- add rocprofiler-register to CPACK_DEBIAN_BINARY_PACKAGE_DEPENDS when found
- add rocprofiler-register to CPACK_RPM_BINARY_PACKAGE_REQUIRES when found
- remove report_tool_load_failures_explicit_
- add HSA_TOOLS_DISABLE_REGISTER flag
- add HSA_TOOLS_REPORT_REGISTER_FAILURE
- use HSA_TOOLS_REPORT_REGISTER_FAILURE instead of HSA_TOOLS_REPORT_LOAD_FAILURE
- changed rocprofiler-register message to not include the word "error"
Change-Id: Ib7fd7f14c42758a54c347874018281bb1b5477a6
If two attach requests to the same piece of shared memory occur,
a double export or premature dmabuf fd close can occur since the export
and close on demand calls are not atomic.
Use a reference counter on shared memory dmabuf FDs that have
already been opened to avoid this problem.
Change-Id: I14a59209c0385e32582af42a57b33b1c6838a9b1
Instead of caching shared memory fds for export on the exporter side,
only export the FD in the async handler when requested.
The importer should request export fd closure once import is done.
Change-Id: I469e0cd1749beeb9c506c8a6461745fb039d9c3b
Adjust code to allow the use of non-contiguous chunks of memory to be
mapped within a single VA range.
Change-Id: Ida21ba202927229347b3a32d9b7106df10819cf5
Users can import device memory without specifying the target node.
DMA buf imports return a Thunk handle that's not useful for
gpu mapping calls.
Fix this by using the import node information to re-import and
map with the correct target GPU.
Also fix IPC detach calls by deregistering the Thunk handle
import immediately during attach instead of failing to do it later
on detach since Thunk handles aren't placed into ROCr allocation
map.
Finally refactor the IPC attach function for cleaner logic flow.
Change-Id: Ib2bf178110b2be98bd6917c765f724e4e613f5f2
Optimizations include:
- Greedy gang by placing gang leaders on first D2D sdma blit context
to avoid dead locking with other gang leaders and items. Note that
this is fine since we can't avoid an oversubscription problem when
there is only 1 xGMI link anyways, so treat all xGMI links as a single
pipe for ganging.
- Non-leader gang items don't have to poll on dependency signals so this
opens up more non-blocking SDMA channels.
- unlock gang lock when gangs are not needed.
- Change gang factor lookup from vector pair to map and register all
gpus in gang factor lookup regardless of link type so that we can take
advantage of the O(logN) direct key/value lookup time.
Fixes include:
- HSA_PAGE_SIZE_4KB was an incorrect macro to use for gang size limit.
As a result, small copies ended up ganging and hitting latency limit.
Use hardcoded 4096 bytes instead.
- Cap auxillary gang factor to the number of non-XGMI SDMA engines.
Change-Id: Ic23fde131502906a807134a04599aa6d012e8cbb
Use emplace to prevent copying the MappedHandle objects when inserting
entries into mapped_handle_map_.
Change-Id: Id3f40f1eb73ce30e62da53c5aea4dd715e83ac59
When allocating a memory handle, the NoAddress thunk flag should be set
so that this allocation does not have a virtual address range.
Also, skip mapping the memory when allocating a memory handle
Change-Id: I1c168bc00ddbc158d447197c4dc25f96bad02b19
After a memory handle is created. hsa_amd_vmem_get_access should return
HSA_ACCESS_PERMISSION_NONE insread of reporting the allocation as
invalid.
Change-Id: I1a09d15c220d48497d09c89059493e538f82aeb9
As the KFD IPC IOCTLs will not be upstreamed, change runtime
implementation to use DMA bufs.
DMA buf fds will be passed over abstract unix domain sockets.
The exporter spins a thread that creates a socket server.
The importer connects to the server to fetch the fd.
libDRM will be required to do a manual import and GPU map for
memory that is not already imported and mapped.
For now, use the legacy IPC implementation by default as a
follow on patch will disable the HSA_ENABLE_IPC_MODE_LEGACY
environment variable.
Change-Id: Ifd8469e9adfc81f8a1ea78d6010fb10b515ba1b4
Implement HybridMutex to improve latencies compared to KernelMutex when
there is contention between several threads calling hsa_signal_create
and hsa_amd_signal_async_handler.
Change-Id: If53377033e749b0050727964c9303f09b02527cc
Fixes issue where t1_ counters may not be updated when doing dispatch
profiling, causing a divide by 0.
Change-Id: I91060ac3f9fd2183d277e6e7cd810398a453a87f
KFD had some fixes for handling of virtual memory APIs. These fixes are
included in interface version 1.15.
Change-Id: Ie701eccf6e032f9ec0a1f4e8a43718964eebddc6
This reverts commit c5db063b2f. This
change is required for the runtime to generate reliable core dump files,
but this feature has been disabled for now by
5e3be9c28a. Until it is needed, revert
the ABI change in the trap handler to maintain compatibility with older
debugger.
Change-Id: I77a1562dc7962befe2bf88442df858e2d2b1c5ab
This reverts commit 803e37ded5.
This commit disables core dump feature. Apparently, gfx1101 SA1 waves
can not enter the trap handler because they receive an invalid
address. However, core dump at the debugger has been moved to rocm
6.2.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Change-Id: I7915caf58118658e5e7f435f91a0a6216d2fdb42
On some systems, pthread_addr_setaffinity_np does not exist, so we need
to use pthread_setaffinity_np on thread after pthread_create
Provided by Julian Samaroo on github
https: //github.com/RadeonOpenCompute/ROCR-Runtime/pull/143
Change-Id: I4649f94333f2d7b0a5993b370a4bfc48d92acecb
Add query to return flags for GPU agent memory properties and AQL
extensions.
Implement flag to determine that GPU agent is an APU
Change-Id: Ic04c51290b2b9763e14989c117f35a2e22297453
When inspecting waves on architectures where SPI may not initialize TTMP
registers, the debugger cannot reliably know if the trap handler was
entered and if it saved valuable information in TTMP registers.
This patch uses the status.skip_export bit (unused by the compute
shaders) to indicate that it got executed before halting a wave.
This is done except for gfx940, where ttmp11[31] can be used (as long as
TTMP registers are always initialized by SPI for this architecture). It
could be possible to be more selective as architectures always
initializing TTMP registers do not require this step, but always doing
is makes maintenance simpler.
Change-Id: I314db6b37772f7daa8bd405e6662a86658d3f5e0
Extracts and creates a core dump ELF file from a fault event, using
core dump front end.
Signed-off-by: Alex Sierra <Alex.Sierra@amd.com>
Change-Id: Ibbbe41b3d13dd3fcb90161e927d48c329cf513a9
Member added to KFDVersion to report if KFD supports core dump
mechanism. This is done through hsaKmtRuntimeEnable API call while
the topology is being built. It also dictates if core dump will be
generated by either KFD or hsa-runtime.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Change-Id: I2e9d4166563402f78613d728446feb692c52d9d1
This API consists in one function to be called from a fault event at the
hsa-runtime to generate a core dump.
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Change-Id: Ib1b90d5beb13f93c4e8ebd21fd61705ebb12ca5d
- fix logic for using HSA_TOOLS_LIB when rocprofiler-register support is enabled
- report tool load failure for rocprofiler-register
Change-Id: Ife23aa3e6ed19174376cd694764583b73f8976cd
The alternate scratch memory is used for dispatches that have a low
number of waves but relatively large wave size.
This allows us to keep the tmpring_size.bits.WAVES field of the main
scratch to full occupancy.
Change-Id: I32d240fac4b7d38200d1eebc1b0fdc8a823920d3
For devices where the CP FW supports asynchronous scratch reclaim, ROCr
is able to claw-back scratch memory that was assigned to an AQL queue.
With that ability, ROCr does not have to rely on using USO
(use-scratch-once) when assigning large amounts of memory to a queue.
If we reach a situation where we are running low on device memory, ROCr
will attempt to claw-back the scratch memory.
Change-Id: Iddf8ec84e37ab8b9fdc58bafbe2b61fe2acb6eb7
For gfx11, the trap_handler fails to recognize a trap id 3 and report
the exception to the debugger if the debugger is attached.
This is because the 2nd level trap handler looks for the DEBUG_ENABLED
bit in ttmp13 instead of ttmp11. This bit is set by the 1st level trap
handler and is part of the 1st/2nd level trap handler ABI.
Change-Id: Ib36361f53d9bcbbed52320d8c3a9ab2c0b28c7cd
SDMA4.4 and SDMA5.2+ has increased it's available copy size to 2^30 bytes
represented by exponent as bits set in the COUNT field of the
linear copy.
Also note that the full 2^22 byte limit is available from SDMA4 onwards
as it has corrected the 0x3fffe0 HW limitation from SDMA3.
As copy limit has increase, this can change system performance
so provide env var HSA_ENABLE_SDMA_COPY_SIZE_OVERRIDE=0 to fall
back to the original 0x3fffe0 limit for debugging purposes.
Change-Id: I0fb6e5378f68e5b8a00ff559271691a943ee06ee
To be able to trace memcpy asynchronously, both dst and src agents need to have profiling enabled and the api for enabling profiling was only enabling for gpu agents. CPU agents didn't have profiling enabled so the signal owner could not be known. hsa_amd_profiling_get_async_copy_time will fail with an HSA status error because it can't read the agent for the given signal.
Change-Id: Ie165e0e39b8fcd6992a55695b9ffcead10a8e812
- Update CMakeLists.txt
- find_package for rocprofiler-register
- this is an optional package until rocprofiler-register is added to the CI
- define HSA_VERSION_{MAJOR,MINOR,PATCH} ppdefs
- Update runtime.cpp
- include <rocprofiler-register/rocprofiler-register.h>
- if rocprofiler-register succeeds, do not support v1 unless explicitly requested
Change-Id: I8f48bbf3f6b52fb91ddade2f198491a1256035fe
Add handler to handle HW exception events reported by underlying
drivers. These events are generally caused by GPU resets and need the
application to abort.
As an improvement, in the future, we can provide additional information
about the exception (e.g mode-reset level)
Change-Id: If3fb5f19f9fce181a9d3b5e34a5506725856e7b0
An AQL packet header field is stored using an atomic release, and needs
to be read using atomic acquire if it may be written by another thread.
Change-Id: I1d75587fd93f9c6216deebffc9a627b404a7e749
Define AMD_AQL_FORMAT_INTERCEPT_MARKER AMD vendor AQL packet. Add
support to intercept queue to invoke a callback for these packets.
Change-Id: Ia58d5fe2171f563632b4edd6343e02585f49d149
When the intecept queue copies packets from the proxy queue to the
wrapped queue, it should not attempt to copy packets that are outside
the proxy queue. This could happen if the user of the proxy queue
advances the write pointer beyond the number of free slots and the
packet rewriter reduces the number of packets.
Change-Id: Id02f5df8aee0ed7269f4de813731d507cf2126b3
Jonathan R. Madsen
Shweta Khatri
Jonathan Kim
David Yat Sin
Mythreya
Lancelot SIX
Ruili Ji
Alex Sierra
Lang Yu
Youssef Aly
Tony Tye