| Message ID | 20230214171330.2722188-6-dhowells@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | iov_iter: Improve page extraction (pin or just list) | expand |
On Tue, 14 Feb 2023 at 18:14, David Howells <dhowells@redhat.com> wrote: > > Implement splice-read for overlayfs by passing the request down a layer > rather than going through generic_file_splice_read() which is going to be > changed to assume that ->read_folio() is present on buffered files. > > Signed-off-by: David Howells <dhowells@redhat.com> > cc: Christoph Hellwig <hch@lst.de> > cc: Jens Axboe <axboe@kernel.dk> > cc: Al Viro <viro@zeniv.linux.org.uk> > cc: John Hubbard <jhubbard@nvidia.com> > cc: David Hildenbrand <david@redhat.com> > cc: Matthew Wilcox <willy@infradead.org> > cc: Miklos Szeredi <miklos@szeredi.hu> > cc: linux-unionfs@vger.kernel.org > cc: linux-block@vger.kernel.org > cc: linux-fsdevel@vger.kernel.org > cc: linux-mm@kvack.org > --- > fs/overlayfs/file.c | 36 +++++++++++++++++++++++++++++++++++- > 1 file changed, 35 insertions(+), 1 deletion(-) > > diff --git a/fs/overlayfs/file.c b/fs/overlayfs/file.c > index c9d0c362c7ef..267b61df6fcd 100644 > --- a/fs/overlayfs/file.c > +++ b/fs/overlayfs/file.c > @@ -419,6 +419,40 @@ static ssize_t ovl_write_iter(struct kiocb *iocb, struct iov_iter *iter) > return ret; > } > > +static ssize_t ovl_splice_read(struct file *in, loff_t *ppos, > + struct pipe_inode_info *pipe, size_t len, > + unsigned int flags) > +{ > + const struct cred *old_cred; > + struct fd real; > + ssize_t ret; > + > + ret = ovl_real_fdget(in, &real); > + if (ret) > + return ret; > + > + ret = -EINVAL; > + if (in->f_flags & O_DIRECT && > + !(real.file->f_mode & FMODE_CAN_ODIRECT)) > + goto out_fdput; This is unnecessary, as it was already done in ovl_real_fdget() -> ovl_real_fdget_meta() -> ovl_change_flags(). > + if (!real.file->f_op->splice_read) > + goto out_fdput; > + > + ret = rw_verify_area(READ, in, ppos, len); Should be on real.file. > + if (unlikely(ret < 0)) > + return ret; Leaks fd. > + > + old_cred = ovl_override_creds(file_inode(in)->i_sb); > + ret = real.file->f_op->splice_read(real.file, ppos, pipe, len, flags); > + > + revert_creds(old_cred); > + ovl_file_accessed(in); > +out_fdput: > + fdput(real); > + > + return ret; > +} > + > /* > * Calling iter_file_splice_write() directly from overlay's f_op may deadlock > * due to lock order inversion between pipe->mutex in iter_file_splice_write() > @@ -695,7 +729,7 @@ const struct file_operations ovl_file_operations = { > .fallocate = ovl_fallocate, > .fadvise = ovl_fadvise, > .flush = ovl_flush, > - .splice_read = generic_file_splice_read, > + .splice_read = ovl_splice_read, > .splice_write = ovl_splice_write, > > .copy_file_range = ovl_copy_file_range, >
Miklos Szeredi <miklos@szeredi.hu> wrote: > > + ret = -EINVAL; > > + if (in->f_flags & O_DIRECT && > > + !(real.file->f_mode & FMODE_CAN_ODIRECT)) > > + goto out_fdput; > > This is unnecessary, as it was already done in ovl_real_fdget() -> > ovl_real_fdget_meta() -> ovl_change_flags(). Does that mean ovl_read_iter() and ovl_write_iter() shouldn't be doing it, then? David
On Wed, 15 Feb 2023 at 16:04, David Howells <dhowells@redhat.com> wrote: > > Miklos Szeredi <miklos@szeredi.hu> wrote: > > > > + ret = -EINVAL; > > > + if (in->f_flags & O_DIRECT && > > > + !(real.file->f_mode & FMODE_CAN_ODIRECT)) > > > + goto out_fdput; > > > > This is unnecessary, as it was already done in ovl_real_fdget() -> > > ovl_real_fdget_meta() -> ovl_change_flags(). > > Does that mean ovl_read_iter() and ovl_write_iter() shouldn't be doing it, > then? That's a different thing, because ovl_*_iter() are checking on ki->flags, not f_flags. Thanks, Miklos
diff --git a/fs/overlayfs/file.c b/fs/overlayfs/file.c index c9d0c362c7ef..267b61df6fcd 100644 --- a/fs/overlayfs/file.c +++ b/fs/overlayfs/file.c @@ -419,6 +419,40 @@ static ssize_t ovl_write_iter(struct kiocb *iocb, struct iov_iter *iter) return ret; } +static ssize_t ovl_splice_read(struct file *in, loff_t *ppos, + struct pipe_inode_info *pipe, size_t len, + unsigned int flags) +{ + const struct cred *old_cred; + struct fd real; + ssize_t ret; + + ret = ovl_real_fdget(in, &real); + if (ret) + return ret; + + ret = -EINVAL; + if (in->f_flags & O_DIRECT && + !(real.file->f_mode & FMODE_CAN_ODIRECT)) + goto out_fdput; + if (!real.file->f_op->splice_read) + goto out_fdput; + + ret = rw_verify_area(READ, in, ppos, len); + if (unlikely(ret < 0)) + return ret; + + old_cred = ovl_override_creds(file_inode(in)->i_sb); + ret = real.file->f_op->splice_read(real.file, ppos, pipe, len, flags); + + revert_creds(old_cred); + ovl_file_accessed(in); +out_fdput: + fdput(real); + + return ret; +} + /* * Calling iter_file_splice_write() directly from overlay's f_op may deadlock * due to lock order inversion between pipe->mutex in iter_file_splice_write() @@ -695,7 +729,7 @@ const struct file_operations ovl_file_operations = { .fallocate = ovl_fallocate, .fadvise = ovl_fadvise, .flush = ovl_flush, - .splice_read = generic_file_splice_read, + .splice_read = ovl_splice_read, .splice_write = ovl_splice_write, .copy_file_range = ovl_copy_file_range,
Implement splice-read for overlayfs by passing the request down a layer rather than going through generic_file_splice_read() which is going to be changed to assume that ->read_folio() is present on buffered files. Signed-off-by: David Howells <dhowells@redhat.com> cc: Christoph Hellwig <hch@lst.de> cc: Jens Axboe <axboe@kernel.dk> cc: Al Viro <viro@zeniv.linux.org.uk> cc: John Hubbard <jhubbard@nvidia.com> cc: David Hildenbrand <david@redhat.com> cc: Matthew Wilcox <willy@infradead.org> cc: Miklos Szeredi <miklos@szeredi.hu> cc: linux-unionfs@vger.kernel.org cc: linux-block@vger.kernel.org cc: linux-fsdevel@vger.kernel.org cc: linux-mm@kvack.org --- fs/overlayfs/file.c | 36 +++++++++++++++++++++++++++++++++++- 1 file changed, 35 insertions(+), 1 deletion(-)