To make the functional tests more useful for benchmarking, allow user to
specify the number of loops and related parameters via command options.
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
somehow the test whether we requested MPI support or not stopped
working, although no obvious code change can be located.
Make the if-statement more stringent by explicitely testing whether
USE_MPI_SUPPORT is "ON".
* SyncAll test case would run Sync
* Despecialized name for argument reader
* Rename sync-test to team-sync-test as it uses teams
* Another stab at probing NUM_GPUS
* add gfx1100 support
Add support for Radeon 7900 GPUs (RX and PRO), and 7800 PRO.
I was contemplating to add gfx1101 and gfx1102 GPUs as well, but those are the lower end models that are more unlikely to be used for compute intensive jobs. In addition, I do not have access to them to test the support.
* update WF_SIZe for different options
Radeon systems use a WarpSize of 32, unlike current Instinct systems,
which use a warp size of 64. For the device side, a gfx specific ifdef
is sufficient. For the host side, we need to query the device
properties.
* adjust functional tests to wf_size of 32
* update unit tests to handle wf_size of 32
* address reviewer comments
* Add `ROCSHMEM_CTX_INVALID` for invalid context handling
- Define `ROCSHMEM_CTX_INVALID` as {nullptr, nullptr}
- Add == and != operators to rocshmem_ctx_t
- Use `ROCSHMEM_CTX_INVALID` on failed context creation
- Skip ctx destroy if context is invalid
* Update docs for context create and destroy APIs usage and behavior
update the tester script to only tests the amo functions on RO that are
expected to pass. We can revisit the non-passing tests later, but this
prevents us from having passing CIs at the moment, while RO is simply
lower priority than other asks.
- make adding PMIx library to compile time based on the result of
finding PMIx support. This is required eg if compiling rocSHMEM with ompi
4.0/4.1, which do not have a built-in PMIx version.
- when setting USE_EXTERNAL_MPI=OFF which ensures that we do not
check for external MPI libraries (even if one would be available).
* Split ionic code to a subdirectory; dyld libionicl; move the fntable to provider_gda_xxx.hpp
pass the pattr to ionic_setup_pd, include endian.hpp
Enable building IONIC conduit for runtime selection
* Uniform style for the fntable between ionic and the rest
* Move mlx5 gda conduit to a subdir; resolve conflict with backend_can_run
function
* Don't forget to init qp for ionic, move mlx5 specialized init qp code to
the mlx5 subdir
* Don't add cmakecaches...
Typo: GDA_BXNT
* Add gda-ionic to all_backends build scripts
* Apply suggestion from reviews
Co-authored-by: Omri Mor <omri50@gmail.com>
Co-authored-by: Edgar Gabriel <edgar.gabriel@amd.com>
* Remove duplicate definitiion of DLSYM macros
---------
Co-authored-by: Omri Mor <omri50@gmail.com>
Co-authored-by: Edgar Gabriel <edgar.gabriel@amd.com>
* add support for compiling all backends
also include the logic to select backends either based on user requests
or through some heuristics
* checkpoint for compiling all backends
* final checkpoint
all tests seem to pass when compiling all three backends simultaneasly
and forcing to use any of the three Backends.
* update PR to new envvar system
* Revamp findibverbs to find ionic again
* gda ionic: rename ionic_sq_buf ionic_cq_buf
Avoid duplicating member names used by mlx5 gda.
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
* gda: move spin lock to util.hpp
Move spin lock out of ionic gda to util.hpp.
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
* gda ionic: assume latest fwabi changes
There is no firmware abi compatibility in this ionic gda code yet, so
assume we are using the latest firmware abi as of now.
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
* gda ionic: allow doorbell with incomplete wqes
Use spin lock to ensure doorbell is only written with an increasing
producer index. Ring the doorbell after this wave has initialized its
wqes. Wqes of other waves might not be fully initialized, but firmware
will not process them until the phase/color flag is updated in the
respecitve wqes.
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
* gda ionic: poll cq for additional completions
Keep polling the cq for more than just the minimum number of completions
for this wave of threads to make progress, as long as the cq is not
empty. A part of wave-optimized cq polling, at the expense of one wave
polling additional completions, it was observed that nearly all other
waves avoid taking the cq lock at all.
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
* gda: max_rd_atomic in rts transition
In modify_qp(RTS), specify max_rd_atomic, not max_dest_rd_atomic.
By not speicfying max_rd_atomic (rather, max_rd_atomic=zero), the local
nic may get stuck transmitting the first read or atomic request. One
read or atomic request is greater than the initiator depth of zero.
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
* gda ionic: allow specifying traffic class
Allow specifying a traffic class. The network might have a specific
traffic class configured as no-drop, for example.
Co-authored-by: Aurelien Bouteiller <aurelien.bouteiller@amd.com>
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
* gda ionic: tweak uxdma assignment
The ideal arrangement will have an equal number of QPs active on each
uxdma pipeline.
Pre-rebase, the better arrangement for rocshmem funcitonal test
benchmarks was [0, 1], [1, 0], [0, 1], [1, 0], ...
Now, following changes that add 'ROCSHMEM_GDA_ALTERNATE_QP_PORTS=1' by
default, the better arrangement is [0, 1], [0, 1], [0, 1], [0, 1], ...
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
---------
Signed-off-by: Allen Hubbe <allen.hubbe@amd.com>
Co-authored-by: Aurelien Bouteiller <abouteil@amd.com>
Co-authored-by: Aurelien Bouteiller <aurelien.bouteiller@amd.com>
* Add environment variable configuration infrastructure
- Namespace rocshmem::envvar
- Track all config env vars in per-category lists
- Remove duplicates from list of allowed env var types
- Reject negative inputs for unsigned integer types
- Accept empty strings for std::string
- Print error source location using C++20 std::source_location
- Unit tests
* Port environment variables
- ROCSHMEM_UNIQUEID_WITH_MPI
- ROCSHMEM_RO_DISABLE_IPC
- ROCSHMEM_BOOTSTRAP_TIMEOUT
- ROCSHMEM_BOOTSTRAP_HOSTID
- ROCSHMEM_BOOTSTRAP_SOCKET_IFNAME
- ROCSHMEM_RO_PROGRESS_DELAY
- ROCSHMEM_BOOTSTRAP_SOCKET_FAMILY
- ROCSHMEM_MAX_NUM_CONTEXTS
+ Merge the independent per-backend copies into a single variable
that is used by all three backends (IPC, RO, GDA).
+ Set default to 32 (for GDA); prior default for IPC and RO was 1024.
- ROCSHMEM_MAX_NUM_HOST_CONTEXTS
- ROCSHMEM_MAX_WF_BUFFERS
- ROCSHMEM_SQ_SIZE
- ROCSHMEM_RO_NET_CPU_QUEUE
+ Renamed from RO_NET_CPU_QUEUE
+ Change env var input type to bool, default to false
+ Invert code logic: setting RO_NET_CPU_QUEUE to anything
would /disable/ a variable gpu_queue, which defaulted to true.
Variable is now named config::ro::net_cpu_queue,
with all prior checks for gpu_queue inverted.
- ROCSHMEM_USE_IB_HCA
- ROCSHMEM_HEAP_SIZE
+ Defaults to 1L << 30 i.e. 1 GiB,
from default heap size in memory/heap_memory.hpp.
- ROCSHMEM_MAX_NUM_TEAMS
+ Unlike other env vars, this can be referenced from devices.
+ Function currently narrows from size_t to int: uses need to be audited
for safety and correctness in using size_t directly.
- ROCSHMEM_GDA_ALTERNATE_QP_PORTS
* New env var ROCSHMEM_DEBUG
- Debug levels:
+ NONE
+ VERSION
+ WARN
+ INFO
+ TRACE
- Currently unused - will be added later
- Mirrors RCCL debug control
* Remove rocshmem::rocshmem_env_config
* Change interface for GetClosestNicToGpu
to accept const char** instead of char**:
the pointed-to string does not need to be modified
- Files were not audited for inclusion of util.hpp only for env vars
---------
Signed-off-by: Omri Mor <Omri.Mor@amd.com>
* feat: implement function to return number of blocks in grid.
* test: update atomics functional tests
- Standard atomic tests: `atomic_add`, `atomic_inc`, `fetch_atomic_add`, `fetch_atomic_inc`, and `fetch_compare_and_swap`
- Bitwise atomic tests: `atomic_and`, `atomic_or`, `atomic_xor`, fetch_atomic_and`, `fetch_atomic_or`, and `fetch_atomic_xor`
- Extended atomic tests: `atomic_fetch`, `atomic_set`, and `atomic_swap`
* Added two different address modes for atomics.
* Added all supported data types for atomics tests.
* use dlsym for MPI functions
to allow compiling without MPI support, convert the usage of MPI functions and symbols to be based on a dlopen/dlsym based mechanism. Turns out this cannot be done entirely vendor neutral, slightly different solutions might be required for Open MPI, MPICH and the new MPI ABI.
* checkpoint
more work to be done.
* checkpoint 2
* checkpoint 3
* checkpoint 4
examples compile and link correctly
* checkpoitn 5 (I think)
* Checkpoitn 6
* dyld-mpi: adapt GDA
* dyldmpi: tests that depend on MPI need to link with it themselves
* do not ../mpi_instance.h
* dyldmpi: make the symetricHeapTestFixture compile
* dyldmpi: Change cmakery, compiles and run gda w/o external MPI
* Make it also compile in external MPI mode
* dyldmpi: ipc unit tests compile but do not link
* dyldmpi: new approach, if external mpi required, link with mpi,
otherwise use ompi5 abi
* C-style comments in cmakelist..
* dyldmpi: examples: do not fail compiling if MPI not found at build time,
instead do not compile the MPI required examples
* more updates to CMake logic
* convert RO backend
and a few other cleanups
* update some unit tests
to work with the dlopen MPI environment correctly.
---------
Co-authored-by: Aurelien Bouteiller <abouteil@amd.com>
* Enable GDA+IPC
Fix ROCSHMEM_DISABLE_IPC for both RO and GDA
* add more functionality to bootstrap class
we need a few more functions in the boostrap class to be able to fully
handle the rocshmem requirements:
- add a function to return the list of local ranks
- provide a groupAllgather operation which takes a vector of ranks
participating
- provide a groupAlltoall operation which takes a vector of ranks
participating
Also, update the functionality of the gda-Alltoall and gda-Allreduce
operations to take advantage of these functions.
* ipc_policy adapted to use bootstrap groupallgather
* bugfix: there was a mistake in computing sendto in groupallgather
* bugfix: shm_size and shm_rank were set in a local variable rather than
the class member
* mpi-bootstrap: remove an unecessary allgather
---------
Co-authored-by: Edgar Gabriel <Edgar.Gabriel@amd.com>
* The QueuePair object was out of scope at the end of the for loop.
So the deconstructor was called.
* Although correct for C++ to do this, it ignores that we copied the QueuePair object into
device memory and have an instance there.
* Early deconstruction resulted in calling ibv_dereg_mr on the atomics memory region.
So when the GPU kernel tried to use the memory region it wasn't
registered which resulted in a protection domain error.
* The solution was to allocate our QueuePair obj with the new operator which leaves memory
management to us, then we can manually call the deconstructor.
abstract out the usage of direct verbs functionality to use tables with
the function pointers. This will allow in a second step the library to
be simultaniously be compiled for multiple NIC vendors/DV libraries and
interfaces. For now, the conversion has been done for IB MLX5 and BCOM
DV, the Pensando AINIC is to follow soon.
* feat(GDA): add `get*` and `get*_nbi` APIs for mlx and bnxt NICs
- implemented thread, wave and wg variants of `get*` and `get_nbi`.
* test(GDA): enable functional tests for `get*` and `get*_nbi` APIs
Allen Hubbe
Edgar Gabriel
akolliasAMD
Yiltan
Aurelien Bouteiller
Avinash Kethineedi
Dimple Prajapati
Allen Hubbe
Omri Mor
yugang-amd