If M0[23] is set then the driver will interpret the interrupt as a
debug event, rather than a signal event.
Clear M0 before sending the interrupt. All paths here are terminal so
it's not necessary to save/restore M0.
Change-Id: Ibd85b8cc6f8556941f2308a2c3fa3c68702cd606
Report traps and fatal exceptions through a wavefront's
amd_queue_t.queue_inactive_signal. Previously, only traps were
reported and requireed the compiler to pass in the signal pointer
in s[0:1].
The signal is obtained through a mapping from doorbell index to
amd_queue_t*. The doorbell is fetched within a wavefront through
the gfx9+ S_SENDMSG(MSG_GET_DOORBELL) instruction.
Change-Id: I319b45f2e15dfcfe4db8f4065da1136e9539a42b
Assembler toolchains are moving from SP3 to LLVM. Replace trap handler
source code with LLVM equivalent.
Fix a trap issue with SQ_WAVE_IB_STS restore. Mostly harmless as all
traps are currently considered fatal to the wavefront.
Change-Id: Iacecd9dd31a1d96a083c8b8327f442f33c861f9f
Adds hsa_amd_register_deallocation_callback and hsa_amd_deregister_deallocation_callback
to notify when HSA memory has been released.
Change-Id: I1f33cee250ca890e5c2e7fddfa4479aa5874651d
Small times may be given to time conversion if GPU clocks are used to
accumulate elapsed time. Because HSA APIs deal in absolute time this
leads to large conversion offsets of order system uptime. Variation
in relative clock ratio estimation may be amplified in this case,
destroying elapsed time measurements.
This patch fixes the relative clock ratio used for times which predate
the call to hsa_init. This correlates errors in such times allowing
the elapsed time to be correctly computed.
The effective maximum system uptime before elapsed time conversion becomes
inaccurate is ~3.5 months. GPU event timestamps are good for process uptime
of ~3.5 months. These are limited by double's mantissa precision.
Change-Id: I48752ff354920439d91016d6f2b0c8ddfa60b445
At the moment it is not possible to build ROCr with Clang. This is
a spurious limitation. The present PR addresses it by guarding GCC
only flags and by fixing some additional warnings that Clang triggers;
one of said warnings did outline a rather interesting issue with math
being done on void*s. - AlexVlx
Void ptr arithmetic had already been fixed in amd-master branch.
Change-Id: I5ee97e20b5c40b10dd73facecabe75f02ba46462
Non-paged memory can be IPC-shared even when HSA_USERPTR_FOR_PAGED_MEM
is enabled.
Change-Id: I8b1fa6d7a4a9327c78a77b3679697fbf55397093
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Makes malloc memory accessible to GPUs so that the memory has the
capabilities of the pool it is locked to.
This admits fine grained locked memory and reserves API space for any future
special CPU pools.
Change-Id: If8c3dd8582a43f19d3d36b3763c1a688cc419ef0
Also rename blit_agent to region_gpu and add comments to clarify
its role in deprecated region API support rather than to do blits.
Change-Id: I80b1043db2e1c5d40a58fc801eef70a688ea9169
Debug agent requires handles to internal queues for single step debugging.
Added tools only API hsa_amd_runtime_queue_create_register for reporting.
hsa_amd_runtime_queue_create_register sets a callback which is invoked
when internal queues are created.
Change-Id: Ia5190ae724fadba686c15f25b2cd085350eeff0e
Apertures now overlap with the change to 48bit addressing which
precludes using aperture checks to discover buffer ownership.
Switches to ptrinfo to decide which device a buffer owned by.
This corrects faults in the legacy hsa_memory_copy api.
Change-Id: I5c7ce0216e1cdc96f836fc6fec9c3defdf4b9d90
Remove fence pool and use two signals. Two signals allows overlapped
submission and copy while reducing thread busy polling.
Change-Id: Idb5f8e4c7f482a596ffce9e7799191fdd785a216
Fix pitch overflow due to small element detection.
Add wide pitch 2D copy handling.
Cleanup code duplication.
Change-Id: I93b1584aba8e5964957eb7ab3544df806ca3e2f9
llvm.debugtrap and other trap IDs are reserved and should not place
the queue into an error state.
Change-Id: I98193a35ac7da94c4a42ee75d87754ee552ebea0
1/ Revised debug event handler to handle different events.
2/ Added queue error handler using the callback in queue create, which will print out wave info when queue in error state.
3/ Preempt queue instead of destory queue when queue error state.
Change-Id: Ib727d208de9caf1c72c76d42268483b24aaebde8
- Ignore exceptions passed to the second-level handler
- Restore SQ_WAVE_IB_STS and SQ_WAVE_STATUS before exiting trap
Change-Id: I872c111c030d94eae644ae073df3c2e508f42f45
Spec requires GPU release fences and CPU acquire fences at queue destroy.
Also update the recognized status codes.
Change-Id: If9166f5149f65417c7057ff7c0f69f6ac094d6ab
Remove unused function (FenceRelease), add comments to barrier packet settings,
correct profiling controls to work with queue wrappers.
Change-Id: I45bb26227bcc2b78edb8ad5dc497603c33234e18
Prefer using memfd_create() for the ring buffer.
We were using /dev/shm, but this won't work on systems that
either don't have /dev/shm or have mounted it with noexec, because
for everything other than gfx700 we map the ring buffer with PROT_EXEC.
memfd_create() is Linux specific and was added in Linux 3.17, so we
will fallback to using /dev/shm on systems where memfd_create() is
not available.
Change-Id: I58fb533eebc362f6d29dc3e316a80801014d50e8
Remove "zombie" queue state and report queue creation failure via
exceptions. Make Shared object a final container and support array
objects with Shared. Add message printing to hsa_exception in
debug builds.
Change-Id: I459f38c80846018acbf45538874e95f91dd6b195
Jay Cornwall
Ramesh Errabolu
Sean Keely
Felix Kuehling
Konstantin Zhuravlyov
Qingchuan Shi
Wilkin
Evgeny