On Thu, Jul 25, 2024 at 3:00 PM Barry Song <21cnbao@xxxxxxxxx> wrote:
>
> On Thu, Jul 25, 2024 at 6:08 PM Michal Hocko <mhocko@xxxxxxxx> wrote:
> >
> > On Thu 25-07-24 10:50:45, Barry Song wrote:
> > > On Thu, Jul 25, 2024 at 12:27 AM Michal Hocko <mhocko@xxxxxxxx> wrote:
> > > >
> > > > On Wed 24-07-24 20:55:40, Barry Song wrote:
> > [...]
> > > > > diff --git a/drivers/vdpa/vdpa_user/iova_domain.c b/drivers/vdpa/vdpa_user/iova_domain.c
> > > > > index 791d38d6284c..eff700e5f7a2 100644
> > > > > --- a/drivers/vdpa/vdpa_user/iova_domain.c
> > > > > +++ b/drivers/vdpa/vdpa_user/iova_domain.c
> > > > > @@ -287,28 +287,44 @@ void vduse_domain_remove_user_bounce_pages(struct vduse_iova_domain *domain)
> > > > > {
> > > > > struct vduse_bounce_map *map;
> > > > > unsigned long i, count;
> > > > > + struct page **pages = NULL;
> > > > >
> > > > > write_lock(&domain->bounce_lock);
> > > > > if (!domain->user_bounce_pages)
> > > > > goto out;
> > > > > -
> > > > > count = domain->bounce_size >> PAGE_SHIFT;
> > > > > + write_unlock(&domain->bounce_lock);
> > > > > +
> > > > > + pages = kmalloc_array(count, sizeof(*pages), GFP_KERNEL | __GFP_NOFAIL);
> > > > > + for (i = 0; i < count; i++)
> > > > > + pages[i] = alloc_page(GFP_KERNEL | __GFP_NOFAIL);
> > > >
> > > > AFAICS vduse_domain_release calls this function with
> > > > spin_lock(&domain->iotlb_lock) so dropping &domain->bounce_lock is not
> > > > sufficient.
> > >
> > > yes. this is true:
> > >
> > > static int vduse_domain_release(struct inode *inode, struct file *file)
> > > {
> > > struct vduse_iova_domain *domain = file->private_data;
> > >
> > > spin_lock(&domain->iotlb_lock);
> > > vduse_iotlb_del_range(domain, 0, ULLONG_MAX);
> > > vduse_domain_remove_user_bounce_pages(domain);
> > > vduse_domain_free_kernel_bounce_pages(domain);
> > > spin_unlock(&domain->iotlb_lock);
> > > put_iova_domain(&domain->stream_iovad);
> > > put_iova_domain(&domain->consistent_iovad);
> > > vhost_iotlb_free(domain->iotlb);
> > > vfree(domain->bounce_maps);
> > > kfree(domain);
> > >
> > > return 0;
> > > }
> > >
> > > This is quite a pain. I admit I don't have knowledge of this driver, and I don't
> > > think it's safe to release two locks and then reacquire them. The situation is
> > > rather complex. Therefore, I would prefer if the VDPA maintainers could
> > > take the lead in implementing a proper fix.
> >
> > Would it be possible to move all that work to a deferred context?
>
> My understanding is that we need to be aware of both the iotlb_lock and
> bounce_lock to implement the correct changes. As long as we still need
> to acquire these two locks in a deferred context, there doesn't seem to
> be any difference.
>
> I can do the memory pre-allocation before spin_lock(&domain->iotlb_lock),
> but I have no knowledge whether the "count" will change after I make
> the preallocation.
>
> diff --git a/drivers/vdpa/vdpa_user/iova_domain.c
> b/drivers/vdpa/vdpa_user/iova_domain.c
> index 791d38d6284c..7ec87ef33d42 100644
> --- a/drivers/vdpa/vdpa_user/iova_domain.c
> +++ b/drivers/vdpa/vdpa_user/iova_domain.c
> @@ -544,9 +544,12 @@ static int vduse_domain_release(struct inode
> *inode, struct file *file)
> {
> struct vduse_iova_domain *domain = file->private_data;
>
> + struct page **pages;
> + spin_lock(&domain->iotlb_lock); maybe also + bounce_lock?
> + count = domain->bounce_size >> PAGE_SHIFT;
> + spin_unlock(&domain->iotlb_lock);
We probably don't need any lock here as bounce_size won't be changed .
> +
> + preallocate_count_pages(pages, count);
> +
> ....
> spin_lock(&domain->iotlb_lock);
> vduse_iotlb_del_range(domain, 0, ULLONG_MAX);
> - vduse_domain_remove_user_bounce_pages(domain);
> + vduse_domain_remove_user_bounce_pages(domain, pages);
> vduse_domain_free_kernel_bounce_pages(domain);
> spin_unlock(&domain->iotlb_lock);
> put_iova_domain(&domain->stream_iovad);
This seems to work.
Thanks
>
>
> > --
> > Michal Hocko
> > SUSE Labs
>
