*Lay foundation to batch packets efficiently for graphs
*Dynamically copy packets with max threshold set with
DEBUG_HIP_GRAPH_BATCH_SIZE, if not stagger packet copy with pow2
*Default threshold for DEBUG_HIP_GRAPH_BATCH_SIZE is 256
*If TS are not collected for a signal for reuse, create a new signal.
This can potentially increase signal footprint if the handler doesn't run
fast enough.
- Refactor deviceLocalAlloc arguments
- Refactor hostAlloc code, have cleaner interface
- Kern args buffer need to have execute flag set as CP enforces this on
certain newer HW.
* SWDEV-532479 - Add tracking of hostcall memory allocations
* SWDEV-532479 - Remove hostcall allocations if request is received
* SWDEV-532479 - Cleanup
* SWDEV-532479 - Naming fix
* SWDEV-532479 - Add new separator after each new function
[ROCm/clr commit: b58faa2f37]
* SWDEV-465041 - Add support for user events with DD
User events can be replaced with HSA signals. Add the interface
to allocate HSA signal for user events and update the status on
CL_COMPLETE.
Force pinned path with DD to avoid blocking calls. Pinned memory
can be released only when the command is complete.
Simplify device enqueue path to use generic kernel arg buffer and
signals
* Fix notifyCmdQueue() logic for OCL
* Avoid blocking calls in OCL with DD
* Add event destruciton in a case of the failure.
[ROCm/clr commit: 2305f8ae56]
Clarify some VGPRs terms description.
Fix some wrong query logics of availableVGPRs_ and
availableRegistersPerCU_ in device info.
Add hipDeviceAttributeMaxAvailableVgprsPerThread
attribute query.
Remove hardcoding of following
info_.vgprAllocGranularity_
info_.vgprsPerSimd_
[ROCm/clr commit: 397f303d97]
* Add hipHostMalloc() new flag hipHostMallocUncached which will force to allocate pinned
host memory on extended fine grained system memory pool.
* Add hipHostAlloc() new flag hipHostAllocUncached which will force to allocate pinned
host memory on extended fine grained system memory pool.
* Add hipHostRegister() new flag hipHostRegisterUncached which will force to map
host memory onto extended fine grained system momory pool.
[ROCm/clr commit: a7d7687b8f]
Support programmatic query and change of scratch limit on
AMD devices.
Change-Id: Id5da355a77366f97868e462847f3916e87fd2af6
[ROCm/clr commit: 1113eff3f9]
* 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]
* SWDEV-505795 - Return the same ptr from hipIpcOpenMemHandle if it is called multiple times
* Move initialization outside of if statement
[ROCm/clr commit: e91cb4f320]
- Remove binning logic, although useful it doesnt work in current
scenario as there is no upper limit on the size of allocation. If an
app or framework uses entire VRAM and then creates suballocs, binning
would result in failure.
Change-Id: Icc27c13e433bb4a1f03e82028d8718488b43bfa5
[ROCm/clr commit: e4d29e228d]
- Create bins each with its own map and lock. This would help cases
where the hash of a VA is differnet than ther one which falls in
different bin, and there is no lock contention
- Use STL shared mutexes, that way we can unique_lock for map updates
vs simple reads which can use shared_lock
Change-Id: I118818be65c6373700f5e511045babb6a398938a
[ROCm/clr commit: e23ff0520b]
Add an atomic counter to track the outstanding HSA handlers.
Wait on CPU for the callbacks if the number exceeds the value
in DEBUG_HIP_BLOCK_SYNC env variable.
Change-Id: I95dc8c4bf0258c7e59411b7504220709ed6898c5
[ROCm/clr commit: 403f624bf8]
- Don't generate callbacks for HIP events
- Don't process profiling info in the callback for HIP events
- Wait for CPU status update of the submitted commands
every 50 calls. That will allow to drain the commands and
destroy HSA signals.
Change-Id: Ib601a350e7e7c2b6c6209a172385389baccf73a9
[ROCm/clr commit: 364dfb0ed1]
Although unpinned copies require synchronizations
in HIP, runtime can avoid syncs for H2D copies with
a staging buffer
Change-Id: If2203c6bc0cbd89742823688dc8e89e9acd873b2
[ROCm/clr commit: 29cc678d8d]
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]
If the graph has kernels that does device side allocation, during packet capture, heap is
allocated because heap pointer has to be added to the AQL packet, and initialized during
graph launch.
Handle race with wait when 2 kernels with device heap are enqueued on multiple streams.
Change-Id: I45933b77fcaf7bc8fdf1bc906462e32b5d8d3688
[ROCm/clr commit: 57156c524d]
Adding a safety check prevents an invalid memory access
if timestamps and kernelNames vectors are of different size.
The patch also moves the addKernelNames for the accumulate command
into dispatchAqlPacket function.
Change-Id: Iea0927e1253800403a1ae3f3d72de1e7d96476c3
[ROCm/clr commit: d44f44a5b1]
Ioannis Assiouras
SaleelK
Danylo Lytovchenko
Betigeri, Sourabh
Manocha, Rahul
Arandjelovic, Marko
Xie, Pengda
Andryeyev, German
Sang, Tao
Lin, Qun
Jayaprakash, Karthik
Assiouras, Ioannis
Stojiljkovic, Vladana
Patel, Jaydeepkumar
agunashe
Jimbo Xie
Anusha GodavarthySurya
Julia Jiang