Re: [PATCH v5 09/12] arm64: mm: replace TIF_LAZY_MMU with in_lazy_mmu_mode()

["David Hildenbrand (Red Hat)" <david@xxxxxxxxxx>]

On 12/4/25 07:52, Anshuman Khandual wrote:
> On 24/11/25 6:52 PM, Kevin Brodsky wrote:
> > The generic lazy_mmu layer now tracks whether a task is in lazy MMU
> > mode. As a result we no longer need a TIF flag for that purpose -
> > let's use the new in_lazy_mmu_mode() helper instead.
> >
> > The explicit check for in_interrupt() is no longer necessary either
> > as in_lazy_mmu_mode() always returns false in interrupt context.
> >
> > Signed-off-by: Kevin Brodsky <kevin.brodsky@xxxxxxx>
> > ---
> >   arch/arm64/include/asm/pgtable.h     | 19 +++----------------
> >   arch/arm64/include/asm/thread_info.h |  3 +--
> >   2 files changed, 4 insertions(+), 18 deletions(-)
> >
> > diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
> > index a7d99dee3dc4..dd7ed653a20d 100644
> > --- a/arch/arm64/include/asm/pgtable.h
> > +++ b/arch/arm64/include/asm/pgtable.h
> > @@ -62,28 +62,16 @@ static inline void emit_pte_barriers(void)
> >   
> >   static inline void queue_pte_barriers(void)
> >   {
> > -       unsigned long flags;
> > -
> > -       if (in_interrupt()) {
> > -               emit_pte_barriers();
> > -               return;
> > -       }
> > -
> > -       flags = read_thread_flags();
> > -
> > -       if (flags & BIT(TIF_LAZY_MMU)) {
> > +       if (in_lazy_mmu_mode()) {
> >                 /* Avoid the atomic op if already set. */
> > -               if (!(flags & BIT(TIF_LAZY_MMU_PENDING)))
> > +               if (!test_thread_flag(TIF_LAZY_MMU_PENDING))
> A small nit - will it be better not to use test_thread_flag() here and just
> keep checking flags like earlier to avoid non-related changes. Although not
> a problem TBH.
I'd assume the existing code wanted to avoid fetching the flags two 
times? So switching to test_thread_flag() should be fine now.
--
Cheers

David