Re: [PATCH v1 3/6] perf build: Add mypy build tests

[Namhyung Kim <namhyung@kernel.org>]

On Fri, Jan 24, 2025 at 06:10:09PM -0800, Ian Rogers wrote:
> On Fri, Jan 24, 2025 at 3:51 PM Namhyung Kim <namhyung@kernel.org> wrote:
> >
> > On Fri, Oct 25, 2024 at 10:23:00AM -0700, Ian Rogers wrote:
> > > If MYPY=1 is passed to the build then run mypy over python code in
> > > perf. Unlike shellcheck this isn't default on as there are currently
> > > too many errors.
> >
> > Can you please add an example output in the commit log?  Also you need
> > to add a comment to describe the build option.
> 
> It feels more of a developer option, so I'm not sure we should
> advertise it. Ideally we'd just default it like the shellcheck case.

Yep, but I think we need some (brief) explanation even for a default
developer option.

> I'm not sure of the benefit of snapshotting mypy errors, a quick
> internet search will show what mypy does and I'd prefer the patch set
> focus on the build infrastructure changes.

You are introducing a new thing so you need to explain what it is, how
to use it and how the output look like.

> 
> > Is it ok to pass the option on a system that doesn't have mypy?
> 
> I'd expect the shell/which assignment to MYPY would leave it
> undefined, in which case the behavior would match not building with
> MYPY=1. It's basically the same logic as with shellchecks. I didn't
> test/optimize for people requesting a build option with dependencies
> missing.

Ok, I'm good as long as it handles the case.

Thanks,
Namhyung