Introducing experimental OpenCL backend with support for Qualcomm Adreno GPUs#10693

Merged

max-krasnyansky merged 28 commits intoggml-org:masterfrom

CodeLinaro:adreno-support

Dec 13, 2024

Commits on Dec 13, 2024

[cl][adreno] Add Adreno GPU support

committed
[cl][ci] Add workflow for CL

lhez
authored and
max-krasnyansky
committed
[cl][adreno] Fix memory leak for non SMALL_ALLOC path

lhez
authored and
max-krasnyansky
committed
opencl: integrate backend dyn.load interface and fix compiler and format warnings
max-krasnyansky
committed
opencl: remove small-alloc support and fix build errors for non-opencl platforms
max-krasnyansky
committed
opencl: fixed merge conflict (MUSA added twice in cmake)
max-krasnyansky
committed
opencl-ci: use RUNNER_TEMP instead of github.workspace
max-krasnyansky
committed
opencl: fix embed tool invocation with python3
max-krasnyansky
committed
opencl: CI workflow fixes
max-krasnyansky
committed
opencl: Clean up small-alloc in CMake files

lhez
authored and
max-krasnyansky
committed
opencl: cleanup ggml-opencl2 header file
max-krasnyansky
committed
opencl: use ulong for offsets and strides in ADD kernel
max-krasnyansky
committed
opencl: use cl_ulong for all offsets
max-krasnyansky
committed
opencl: use cl_ulong for sizes and strides
max-krasnyansky
committed
opencl: use `GGML_LOG_xxx` instead of `fprintf(stderr, ...)`

lhez
authored and
max-krasnyansky
committed
opencl: rename backend `opencl2` -> `opencl`

lhez
authored and
max-krasnyansky
committed
opencl: rename kernel files `ggml-opencl2` -> `ggml-opencl`

lhez
authored and
max-krasnyansky
committed
opencl: make OpenCL required, remove redundant lib and inc directories

lhez
authored and
max-krasnyansky
committed
opencl: rename backend - funcs, structs, etc `opencl2` -> `opencl`

lhez
authored and
max-krasnyansky
committed
opencl: remove copyright marker since main license already covers

lhez
authored and
max-krasnyansky
committed
opencl: replace some more OPENCL2 leftovers
max-krasnyansky
committed
opencl: remove limits on `tensor_extra`

lhez
authored and
max-krasnyansky
committed
opencl: use pools for `tensor_extra`

lhez
authored and
max-krasnyansky
committed
opencl: fix compiler warnings with GCC and Clang
max-krasnyansky
committed
opencl: fail gracefully if opencl devices are not available
max-krasnyansky
committed
opencl: fix MSVC builds (string length error)
max-krasnyansky
committed
opencl: check for various requirements, allow deprecated API

lhez
authored and
max-krasnyansky
committed
opencl: update log message for unsupported GPUs
max-krasnyansky
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Introducing experimental OpenCL backend with support for Qualcomm Adreno GPUs#10693

Introducing experimental OpenCL backend with support for Qualcomm Adreno GPUs#10693
max-krasnyansky merged 28 commits intoggml-org:masterfrom
CodeLinaro:adreno-support

Commits on Dec 13, 2024

[cl][adreno] Add Adreno GPU support

[cl][ci] Add workflow for CL

[cl][adreno] Fix memory leak for non SMALL_ALLOC path

opencl: integrate backend dyn.load interface and fix compiler and format warnings

opencl: remove small-alloc support and fix build errors for non-opencl platforms

opencl: fixed merge conflict (MUSA added twice in cmake)

opencl-ci: use RUNNER_TEMP instead of github.workspace

opencl: fix embed tool invocation with python3

opencl: CI workflow fixes

opencl: Clean up small-alloc in CMake files

opencl: cleanup ggml-opencl2 header file

opencl: use ulong for offsets and strides in ADD kernel

opencl: use cl_ulong for all offsets

opencl: use cl_ulong for sizes and strides

opencl: use `GGML_LOG_xxx` instead of `fprintf(stderr, ...)`

opencl: rename backend `opencl2` -> `opencl`

opencl: rename kernel files `ggml-opencl2` -> `ggml-opencl`

opencl: make OpenCL required, remove redundant lib and inc directories

opencl: rename backend - funcs, structs, etc `opencl2` -> `opencl`

opencl: remove copyright marker since main license already covers

opencl: replace some more OPENCL2 leftovers

opencl: remove limits on `tensor_extra`

opencl: use pools for `tensor_extra`

opencl: fix compiler warnings with GCC and Clang

opencl: fail gracefully if opencl devices are not available

opencl: fix MSVC builds (string length error)

opencl: check for various requirements, allow deprecated API

opencl: update log message for unsupported GPUs

Introducing experimental OpenCL backend with support for Qualcomm Adreno GPUs#10693max-krasnyansky merged 28 commits intoggml-org:masterggml-org/llama.cpp:masterfrom CodeLinaro:adreno-supportCodeLinaro/llama.cpp:adreno-supportCopy head branch name to clipboard

Commits

Commits on Dec 13, 2024

Introducing experimental OpenCL backend with support for Qualcomm Adreno GPUs#10693
max-krasnyansky merged 28 commits intoggml-org:masterfrom
CodeLinaro:adreno-support