- hipDeviceGet/SetGraphMemAttr
- hipDeviceGraphMemTrim
- there is no memory pool for graphs currently
Signed-off-by: sdashmiz <shadi.dashmiz@amd.com>
Change-Id: I11db76ea7ea1c7732175fc93264448052357e8dc
[ROCm/hipother commit: 9e3c65ccbd]
- Memcpy1D node type is not complying with cuda
Signed-off-by: sdashmiz <shadi.dashmiz@amd.com>
Change-Id: If8113f5e699de0c62d98effc4580a2e0fee9a950
[ROCm/hipother commit: 9583567388]
Let more types support hipResourceViewDesc.
Add missing symbols to fix cuda build failing.
Change-Id: Ife694cc6491427093863252e257e820b0bb4fa0f
[ROCm/hipother commit: 3fc1dca59f]
Update HIP's unsafeAtomicAdd to:
- Compile properly even when not compiling for gfx90a
- Fall back to safe atomic add on non-gfx90a architectures
- use flat atomic add for FP64 on gfx90a, instead of dynamically
checking memory spaces.
In addition, when the compiler is passed -munsafe-fp-atomics, it
will define __AMDGCN_UNSAFE_FP_ATOMICS__. When this happens, the
compiler is requesting that the HIP headers force all HIP
atomicAdd() calls on floats or doubles to use their unsafe versions.
This patch thus causes unsafeAtomicAdd() calls when that define
is seen. This call to unsafeAtomicAdd() is also done for atomicSub(),
since that calls atomicAdd underneath. This is not done for
system-scope atomicAdd because, on gfx90a, system-scope atomic FP
add instructions would need to target fine-grained memory, which is
always unsafe.
This patch also creates safeAtomicAdd() functions for float and double.
These functions will create a standalone safe atomic, even when the
application is compiled with -munsafe-fp-atomics.
Finally, this patch adds wrappers in the Nvidia path of HIP so that
these HIP functions call through to atomicAdd there as well.
Change-Id: I8af0621d3d28ea30c9278bfeea7393d03bbdac6d
[ROCm/hipother commit: b712a51e74]
This reverts commit 09432f798b.
Reason for revert: hip changes not in yet
Change-Id: I89e35d171cf01fb82d361279998efcc90e54ef33
[ROCm/hipother commit: 16dbf85da0]
HIP apps running on CUDA 11.0 are failing due to some graph APIs/enums
added only in CUDA 11.1 or CUDA 11.3
Change-Id: I0d32b412cb76c42c7b3a9c612d750990f9e89908
[ROCm/hipother commit: f607f01311]
Replace return value of hipPointerGetAttributes() on NV from
hipErrorUnknown to hipErrorInvalidValue if memory has been freed.
Change-Id: I3fe6dbc35a7a14aa9109df297b7885df83d28149
[ROCm/hipother commit: 3efa0aa254]
Change-Id: I5daeacd9dd5c6ce7f914d6e6e45dd41fb2a675a5
hipMemRangeGetAttributes was returning hipErrorInvalidValue due to improper
mapping of the arguments to cudaMemRangeGetAttributes.
[ROCm/hipother commit: 5165e3650a]
Add concurrentManagedAccess detection in hipMallocManaged test.
Skip test when device doesn't support concurrentManagedAccess.
Change-Id: Ie54046feef3baba857a7068972ec1fc1a60c2dfd
[ROCm/hipother commit: 4fc75f9536]
1.Rename include/hip/hcc_detail/ as include/hip/amd_detail/
2.Rename include/hip/nvcc_detail/ as include/hip/nvidia_detail/
3.Create __HIP_PLATFORM_AMD__ to replace __HIP_PLATFORM_HCC__
4.Create __HIP_PLATFORM_NVIDIA__ to replace __HIP_PLATFORM_NVCC__
After hcc_detail, nvcc_detail, __HIP_PLATFORM_HCC__ and __HIP_PLATFORM_NVCC__
have been removed from upstream, they will be removed from hip runtime.
Change-Id: I1ae457effd739d6c25bca203c1724b026be21fce
[ROCm/hipother commit: 49e3b900bb]
Jaydeep Patel
sdashmiz
Tao Sang
Joseph Greathouse
Rakesh Roy
German Andryeyev
Satyanvesh Dittakavi
haoyuan2
anusha GodavarthySurya
Jatin Chaudhary
Sarbojit Sarkar
agunashe
Rahul Garg
Maneesh Gupta
Julia Jiang