linxt20 opened a new issue, #468: URL: https://github.com/apache/tsfile/issues/468
bug1-nullPointer-tsfile_reader_table_test.cc-159 bug2-nullPointer-tsfile_reader_table_test.cc-247: bug3-nullPointer-tsfile_reader_table_test.cc-358: bug4-nullPointer-tsfile_reader_test.cc-144: bug5-nullPointer-tsfile_writer_test.cc-192: bug6-nullPointer-tsfile_writer_test.cc-371: bug7-nullPointer-tsfile_writer_test.cc-623: bug8-nullPointer-tsfile_writer_test.cc-757: bug9-nullPointer-tsfile_writer_test.cc-827: bug10-nullPointer-tsfile_writer_test.cc-897. These 10 bugs belong to the same problem, that is, the query result may be nullptr, and the API needs to be checked before calling. Take bug1 as an example: Problem description: The table_result_set here may be a null pointer, and the next operation cannot be called. Cause analysis: The table_result_set here comes from the reference parameter tmp_result_set passed in by reader.query. In the query of the reader, the query function of table_query_executor will be called to complete the actual result return, but in the query function of TableQueryExecutor, the return result ret_qds can be nullptr (nullptr is assigned in line 41). Therefore, when performing next on the query result, you need to check whether it is nullptr first. bug11-nullPointerArithmetic-gzip_compressor_test.cc-90: bug12-nullPointerArithmetic-gzip_compressor_test.cc-124: bug13-nullPointerArithmetic-lz4_compressor_test.cc-90: bug14-nullPointerArithmetic-lz4_compressor_test.cc-123: bug15-nullPointerArithmetic-lzo_compressor_test.cc-90: bug16-nullPointerArithmetic-lzo_compressor_test.cc-123: bug17-nullPointerArithmetic-snappy_compressor_test.cc-90: bug18-nullPointerArithmetic-snappy_compressor_test.cc-123: . These 8 bugs belong to the same problem, that is, the result of decompression may be nullptr, which needs to be checked before it can be used as a pointer for calculation. Take bug11 as an example: Problem description: The decpmressed_buf here may be a null pointer and cannot be directly involved in pointer calculation. Cause analysis: The decpmressed_buf here comes from the reference parameter decpmressed_buf of the uncompress operation of the compressor, and the uncompress function will call the corresponding decompression function. In the decompression function of gzip, the passed parameter will not be assigned when the decompression fails, and will be assigned to nullptr when the byte stream is empty. Therefore, there is a possibility of a null pointer, and it needs to be checked before use. At the same time, when observing the code, it is found that the status code of whether the decompression is successful is not returned and checked, which is also a risk point. bug19-containerOutOfBounds-expression.cc-115. Problem description: When the size is 0, an exception should be thrown, but it is not thrown here, causing the program to continue running. In an abnormal situation, at(0) was called, and an out-of-bounds situation occurred. From the code, if the execution continues in the abnormal situation, there will be more errors, so it is necessary to interrupt in the abnormal situation. bug20-missingMemberCopy-schema.h-236. Problem description: In the move constructor, the variable column_pos_index_ is omitted when assigning a value, which will cause an abnormal result. Other serious abnormality investigations: - memleak 3: confirmed as a false positive after investigation. - virtualCallInConstructor 1: confirmed as a false positive after investigation. - uninitMemberVar 3: confirmed as a false positive after investigation. Among other problems, the test unit is marked as an unused function, so it accounts for a large number. Most of the others are code style and optimization suggestion issues, such as static variable and constant variable optimization, redundant assignment, check and variable issues. the cppcheck result file link: https://pan.baidu.com/s/1EisYqq5Iey9Q0M9rT5c_Gw?pwd=vsev -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
