RegisterFormatErrorMessage

Description

Registers the predefined error code information with CANN in the specified JSON format. After registration, call the ReportPredefinedErrMsg API to report error codes.

Prototype

int32_t RegisterFormatErrorMessage(const char *error_msg, size_t error_msg_len)

To facilitate usage, the macro REG_FORMAT_ERROR_MSG is encapsulated. You can directly use this macro to register error code information. The macro directly defines static variable registration, enabling the registration to complete upon process loading.

#define REG_FORMAT_ERROR_MSG(error_msg, error_msg_len) \
  REG_FORMAT_ERROR_MSG_UNIQ_HELPER((error_msg), (error_msg_len), __COUNTER__)
#define REG_FORMAT_ERROR_MSG_UNIQ_HELPER(error_msg, error_msg_len, counter) \
  REG_FORMAT_ERROR_MSG_UNIQ((error_msg), (error_msg_len), counter)
#define REG_FORMAT_ERROR_MSG_UNIQ(error_msg, error_msg_len, counter) \
  static const auto register_error_msg_##counter ATTRIBUTE_USED = []() -> int32_t { \
    return error_message::RegisterFormatErrorMessage((error_msg), (error_msg_len)); \
  }()

Parameters

Parameter	Input/Output	Description
error_msg	Input	Error code information. Multiple error codes can be registered at a time. The error code information must be organized in JSON format. For details, see Example.
error_msg_len	Input	Length of error_msg, excluding '\0' at the end.

Parameter

Input/Output

Description

error_msg

Input

Error code information. Multiple error codes can be registered at a time.

The error code information must be organized in JSON format. For details, see Example.

error_msg_len

Input

Length of error_msg, excluding '\0' at the end.

Returns

0: success
-1: failure

Example

error_msg must be organized in JSON format. error_info_list is an array that contains error information objects and must contain at least one element. The fields are described as follows:

errClass: error classification.
errTitle: error title.
ErrCode: error code. The error code must be different from existing ones. For details about existing error codes, see Error Codes in Troubleshooting.
ErrMessage: error message, which can contain the formatted placeholders (%s).
Arglist: parameter list, which describes the parameters corresponding to the placeholders in ErrMessage. The length of the parameter list must be the same as the number of formatted placeholders in ErrMessage.
suggestion: suggestion information, including:
- Possible Cause: possible cause.
- Solution: solution.

#include <string>
#include "base/err_msg.h"

const std::string error_msg = R"(
{
    "error_info_list": 
    [
        {
            "errClass": "GE Errors",
            "errTitle": "Invalid_Dynamic_Shape_Argument",
            "ErrCode": "E10018",
            "ErrMessage": "Value [%s] for shape [%s] is invalid. When [--dynamic_batch_size] is included, only batch size N can be -1 in [--input_shape].",
            "Arglist": "shape,index",
            "suggestion": {
                "Possible Cause": "When [--dynamic_batch_size] is included, only batch size N can be -1 in the shape.",
                "Solution": "Try again with a valid [--input_shape] argument. Make sure that non-batch size axes are not -1."
            }
        },		
        {
           "errClass": "GE Errors",
           "errTitle": "Invalid_--input_shape_Argument",
           "ErrCode": "E10019",
           "ErrMessage": "When [--dynamic_image_size] is included, only the height and width axes can be -1 in [--input_shape].",
           "Arglist": "",
           "suggestion": {
                "Possible Cause": "When [--dynamic_image_size] is included, only the height and width axes can be -1 in the shape.",
                "Solution": "Try again with a valid [--input_shape] argument. Make sure that axes other than height and width are not -1."
            }
        }
    ]
}
)";

REG_FORMAT_ERROR_MSG(error_msg.c_str(), error_msg.size());

Parent topic: Error Reporting APIs