This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: [PATCH] Add new model m32r2 of Renesas M32R.
Hi Kazuhiro,
> I send a miss patch about -error-explicit-parallel-conflicts.
> I made the -error-explicit-parallel-conflicts option to generate an
> illegal parallel code for chip test only. It's not for general
> users.
In which case shouldn't the switch be removed entirely ?
I am not going to apply your patch yet, because I have some questions
about it:
> case OPTION_WARN_PARALLEL:
> warn_explicit_parallel_conflicts = 1;
> - error_explicit_parallel_conflicts = 0;
> break;
>
> case OPTION_NO_WARN_PARALLEL:
> warn_explicit_parallel_conflicts = 0;
> - error_explicit_parallel_conflicts = 0;
> break;
>
> case OPTION_ERROR_PARALLEL:
> - warn_explicit_parallel_conflicts = 1;
> error_explicit_parallel_conflicts = 1;
> break;
>
> case OPTION_NO_ERROR_PARALLEL:
> error_explicit_parallel_conflicts = 0;
> - warn_explicit_parallel_conflicts = 0;
> break;
Why have you removed these lines ? Surely the --warn... and
--error... options are incompatible ? You cannot both warn about a
parallel conflict and generate an error for it.
> + if ( check_outputs && !error_explicit_parallel_conflicts )
> + return 0;
What is the purpose of this ? It appears to mean that if
error_explicit_parallel_conflicts is not true, then the code will not
check for parallel conflicts, even if warn_explicit_parallel_conflicts
it true.
> if (parallel_p && warn_explicit_parallel_conflicts)
> {
> if (first_writes_to_seconds_operands (&first, &second, FALSE))
> /* xgettext:c-format */
> ! as_warn (_("%s: output of 1st instruction is the same as an input to 2nd instruction - is this intentional ?"), str2);
>
> if (first_writes_to_seconds_operands (&second, &first, FALSE))
> /* xgettext:c-format */
> ! as_warn (_("%s: output of 2nd instruction is the same as an input to 1st instruction - is this intentional ?"), str2);
> }
Where are the error messages ? If error_explicit_parallel_conflicts
is true, then shouldn't an error message be generated rather than a
warning ?
Cheers
Nick