Re: Patch: make --enable-mapped-location work

[Tom Tromey <tromey at redhat dot com>]

>>>>> "Diego" == Diego Novillo <dnovillo@google.com> writes:

Diego> Once Ada and Fortran are fixed, I would prefer that we remove
Diego> the switch completely, to avoid all the ifdef pollution.

Mark suggested enabling it by default for 4.3 and then removing it in
4.4.  That seemed reasonable, in a cautious sort of way, to me.  I'm
happy to implement whatever folks agree on, though; I don't have a
strong feeling about this.

>> +const cpp_token *
>> +cpp_get_token_with_location (cpp_reader *pfile, source_location *loc)

Diego> Comments.

Thanks.  Sorry for forgetting those.

>> int *
>> expr_lineno (const_tree node)

Diego> Likewise.

Tricky of you, since the original didn't have comments here :-)

>> +#define EXPR_FILENAME(NODE) (expr_filename ((NODE)))
>> +#define EXPR_LINENO(NODE) (expr_lineno ((NODE)))
>> +#else
>> +/* Note that these can only be used as lvalues when
>> +   #!USE_MAPPED_LOCATION.  */

Diego> Perhaps it's not terribly important, but what about making both
Diego> versions non-lvalues?  For consistency, mostly.

Good idea.  I did this and it does clean up the code a bit.

Diego> The rest looks OK.  Thanks for fixing this.

I'm testing a new patch.  I'll send it when that is done.

Tom