Re: [PATCH 0/6] unused function parameter potpourri

From: Elijah Newren <newren@gmail.com>
Date: Aug 19, 2022, 23:07 UTC
Message-ID: <CABPp-BFoFBE4o1-M258-DHQSbW8WS2-EQCipCdmq2ECi9BL7pg@mail.gmail.com>
In-Reply-To: <892b2a45-1b6b-aa87-4902-471793135183@github.com>

On Fri, Aug 19, 2022 at 12:35 PM Derrick Stolee <derrickstolee@github.com> wrote:

Show 21 quoted lines

>
> On 8/19/2022 4:48 AM, Jeff King wrote:
> > Here are a few small cleanups of unused function parameters. The first
> > five just drop the unused parameters. These are all trivially correct,
> > since otherwise the compiler would complain. But I tried to make sure
> > that dropping was the right thing in each (rather than it being a bug
> > where the parameter should have been used).
> >
> > The final one just uses the parameters for an assertion, following a
> > pattern we've used before.
> >
> > I'll try to cc the individual authors for each patch.
> >
> >   [1/6]: xdiff: drop unused mmfile parameters from xdl_do_histogram_diff()
> >   [2/6]: log-tree: drop unused commit param in remerge_diff()
> >   [3/6]: match_pathname(): drop unused "flags" parameter
> >   [4/6]: verify_one_sparse(): drop unused repository parameter
> >   [5/6]: reftable: drop unused parameter from reader_seek_linear()
> >   [6/6]: reflog: assert PARSE_OPT_NONEG in parse-options callbacks
>
> Thanks for doing this cleanup. It all looks correct to me.

Like Phillip, I seem to also have a blind spot for unused parameters -- this isn't the first one I created either.

Anyway, this series looks good to me as well.

> Patch 5 mentioned some choice as to modifying the parameter list
> versus using the UNUSED() macro. I think renaming the method can
> solve the uncertainty there, but it's also not necessary for the
> change to be correct.