* Add HasExpertSchedMode device prop
* Add unit tests for HasExpertSchedMode
* Add gfx12 check for HasExpertSchedMode prop
* Update gfx major version check and test for ExpertSchedMode
* Minor fix and ROCr version bump
* Update projects/rocr-runtime/runtime/hsa-runtime/inc/hsa_ext_amd.h
* Update projects/rocr-runtime/runtime/hsa-runtime/inc/hsa_ext_amd.h
* Apply suggestion from @dayatsin-amd
* Apply suggestion from @dayatsin-amd
---------
Co-authored-by: Stefan Sokolovic <stefan.sokolovic2@amd.com>
Co-authored-by: David Yat Sin <77975354+dayatsin-amd@users.noreply.github.com>
- When doing device/stream sync, we can submit a handler which may
introduce some host side delays. Use DEBUG_CLR_BATCH_CPU_SYNC_SIZE to
batch commands for host wait. Default for HIP is 8 commands.
- Investigation is underway in ROCr but need to address this for now in
HIP runtime.
[ROCm/clr commit: 9b045922a8]
* SWDEV-518831 - fix streams' sync issue in mthreads
1. Fix sync issue of null stream and non-null streams in
multithreads.
2. Remove assert(GetSubmissionBatch() == nullptr) as it
is invalid in multithreads.
3. Update getActiveQueues() to deal with the state of
being terminated.
[ROCm/clr commit: 27aad09bd4]
Explicitly nulling the pointer causes us to report the error below
instead of keeping a dangling pointer around that will most likely lead
to a subsequent segfault.
[ROCm/clr commit: 199b0f1086]
Add initial implementation of virtual memory heap with
dynamic virtual memory mapping support for memory pools.
DEBUG_HIP_MEM_POOL_VMHEAP controls the new method.
Change-Id: I8dc5be2e0f34ab472f1800f43bb6243639a5e500
[ROCm/clr commit: 296dce5570]
Shared mutex allows to have access to the list of streams from
multiple threads at the same time.
Change-Id: Ibee64b846cde03321d5b17dbee2829c0bab7e7d6
[ROCm/clr commit: efd3ea4b30]
=> If null stream is not created during sync skip nullstrm creation
=> Do cpu wait on blocking & null stream if it exists
Change-Id: I90d6ced6a2dd1782ba58f3fed4e3608fc0efa55a
[ROCm/clr commit: 17e7b7c2ef]
=> GraphExec instance is destroyed before async launch completes,
destroy after all pending graph launches
=> Remove GraphExec destroy during next sync point(hipStreamSync,
hipDeviceSync etc..)
Change-Id: I4df682aae5787fd6e5240a7be936ce50361345d0
[ROCm/clr commit: f9f995c6d0]
The new set tracks only the queues that have a command
submitted to them. This allows for fast iteration
in waitActiveStreams.
Change-Id: I2c832eefa01280d9a87a5f57874d36d2e9441de7
[ROCm/clr commit: bcc545e6b8]
SPT is destroyed with hipDeviceReset(). If a
stream is created right after reset, the same
object id could be reused. Later SPT destructor
incorrectly verifies that the stream is valid
referring to the reused object id causing the
corruption.
Change-Id: I3b1f7ffdf8bab874dca7b8fde22318162997b8f6
[ROCm/clr commit: f6a68b3c2e]
On some platforms user can ask for extended shared memory for a
particular kernel in some cases. This feature does not exist on HIP at
the moment. So we are setting it to sharedMemPerBlock which is the
maximum user can expect for their kernels.
Change-Id: I81005cf0d1c9fb941e77d34fb8385241ffe5bdd0
[ROCm/clr commit: 4b95e7bc87]
Handle GraphExec instance is destroyed before async launch completes
GraphExec instance is destroyed after async launch completes
GraphExec instance is destroyed without a launch
Change-Id: I45a7c82295fea916c7559bd8f796df710513aea1
[ROCm/clr commit: bf4d10ff61]
Under ROCr physical allocations don't have initial VA and require extra
flag in ROCclr. Add an option to have a mempool of physical allocations.
Change-Id: I4d062fe0dd8113d4eaf6e8b51749ed56d8701d1e
[ROCm/clr commit: f296159f62]
Mempool has capability to track dependency between streams for
faster memory reuse. Enable that capability.
Change-Id: I28266a7e38d0fc4c5d027b9542d3719653840821
[ROCm/clr commit: 17d0c166d2]
Windows path still uses multi threading implementation. Hence, in graphs
all nodes are executed in a queue thread and that requires to manage
mempool in the queue thread. However, the spec allows to destory memory,
allocated in a graph, outside of the graph's execution. That may cause
mempool management to go out of sync.
Change-Id: I0ffb2244b3cb720455ed44d1b3e2487fa8959a77
[ROCm/clr commit: 4f6b0da028]
This should be used in place of dlsym or GetProcAddress (linux and windows respectively)
Change-Id: I5501b538e03892e8e5a2282678d848fcaf21d911
[ROCm/clr commit: 0479cdb3dd]
Fetching null stream's logic has changed earlier from amd::HostQueue
to hip::Stream. This seem to cause some timing difference between
checking for null stream and creating it due to which issues are
observed in multithreaded applications using default stream.
Change-Id: Ie02365dec537275d23a1d225de9811e2fd3a9c55
[ROCm/clr commit: 8d265838cb]
During hipGraphExecKernelNodeSetParams kernel function can also be updated.
Hence size required for kernel parameters differs from what is allocated during graphInstantiation.
So, create new 128KB kernel pool and allocate kernel args from the pool.
If the pool is full create new 128KB pool. Release kernel pools when graph exec object is destroyed.
Change-Id: I9567946d63400c79cbfd4c5439c654c92557ceae
[ROCm/clr commit: e9957151f3]
Runtime has to release extra memory, held by the pools,
in synchronization points for event, stream or device.
Change-Id: Id533a5e1d137812aa72bdfe101b4b333c6a43d66
[ROCm/clr commit: 3fa4e31180]
If uuid is copied via strncpy it will stop at first null character. We
need to copy all 16 bytes which might have a null on windows.
Change-Id: I8667919cb251133eec3333a23768c356879727e8
[ROCm/clr commit: db909b2a1e]
Pass LUID infromation from PAL to HIP device properties.
NodeMask is calculated based on the order of devices in PAL
Change-Id: Ic53707b68b16f392e5284a0105bf1f3941b46653
[ROCm/clr commit: 694fe4bedb]
Maneesh Gupta
Filip Jankovic
Todd tiantuo Li
Scott Todd
Shadi Dashmiz
Ioannis Assiouras
Sam Ruscica
SaleelK
Danylo Lytovchenko
Chaudhary, Jatin Jaikishan
Arandjelovic, Marko
Jiang, Julia
Sang, Tao
Haehnle, Nicolai
Stojiljkovic, Vladana
Brzak, Branislav
German Andryeyev
Anusha GodavarthySurya
Jaydeep Patel
Ioannis Assiouras
Sourabh Betigeri
jiabaxie
kjayapra-amd
Satyanvesh Dittakavi
Rahul Garg