[downstream patch FYI] workaround stale frame_info * (PR 13866)

[Tom Tromey]

Pedro> I've then ran the whole testsuite with this on, and that didn't catch
Pedro> any other problem.  You mention in PR13866 many situations with stale
Pedro> frame_info; did you have some other way to catch those,

Jan> No.  This single PR13866 crash and GDB code in general gives me enough
Jan> reasoning to believe:

Jan> (a) There exist other such crashes in GDB, just not tested by the
Jan> testsuite.
Jan> (b) More of such crashes will regress by changes in the future.

Jan> This is just my personal opinion you may not agree with.

Despite the difficulties perhaps we should try to write a static
analyzer for this.  I am not sure if it could be made reliable enough,
but maybe it could.  The problem cases are when the analyzer must
account for cleanups -- but in this case we might be able to get away
with ignoring them.

Another idea is to simply get rid of frame_info and have only a
frame_id-based API.

Tom