cifs: use discard iterator to discard unneeded network data more efficiently

(cleanup patch, to make netfs merge easier)

The iterator, ITER_DISCARD, that can only be used in READ mode and
just discards any data copied to it, was added to allow a network
filesystem to discard any unwanted data sent by a server.
Convert cifs_discard_from_socket() to use this.

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
---
 fs/cifs/cifsproto.h |  2 ++
 fs/cifs/cifssmb.c   |  6 +++---
 fs/cifs/connect.c   | 10 ++++++++++
 3 files changed, 15 insertions(+), 3 deletions(-)

  unsigned int page_offset, unsigned int to_read)

Steve French <smfrench@gmail.com> writes:
> +ssize_t
> +cifs_discard_from_socket(struct TCP_Server_Info *server, size_t to_read)
> +{
> +	struct msghdr smb_msg;
> +
> +	iov_iter_discard(&smb_msg.msg_iter, READ, to_read);
> +
> +	return cifs_readv_from_socket(server, &smb_msg);
> +}
> +

Shouldn't smb_msg be initialized to zeroes? Looking around this needs to
be done for cifs_read_from_socket() and cifs_read_page_from_socket() too.

Cheers,

Aurélien Aptel <aaptel@suse.com> wrote:

> > +{
> > +	struct msghdr smb_msg;
> > +
> > +	iov_iter_discard(&smb_msg.msg_iter, READ, to_read);
> > +
> > +	return cifs_readv_from_socket(server, &smb_msg);
> > +}
> > +
> 
> Shouldn't smb_msg be initialized to zeroes? Looking around this needs to
> be done for cifs_read_from_socket() and cifs_read_page_from_socket() too.

Yeah - I think you're right.  I didn't manage to finish making the changes,
so what I gave to Steve wasn't tested.

David

The other two routines initialize in iov_iter_bvec

iov->type
iov->bvec
iov->offset
iov->count

but iov_iter_discard already does the initialization:
iov_type
iov_offset
iov_count

and then we call cifs_readv_from_socket in all 3
which sets:
    iov->msg_control = NULL
    iov->msg_controllen = NULL

I will set the two additional ones to null
    iov->msg_name
and
    iov->msg_namelen



On Thu, Feb 4, 2021 at 4:29 AM Aurélien Aptel <aaptel@suse.com> wrote:
>
> Steve French <smfrench@gmail.com> writes:
> > +ssize_t
> > +cifs_discard_from_socket(struct TCP_Server_Info *server, size_t to_read)
> > +{
> > +     struct msghdr smb_msg;
> > +
> > +     iov_iter_discard(&smb_msg.msg_iter, READ, to_read);
> > +
> > +     return cifs_readv_from_socket(server, &smb_msg);
> > +}
> > +
>
> Shouldn't smb_msg be initialized to zeroes? Looking around this needs to
> be done for cifs_read_from_socket() and cifs_read_page_from_socket() too.
>
> Cheers,
> --
> Aurélien Aptel / SUSE Labs Samba Team
> GPG: 1839 CB5F 9F5B FB9B AA97  8C99 03C8 A49B 521B D5D3
> SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg, DE
> GF: Felix Imendörffer, Mary Higgins, Sri Rasiah HRB 247165 (AG München)
>

lightly updated patch


On Thu, Feb 25, 2021 at 1:29 PM Steve French <smfrench@gmail.com> wrote:
>
> The other two routines initialize in iov_iter_bvec
>
> iov->type
> iov->bvec
> iov->offset
> iov->count
>
> but iov_iter_discard already does the initialization:
> iov_type
> iov_offset
> iov_count
>
> and then we call cifs_readv_from_socket in all 3
> which sets:
>     iov->msg_control = NULL
>     iov->msg_controllen = NULL
>
> I will set the two additional ones to null
>     iov->msg_name
> and
>     iov->msg_namelen
>
>
>
> On Thu, Feb 4, 2021 at 4:29 AM Aurélien Aptel <aaptel@suse.com> wrote:
> >
> > Steve French <smfrench@gmail.com> writes:
> > > +ssize_t
> > > +cifs_discard_from_socket(struct TCP_Server_Info *server, size_t to_read)
> > > +{
> > > +     struct msghdr smb_msg;
> > > +
> > > +     iov_iter_discard(&smb_msg.msg_iter, READ, to_read);
> > > +
> > > +     return cifs_readv_from_socket(server, &smb_msg);
> > > +}
> > > +
> >
> > Shouldn't smb_msg be initialized to zeroes? Looking around this needs to
> > be done for cifs_read_from_socket() and cifs_read_page_from_socket() too.
> >
> > Cheers,
> > --
> > Aurélien Aptel / SUSE Labs Samba Team
> > GPG: 1839 CB5F 9F5B FB9B AA97  8C99 03C8 A49B 521B D5D3
> > SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg, DE
> > GF: Felix Imendörffer, Mary Higgins, Sri Rasiah HRB 247165 (AG München)
> >
>
>
> --
> Thanks,
>
> Steve

Had a typo in the previous attachment - reattached

The iterator, ITER_DISCARD, that can only be used in READ mode and
just discards any data copied to it, was added to allow a network
filesystem to discard any unwanted data sent by a server.
Convert cifs_discard_from_socket() to use this.

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
---
 fs/cifs/cifsproto.h |  2 ++
 fs/cifs/cifssmb.c   |  6 +++---
 fs/cifs/connect.c   | 17 +++++++++++++++++
 3 files changed, 22 insertions(+), 3 deletions(-)

diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h
index d4b2cc7e6b1e..64eb5c817712 100644
--- a/fs/cifs/cifsproto.h
+++ b/fs/cifs/cifsproto.h
@@ -232,6 +232,8 @@ extern unsigned int
setup_special_user_owner_ACE(struct cifs_ace *pace);
 extern void dequeue_mid(struct mid_q_entry *mid, bool malformed);
 extern int cifs_read_from_socket(struct TCP_Server_Info *server, char *buf,
           unsigned int to_read);
+extern ssize_t cifs_discard_from_socket(struct TCP_Server_Info *server,
+ size_t to_read);
 extern int cifs_read_page_from_socket(struct TCP_Server_Info *server,
  struct page *page,
  unsigned int page_offset,
diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c
index 0496934feecb..c279527aae92 100644
--- a/fs/cifs/cifssmb.c
+++ b/fs/cifs/cifssmb.c
@@ -1451,9 +1451,9 @@ cifs_discard_remaining_data(struct
TCP_Server_Info *server)
  while (remaining > 0) {
  int length;

- length = cifs_read_from_socket(server, server->bigbuf,
- min_t(unsigned int, remaining,
-     CIFSMaxBufSize + MAX_HEADER_SIZE(server)));
+ length = cifs_discard_from_socket(server,
+ min_t(size_t, remaining,
+       CIFSMaxBufSize + MAX_HEADER_SIZE(server)));
  if (length < 0)
  return length;
  server->total_read += length;
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index 55b1ec731d52..b9085625334e 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -564,6 +564,23 @@ cifs_read_from_socket(struct TCP_Server_Info
*server, char *buf,
  return cifs_readv_from_socket(server, &smb_msg);
 }

+ssize_t
+cifs_discard_from_socket(struct TCP_Server_Info *server, size_t to_read)
+{
+ struct msghdr smb_msg;
+
+ /*
+ *  iov_iter_discard already sets smb_msg.type and count and iov_offset
+ *  and cifs_readv_from_socket sets msg_control and msg_controllen
+ *  so little to initialize in struct msghdr
+ */
+ smb_msg.msg_name = NULL;
+ smb_msg.msg_namelen = 0;
+ iov_iter_discard(&smb_msg.msg_iter, READ, to_read);
+
+ return cifs_readv_from_socket(server, &smb_msg);
+}
+
 int
 cifs_read_page_from_socket(struct TCP_Server_Info *server, struct page *page,
  unsigned int page_offset, unsigned int to_read)

On Thu, Feb 25, 2021 at 1:43 PM Steve French <smfrench@gmail.com> wrote:
>
> lightly updated patch
>
>
> On Thu, Feb 25, 2021 at 1:29 PM Steve French <smfrench@gmail.com> wrote:
> >
> > The other two routines initialize in iov_iter_bvec
> >
> > iov->type
> > iov->bvec
> > iov->offset
> > iov->count
> >
> > but iov_iter_discard already does the initialization:
> > iov_type
> > iov_offset
> > iov_count
> >
> > and then we call cifs_readv_from_socket in all 3
> > which sets:
> >     iov->msg_control = NULL
> >     iov->msg_controllen = NULL
> >
> > I will set the two additional ones to null
> >     iov->msg_name
> > and
> >     iov->msg_namelen
> >
> >
> >
> > On Thu, Feb 4, 2021 at 4:29 AM Aurélien Aptel <aaptel@suse.com> wrote:
> > >
> > > Steve French <smfrench@gmail.com> writes:
> > > > +ssize_t
> > > > +cifs_discard_from_socket(struct TCP_Server_Info *server, size_t to_read)
> > > > +{
> > > > +     struct msghdr smb_msg;
> > > > +
> > > > +     iov_iter_discard(&smb_msg.msg_iter, READ, to_read);
> > > > +
> > > > +     return cifs_readv_from_socket(server, &smb_msg);
> > > > +}
> > > > +
> > >
> > > Shouldn't smb_msg be initialized to zeroes? Looking around this needs to
> > > be done for cifs_read_from_socket() and cifs_read_page_from_socket() too.
> > >
> > > Cheers,
> > > --
> > > Aurélien Aptel / SUSE Labs Samba Team
> > > GPG: 1839 CB5F 9F5B FB9B AA97  8C99 03C8 A49B 521B D5D3
> > > SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg, DE
> > > GF: Felix Imendörffer, Mary Higgins, Sri Rasiah HRB 247165 (AG München)
> > >
> >
> >
> > --
> > Thanks,
> >
> > Steve
>
>
>
> --
> Thanks,
>
> Steve