[PATCH 2/4] Don't erase empty indices in DWARF reader

[Tom Tromey]

Joel> If we believe the code is dead, let's drop it.

Joel> Maybe Andrew's assert suggestion could be helpful, although perhaps
Joel> a complaint might be a better compromise in this case, as I would say
Joel> a violation of this expectation would still mean a usable debugging
Joel> experience for the user, so let's not break things with an assert?

I think a complaint would be incorrect here.  Normally complaints are
reserved for bad debug information, but this is just an internal aspect
of the code -- the contract of parallel-for.

I think it may be best to just drop this patch and leave the current
code in place.  It doesn't do anything, but on the other hand, this
seems to be non-obvious, so it may serve a useful assurance purpose.

Tom