
(to save the global memory). Those units are organized into
2 x dimensional arrays, where the first dimension is
the number of blocks per SM and the second one is the number of SMs.
Runtime functions manage this global memory space between the executing
teams.
llvm-svn: 345978
Added support for memory coalescing for better performance for
globalized variables. From now on all the globalized variables are
represented as arrays of 32 elements and each thread accesses these
elements using `tid & 31` as index.
llvm-svn: 344049
Summary: In the SPMD case, we need to initialize the data sharing and globalization infrastructure. This covers the case when an SPMD region calls a function in a different compilation unit.
Reviewers: ABataev, carlo.bertolli, caomhin
Reviewed By: ABataev
Subscribers: Hahnfeld, jholewinski, guansong, cfe-commits
Differential Revision: https://reviews.llvm.org/D49188
llvm-svn: 337015
In generic data-sharing mode we are allowed to not globalize local
variables that escape their declaration context iff they are declared
inside of the parallel region. We can do this because L2 parallel
regions are executed sequentially and, thus, we do not need to put
shared local variables in the global memory.
llvm-svn: 336567
The linkage of the global entries must be weak to enable support of
redefinition of the same target regions in multiple compilation units.
llvm-svn: 331768
Summary: The workers also need to initialize the global stack. The call to the initialization function needs to happen after the kernel_init() function is called by the master. This ensures that the per-team data structures of the runtime have been initialized.
Reviewers: ABataev, grokos, carlo.bertolli, caomhin
Reviewed By: ABataev
Subscribers: jholewinski, guansong, cfe-commits
Differential Revision: https://reviews.llvm.org/D44749
llvm-svn: 328219
If the generic codegen is enabled and private copy of the original
variable escapes the declaration context, this private copy should be
globalized just like it was the original variable.
llvm-svn: 327985
Summary: Remove this scheme for now since it will be covered by another more generic scheme using global memory. This code will be worked into an optimization for the generic data sharing scheme. Removing this completely and then adding it via future patches will make all future data sharing patches cleaner.
Reviewers: ABataev, carlo.bertolli, caomhin
Reviewed By: ABataev
Subscribers: jholewinski, guansong, cfe-commits
Differential Revision: https://reviews.llvm.org/D43625
llvm-svn: 326948
Summary:
The backend should only emit data sharing code for the cases where it is needed.
A new function attribute is used by Clang to enable data sharing only for the cases where OpenMP semantics require it and there are variables that need to be shared.
Reviewers: hfinkel, Hahnfeld, ABataev, carlo.bertolli, caomhin
Reviewed By: ABataev
Subscribers: cfe-commits, jholewinski
Differential Revision: https://reviews.llvm.org/D41123
llvm-svn: 320527
Summary:
This patch is part of the development effort to add support in the current OpenMP GPU offloading implementation for implicitly sharing variables between a target region executed by the team master thread and the worker threads within that team.
This patch is the first of three required for successfully performing the implicit sharing of master thread variables with the worker threads within a team. The remaining two patches are:
- Patch D38978 to the LLVM NVPTX backend which ensures the lowering of shared variables to an device memory which allows the sharing of references;
- Patch (coming soon) is a patch to libomptarget runtime library which ensures that a list of references to shared variables is properly maintained.
A simple code snippet which illustrates an implicit data sharing situation is as follows:
```
#pragma omp target
{
// master thread only
int v;
#pragma omp parallel
{
// worker threads
// use v
}
}
```
Variable v is implicitly shared from the team master thread which executes the code in between the target and parallel directives. The worker threads must operate on the latest version of v, including any updates performed by the master.
The code generated in this patch relies on the LLVM NVPTX patch (mentioned above) which prevents v from being lowered in the thread local memory of the master thread thus making the reference to this variable un-shareable with the workers. This ensures that the code generated by this patch is correct.
Since the parallel region is outlined the passing of arguments to the outlined regions must preserve the original order of arguments. The runtime therefore maintains a list of references to shared variables thus ensuring their passing in the correct order. The passing of arguments to the outlined parallel function is performed in a separate function which the data sharing infrastructure constructs in this patch. The function is inlined when optimizations are enabled.
Reviewers: hfinkel, carlo.bertolli, arpith-jacob, Hahnfeld, ABataev, caomhin
Reviewed By: ABataev
Subscribers: cfe-commits, jholewinski
Differential Revision: https://reviews.llvm.org/D38976
llvm-svn: 318773
Joseph Huber