Граф коммитов
76333 Коммитов
Автор SHA1 Сообщение Дата
Alex Sierra 178a619b80 src: use SVM mechanism to register userptr memory
Register and map userptrs through Shared Virtual Memory(SVM) API at
the Kernel level when available. Using this approach, performance
will be improve as register/unregister memory will not trigger any
system call to KFD driver.

Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Change-Id: I20723cbeb340bf48b95e1115f0102c031397bc14
2022-10-21 15:32:02 -04:00
Alex Sierra ea19fbb646 src: use SVM mechanism to register userptr memory
Register and map userptrs through Shared Virtual Memory(SVM) API at
the Kernel level when available. Using this approach, performance
will be improve as register/unregister memory will not trigger any
system call to KFD driver.

Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Change-Id: I20723cbeb340bf48b95e1115f0102c031397bc14


[ROCm/ROCR-Runtime commit: 178a619b80]
2022-10-21 15:32:02 -04:00
David Yat Sin 18547173e9 Early return for invalid pointer queries
If a user queries the pointer info on an invalid pointer,
hsaKmtQueryPointerInfo will return error or unknown pointer. The other
fields in HsaPointerInfo are invalid, so we do not return them to the
user.
Also removing the assert and returning unknown pointer instead. As the
assert will not trigger in release builds.
hsaKmtQueryPointerInfo may also return unknown pointer for userptrs as
they are not always tracked by thunk. Adjusting code to still treat
these pointers as valid in this case.

Change-Id: Idf5cd8b61cd532d31b072f449839d223369bb138
2022-10-21 15:28:48 -04:00
David Yat Sin 8b9275bb36 Early return for invalid pointer queries
If a user queries the pointer info on an invalid pointer,
hsaKmtQueryPointerInfo will return error or unknown pointer. The other
fields in HsaPointerInfo are invalid, so we do not return them to the
user.
Also removing the assert and returning unknown pointer instead. As the
assert will not trigger in release builds.
hsaKmtQueryPointerInfo may also return unknown pointer for userptrs as
they are not always tracked by thunk. Adjusting code to still treat
these pointers as valid in this case.

Change-Id: Idf5cd8b61cd532d31b072f449839d223369bb138


[ROCm/ROCR-Runtime commit: 18547173e9]
2022-10-21 15:28:48 -04:00
Laurent Morichetti bec405981e SWDEV-361847 - Print 'char' parameters as 'integer'
When printing HIP API function parameters, use the integer format to
print 'char' arguments to avoid printing invalid ASCII characters
(value > 127).

Make sure the roctracer::hip_support::detail operator<< overloads are
used when printing arguments.

Change-Id: Id072c2ed19b1b4166108599e393d1cae6c54b6b0


[ROCm/clr commit: 19107858d6]
2022-10-21 14:55:33 -04:00
Laurent Morichetti 19107858d6 SWDEV-361847 - Print 'char' parameters as 'integer'
When printing HIP API function parameters, use the integer format to
print 'char' arguments to avoid printing invalid ASCII characters
(value > 127).

Make sure the roctracer::hip_support::detail operator<< overloads are
used when printing arguments.

Change-Id: Id072c2ed19b1b4166108599e393d1cae6c54b6b0
2022-10-21 14:55:33 -04:00
Ajay 09e359f925 SWDEV-357207 - Linux: enable StreamWrite/Wait APIs in deviceAttribute
Change-Id: I593df2dbb7a0233dd42f8d873510d85bbb27feff


[ROCm/clr commit: 261620a2ef]
2022-10-21 12:44:35 -04:00
Ajay 261620a2ef SWDEV-357207 - Linux: enable StreamWrite/Wait APIs in deviceAttribute
Change-Id: I593df2dbb7a0233dd42f8d873510d85bbb27feff
2022-10-21 12:44:35 -04:00
Ajay 548b3e0dab SWDEV-332251 - Remove tests packaging from hipamd
Change-Id: I98c0d19ce2eea921d214661b8a1588cb5dbf4268


[ROCm/clr commit: f2fbe4938e]
2022-10-21 11:48:30 -04:00
Ajay f2fbe4938e SWDEV-332251 - Remove tests packaging from hipamd
Change-Id: I98c0d19ce2eea921d214661b8a1588cb5dbf4268
2022-10-21 11:48:30 -04:00
Freddy Paul ac66865385 Remove RPATH/RUNPATH from ROCm libraries
:Since all public interface libraries are present in
same folder RUNPATH/RPATH is not required in the library itself.
Application shall provide the required RPATH/RUNPATH to load all
libraries.

Change-Id: I1d1ba920bf291eb89bd1f4c0fd0cfd80c7d739bd
2022-10-21 11:05:06 -04:00
Freddy Paul 993b1dee7e Remove RPATH/RUNPATH from ROCm libraries
:Since all public interface libraries are present in
same folder RUNPATH/RPATH is not required in the library itself.
Application shall provide the required RPATH/RUNPATH to load all
libraries.

Change-Id: I1d1ba920bf291eb89bd1f4c0fd0cfd80c7d739bd


[ROCm/ROCR-Runtime commit: ac66865385]
2022-10-21 11:05:06 -04:00
German 6ab9293146 SWDEV-360183 - Finish the trace if queue was destroyed
- Make sure SQQT trace is captured for RGP server if the queue is destroyed before normal capture is done.
- Remove prepare queue from the logic. It's not really used for any HW capture and can cause RGP server abort if destroyed before capture is even started(delayed capture)

Change-Id: I6eb19963190a5769c6477a5496c1b831a6d59b89


[ROCm/clr commit: c1c5127875]
2022-10-21 10:44:10 -04:00
German c1c5127875 SWDEV-360183 - Finish the trace if queue was destroyed
- Make sure SQQT trace is captured for RGP server if the queue is destroyed before normal capture is done.
- Remove prepare queue from the logic. It's not really used for any HW capture and can cause RGP server abort if destroyed before capture is even started(delayed capture)

Change-Id: I6eb19963190a5769c6477a5496c1b831a6d59b89
2022-10-21 10:44:10 -04:00
Jonathan R. Madsen 18128adfd3 OMNITRACE_ROCTRACER_DISCARD_INVALID=N (#186)
- only print N warnings that end < begin
- if N is zero, swap(begin, end) and include data

[ROCm/rocprofiler-systems commit: f2a0485d60]
2022-10-21 09:24:13 -05:00
Jonathan R. Madsen f2a0485d60 OMNITRACE_ROCTRACER_DISCARD_INVALID=N (#186)
- only print N warnings that end < begin
- if N is zero, swap(begin, end) and include data
2022-10-21 09:24:13 -05:00
Alex (Bin) Xie d77a150bf9 SWDEV-361591 - clean up header files
Change-Id: I25ca37a94524da0521f5b856a9fa9bab73f103c1


[ROCm/clr commit: 06867f0c04]
2022-10-21 10:23:55 -04:00
Alex (Bin) Xie 06867f0c04 SWDEV-361591 - clean up header files
Change-Id: I25ca37a94524da0521f5b856a9fa9bab73f103c1
2022-10-21 10:23:55 -04:00
Jonathan R. Madsen 25d2a0e359 Dramatic improvement of post-processing critical trace data (#185)
- several orders of magnitude faster

[ROCm/rocprofiler-systems commit: 9ff4b6b624]
2022-10-21 09:23:52 -05:00
Jonathan R. Madsen 9ff4b6b624 Dramatic improvement of post-processing critical trace data (#185)
- several orders of magnitude faster
2022-10-21 09:23:52 -05:00
Jonathan R. Madsen 3073a5095a Disable HSA API and activity by default (#183)
- OMNITRACE_ROCTRACER_HSA_ACTIVITY is OFF by default
- OMNITRACE_ROCTRACER_HSA_API is OFF by default
- in real applications, this adds way too much tracing data to perfetto

[ROCm/rocprofiler-systems commit: 271f851896]
2022-10-21 09:23:29 -05:00
Jonathan R. Madsen 271f851896 Disable HSA API and activity by default (#183)
- OMNITRACE_ROCTRACER_HSA_ACTIVITY is OFF by default
- OMNITRACE_ROCTRACER_HSA_API is OFF by default
- in real applications, this adds way too much tracing data to perfetto
2022-10-21 09:23:29 -05:00
Jonathan R. Madsen 665b0fa471 Fix perfetto debug annotations of function parameters (#181)
Fix debug annotations of function parameters

[ROCm/rocprofiler-systems commit: 1bead56ce8]
2022-10-21 07:24:21 -05:00
Jonathan R. Madsen 1bead56ce8 Fix perfetto debug annotations of function parameters (#181)
Fix debug annotations of function parameters
2022-10-21 07:24:21 -05:00
Jonathan R. Madsen 8f9dcffae4 Update containers workflow (#182)
* Bump version to 1.7.2
* Update containers workflow

[ROCm/rocprofiler-systems commit: 5bb7359b81]
2022-10-21 06:21:00 -05:00
Jonathan R. Madsen 5bb7359b81 Update containers workflow (#182)
* Bump version to 1.7.2
* Update containers workflow
2022-10-21 06:21:00 -05:00
Chauncey Hui 7347172bc3 SWDEV-2 - Change OpenCL version number from 3516 to 3517
[ROCm/clr commit: 904f3e3d75]
2022-10-21 03:00:52 -04:00
Chauncey Hui 904f3e3d75 SWDEV-2 - Change OpenCL version number from 3516 to 3517 2022-10-21 03:00:52 -04:00
Satyanvesh Dittakavi 25ff252d73 SWDEV-362371 - Uninitialized stream error checks
Change-Id: I7e04ef46f0cf2b0c45caf42d5b2ee31cb68e93eb


[ROCm/clr commit: bded3d8d1f]
2022-10-20 21:57:11 -04:00
Satyanvesh Dittakavi bded3d8d1f SWDEV-362371 - Uninitialized stream error checks
Change-Id: I7e04ef46f0cf2b0c45caf42d5b2ee31cb68e93eb
2022-10-20 21:57:11 -04:00
Tao Sang 326049efb8 SWDEV-360706 - Fix WriteAqlArgAt() crash in X86
In X86 on Windows, sizeof(size_t)=4, but size=8, for
amd::KernelParameterDescriptor::HiddenGlobalOffsetX/Y/Z items.
Loose the condition to prevent crash.

Change-Id: I2216f71f4d4fd6dd3766023b1c821cb3d35d7848


[ROCm/clr commit: 3d281114fb]
2022-10-20 18:04:49 -04:00
Tao Sang 3d281114fb SWDEV-360706 - Fix WriteAqlArgAt() crash in X86
In X86 on Windows, sizeof(size_t)=4, but size=8, for
amd::KernelParameterDescriptor::HiddenGlobalOffsetX/Y/Z items.
Loose the condition to prevent crash.

Change-Id: I2216f71f4d4fd6dd3766023b1c821cb3d35d7848
2022-10-20 18:04:49 -04:00
jatang 17948751c9 SWDEV-362451 - Disable OCLPinnedMemory for now.
Change-Id: I9a771cfc5751e06c5a494ad2e4036f9aa3efbe9a


[ROCm/clr commit: b7d64f9f5f]
2022-10-20 16:03:36 -04:00
jatang b7d64f9f5f SWDEV-362451 - Disable OCLPinnedMemory for now.
Change-Id: I9a771cfc5751e06c5a494ad2e4036f9aa3efbe9a
2022-10-20 16:03:36 -04:00
Sourabh Betigeri 644fc9421c SWDEV-361730 - Fix to match the kernel arg allocation sizes between gpu visible mem and host/runtime mem
Change-Id: I6495aa57e97b46373c1d90cc230435780fdcf3ef


[ROCm/clr commit: b2665ccc4b]
2022-10-20 14:20:37 -04:00
Sourabh Betigeri b2665ccc4b SWDEV-361730 - Fix to match the kernel arg allocation sizes between gpu visible mem and host/runtime mem
Change-Id: I6495aa57e97b46373c1d90cc230435780fdcf3ef
2022-10-20 14:20:37 -04:00
Ajay e81e8965c9 SWDEV-360944 - palBlit setArgument desc.size_ --> argSize
Fixes AMF tests that were impacted on Windows

Change-Id: I0326e794ec5a34fdec7a5ff4ed79d5bd055997d2


[ROCm/clr commit: a5b550e828]
2022-10-20 12:44:15 -04:00
Ajay a5b550e828 SWDEV-360944 - palBlit setArgument desc.size_ --> argSize
Fixes AMF tests that were impacted on Windows

Change-Id: I0326e794ec5a34fdec7a5ff4ed79d5bd055997d2
2022-10-20 12:44:15 -04:00
Laurent Morichetti 860e33dbed SWDEV-362046 - Report HIP_OPS activities using the ROCr driver_node_id instead of the device's index
The ROCclr assigns zero-based IDs to GPUs in the order they are
discovered. That zero-based ID is what is used to identify the GPU
on which the HIP_OPS activity took place.

When multiple ranks are used, each rank's first logical device always
has GPU ID 0, regardless of which physical device is selected with
CUDA_VISIBLE_DEVICES. Because of this, when merging trace files from
multiple ranks, GPU IDs from different processes may overlap.

The long term solution is to use the KFD's gpu_id which is stable
across APIs and processes. Unfortunately the gpu_id is not yet exposed
by the ROCr, so for now use the driver's node id.

Change-Id: Ib78854527d600d175bb76e2df0747c33f898c615


[ROCm/clr commit: 9a82118c85]
2022-10-20 12:31:30 -04:00
Laurent Morichetti 9a82118c85 SWDEV-362046 - Report HIP_OPS activities using the ROCr driver_node_id instead of the device's index
The ROCclr assigns zero-based IDs to GPUs in the order they are
discovered. That zero-based ID is what is used to identify the GPU
on which the HIP_OPS activity took place.

When multiple ranks are used, each rank's first logical device always
has GPU ID 0, regardless of which physical device is selected with
CUDA_VISIBLE_DEVICES. Because of this, when merging trace files from
multiple ranks, GPU IDs from different processes may overlap.

The long term solution is to use the KFD's gpu_id which is stable
across APIs and processes. Unfortunately the gpu_id is not yet exposed
by the ROCr, so for now use the driver's node id.

Change-Id: Ib78854527d600d175bb76e2df0747c33f898c615
2022-10-20 12:31:30 -04:00
Julia Jiang a8c33d177a SWDEV-357122 - fix failure in vdi so as to handle unreasonable input size in MallocManagedNegativeTest
Change-Id: I1ed1916b652afc67327b0935c3c60fc2a404df30


[ROCm/clr commit: dacd55f3d7]
2022-10-20 12:19:22 -04:00
Julia Jiang dacd55f3d7 SWDEV-357122 - fix failure in vdi so as to handle unreasonable input size in MallocManagedNegativeTest
Change-Id: I1ed1916b652afc67327b0935c3c60fc2a404df30
2022-10-20 12:19:22 -04:00
Wenkai Du 36e5e02e46 Merge remote-tracking branch 'nccl/master' into develop
[ROCm/rccl commit: 4f0e223db4]
2022-10-20 15:41:29 +00:00
Wenkai Du 4f0e223db4 Merge remote-tracking branch 'nccl/master' into develop 2022-10-20 15:41:29 +00:00
Bill(Shuzhou) Liu a25c71b730 Fix unit test compile error
Update the unit test as header enum type name is changed.

Change-Id: Ie965462da7d46259883650b15644003cf936982a
2022-10-20 09:54:11 -05:00
Bill(Shuzhou) Liu 33f7fba549 Fix unit test compile error
Update the unit test as header enum type name is changed.

Change-Id: Ie965462da7d46259883650b15644003cf936982a


[ROCm/amdsmi commit: a25c71b730]
2022-10-20 09:54:11 -05:00
Bill(Shuzhou) Liu 93bc47a5c3 Update the document
The README.md is updated for amdsmi

Change-Id: Ic027476d9f6b62e8972793e78db28b12644036ea
2022-10-20 09:25:02 -05:00
Bill(Shuzhou) Liu d944c4e11b Update the document
The README.md is updated for amdsmi

Change-Id: Ic027476d9f6b62e8972793e78db28b12644036ea


[ROCm/amdsmi commit: 93bc47a5c3]
2022-10-20 09:25:02 -05:00
Dejan Andjelkovic 6064f160a3 SWDEV-361376 - Add python wrapper
- Add generator for python wrapper
- Add interface, exception and init files
- Add CMake custom targets

Change-Id: I63c1d94fbb587387c22f559a3db79987eb214a2e
Signed-off-by: Dejan Andjelkovic <Dejan.Andjelkovic@amd.com>
2022-10-20 09:24:53 -05:00
Dejan Andjelkovic 14d9160565 SWDEV-361376 - Add python wrapper
- Add generator for python wrapper
- Add interface, exception and init files
- Add CMake custom targets

Change-Id: I63c1d94fbb587387c22f559a3db79987eb214a2e
Signed-off-by: Dejan Andjelkovic <Dejan.Andjelkovic@amd.com>


[ROCm/amdsmi commit: 6064f160a3]
2022-10-20 09:24:53 -05:00