Silence Clang’s -Wunreachable-code warnings

[johnsonjh]

Double-check these — the changes here only silences the warnings.

Silence Clang’s -Wunreachable-code warnings:

src/loader.c:61:17: warning: code will never be executed [-Wunreachable-code]
   61 |             c = '+';

src/loader.c:149:30: warning: code will never be executed [-Wunreachable-code]
  149 |                     offset = fcb2 + 1;

src/loader.c:156:30: warning: code will never be executed [-Wunreachable-code]
  156 |                     offset = fcb2 + 1;

src/loader.c:195:30: warning: code will never be executed [-Wunreachable-code]
  195 |                     offset = fcb2 + 9;

src/loader.c:214:30: warning: code will never be executed [-Wunreachable-code]
  214 |                     offset = fcb2 + 1;

src/loader.c:221:30: warning: code will never be executed [-Wunreachable-code]
  221 |                     offset = fcb2 + 1;

[dmsc]

Don't really like this - as the code is left unreachable on purpose, depending on the value of the macro. The (1) seems a hack that will break on other compilers.

The question should be - it is useful for the emulator to parse the FCB in DOS 1, 2 or 3 modes? If it is a useful feature, perhaps we should add a command line option to select the mode.

Thanks!

[johnsonjh]

Doing (1)+(1) and similar, or comparing against ((0)) instead of just 0, does work on all compilers, back to K&R. The other option would be to use #if in the preprocessor, but that is probably even uglier looking. Whatever you want to do is fine of course. This warning isn’t part of Clang’s -Wall at this time (but likely will be in the future). Edit: If you plan on making this a runtime switch, probably best to just close this out, but merging as-is does clear the warnings so you can use that flag to find new issues in the future.

[johnsonjh]

@dmsc should I close this one out?

[dmsc]

@dmsc should I close this one out?

I will keep it open until I do a proper fix :)