On Mon, 20 Dec 2021, Mark Hemment wrote:
> On Thu, 16 Dec 2021 at 23:57, NeilBrown <neilb@xxxxxxx> wrote:
> >
> > Currently various places test if direct IO is possible on a file by
> > checking for the existence of the direct_IO address space operation.
> > This is a poor choice, as the direct_IO operation may not be used - it is
> > only used if the generic_file_*_iter functions are called for direct IO
> > and some filesystems - particularly NFS - don't do this.
> >
> > Instead, introduce a new mapping flag: AS_CAN_DIO and change the various
> > places to check this (avoiding a pointer dereference).
> > unlock_new_inode() will set this flag if ->direct_IO is present, so
> > filesystems do not need to be changed.
> ...
> > diff --git a/fs/inode.c b/fs/inode.c
> > index 6b80a51129d5..bae65ccecdb1 100644
> > --- a/fs/inode.c
> > +++ b/fs/inode.c
> > @@ -1008,6 +1008,9 @@ EXPORT_SYMBOL(lockdep_annotate_inode_mutex_key);
> > void unlock_new_inode(struct inode *inode)
> > {
> > lockdep_annotate_inode_mutex_key(inode);
> > + if (inode->i_mapping->a_ops &&
> > + inode->i_mapping->a_ops->direct_IO)
> > + set_bit(AS_CAN_DIO, &inode->i_mapping->flags);
> > spin_lock(&inode->i_lock);
> > WARN_ON(!(inode->i_state & I_NEW));
> > inode->i_state &= ~I_NEW & ~I_CREATING;
>
> Does d_instantiate_new() also need to set AS_CAN_DIO?
Yes it does - thanks for catching that. I'll update my patch.
Thanks,
NeilBrown
