The generic entry points for static loop scheduling previously
hardcoded that the runtime was initialized. This can be wrong if
the compiler analyzes that the runtime is not needed and calls
the init functions accordingly.
This didn't affect clang-ykt because they have entry points for
different combinations of SPMD x Runtime not needed. I didn't do
measurements yet but with inlining we might get away with always
calling the generic interface and letting compiler and runtime
figure out the rest.
In any case, a correct runtime is always better than having
functions that may only be called if previous calls passed in
a specific set of arguments!
Differential Revision: https://reviews.llvm.org/D47131
llvm-svn: 333285
Introduce OPENMP_INSTALL_LIBDIR and use in all install() commands.
This also fixes installation of libomptarget-nvptx that previously
didn't honor {OPENMP,LLVM}_LIBDIR_SUFFIX.
Differential Revision: https://reviews.llvm.org/D47130
llvm-svn: 333284
The existing implementation of the dynamic scheduling
breaks the contract introduced by the original openmp
runtime and, thus, is incorrect. Patch fixes it and
introduces correct dynamic scheduling model.
Thanks to Alexey Bataev for submitting this patch.
Differential Revision: https://reviews.llvm.org/D47333
llvm-svn: 333225
We already know where the CUDA SDK is, so there is no point in
letting Clang search for it again and possibly finding no or
a different installation.
--cuda-path is supported since the beginning of CUDA support in
Clang, so making this required doesn't impose additional restrictions.
Differential Revision: https://reviews.llvm.org/D46930
llvm-svn: 332495
Move all logic related to selecting the bitcode compiler and linker
into a new file and dynamically test required compiler flags. This
also adds -fcuda-rdc for Clang trunk as previously attempted in D44992
which fixes the build.
As a result this change also enables building the library by default
if all prerequisites are met.
Differential Revision: https://reviews.llvm.org/D46901
llvm-svn: 332494
Summary: Add function to the NVPTX libomptarget library that will return true if the current target region is being executed in SPMD mode.
Reviewers: ABataev, grokos, carlo.bertolli, caomhin
Reviewed By: grokos
Subscribers: guansong, openmp-commits
Differential Revision: https://reviews.llvm.org/D46840
llvm-svn: 332360
implicit_task_end callbacks in nested parallel regions did not always give the
correct thread_num, since the inner parallel region may have already been
finalized.
Now, the thread_num is stored at the beginning of the implicit task and
retrieved at the end, whenever necessary.
A testcase was added as well.
Differential Revision: https://reviews.llvm.org/D46260
llvm-svn: 331632
The api_calls_misc.c testcase tests the following api calls:
ompt_get_callback()
ompt_get_state()
ompt_enumerate_states()
ompt_enumerate_mutex_impls()
These have not been tested previously.
The api_calls.c testcase has been renamed to api_calls_places.c because it only tests api calls that are related to places.
Differential Revision: https://reviews.llvm.org/D42523
llvm-svn: 331631
Summary:
Enable the device side debug messages at compile time, use env var to control at runtime.
To achieve this, an environment data block is passed to the device lib when it is loaded.
By default, the message is off, to enable it, a user need to set LIBOMPDEVICE_DEBUG=1.
Reviewers: grokos
Reviewed By: grokos
Subscribers: openmp-commits
Tags: #openmp
Differential Revision: https://reviews.llvm.org/D46210
llvm-svn: 331550
Summary: Minor printf format correction. NVCC ignore those. Clang will give warning on these if debug is enabled.
Reviewers: grokos
Reviewed By: grokos
Subscribers: openmp-commits
Tags: #openmp
Differential Revision: https://reviews.llvm.org/D45528
llvm-svn: 330944
Summary: The LIBOMPTARGET_NVPTX_DEBUG flag is inconsistent between using nvcc to generate .a file and clang to generate .bc file. Sync the two setting so we can get debug messages from the bc file path as well.
Reviewers: grokos
Subscribers: Hahnfeld, openmp-commits, mgorny
Tags: #openmp
Differential Revision: https://reviews.llvm.org/D45530
llvm-svn: 330477
Currently, the affinity API reports garbage for the initial place list and any
thread's place lists when using KMP_AFFINITY=none|compact|scatter.
This patch does two things:
for KMP_AFFINITY=none, Creates a one entry table for the places, this way, the
initial place list is just a single place with all the proc ids in it. We also
set the initial place of any thread to 0 instead of KMP_PLACE_ALL so that the
thread reports that single place (place 0) instead of garbage (-1) when using
the affinity API.
When non-OMP_PROC_BIND affinity is used
(including KMP_AFFINITY=compact|scatter), a thread's place list is populated
correctly. We assume that each thread is assigned to a single place. This is
implemented in two of the affinity API functions
Differential Revision: https://reviews.llvm.org/D45527
llvm-svn: 330283
This patch introduces GOMP_taskloop to our API. It adds GOMP_4.5 to our
version symbols. Being a wrapper around __kmpc_taskloop, the function
creates a task with the loop bounds properly nested in the shareds so that
the GOMP task thunk will work properly. Also, the firstprivate copy constructors
are properly handled using the __kmp_gomp_task_dup() auxiliary function.
Currently, only linear spawning of tasks is supported
for the GOMP_taskloop interface.
Differential Revision: https://reviews.llvm.org/D45327
llvm-svn: 330282
Summary:
This one line change is to remove this warning message
"warning: integer conversion resulted in a change of sign"
Reviewers: grokos
Reviewed By: grokos
Subscribers: openmp-commits
Tags: #openmp
Differential Revision: https://reviews.llvm.org/D45415
llvm-svn: 329713
This change removes the unnecessary lock operation on __kmp_initz_lock inside
the __kmp_atfork_child() function for Linux; the lock variable is initialized
in the same function later.
Patch by Hansang Bae
Differential Revision: https://reviews.llvm.org/D44949
llvm-svn: 328900
The summarizeStats.py script processes raw data provided by the
instrumented (stats-gathering) OpenMP* runtime library. It provides:
1) A radar chart which plots counters as frequency (per GigaTick) of use within
the program. The frequencies are plotted as log10, however values less than
one are kept as it is and represented in red color. This was done to help
visualize the differences better.
2) Pie charts separating total time as compute and non-compute. The compute and
non-compute times have their own pie charts showing the constructs that
contributed to them. The percentages listed are with respect to the total
time.
3) '.csv' file with percentage of time spent within the different constructs.
The script can be used as:
$ python $PATH_TO_SCRIPT/summarizeStats.py instrumented1.csv instrumented2.csv
Patch by Taru Doodi
Differential Revision: https://reviews.llvm.org/D41838
llvm-svn: 328568
Summary: The global stack initialization function may be called multiple times. The initialization of the shared memory slots should only happen when the function is called for the first time for a given warp master thread.
Reviewers: grokos, carlo.bertolli, ABataev, caomhin
Reviewed By: grokos
Subscribers: guansong, openmp-commits
Differential Revision: https://reviews.llvm.org/D44754
llvm-svn: 328148
Summary: The check for the master warp must take into consideration the actual number of warps: the master warp is equal to the last active warp not necessarily WARPSIZE - 1.
Reviewers: grokos, carlo.bertolli, ABataev, caomhin
Reviewed By: grokos
Subscribers: guansong, openmp-commits
Differential Revision: https://reviews.llvm.org/D44537
llvm-svn: 328146
Summary:
This patch allows worker to have a global memory stack managed by the runtime. This patch is needed for completeness and consistency with the globalization policy: if a worker-side variable escapes the current context it then needs to be globalized.
Until now, only the master thread was allowed to have such a stack. These global values can now potentially be shared amongst workers if the semantics of the OpenMP program require it.
Reviewers: ABataev, grokos, carlo.bertolli, caomhin
Reviewed By: grokos
Subscribers: guansong, openmp-commits
Differential Revision: https://reviews.llvm.org/D44487
llvm-svn: 328144
Added settings code to read OMP_TARGET_OFFLOAD environment variable. Added
target-offload-var ICV as __kmp_target_offload, set via OMP_TARGET_OFFLOAD,
if available, otherwise defaulting to DEFAULT. Valid values for the ICV are
specified as enum values {0,1,2} for disabled, default, and mandatory. An
internal API access function __kmpc_get_target_offload is provided.
Patch by Terry Wilmarth
Differential Revision: https://reviews.llvm.org/D44577
llvm-svn: 328046
Summary:
Allow the runtime to use the existing shared memory statically allocated slots.
When a variable is globalized, the underlying memory can be either shared or global memory (both have block-wide visibility). In this case, we allow that the storage to use a limited amount of shared memory that has been statically allocated already. Only if shared memory doesn't prove to be enough do we then invoke malloc() to create a new global memory slot.
Reviewers: ABataev, carlo.bertolli, grokos, caomhin
Reviewed By: grokos
Subscribers: guansong, openmp-commits
Differential Revision: https://reviews.llvm.org/D44486
llvm-svn: 327639
Summary: To save on calls to malloc, this patch enables the re-use of pre-allocated global memory slots.
Reviewers: ABataev, grokos, carlo.bertolli, caomhin
Reviewed By: grokos
Subscribers: guansong, openmp-commits
Differential Revision: https://reviews.llvm.org/D44470
llvm-svn: 327637
Summary:
This patch adds support for the sharing of variables from the master thread of a team to the worker threads of the team.
The runtime uses a stack structure implemented as a doubly-linked list of slots with each slot having the exact same size as the size requested. This implementation leverages existing data structures. The runtime functions are added as separate functions to avoid interfering with the current interface.
Limitations to be addressed in future patches:
- This current patch only employs global memory. In a future patch we will enable to usage for shared memory as an optimization.
- Allow the allocation of several requested sizes in the same slot.
Reviewers: ABataev, grokos, caomhin, carlo.bertolli
Reviewed By: grokos
Subscribers: Hahnfeld, guansong, openmp-commits
Differential Revision: https://reviews.llvm.org/D44260
llvm-svn: 327440
Summary: To make the two parts of the union have the same size, the size of vect needs to be increased by 16 bits.
Reviewers: grokos, carlo.bertolli, caomhin, ABataev
Reviewed By: grokos, ABataev
Subscribers: fedor.sergeev, guansong, openmp-commits
Differential Revision: https://reviews.llvm.org/D44254
llvm-svn: 327040
Summary:
This patch reverts the changes to libomptarget that were coupled with the changes to Clang code gen for data sharing using shared memory. A similar patch exists for Clang: D43625
Shared memory is meant to be used as an optimization on top of a more general scheme. So far we didn't have a global memory implementation ready so shared memory was a solution which applied to the current level of OpenMP complexity supported by trunk on GPU devices (due to the missing NVPTX backend patch this functionality has never been exercised). Now that we have a global memory solution this patch is "in the way" and needs to be removed (for now). This patch (or an equivalent version of it) will be put out for review once the global memory scheme is in place.
Reviewers: ABataev, grokos, carlo.bertolli, caomhin
Reviewed By: grokos
Subscribers: Hahnfeld, guansong, openmp-commits
Differential Revision: https://reviews.llvm.org/D43626
llvm-svn: 326950
We have to ensure that the runtime is initialized _before_ waiting
for the two started threads to guarantee that the master threads
post their ompt_event_thread_begin before the worker threads. This
is not guaranteed in the parallel region where one worker thread
could start before the other master thread has invoked the callback.
The problem did not happen with Clang becauses the generated code
calls __kmpc_global_thread_num() and cashes its result for functions
that contain OpenMP pragmas.
Differential Revision: https://reviews.llvm.org/D43882
llvm-svn: 326435
The thread_num parameter of ompt_get_task_info() was not being used previously,
but need to be set.
The print_task_type() function (form the task-types.c testcase) was merged into
the print_ids() function (in callback.h). Testing of ompt_get_task_info() was
added to the task-types.c testcase. It was not tested extensively previously.
Differential Revision: https://reviews.llvm.org/D42472
llvm-svn: 326338
The main change of this patch is to insert {{.*}} in current_address=[[RETURN_ADDRESS_END]].
This is needed to match any of the alternatively printed addresses.
Additionally, clang-format is applied to the two tests.
Differential Revision: https://reviews.llvm.org/D43115
llvm-svn: 326312
This is required to be NULL for implicit barriers at the end of a
parallel region. Noticed in review of D43191.
Differential Revision: https://reviews.llvm.org/D43308
llvm-svn: 325922
The compiler inlines the user code in the task. Check for that case at
runtime by comparing the frame addresses and print the expected exit
address.
Also showcase how I think the OMPT tests could be reformatted to match
LLVM's code style. In my opinion it would be great to that kind of change
to all tests that need to be touched for whatever reason...
Differential Revision: https://reviews.llvm.org/D43191
llvm-svn: 325921
Test whether OMPT-callbacks for two threads that initiate a parallel region are correct.
Differential Revision: https://reviews.llvm.org/D41942
llvm-svn: 325423
Only use ompt_ functions when testing OMPT in api_calls testcase.
Add size parameter to print_list.
Fix small bug in implementation of ompt_get_partition_place_nums(): return correct length.
Differential Revision: https://reviews.llvm.org/D42162
llvm-svn: 325422
GlobalISel doesn't yet implement blockaddress and falls back to
SelectionDAG. This results in additional branch instruction to
the next basic block which breaks the OMPT tests.
Disable GlobalISel for now when compiling the tests because fixing
them is not easily possible. See http://llvm.org/PR36313 for full
discussion history.
Differential Revision: https://reviews.llvm.org/D43195
llvm-svn: 325218
This affects all outlined functions, not just tasks! Only show warning
when using Clang 5.0 or later.
Differential Revision: https://reviews.llvm.org/D43190
llvm-svn: 325131
Summary:
Different NVIDIA GPUs support different compute capabilities. To enable the inlining of runtime functions and the best performance on different generations of NVIDIA GPUs, a bc library for each compute capability needs to be compiled. The same compiler build will then be usable in conjunction with multiple generations of NVIDIA GPUs.
To differentiate between versions of the same bc lib, the output file name will contain the compute capability ID.
Depends on D14254
Reviewers: Hahnfeld, hfinkel, carlo.bertolli, caomhin, ABataev, grokos
Reviewed By: Hahnfeld, grokos
Subscribers: guansong, mgorny, openmp-commits
Differential Revision: https://reviews.llvm.org/D41724
llvm-svn: 324904
Tests the search for tools as defined in the spec. The OMP_TOOL_LIBRARIES
environment variable contains paths to the following files(in that order)
-to a nonexisting file
-to a shared library that does not have a ompt_start_tool function
-to a shared library that has an ompt_start_tool implementation returning NULL
-to a shared library that has an ompt_start_tool implementation returning a
pointer to a valid instance of ompt_start_tool_result_t
The expected result is that the last tool gets active and can print in the
thread-begin callback.
Differential Revision: https://reviews.llvm.org/D42166
llvm-svn: 324588
Add a testcase that checks wheter the runtime can handle an ompt_start_tool
method that returns NULL indicating that no tool shall be loaded.
All tool_available testcases need a separate folder to avoid file conflicts for
the generated tools.
Differential Revision: https://reviews.llvm.org/D41904
llvm-svn: 324587
Summary: This patch extends the libomptarget functionality in patch D14254 with support for the data sharing scheme for supporting implicitly shared variables. The runtime therefore maintains a list of references to shared variables.
Reviewers: carlo.bertolli, ABataev, Hahnfeld, grokos, caomhin, hfinkel
Reviewed By: Hahnfeld, grokos
Subscribers: guansong, llvm-commits, openmp-commits
Differential Revision: https://reviews.llvm.org/D41485
llvm-svn: 324495
If tool initialization returns 0, OMPT should not be active. The current
implementation provided some callback invocations in this case.
Differential Revision: https://reviews.llvm.org/D42709
llvm-svn: 324320
https://reviews.llvm.org/D42757
The method ThreadsInTeam is used to determine the number of threads to be used in a parallel region under SPMD mode (see line 127 of supporti.h in libomptarget/deviceRTLs/nvptx/src/). This patch fixes the corresponding debug print upon initialization of the kernel in SPMD mode.
llvm-svn: 323978
That's what we really need to link the CUDA plugin against,
not the CUDA runtime API in CUDA_LIBRARIES! While the latter
comes with the CUDA SDK, the Driver API is installed with
the kernel driver and there is at most one per system. As
fallback we can use the stubs library distributed with the
CUDA SDK for linking.
Differential Revision: https://reviews.llvm.org/D42643
llvm-svn: 323787
Use equivalents for the last calls to the Runtime API. Remove
stray assert in case of an error found during review, we should
only return OFFLOAD_FAIL.
Differential Revision: https://reviews.llvm.org/D42686
llvm-svn: 323786
This patch implements the device runtime library whose interface is used in the code generation for OpenMP offloading devices.
Currently there is a single device RTL written in CUDA meant to CUDA enabled GPUs.
The interface is a variation of the kmpc interface that includes some extra calls to do thread and storage management that only make sense for a GPU target.
Differential revision: https://reviews.llvm.org/D14254
llvm-svn: 323649
Use fuzzy return addresses in lock testcases so that these
testcases can also be run using the Intel Compiler.
Patch by Simon Convent!
Differential Revision: https://reviews.llvm.org/D41896
llvm-svn: 323529
Add Workaround for Intel Compiler Bug with Case#: 03138964
A critical region within a nested task causes a segfault in icc 14-18:
int main()
{
#pragma omp parallel num_threads(2)
#pragma omp master
#pragma omp task
#pragma omp task
#pragma omp critical
printf("test\n");
}
When the critical region is in a separate function, the segault does not occur.
So we add noinline to make sure that the function call stays there.
Differential Revision: https://reviews.llvm.org/D41182
llvm-svn: 322622
The defintion is not part of the spec and thus should not have the prefix
"ompt_" but rather a prefix that indicates that this is implementation
specific.
Differential Revision: https://reviews.llvm.org/D41166
llvm-svn: 322621
When the current thread is not an (initialized) OpenMP thread, the runtime
entry points return values that correspond to "not available" or similar
Differential Revision: https://reviews.llvm.org/D41167
llvm-svn: 322620
If user requested affinity with granularity=tile we need to either use HWLOC
or ignore the request. The change allows user to not specify
KMP_TOPOLOGY_METHOD=hwloc and choose it automatically instead.
Patch by Andrey Churbanov
Differential Revision: https://reviews.llvm.org/D40905
llvm-svn: 322205
This change simplifies __kmp_expand_threads to take a single argument.
Previously, it allowed two arguments and had logic to decide on different
potential expansion sizes. However, no calls to __kmp_expand_threads in the
runtime make use of this extra logic. Thus the extra argument and logic is
removed here.
Patch by Terry Wilmarth
Differential Revision: https://reviews.llvm.org/D41836
llvm-svn: 322204
This change improves stability of the runtime when the application forks child
processes. Acquiring/releasing __kmp_initz_lock and __kmp_forkjoin_lock in the
atfork handlers insures that the actual fork does not occur while those two
locks are held, and __kmp_itt_reset() reverts the itt's global state to the
initial state which also initializes the mutex stored in the global state.
Some missing initialization code was also inserted in the child's atfork handler.
Patch by Hansang Bae
Differential Revision: https://reviews.llvm.org/D41462
llvm-svn: 322202
This field is defined as kmp_int32, so we should use neither
pointers to kmp_int64 nor 64 bit atomic instructions.
(Found while testing on a Raspberry Pi, 32 bit ARM)
Differential Revision: https://reviews.llvm.org/D41656
llvm-svn: 321964
This patch enables OMPT by default if version 50 or later is built and the config says, that OMPT will be supported.
Differential Revision: https://reviews.llvm.org/D41508
llvm-svn: 321675
We now have several options that apply for both libraries and they
shouldn't be documented in multiple files. When already merging
the two Build_With_CMake.txt documents, convert them to
reStructuredText which is used for all of LLVM's documentation.
Differential Revision: https://reviews.llvm.org/D40920
llvm-svn: 321481
As for normal task creation, the task frame addresses need to be stored
for the encountering task.
Differential Revision: https://reviews.llvm.org/D41165
llvm-svn: 321421
The compiler warns that _BSD_SOURCE is deprecated and _DEFAULT_SOURCE should
be used instead. We keep _BSD_SOURCE for older compilers, that don't know
about _DEFAULT_SOURCE.
The linker drops the tool when linking, since there is no visible need for
the library. So we need to tell the linker, that the tool should be linked
anyway.
Differential Revision: https://reviews.llvm.org/D41499
llvm-svn: 321362
The format string for hints only prints the second argument (string) and drops
the first argument (hint id). Depending on how you read the POSIX text for
printf, this could be valid. But for practical reason, i.e., unpacking the
va_list passed to printf based on the formating information, it makes sense
to fix the implementation and not pass the id for hint.
Failing testcases were:
misc_bugs/teams-reduction.c
ompt/parallel/not_enough_threads.c
Differential Revision: https://reviews.llvm.org/D41504
llvm-svn: 321361
This function is defined in OpenMP-TR6 section 4.1.5.1.6
The functions was not implemented yet.
Since ompt-functions can only be called after the runtime was initialized and
has loaded a tool, it can assume the runtime to be initialized. In contrast
to omp_get_num_procs which needs to check whether the runtime is initialized.
Differential Revision: https://reviews.llvm.org/D40949
llvm-svn: 321269
This revision fixes failing testcases with parallel for loops and the gomp
interface. The return address needs to be stored at entry to runtime.
The storage is cleared on usage, so we need to update the storage before
calling again internal functions, that will trigger event callbacks.
Differential Revision: https://reviews.llvm.org/D41181
llvm-svn: 321265
We use the bitmap ompt_enabled thoughout the runtime, to avoid loading the
vector of callback functions when testing if specific code should be executed.
Before invoking an event callback function, the pointer is tested for NULL.
This revision resets the corresponding bit in ompt_enabled to 0 if
NULL is passed in set_callback.
Differential Revision: https://reviews.llvm.org/D41171
llvm-svn: 321264
Clang 5 or higher adds an intermediate function call in certain cases when
compiling with debug flag. This revision updates the testcases to work
correctly.
Differential Revision: https://reviews.llvm.org/D40595
llvm-svn: 321263
Reasons for expected failures are mainly bugs when using lables in OpenMP regions
or missing support of some OpenMP features.
For some worksharing clauses, support to distinguish the kind of workshare was
added just recently.
If an issue was fixed in a minor release version of a compiler, we flag the
test as unsupported for this compiler version to avoid false positives.
Same for fixes that where backported to older compiler versions.
Differential Revision: https://reviews.llvm.org/D40384
llvm-svn: 321262
There are two /proc/cpuinfo layots in use for AArch64: old and new.
The old one has all 'processor : n' lines in one section, hence
checking for duplications does not make sense.
Differential Revision: https://reviews.llvm.org/D41000
llvm-svn: 320593
All architectures except x86_64 used the linear barrier implementation
by default which doesn't give good performance for a larger number
of threads.
Improvements for PARALLEL overhead (EPCC) with this patch on a Power8
system (2 sockets x 10 cores x 8 threads, OMP_PLACES=cores)
20 threads: 4.55us -> 3.49us
40 threads: 8.84us -> 4.06us
80 threads: 19.18us -> 4.74us
160 threads: 54.22us -> 6.73us
Differential Revision: https://reviews.llvm.org/D40358
llvm-svn: 320152
To make thread affinity work according to the OpenMP spec, the
runtime needs information about the hardware topology. On Linux
the default way is to parse /proc/cpuinfo which contains this
information for x86 machines but (at least) not for AArch64 and
Power architectures.
Fortunately, there is a different code path which is able to get
that data from sysfs. The needed patch has landed in 2006 for
Linux 2.6.16 which is safe to assume nowadays (even RHEL 5 had
a kernel version derived from 2.6.18, and we are now at RHEL 7!).
Differential Revision: https://reviews.llvm.org/D40357
llvm-svn: 320151
Otherwise I see hangs in the omp_single_copyprivate test when
compiling in release mode. With the debug assertions, I get a
failure `head > 0 && tail > 0`.
Differential Revision: https://reviews.llvm.org/D40722
llvm-svn: 320150
This last of four patches adds a new file for the interface
functions that Clang uses during code generation. The only
change except simply moving the current code is renaming the
function CheckDeviceAndCtors() and using the correct type for
64bit device ids.
Differential Revision: https://reviews.llvm.org/D40801
llvm-svn: 319972
Jonas Hahnfeld