Re: [PATCH v4 09/10] mm/vma: update vma_assert_locked() to use lockdep

[Lorenzo Stoakes <lorenzo.stoakes@oracle.com>]

On Mon, Jan 26, 2026 at 02:42:00PM +0100, Vlastimil Babka wrote:
> On 1/23/26 21:12, Lorenzo Stoakes wrote:
> > We can use lockdep to avoid unnecessary work here, otherwise update the
> > code to logically evaluate all pertinent cases and share code with
> > vma_assert_write_locked().
> >
> > Make it clear here that we treat the VMA being detached at this point as a
> > bug, this was only implicit before.
> >
> > Reviewed-by: Suren Baghdasaryan <surenb@google.com>
> > Signed-off-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
>
> Reviewed-by: Vlastimil Babka <vbabka@suse.cz>

Thanks!

>
> Nit:
>
> > ---
> >  include/linux/mmap_lock.h | 41 +++++++++++++++++++++++++++++++++++++--
> >  1 file changed, 39 insertions(+), 2 deletions(-)
> >
> > diff --git a/include/linux/mmap_lock.h b/include/linux/mmap_lock.h
> > index 23bde4bd5a85..4a0aafc66c5d 100644
> > --- a/include/linux/mmap_lock.h
> > +++ b/include/linux/mmap_lock.h
> > @@ -322,19 +322,56 @@ int vma_start_write_killable(struct vm_area_struct *vma)
> >  	return __vma_start_write(vma, __vma_raw_mm_seqnum(vma), TASK_KILLABLE);
> >  }
> >
> > +/**
> > + * vma_assert_write_locked() - assert that @vma holds a VMA write lock.
> > + * @vma: The VMA to assert.
> > + */
> >  static inline void vma_assert_write_locked(struct vm_area_struct *vma)
> >  {
> >  	VM_WARN_ON_ONCE_VMA(!__is_vma_write_locked(vma), vma);
> >  }
> >
> > +/**
> > + * vma_assert_locked() - assert that @vma holds either a VMA read or a VMA write
> > + * lock and is not detached.
> > + * @vma: The VMA to assert.
> > + */
> >  static inline void vma_assert_locked(struct vm_area_struct *vma)
> >  {
> > +	unsigned int refcnt;
> > +
> > +	/*
> > +	 * If read-locked or currently excluding readers, then the VMA is
> > +	 * locked.
> > +	 */
> > +#ifdef CONFIG_LOCKDEP
> > +	if (lock_is_held(&vma->vmlock_dep_map))
> > +		return;
>
> Wouldn't this work a tiny bit better?
>
> 	if (!lock_is_held(&vma->vmlock_dep_map))
> 		vma_assert_write_locked(vma);
> 	return;

Hm yeah could do, I guess we don't need to mix the 'uncertain' stuff below with
the lockdep-certainty, at this point we _know_ there is no read lock/readers
being excluded exclusive lock so it can only be a write lock.

Will test locally to make sure sane then send a fix-patch. :)

>
> > +#endif
> > +
> >  	/*
> >  	 * See the comment describing the vm_area_struct->vm_refcnt field for
> >  	 * details of possible refcnt values.
> >  	 */
> > -	VM_WARN_ON_ONCE_VMA(refcount_read(&vma->vm_refcnt) <= 1 &&
> > -			    !__is_vma_write_locked(vma), vma);
> > +	refcnt = refcount_read(&vma->vm_refcnt);
> > +
> > +	/*
> > +	 * In this case we're either read-locked, write-locked with temporary
> > +	 * readers, or in the midst of excluding readers, all of which means
> > +	 * we're locked.
> > +	 */
> > +	if (refcnt > 1)
> > +		return;
> > +
> > +	/* It is a bug for the VMA to be detached here. */
> > +	VM_WARN_ON_ONCE_VMA(!refcnt, vma);
> > +
> > +	/*
> > +	 * OK, the VMA has a reference count of 1 which means it is either
> > +	 * unlocked and attached or write-locked, so assert that it is
> > +	 * write-locked.
> > +	 */
> > +	vma_assert_write_locked(vma);
> >  }
> >
> >  static inline bool vma_is_attached(struct vm_area_struct *vma)
>

Cheers, Lorenzo