* Revert disabling of context tracking for Radeon
Original commit 6fc228e2
`Disable context tracking for the current version. (#1839)`
* Add env variable for disabling of context tracking for Radeon
`export NCCL_DISABLE_CONTEXT_TRACKING=1` to force disable of context tracking
* Update docs/how-to/rccl-usage-tips.rst
Fix grammar, thanks @amd-jnovotny
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
* Rename NCCL_DISABLE_CONTEXT_TRACKING -> RCCL_DISABLE_CONTEXT_TRACKING
* Revert changes in includes and rename util function
---------
Co-authored-by: Jeffrey Novotny <jnovotny@amd.com>
* conf: update RTD config to ub24.04 (doxygen 1.9.8) and py3.12
* update generate-docs workflow
* Update "modules" to "topics" due to Doxygen 1.9.8
* bump rocm-docs-core to 1.25.0 and pip-compile requirements.txt
* doxygen: fill in version string in Doxyfile from conf.py
* remove unneeded rocm-smi-lib tutorials
* remove wikipedia references in doxyfile to satisfy ci check
Signed-off-by: Park, Peter <Peter.Park@amd.com>
* conf: update RTD config to ub24.04 (doxygen 1.9.8) and py3.12
* update generate-docs workflow
* Update "modules" to "topics" due to Doxygen 1.9.8
* bump rocm-docs-core to 1.25.0 and pip-compile requirements.txt
* doxygen: fill in version string in Doxyfile from conf.py
* remove unneeded rocm-smi-lib tutorials
* remove wikipedia references in doxyfile to satisfy ci check
Signed-off-by: Park, Peter <Peter.Park@amd.com>
[ROCm/amdsmi commit: 311eade5b1]
This unbreaks having sources on one mount point and builds at another.
Signed-off-by: Marius Brehler <marius.brehler@amd.com>
Change-Id: I68363112382a95baaa867cad91e09bdec2b30d90
This unbreaks having sources on one mount point and builds at another.
Signed-off-by: Marius Brehler <marius.brehler@amd.com>
Change-Id: I68363112382a95baaa867cad91e09bdec2b30d90
[ROCm/amdsmi commit: bd3579a1ac]
* SWDEV-1 - if platform auto-detection via hipconfig fails, provide a meaningful error and do not try to parse the output
* SWDEV-1 - if getting HIP_VERSION via hipconfig fails, provide a meaningful error and do not try to parse the output
This will update the CI to match TheRocks CI by introducing to use a
python script to report on health status. Commands that were in the
section that modified the status were moved to separate sections
(ccache, git config, ..).
Related PR:
https://github.com/ROCm/TheRock/pull/1516
Having the SOVERSION derived from the git tags doesn't scale well
for distributions that don't have the git history while building
(such as a tarball).
As part of e7d6590 the strings are parsed from a header. Re-use
those.
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Having the SOVERSION derived from the git tags doesn't scale well
for distributions that don't have the git history while building
(such as a tarball).
As part of 8b96ee5 the strings are parsed from a header. Re-use
those.
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
[ROCm/amdsmi commit: ccfdb65b6f]
Added back the temp-type map initialization to
RSMI_TEMP_TYPE_INVALID before probing hwmon files. This
prevents std::out_of_range for unsupported or absent
temperature sensor types.
Signed-off-by: Bindhiya Kanangot Balakrishnan <Bindhiya.KanangotBalakrishnan@amd.com>
Added back the temp-type map initialization to
RSMI_TEMP_TYPE_INVALID before probing hwmon files. This
prevents std::out_of_range for unsupported or absent
temperature sensor types.
Signed-off-by: Bindhiya Kanangot Balakrishnan <Bindhiya.KanangotBalakrishnan@amd.com>
[ROCm/amdsmi commit: 3e7e4ab1ac]
- Updated kernel arg manager to support allocating kernel args on multiple devices for single graph.
- Updated AQL path to capture on the device where graph node is added.
Co-authored-by: Anusha GodavarthySurya <Anusha.GodavarthySurya@amd.com>
* SWDEV-550626 - Refactor atomics header and tests
1. Introduce __HIP_ATOMIC_BACKWARD_COMPAT.
By default we define __HIP_ATOMIC_BACKWARD_COMPAT=1 to
let hip atomic functions maintain old assumptions. if
users want to adopt the new behavior, that is , by default
assume no-fine-grained no-remote-memory, then they can
define __HIP_ATOMIC_BACKWARD_COMPAT=0 and get the new
behaviour.
2. Use __HIP_ATOMIC_BACKWARD_COMPAT_MEMORY to replace
original __HIP_FINE_GRAINED_MEMORY in atomic header.
And apply __HIP_FINE_GRAINED_MEMORY onto all
atomicXXX_system() functions to prevent failure on memory
allocated by hipHostMalloc().
3. Replace HIP_TEST_FINE_GRAINED_MEMORY with
HIP_TEST_ATOMIC_BACKWARD_COMPAT_MEMORY in hip-tests.
4. Fix negative test errors.
Fix managed memory test error on memory order.
some other minor changes.
As a result all originally disabled tests are enabled.
5. Add more atomics tests in some cases.
6. Reduce test time in each case.
Reduce iteration number to 1 for tests that cost too much time.
8. Put common codes into hip_test_common.hh
* Fix formatting & Examples for amdsmi_get_afids_from_cper CPER record examples in documentation
---------
Change-Id: Ib5e268dc818adc633541652a0eb982641385bf7d
Signed-off-by: AL Musaffar, Yazen <Yazen.ALMusaffar@amd.com>
David DeBonis
Nilesh M Negi
Bhuvan Mital
Artem Kuzmitckii
Ajay GunaShekar
Arif, Maisam
Park, Peter
Mustafa Abduljabbar
alex-breslow-amd
Laura Promberger
amd-hsivasun
Rahul Manocha
Gerardo Hernandez
German Andryeyev
Aurelien Bouteiller
Mario Limonciello
Larry Meadows
Bindhiya Kanangot Balakrishnan
Sourabh U Betigeri
David Yat Sin
Godavarthy Surya, Anusha
MachineTom
Yiltan
systems-assistant[bot]
AL Musaffar, Yazen
Stella Laurenzo