Historical lunar landing crash about raceintospace

[mkmccarty]

Variable sizes are defined as MA[2][4] and MANNED[2]. This puts MANNED[0] == 9, which is just a count populated from Data->P[plr].Mission[mis].Men;

[mkmccarty]

Maybe some bounds/sanity checking above this to trap this sort of thing.

[willglynn]

Maybe we should have a game_state_sanity_check() function which runs assert()s on as many things as we can imagine. Then we pepper calls to this all over: each return to the port screen, start of turn, saving, loading, etc.

[mkmccarty]

A perfect place for:

  // Verify that the version of the library that we linked against is
  // compatible with the version of the headers we compiled against.
  GOOGLE_PROTOBUF_VERIFY_VERSION;

[rnyoakum]

Reviewing this, I'm pretty sure the problem was this issue.

This issue has two parts:

1. The sudden change in mission step execution, and
2. The segfault caused by the bad Manned[0] value

The second part is exactly the problem caused by #200. As for the weird mission ordering, I suspect the pAry overflow is also the issue - after MANNED[] it would eventually overflow into STEP. Also possible is that modifying MANNED[] created its own out-of-range issues that compounded the chaos - such as an MA[][] out-of-range access into Mev.

Until someone else can reproduce the problem, I'm closing this.

[peyre]

Ryan, thank you! I love that you've managed to close one of our old Issues. I'd sure like to clear some of those out.