introduce LockedAccessor option so destructor does not unlock.
Allows locks to exist across function boundaries, required
for hipLaunchKernel macro which has several unusual requirements.
(including C comppatibility, must use variadic macro, more).
[ROCm/hip commit: 8635863724]
Move critical data into separate class and protect with LockAccessor
wrapper class.
For device, the streams list is the critical data since it is modified when
streams are created or destroyed. The streams list is accessed in
several places including when synchronizing across all streams on the
device (ie from the default stream).
Other device data is set once by the device cosntructor and is not critical
so
All functions which acquire the LockAccessor now named with "locked_" prefix.
[ROCm/hip commit: 530ab9434a]
- Some comments in hip_hcc.cpp
- document hipHostRegister* flags.
- expand docs on hipHostRegister/Unregister. Use "register" rather than
"pin" to describe action these take.
- change required CUDA version to 6 (require unified memory)
- remake doxygen.
[ROCm/hip commit: eced013ae4]
- Validate compile-time disables.
- Add README.md section explain how to install/use CodeXL tracing
- Add code docs on trace_helper.h
- fix color on hipLaunchKernel to green.
[ROCm/hip commit: 2569e15375]
- Creates markers in HIP group and they show up in CodeXL trace
- Marker text includes HIP functioin arguments
- (Add trace_helper to convert arguments to strings)
- Still need to add HIP_INIT_API for ~30 HIP functions.
[ROCm/hip commit: b6b3fea05b]
Note hipHostMalloc (not hipHostAlloc or hipMallocHost).
- the hipHost* is used for all HIP APIs dealing with Host memory.
(including hipHostMalloc, hipHostFree, hipHostRegister,
hipHostUnregister, hipHostGetFlags, hipHostGetDevicePointer).
- hipMallocHost is consistent with "hipMalloc" for allocating device
memory. Enumerations hipHostMalloc* also used as optional
flags parm to hipHostMalloc.
[ROCm/hip commit: ab910efb96]
- move waitAllStreams to device member function.
- create separate stream member function for copyAsync, like copySync.
hipMemcpyAsync now calls the copyAsync.
[ROCm/hip commit: 52cc2bb75a]
- NULL stream was waiting for itself to be empty before each command.
- Force "blocking" streams to wait for NULL to empty. This was missing
before.
- async copy was disabling itself via trueAsync=false for common cases.
Refactor:
- rename _null_stream to _default_stream.
- move some null sync function to defaultSync, move to dev member func.
[ROCm/hip commit: 6472c8b915]
-Move staging buffer locks inside the staging buffer code.
-Remove dedicated per-device completion_signal + per-device lock -
instead allocated signal from the per-stream pool. This elimintes
the lock and allows more concurrency.
-remove switch HIP_DISABLE_BIDIR_MEMCPY
[ROCm/hip commit: 0af4d3623f]
- refactor staging buffer to operate on hsa* data structures not
hc::accelerator.
- use hsa_memory_allocate to allocate staging buffers rather than
am_alloc.
- Refactor device reset with single member function. Don't reallocate
staging buffers on reset.
- Properly track dependencies based on command type. Add new deps for
H2D and D2D rather than overloading H2D.
[ROCm/hip commit: 7d500599fa]
Still #include staging_buffer.cpp into hip_hcc.cpp.
Directed tests compile hip_hcc to static library and use the library.
[ROCm/hip commit: 28ee7aff71]
Ben Sander
pensun
Aditya Atluri
Aditya Atluri