Skip to content

Fix for multinode processing #69

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Mar 7, 2025
Merged

Fix for multinode processing #69

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
cddb23e
Fix for multinode processing
razdoburdin Mar 7, 2025
ae632dd
fix build
Mar 7, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion gputreeshap
Open in desktop
Submodule gputreeshap updated 19 files
+0 −14 .github/CODEOWNERS
+0 −4 .github/copy-pr-bot.yaml
+0 −31 .github/workflows/pr.yaml
+1 −1 .gitignore
+0 −26 .pre-commit-config.yaml
+23 −48 CMakeLists.txt
+1 −1 Doxyfile.in
+21 −17 GPUTreeShap/gpu_treeshap.h
+1 −1 README.md
+38 −71 benchmark/benchmark.py
+0 −2 build.sh
+0 −31 ci/build_and_test.sh
+0 −20 ci/check_style.sh
+33 −0 ci/checks/style.sh
+59 −0 ci/gpu/build.sh
+0 −45 ci/release/update-version.sh
+0 −69 dependencies.yaml
+0 −19 fetch_rapids.cmake
+2 −3 tests/test_gpu_treeshap.cu
11 changes: 9 additions & 2 deletions plugin/sycl/device_manager.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,18 +21,25 @@ ::sycl::queue* DeviceManager::GetQueue(const DeviceOrd& device_spec) const {
(collective::IsDistributed());
DeviceRegister& device_register = GetDevicesRegister();
if (not_use_default_selector) {
const int device_idx =
collective::IsDistributed() ? collective::GetRank() : device_spec.ordinal;
if (device_spec.IsSyclDefault()) {
auto& devices = device_register.devices;
const int device_idx = collective::IsDistributed()
? collective::GetRank() % devices.size()
: device_spec.ordinal;
CHECK_LT(device_idx, devices.size());
queue_idx = device_idx;
} else if (device_spec.IsSyclCPU()) {
auto& cpu_devices_idxes = device_register.cpu_devices_idxes;
const int device_idx = collective::IsDistributed()
? collective::GetRank() % cpu_devices_idxes.size()
: device_spec.ordinal;
CHECK_LT(device_idx, cpu_devices_idxes.size());
queue_idx = cpu_devices_idxes[device_idx];
} else if (device_spec.IsSyclGPU()) {
auto& gpu_devices_idxes = device_register.gpu_devices_idxes;
const int device_idx = collective::IsDistributed()
? collective::GetRank() % gpu_devices_idxes.size()
: device_spec.ordinal;
CHECK_LT(device_idx, gpu_devices_idxes.size());
queue_idx = gpu_devices_idxes[device_idx];
} else {
Expand Down
Loading