GPUs excluded by RVD are not expected to have scratch, memory, trap
handling nor memory regions set up. Now that these GPUs are added to
a new list, early return on agent destruction to prevent bad function
calls on destroy.
Also fix up broken memory releases between the gpu lists and ugly braces.
Change-Id: I52fc6e86ceba0a0383cedc63310eb409515eaf9f
[ROCm/ROCR-Runtime commit: 9d2fe1ac2a]
Atomic memory operations on these memory buffers are not guaranteed
to be visible at system scope
Change-Id: I4cccde114632071a000384502a83bc191e77e85b
[ROCm/ROCR-Runtime commit: 364715cbc6]
A work group processor (WGP) require both its CU to be enabled
in order to be enabled.
The KFD will round robin distribute by even-indexed pairs so
enforce this requirement for runtime set mask calls.
Change-Id: Ic46661b01f398aa1fe24d96b5c9c31f122f967a3
[ROCm/ROCR-Runtime commit: f600687537]
Discovered agent handles should only apply to copy routing, not to
copy device selection. The user may not have mapped all allocations
to all GPUs so we must ensure that the copying device is one passed
by the user.
Change-Id: I2532e66d30e6842624e594f235dd144a186220d4
[ROCm/ROCR-Runtime commit: a8603b9397]
Mostly a demo at this point. Logs SVM (aka HMM) info to
HSA_SVM_PROFILE if set.
Example: HSA_SVM_PROFILE=log.txt SomeApp
Change-Id: Ib6fd688f661a21b2c695f586b833be93662a15f4
[ROCm/ROCR-Runtime commit: 965df6eef7]
During hsa initializing stage, ROCr now searches all the loaded libraries
for a symbol "HSA_AMD_TOOL_PRIORITY" and adds all those libraries to
the tools library init list. Tools libraries listed in HSA_TOOLS_LIB
env variable are also loaded in the given order and take priority
over HSA_AMD_TOOL_PRIORITY.
Change-Id: I739af42bbd777c44a9152c11e17dd69979b65e82
[ROCm/ROCR-Runtime commit: e7fc301aa7]
Adds a script to run clang-format on the latest patch so we don't
need to remember the command line.
Also applies missing formatting to the prior commit,
"Add API for available GPU memory".
Change-Id: Ida51aedc38af229f6a26e275072654860748fa93
[ROCm/ROCR-Runtime commit: e7152c8b16]
Add support for AMD Agent to return amount of memory available
Change-Id: I5c32e2cebbaa2993b044250aefe434e4cc02d8c2
Signed-off-by: David Yat Sin <david.yatsin@amd.com>
[ROCm/ROCR-Runtime commit: 4ac840269c]
Discards all user provided async copy agent info and relies on
pointer info discovery.
Change-Id: Ife3e708a49ffccbede4983ab47d5ed0032970857
[ROCm/ROCR-Runtime commit: 3ebe99f96d]
Only block info can return an agent which is disabled in the
process.
Change-Id: I34cb1f9eea9217e10a484726c90d930e3414e769
[ROCm/ROCR-Runtime commit: 13a0cdfa77]
Physical owning agent may not be visible to the current process
due to RVD.
Change-Id: Ib463336a5ed73a479f3aa74eb140932b9e0435fb
[ROCm/ROCR-Runtime commit: 247606c455]
IPC use cases with RVD set can't convey proper agent handles.
Runtime discovery is required to properly route the copy in this
case.
Change-Id: I4c97e132fb4b6ac1040de1cb17fe5a3e36d6be48
[ROCm/ROCR-Runtime commit: c289a43e88]
Simplifies behavior. A memory type now either always generates an
entry or never does.
Change-Id: Ie98cddea01e801308ac0ba650795fdef92b7e47d
[ROCm/ROCR-Runtime commit: 0ba9b162db]
This is consistent with KFD and has significantly better latency.
KFD is taking this as the definition of the SystemClockCounter.
Change-Id: I4c1b3bc58c738206265c55ebefd41356c013bfe5
[ROCm/ROCR-Runtime commit: 0ee82742a7]
Eliminates the need for manually assembling the source of the
second level trap handler to produce the shader binary. Also
separated blit shaders' binary source and version one second
level trap handler binary sources into different header files.
Change-Id: If29a18ee06dc083ec880ea962f234c6b5cac806a
[ROCm/ROCR-Runtime commit: 1b0440e7b3]
Host to device SDMA copies do not require an HDP cache flush when
connected by xGMI since data copies over the data fabric and not HDP.
Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: Sean Keely <sean.keely@amd.com>
Change-Id: I78d73a47edcc1a9c0ba59f33cf91485f13f1c45b
[ROCm/ROCR-Runtime commit: 658b053943]
Declare the type of HSA_AMD_AGENT_INFO_COOPERATIVE_COMPUTE_UNIT_COUNT
and add a missing break statement.
Change-Id: I86ce8a2e620438e046b60cee991ce1fbe07a3e88
[ROCm/ROCR-Runtime commit: 64dae113b1]
On gfx10+ we need to issue a minimum count of active lanes or
groups before ADC moves on. Ensure that scratch allocations
attempt to reach this limit.
Occupancy throttling due to OOM condition may still drop below this
limit.
Change-Id: I0edf2e40fbe1a95e9a262564cebd2b6a82501a0b
[ROCm/ROCR-Runtime commit: 2eedf953f3]
__x86_64__ and __AMD64__ should be already defined by the compiler to
specify the compilation target and shouldn't be defined manually.
I fixed two x86_64 checks to include VS variables, as removing this
might cause it to fail to compile on that compiler.
Signed-off-by: Jeremy Newton <Jeremy.Newton@amd.com>
Change-Id: I600ff449af85bf7d83ecab167d97933922e2d917
[ROCm/ROCR-Runtime commit: 178a7a5cfa]
Each time delay is grown we need to reset elapsed. We want to take
the most accurate sample from the set at fixed delay.
Without this we will hang if there is ever an insufficiently accurate,
high unit clock read.
Change-Id: Ic65f364067789ac85a6572d67af2d77528e265bb
[ROCm/ROCR-Runtime commit: 4e9849034d]
Release staging buffers after loading has completed. The debugger
no longer uses this copy.
Change-Id: I46f36b50033bebe5a9ebc648b291d46f1d09b21d
[ROCm/ROCR-Runtime commit: 03a52655a8]
The loader must use internal interfaces to access page allocation
flags. Code pages should also ensure use of cached memory.
Also relocate i-cache flush after code page copy.
Change-Id: I86d36243b6eebb1d46b991b372a5236baaf941ab
[ROCm/ROCR-Runtime commit: 048700f2e7]
VM faults should not report via the queue error handler.
The system event contains much more useful information.
Change-Id: I744d9b97b23334d7ed2c0f450111c1b8032567e3
[ROCm/ROCR-Runtime commit: fbc48521dc]
Hive ID is used during copy path selection to locate an optimal
pool of SDMA engines. However, for CPU-GPU connections we always
want to use the host port facing engines, known generally as the
PCIe optimzed engines. We want this selection even when the
connection is XGMI hence dropping the hive id for CPUs.
Change-Id: Iffe44174afecfc0bb3272b806fce549c930a49d9
[ROCm/ROCR-Runtime commit: af0f90800d]
Compiler is not promoted to mainline yet.
This reverts commit 7dcccdf452.
Change-Id: I7256aeb3698ee3ae640a9f457a929abe24d5ef17
[ROCm/ROCR-Runtime commit: 7e73760cd0]
Excessive scratch allocations can normally trigger occupancy
reduction. This breaks cooperative groups so if occupancy
reduction is required on a cooperative dispatch fail with OOM.
Change-Id: I64612a2e38bf1286f3b74c1c2a68ab0c85452771
[ROCm/ROCR-Runtime commit: 8a6954c63c]
With asym. harvest hw does not issue groups equally to each SE,
occasionally hw will skip an SE so that the distribution reflects
each SE's CU count. Scratch resources must be allocated to reflect
this asymmetric distribution of groups.
Change-Id: I65e26206500483ea18e6e8796e65ecba5354b029
[ROCm/ROCR-Runtime commit: 552dcead93]
HW does not ignore low bits of the scratch wave count and will
stride beyond the end of the allocation if the wave count is
ever indivisible by SE count. Rather than returning the allocation
size for cached large scratch allocations, use the requested
scratch size in scratch setup. Scratch cache will retain the
cached allocation's size.
Change-Id: I0129ddc99a8940d01d8fbcd0b02d5061f31f456d
[ROCm/ROCR-Runtime commit: cedc3e80a8]
discardBlock may be called multiple times on the same block.
We must not discard the block multiple times or we will corrupt
in-use memory accounting.
Change-Id: Ife9f3162785965a795dcf81887d4d447cc096e62
[ROCm/ROCR-Runtime commit: b9a0c1d313]
Minimum queue size was not enforced at the Agent level. Minimum
size should be one page to give unifority across all asics.
Change-Id: I26394f79458d09fbceb79fc8aaf495e2c26a8ff3
[ROCm/ROCR-Runtime commit: a6742209f7]
Adds asserts for invalid dispatch dims and scratch requests that
don't actually use scratch.
Change-Id: I6e6eef3f17dc38adaf96550fa55bd8625868efa3
[ROCm/ROCR-Runtime commit: a65f3f5b71]
On gfx90a only a reduced number of CUs must be used for cooperative
dispatches due to CWSR and launcher interactions with asymetric
harvest. We must use one fewer CUs per SE than the lowest count of
CUs on any SE.
Also adds env var HSA_COOP_CU_COUNT which enables the cooperative
CU count computation. Set to 1 to enable the new computation.
This is an opt-in feature that will become enabled by default (opt-out)
in a future release.
Change-Id: Ifbb75ced3bbc15876eef44922c6a4f6fde8c4c28
[ROCm/ROCR-Runtime commit: 37942c982a]
The start iterator becomes invalid after it is removed from
std::map prefetch_map_. This was causing a segfault when the iterator is
incremented afterwards.
Signed-off-by: David Yat Sin <david.yatsin@amd.com>
Change-Id: I4b0b763d2cb4ee99c0b8571c2c526b834e74077a
[ROCm/ROCR-Runtime commit: 86164fbfec]
ROCM_VISIBLE_DEVICES was used where ROCR_VISIBLE_DEVICES was
intended.
Change-Id: I644a546f3c9dd0b50898ef8a21dbb8f5c3a36926
[ROCm/ROCR-Runtime commit: fce6ba052e]
Prior solution used a single global lock to protect the memory tracking structures.
This change protects the memory tracking structure with a shared mutex (rw lock) in
shared (r) mode for memory allocations and frees so that long duration processes,
calling to kfd, can be done in parallel. Operations which must modify the memory map
take the mutex in exclusive mode (w) and must not call to the thunk while holding
the mutex.
The fragment allocator now requires separate protection and is protected with a
mutex at the device level. Protecting at the device level, rather than pool,
allows retention of the current recursive design and allows calling Trim from
withing Allocate. This could be made finer (pool level locks) but would
require backing out of Allocate entirely to call Trim. Trim and any retried
Allocation must be done in isolation (per device) or we may report OOM when
memory is actually available in some pool's fragment cache. So some device
level serialization is required in at least some paths.
Change-Id: I7c1e94d6965ffcc602b12fefdd3a6e97b84b5e00
[ROCm/ROCR-Runtime commit: df55cb0450]
Graham Sider
David Yat Sin
Jay Cornwall
Jonathan Kim
skhatri
Sean Keely
Yifan Zhang
Shweta Khatri
Jeremy Newton