Warning that is possibly wrong?

[Maury Markowitz]

First post here, please be gentle!

I have some C code performing a BASIC INPUT statement while reading from a file. The opening bit of this code is...

    char buffer[MAX_INPUT_LENGTH];
    int type;

    // get the stream
    int channel = floor(evaluate_expression(statement->parms.generic.generic_parameter).number);
    FILE* fp = handle_for_channel(channel);
    if (fp == NULL) {
      handle_error(ern_FILE_NOT_OPEN, "Attempt to INPUT from a file that has not been OPENed");
      // and make sure we can read it
      if (!channel_is_readable(channel)) {
        handle_error(ern_FILE_NOT_INPUT, "Attempt to INPUT from a file that is read-only");
      }
    }

    // read one line from the file, if it's empty, warn and return
    if (fgets(buffer, sizeof buffer, fp) == NULL) {
      handle_error(ern_OUT_OF_DATA, "Reached the end-of-file while performing INPUT");
      return;
    }
    long len = strlen(buffer);

The question has to do with the fgets line, which in cppcheck is returning:

Either the condition 'fp==NULL' is redundant or there is possible null pointer dereference: fp

But, to my not very expert eye, I am not comparing fp to null. Am I misinterpreting what is going on here, or is this a spurious warning?

Complete code here, it is small and should easily make on practically anything:

https://github.com/maurymarkowitz/RetroBASIC

[CHR]

Does handle_error() return or abort? If it returns, fp can be used even if it is NULL.

[Maury Markowitz]

handle_error MAY abort, but will not if a TRAP is set, which it often is during I/O.

But the warning is not related to that I don't think?

[Maury Markowitz]

Ohhh, wait now I see what you mean. It's not complaining about fp==NULL on this line, it's saying the one above is the issue. Now I get it!