A GPU-accelerated library containing highly optimized building blocks and an execution engine for data processing to accelerate deep learning training and inference applications.
In file included from /build/python-nvidia-dali/src/python-nvidia-dali/dali/pipeline/workspace/workspace.h:27,
from /build/python-nvidia-dali/src/python-nvidia-dali/dali/pipeline/workspace/sample_workspace.cc:15:
/build/python-nvidia-dali/src/python-nvidia-dali/dali/pipeline/data/tensor_list.h:87:34: warning: template-id not allowed for constructor in C++20 [-Wtemplate-id-cdtor]
87 | DLL_PUBLIC TensorList<Backend>(TensorList<Backend> &&other) noexcept;
| ^~~~~~~~~~~~~~~~~~~
/build/python-nvidia-dali/src/python-nvidia-dali/dali/pipeline/data/tensor_list.h:87:34: note: remove the ‘< >’
[ 38%] Building CXX object dali/CMakeFiles/dali.dir/util/cufile.cc.o
In file included from /build/python-nvidia-dali/src/python-nvidia-dali/dali/util/uri.cc:15:
/build/python-nvidia-dali/src/python-nvidia-dali/dali/util/uri.h:51:8: error: use of enum ‘ParseOpts’ without previous declaration
51 | enum ParseOpts : uint32_t {
| ^~~~~~~~~
/build/python-nvidia-dali/src/python-nvidia-dali/dali/util/uri.h:51:20: error: ‘uint32_t’ was not declared in this scope
51 | enum ParseOpts : uint32_t {
| ^~~~~~~~
/build/python-nvidia-dali/src/python-nvidia-dali/dali/util/uri.h:24:1: note: ‘uint32_t’ is defined in header ‘<cstdint>’; this is probably fixable by adding ‘#include <cstdint>’
23 | #include "dali/core/api_helper.h"
+++ |+#include <cstdint>
24 |
/build/python-nvidia-dali/src/python-nvidia-dali/dali/util/uri.h:51:29: error: default member initializer for unnamed bit-field
51 | enum ParseOpts : uint32_t {
| ^
Version
1.40.0
Describe the bug.
https://github.com/NVIDIA/DALI/blob/v1.40.0/dali/util/uri.h#L51 use
uint32_t
without including the cstdint header, this leads to compilation failure.Minimum reproducible example
Relevant log output
Other/Misc.
quick fix, add one line to https://github.com/NVIDIA/DALI/blob/v1.40.0/dali/util/uri.h#L17
Check for duplicates