Re: [RFC PATCH 01/10] thread_info: add helpers to snapshot thread flags

[Marco Elver]

On Wed, 9 Jun 2021 at 14:20, Mark Rutland <mark.rutland@xxxxxxx> wrote:
>
> We have common helpers to manipulate individual thread flags, but where
> code wants to check several flags at once, it must open code reading
> current_thread_info()->flags and operating on a snapshot.
>
> As some flags can be set remotely it's necessary to use READ_ONCE() to
> get a consistent snapshot even when IRQs are disabled, but some code
> forgets to do this. Generally this is unlike to cause a problem in
> practice, but it is somewhat unsound, and KCSAN will legitimately warn
> that there is a data race.
>
> To make it easier to do the right thing, and to highlight that
> concurrent modification is possible, let's add a new helpers to snapshot
> the flags, which should be used in preference to plain reads.
> Subsequent patches will move existing code to use the new helpers.
>
> Signed-off-by: Mark Rutland <mark.rutland@xxxxxxx>
> Cc: Boqun Feng <boqun.feng@xxxxxxxxx>
> Cc: Dmitry Vyukov <dvyukov@xxxxxxxxxx>
> Cc: Marco Elver <elver@xxxxxxxxxx>
> Cc: Paul E. McKenney <paulmck@xxxxxxxxxx>
> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Cc: Will Deacon <will@xxxxxxxxxx>

Acked-by: Marco Elver <elver@xxxxxxxxxx>


> ---
>  include/linux/thread_info.h | 10 ++++++++++
>  1 file changed, 10 insertions(+)
>
> diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h
> index 157762db9d4b..f3769842046d 100644
> --- a/include/linux/thread_info.h
> +++ b/include/linux/thread_info.h
> @@ -117,6 +117,11 @@ static inline int test_ti_thread_flag(struct thread_info *ti, int flag)
>         return test_bit(flag, (unsigned long *)&ti->flags);
>  }
>
> +static inline unsigned long read_ti_thread_flags(struct thread_info *ti)
> +{
> +       return READ_ONCE(ti->flags);
> +}
> +

Are some of the callers 'noinstr'? I haven't seen it in this series
yet, but if yes, then not inlining (which some compilers may do with
heavier instrumentation) might cause issues and this could be
__always_inline.

Thanks,
-- Marco