Re: [PATCH 1/2] rust: Add initial cpumask abstractions

[Viresh Kumar]

On 25-02-25, 10:55, Alice Ryhl wrote:
> On Tue, Feb 25, 2025 at 10:47 AM Viresh Kumar <viresh.kumar@xxxxxxxxxx> wrote:
> > +impl Drop for Cpumask {
> > +    fn drop(&mut self) {
> > +        if self.owned {
> > +            // SAFETY: `ptr` is guaranteed to be valid for the lifetime of `self`. And it is safe
> > +            // to call `free_cpumask_var()`.
> > +            unsafe { bindings::free_cpumask_var(self.ptr) }
> 
> This is missing a semicolon, but it's not the last statement in the
> block. Did you compile this with CPUMASK_OFFSTACK=n? I don't think it
> compiles with that setting.

I would always add a semicolon here, yeah I missed adding that but ..

I have missed minor things before sending a series a few times in the
past and this one really scared me thinking here I did it again :)

Though I was sure that I have built the code with both
CPUMASK_OFFSTACK=y and =n, I performed the builds again and it worked
(again).  That confused me even more :)

And here is what I think is happening here (which makes it build fine
accidentally):
- free_cpumask_var() has a return type of void.
- The block {} allows it to build fine without a semicolon here.
- I performed a simple test for this [1] and it works too.

> > +            #[cfg(not(CONFIG_CPUMASK_OFFSTACK))]
> > +            // SAFETY: The pointer was earlier initialized from the result of `KBox::into_raw()`.
> > +            unsafe {
> > +                drop(KBox::from_raw(self.ptr))
> > +            };
> 
> This looks like you did not run rustfmt.

I did this:

make CLIPPY=1 rustfmtcheck ARCH=arm64 O=../barm64t/ -j8 CROSS_COMPILE=aarch64-linux-gnu- CONFIG_DEBUG_SECTION_MISMATCH=y

I hope that is all I need ? I checked again with both CONFIG options,
doesn't complain with rustc 1.84.1.

-- 
viresh

[1] https://play.rust-lang.org/?version=stable&mode=debug&edition=2024&gist=70cd7d31633d98774a088fed68ebb00d