This change makes HSAIL usage similar to that of Comgr. By default, the
runtime will statically link against it, however if HSAIL_DYN_DLL is
defined, then the runtime will try to dynamically load HSAIL.
Currently stick to statically linking to HSAIL. In a feature patch the
dynamic loading behaviour will be enabled.
Change-Id: I6a78a4375975cf847f236b200404c8cf941d012b
HIP tests require HIP callbacks to be processed in another thread.
This change will use a thread from HSA signal callbacks to make
sure a HIP callback was done asynchronously.
Also process the callback before changing the status of command
Change-Id: Icef85d0e0f808663882cf6881ff1be3e5eca29ac
- Don't notify if the batch is empty, because that means
the current command was processed already.
- Disable pinning optimization to avoid a race condition on stall.
- TS marker submition requires extra AQL barrier
to track the status.
Change-Id: I17eff4ad12ac66cfe1bb44048bebb1891805279d
Direct disaptch doesn't insert extra barriers for Markers if
AQL barrier was the last issued command already.
Change-Id: I00fbc658547d83dd3ee64ec391ed50e5f8a08e30
- Avoid GPU wait on the marker submission and update the command
batch after HSA signal callback upon HSA barrier completion.
Change-Id: I5c1c97212aefc2ae4b99aa9e2a81627ee9a38c1c
When OCL ROCr backend performs CL_MEM_COPY_HOST_PTR it may attempt
to have access to amd::Memory object it's currently creating,
but it's not ready yet. The logic creates a temporary dummy object
to perform a copy transfer. The new change will make sure runtime
skips allocation of the same device::Memory object second time.
Change-Id: I14c6a00a3941fdcaa6aea299e9f096e4c3f5cadf
Make sure the logic updates the command status when it's done in
HW, but not on submission.
Add the last command tracking, otherwise queue sync logic in the HIP
upper layer may skip synchronization, assuming the queue is empty.
Change-Id: I2d046792553e74df090a10f7d7a78914610f6df2
Use barrier packets for every profile marker that gets submitted
and use the completion signal to get GPU ts. This gives most accurate
dispatch time. Club cache flushes with profile marker if there is a
pending dispatch that needs cache flush. This optimization saves on
extra barrier and helps wall time
Change-Id: Ib62d6d7aabf4743827b561be6c9c5afa813203da
If deferred allocation is disabled, then make sure the image view
is created without a delay. Also reset the allocation state, since
create() method isn't called for a view creation.
Change-Id: I7aa22a62bff18289ade83e56b5d3305ba68c715b
The hack dosn't really track the commands status. It may be not
necessary for HIP, but will cause early resource release.
Change-Id: I791ad36dd8abd3b6b3d2c9b16a210a555c08ca64
OCL can't distinguish different copy types, but ROC profiler
expects SDMA transfer visibility. Add extra code to detect
a transfer with the host memory and substitute OCL command
Change-Id: I5290acd0e10bc082e00c1d4ae1474a075de7f165
Replace amd::Atomic with std::atomic. Remove make_atomic uses by
converting the variable to std::atomic and making sure the memory
order is relaxed when synchronizes-with is not needed.
Delete utils/atomic.hpp.
Change-Id: I0b36db8d604a8510ac6e36b32885fd16a1b8ccfa
Reduce the size of the queueLock and lastCmdLock critical sections
to improve lock contention performance. The smaller the critical
sections are the better.
lasCmdLock is still needed to guarantee that getLastEnqueueCommand_
can retain the command before it is swapped out and released.
Change-Id: Id35d4a77c035b2da0de4c15568b153d49e958bb7
Vladislav Sytchenko
Saleel Kudchadker
pghafari
Sarbojit Sarkar
Satyanvesh Dittakavi
German Andryeyev
Ravi C Akkenapally
Jason Tang
kjayapra-amd
cjatin
agodavar
Evgeny
Laurent Morichetti
Tao Sang