mmueller / clog Goto Github PK

In CGDB, we don't use O_APPEND and we do use O_TRUNC.

Can this be configurable?

In our use of clog we have this in clog_free,
if (_clog_loggers[id]) {
if (_clog_loggers[id]->opened) {
close( _clog_loggers[ id ]->fd );
}
free( _clog_loggers[ id ] );

•    _clog_loggers[ id ] = 0;
  

Does this make sense?

where use the function clog_free to free logger,you forgot to set _clog_loggers[id] to NULL,when you second to init a logger which id equal to before,the function clog_init_fd judge the logger is not NULL and return 1

in function _clog_log()

    /* Format the message text with the argument list. */
    result = vsnprintf(dynbuf, buf_size, fmt, ap);
    if ((size_t) result >= buf_size) {
        buf_size = result + 1;
        dynbuf = (char *) malloc(buf_size);
        result = vsnprintf(dynbuf, buf_size, fmt, ap);
        if ((size_t) result >= buf_size) {
            /* Formatting failed -- too large */
            _clog_err("Formatting failed (1).\n");
            free(dynbuf);
            return;
        }
    }

according to vsnprintf manpage:

The functions vprintf(), vfprintf(), vsprintf(), vsnprintf() are equivalent to the functions printf(), fprintf(), sprintf(), snprintf(), respectively, except that they are called with a va_list instead of a variable number of arguments. These functions do not call the va_end macro. Because they invoke the va_arg macro, the value of ap is undefined after the call. See stdarg(3).

So, the second call to vsnprintf

result = vsnprintf(dynbuf, buf_size, fmt, ap);

may cause segment fault.