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

[Andrew Burgess <aburgess@redhat.com>]

Tom Tromey <tromey@adacore.com> writes:

>>>>>> "Tom" == Tom Tromey <tromey@adacore.com> writes:
>
> Tom> I think it may be best to just drop this patch and leave the current
> Tom> code in place.  It doesn't do anything, but on the other hand, this
> Tom> seems to be non-obvious, so it may serve a useful assurance purpose.
>
> It occurred to me today that another approach might be to add a
> self-test for this -- that is, use parallel-for-each on (say) a single
> task but when multiple threads are available.  Then, it could check that
> there are no NULL responses.

Hi Tom,

Sorry I missed your original reply, and thanks for the extra info.  If
the code is dead then I don't see any point keeping it around.

If you don't believe there's any point adding an assert in its place
then I'm fine with that.

I'm always pleased to see new selftests being added, so if you feel
that's worth doing then +1 from me.

But I don't think you should drop this patch if you are happy with it,
no point in holding on to dead code.

Thanks,
Andrew