Skip to content

[CORE]: Improve filesystem error messages during Linux device discovery #27289

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

Open
theHamsta wants to merge 2 commits into
base: main
Choose a base branch
from
Open

[CORE]: Improve filesystem error messages during Linux device discovery #27289

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
f3bf120
[CORE]: Address review dog lints in #26210 (Linux device discovery)
theHamsta Jan 28, 2026
da15088
[CORE]: Improve filesystem error messages during Linux device discovery
theHamsta Jan 28, 2026
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
18 changes: 11 additions & 7 deletions onnxruntime/core/platform/linux/device_discovery.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
#include <iterator>
#include <optional>
#include <regex>
#include <string>
#include <string_view>

#include "core/common/common.h"
Expand All @@ -21,14 +22,17 @@ namespace onnxruntime {

namespace {

Status ErrorCodeToStatus(const std::error_code& ec) {
Status ErrorCodeToStatus(const std::error_code& ec, const fs::path& path, std::string_view context) {
Copy link
Contributor Author

@theHamsta theHamsta Feb 9, 2026 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

the string "context" could be overkill. I thought it would provide useful context, why we accessed the file. But I would at least include the filesystem path that caused the error code. Otherwise, debugging the error code could be difficult

if (!ec) {
return Status::OK();
}

return Status{common::StatusCategory::ONNXRUNTIME, common::StatusCode::FAIL,
MakeString("Error: std::error_code with category name: ", ec.category().name(),
", value: ", ec.value(), ", message: ", ec.message())};
", value: ", ec.value(),
", message: ", ec.message(),
", filesystem path: ", path,
", context: ", context)};
}

struct GpuSysfsPathInfo {
Expand All @@ -40,7 +44,7 @@ Status DetectGpuSysfsPaths(std::vector<GpuSysfsPathInfo>& gpu_sysfs_paths_out) {
std::error_code error_code{};
const fs::path sysfs_class_drm_path = "/sys/class/drm";
const bool sysfs_class_drm_path_exists = fs::exists(sysfs_class_drm_path, error_code);
ORT_RETURN_IF_ERROR(ErrorCodeToStatus(error_code));
ORT_RETURN_IF_ERROR(ErrorCodeToStatus(error_code, sysfs_class_drm_path, "Checking existance of DRM sysfs path"));

if (!sysfs_class_drm_path_exists) {
gpu_sysfs_paths_out = std::vector<GpuSysfsPathInfo>{};
Expand Down Expand Up @@ -69,7 +73,7 @@ Status DetectGpuSysfsPaths(std::vector<GpuSysfsPathInfo>& gpu_sysfs_paths_out) {
std::vector<GpuSysfsPathInfo> gpu_sysfs_paths{};

auto dir_iterator = fs::directory_iterator{sysfs_class_drm_path, error_code};
ORT_RETURN_IF_ERROR(ErrorCodeToStatus(error_code));
ORT_RETURN_IF_ERROR(ErrorCodeToStatus(error_code, sysfs_class_drm_path, "Iterating over DRM sysfs devices"));

for (const auto& dir_item : dir_iterator) {
const auto& dir_item_path = dir_item.path();
Expand Down Expand Up @@ -121,7 +125,7 @@ Status GetPciBusId(const std::filesystem::path& sysfs_path, std::optional<std::s

std::error_code error_code;
auto pci_bus_id_path = std::filesystem::canonical(sysfs_path / "device", error_code); // resolves symlink to PCI bus id, e.g. 0000:65:00.0
ORT_RETURN_IF_ERROR(ErrorCodeToStatus(error_code));
ORT_RETURN_IF_ERROR(ErrorCodeToStatus(error_code, sysfs_path / "device", "Getting PCI bus id from DRM device by resolving symlink"));

auto pci_bus_id_filename = pci_bus_id_path.filename();
if (std::regex_match(pci_bus_id_filename.string(), pci_bus_id_regex)) {
Expand All @@ -130,9 +134,9 @@ Status GetPciBusId(const std::filesystem::path& sysfs_path, std::optional<std::s
pci_bus_id = {};
LOGS_DEFAULT(WARNING) << MakeString("Skipping pci_bus_id for PCI path at \"",
pci_bus_id_path.string(),
"\" because filename \"", pci_bus_id_filename, "\" dit not match expected pattern of ",
"\" because filename \"", pci_bus_id_filename, "\" did not match expected pattern of ",
regex_pattern);
};
}

return Status::OK();
}
Expand Down