PATHS have lower priority than system paths so if there is a clang installation in /usr/local find_package(Clang)
will pick that instead. To fix this provide the llvm path via the HINTS argument instead.
Change-Id: I57794a9a5c5fc26f12c60798fbc8e826c1880e13
[ROCm/clr commit: 238b973a5a]
Make sure parent_ field is cleared for the internal image views.
The internal image views don't require dependencies tracking.
The issue appeard only when Navi10 pitch workaround was enabled.
Change-Id: I376d212750085a9391f8c32fc2979dcb5d93c89c
[ROCm/clr commit: 6b89980fb2]
In hip-runtime-amd RPM package the autoreqprov field was disabled. So the package provides list will not have any libraries
Other packages that depends on hip libraries will be having the reported issue.
Disable the RPM autoreq field alone, so that the package provides list will have libraries
Change-Id: I40c30d01ea137867bdb64c2509c1fc86827c09dd
[ROCm/clr commit: 19f43aa273]
If cl_khr_fp16 extension is enabled, then OCL runtime should report CL_DEVICE_HALF_FP_CONFIG.
Change-Id: I7c4ac48387f80bc704a475c57e5b52a462090d1b
[ROCm/clr commit: ad2c1e899a]
Update pCopy->srcMemoryType and pCopy->dstMemoryType so
that ihipGetMemcpyParam3DCommand() won't redetect memory
type. This will prevent dstXInBytes and srcXInBytes getting
updated twice with offset in hipMemoryTypeUnified type.
Change-Id: Ief4ea81941e25d22bcafd213b3b84a65eb23cf94
[ROCm/clr commit: 56daa6c489]
- Enable CUs adjacent pairwise for WGP mode
- In HostQueue::terminate() do not segfault if virtual device hasn't been created
Change-Id: I94402ff333308af5824878086cc238b3993d534d
[ROCm/clr commit: 8c1232124e]
- Rename HIP_USE_SDMA_QUERY to DEBUG_CLR_USE_SDMA_QUERY as this is
supposed to be a temporary env var for debug purposes only.
Change-Id: If6ebd52ab87624375a3df24ceccdcc05c60a65af
[ROCm/clr commit: 770b2a4711]
Blit manager requires an image view to reduce the amount
of copy kernels. Creation/destruction of a view in ROCr is
an expensive operation. Thus, runtime can cache views for fast access.
Change-Id: Ia67d775b481cc8326d91215ca22d4a73c1dddb59
[ROCm/clr commit: d29755452b]
- Remove large bar memcpy path. Since we end up waiting for a barrier,
its defeating the true intent of the copy, Also memcpy over PCIE\XGMI is
introducing variability in perf for HPC apps like GROMACS
Change-Id: I3b5c9d9ce93333959c39023bf4f703e2ccb6e3af
[ROCm/clr commit: 0a3d4bd4d4]
rocclr/platform/external_memory.hpp:93:30: warning: class with destructor marked 'final' cannot be inherited from [-Wfinal-dtor-non-final-class]
93 | virtual ~ExternalImage() final {}
Change-Id: I56d760fa6c08544100e3bc03d35129bd16d8a428
[ROCm/clr commit: e9750de215]
Add missing mipmap Apis’ implementation.
Fix some bugs of mimpmap apis.
Use hipmipmappedArray to differentiate cuda
and driver apis on Nvidia.
Change-Id: I6079d9f3b2ddf4e42b9a6f7f3902322cfca02cfd
[ROCm/clr commit: f03c11491b]
let ROCr and addrlib decided whether a pitch is supported.
Then, check if we uses workaround.
Also improve image performance for Navi 1x in some cases.
Change-Id: If8bda46648b23c1dcfa26608ca3bb9cd89fef430
[ROCm/clr commit: b961d4a970]
- Make IPC interfaces generic between devices and rely on the IPC buffer
for attach/detach logic
Change-Id: Id3c18d122030329b7ee532bbb6317de9dd6a0bbe
[ROCm/clr commit: 9aa6f25f1c]
Anusha GodavarthySurya
Ioannis Assiouras
Sourabh Betigeri
sdashmiz
kjayapra-amd
Todd tiantuo Li
German Andryeyev
Alex Xie
Ranjith Ramakrishnan
Maneesh Gupta
victzhan
taosang2
Jaydeep Patel
Rakesh Roy
Saleel Kudchadker
ajay
Jatin Chaudhary
Julia Jiang
Matt Arsenault
Rahul Garg
pghafari