* SWDEV-543981 new kernel latency test with different timing modes and taking multiple iterations of same test
* SWDEV-543981 cleanup
* SWDEV-543981 removed outdated hit test
* SWDEV-543981 Updated timing kernel
[ROCm/hip-tests commit: d227a8110c]
* SWDEV-543981 new kernel latency test with different timing modes and taking multiple iterations of same test
* SWDEV-543981 cleanup
* SWDEV-543981 removed outdated hit test
* SWDEV-543981 Updated timing kernel
Implemented reference counting to manage init and shutdown processes,
allowing for multiple initializations and shutdowns.
[ROCm/amdsmi commit: c45a53d751]
Leverages the traits of extended-scope fine-grain memory to get rid of a device-scope acquire-release fence. This improves throughput for single node workloads on gfx942 and gfx950 for some input sizes (e.g., ~32 MiB to about 256 MiB) when using the simple protocol. Multinode workloads on MI300X see a smaller but statistically significant uplift for some message sizes. Runtime disablement is supported via setting the environment variable RCCL_GFX942_CHEAP_FENCE_ON to 0.
[ROCm/rccl commit: 1aa2570b48]
Leverages the traits of extended-scope fine-grain memory to get rid of a device-scope acquire-release fence. This improves throughput for single node workloads on gfx942 and gfx950 for some input sizes (e.g., ~32 MiB to about 256 MiB) when using the simple protocol. Multinode workloads on MI300X see a smaller but statistically significant uplift for some message sizes. Runtime disablement is supported via setting the environment variable RCCL_GFX942_CHEAP_FENCE_ON to 0.
Also advice parameter of madvise() system call is not a bitmask. So fix
that also
v2: Use MAP_SHARED instead of MAP_PRIVATE. This avoids MMU notifiers and
evictions.
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
* SWDEV-532479 - Add tracking of hostcall memory allocations
* SWDEV-532479 - Remove hostcall allocations if request is received
* SWDEV-532479 - Cleanup
* SWDEV-532479 - Naming fix
* SWDEV-532479 - Add new separator after each new function
[ROCm/clr commit: b58faa2f37]
* SWDEV-532479 - Add tracking of hostcall memory allocations
* SWDEV-532479 - Remove hostcall allocations if request is received
* SWDEV-532479 - Cleanup
* SWDEV-532479 - Naming fix
* SWDEV-532479 - Add new separator after each new function
* Added useAcc as a template parameter to address the 2% performance regression in allreduceWithBias
---------
Co-authored-by: Marzieh Berenjkoub <mberenjk@amd.com>
[ROCm/rccl commit: c61152baa4]
* Added useAcc as a template parameter to address the 2% performance regression in allreduceWithBias
---------
Co-authored-by: Marzieh Berenjkoub <mberenjk@amd.com>
* Updated ubuntu noble workflow
* Actually use ROCm when supposed to; Added Python 3.13; Set ompt flag to ON
* Print compiler version outside of if logic
Co-authored-by: David Galiffi <David.Galiffi@amd.com>
* Move compiler version info
Co-authored-by: David Galiffi <David.Galiffi@amd.com>
* Updated `working-directory`
set to `projects/rocprofiler-systems`
* Update .github/workflows/rocprofiler-systems-ubuntu-noble.yml
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Simplifying some of the build options
---------
Co-authored-by: Kian Cossettini <Kian.Cossettini@amd.com>
Co-authored-by: David Galiffi <David.Galiffi@amd.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* SWDEV-547453 Release the kernel command if the operation returns an error
* SWDEV-547453 - Initialize parameters_ to default value
* SWDEV-547453 - Run clang-format
[ROCm/clr commit: a15957fee9]
* SWDEV-547453 Release the kernel command if the operation returns an error
* SWDEV-547453 - Initialize parameters_ to default value
* SWDEV-547453 - Run clang-format
Previously regardless of hsa_agent passed the prefetch is always driven
to node 0, now the agent of interest may be properly prefetched.
Signed-off-by: Sunday Clement <Sunday.Clement@amd.com>
Co-authored-by: Sunday Clement <Sunday.Clement@amd.com>
Co-authored-by: systems-assistant[bot] <systems-assistant[bot]@users.noreply.github.com>
* SWDEV-549309 - Ensure that the kernel is issued to a stream that is associated to the current device
* SWDEV-549309 - Ensure that the kernel is issued to a stream that is associated to the current device in atomic tests
* SWDEV-549309 - Added fix for Unit_hipMallocFromPoolAsync_Multidevice_Concurrent test
* SWDEV-549309 - Added fix for Unit_hipMemcpyPeerAsync_Positive_Default
* SWDEV-549309 - Added fix for Unit_hipStreamAttachMemAsync_Positive_AttachGlobal and Unit_hipGetProcAddress_GraphAPIs_AddMemsetMemcpyNodes
[ROCm/hip-tests commit: d21a95cea8]
* SWDEV-549309 - Ensure that the kernel is issued to a stream that is associated to the current device
* SWDEV-549309 - Ensure that the kernel is issued to a stream that is associated to the current device in atomic tests
* SWDEV-549309 - Added fix for Unit_hipMallocFromPoolAsync_Multidevice_Concurrent test
* SWDEV-549309 - Added fix for Unit_hipMemcpyPeerAsync_Positive_Default
* SWDEV-549309 - Added fix for Unit_hipStreamAttachMemAsync_Positive_AttachGlobal and Unit_hipGetProcAddress_GraphAPIs_AddMemsetMemcpyNodes
GunaShekar, Ajay
Luo, Phoebe
Manocha, Rahul
Kian Cossettini
Ammar ELWazir
Bill Liu
Arandjelovic, Marko
alex-breslow-amd
Daniel Su
hkasivis
Dittakavi, Satyanvesh
Joseph Macaranas
RahulC
jokim-amd
mberenjk
systems-assistant[bot]
jamessiddeley-amd
Maisam Arif
Giovanni Lenzi Baraldi
Stojiljkovic, Vladana
Adel Johar
Assiouras, Ioannis
Jonathan R. Madsen