On Tue, Dec 05, 2017 at 01:28:55PM -0500, Neil Horman wrote:
> On Tue, Dec 05, 2017 at 11:16:04PM +0800, Xin Long wrote:
> > ulpevent_data is added as a member of sctp_stream_interleave, used to
> > do the most process in ulpq, including to convert data or idata chunk
> > to event, reasm them in reasm queue and put them in lobby queue in
> > right order, and deliver them up to user sk rx queue.
> >
> > This procedure is described in section 2.2.3 of RFC8260.
> >
> > It adds most functions for idata here to do the similar process as
> > the old functions for data. But since the details are very different
> > between them, the old functions can not be reused for idata.
> >
> > event->ssn and event->ppid settings are moved to ulpevent_data from
> > sctp_ulpevent_make_rcvmsg, so that sctp_ulpevent_make_rcvmsg could
> > work for both data and idata.
> >
> > Note that mid is added in sctp_ulpevent for idata, __packed has to
> > be used for defining sctp_ulpevent, or it would exceeds the skb cb
> > that saves a sctp_ulpevent variable for ulp layer process.
> >
> > Signed-off-by: Xin Long <lucien.xin@xxxxxxxxx>
> > ---
> > include/net/sctp/stream_interleave.h | 2 +
> > include/net/sctp/structs.h | 3 +
> > include/net/sctp/ulpevent.h | 20 +-
> > net/sctp/sm_sideeffect.c | 5 +-
> > net/sctp/stream_interleave.c | 418 +++++++++++++++++++++++++++++++++++
> > net/sctp/ulpevent.c | 2 -
> > net/sctp/ulpqueue.c | 12 +-
> > 7 files changed, 451 insertions(+), 11 deletions(-)
> >
> > diff --git a/include/net/sctp/stream_interleave.h b/include/net/sctp/stream_interleave.h
> > index d8d1b51..02f60f5 100644
> > --- a/include/net/sctp/stream_interleave.h
> > +++ b/include/net/sctp/stream_interleave.h
> > @@ -39,6 +39,8 @@ struct sctp_stream_interleave {
> > int len, __u8 flags, gfp_t gfp);
> > void (*assign_number)(struct sctp_chunk *chunk);
> > bool (*validate_data)(struct sctp_chunk *chunk);
> > + int (*ulpevent_data)(struct sctp_ulpq *ulpq,
> > + struct sctp_chunk *chunk, gfp_t gfp);
> > };
> >
> > void sctp_stream_interleave_init(struct sctp_stream *stream);
> > diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h
> > index 348b25e..d7da719 100644
> > --- a/include/net/sctp/structs.h
> > +++ b/include/net/sctp/structs.h
> > @@ -411,6 +411,8 @@ void sctp_stream_update(struct sctp_stream *stream, struct sctp_stream *new);
> > #define sctp_mid_skip(stream, type, sid, mid) \
> > ((stream)->type[sid].mid = mid + 1)
> >
> > +#define sctp_stream_in(asoc, sid) (&(asoc)->stream.in[sid])
> > +
> > /*
> > * Pointers to address related SCTP functions.
> > * (i.e. things that depend on the address family.)
> > @@ -1386,6 +1388,7 @@ struct sctp_stream_in {
> > __u16 ssn;
> > };
> > __u32 fsn;
> > + char pd_mode;
> > };
> >
> > struct sctp_stream {
> > diff --git a/include/net/sctp/ulpevent.h b/include/net/sctp/ulpevent.h
> > index 231dc42..ce4f2aa 100644
> > --- a/include/net/sctp/ulpevent.h
> > +++ b/include/net/sctp/ulpevent.h
> > @@ -45,19 +45,29 @@
> > /* A structure to carry information to the ULP (e.g. Sockets API) */
> > /* Warning: This sits inside an skb.cb[] area. Be very careful of
> > * growing this structure as it is at the maximum limit now.
> > + *
> > + * sctp_ulpevent is saved in sk->cb(48 bytes), whose last 4 bytes
> > + * have been taken by sock_skb_cb, So here it has to use 'packed'
> > + * to make sctp_ulpevent fit into the rest 44 bytes.
> > */
> > struct sctp_ulpevent {
> > struct sctp_association *asoc;
> > struct sctp_chunk *chunk;
> > unsigned int rmem_len;
> > - __u32 ppid;
> > + union {
> > + __u32 mid;
> > + __u16 ssn;
> > + };
> > + union {
> > + __u32 ppid;
> > + __u32 fsn;
> > + };
> > __u32 tsn;
> > __u32 cumtsn;
> > __u16 stream;
> > - __u16 ssn;
> > __u16 flags;
> > __u16 msg_flags;
> > -};
> > +} __packed;
> >
> What kind of performance do you see before and after this patch? I ask because
> it seems like the members between ppid through stream are going to be misaligned
> (not on a 4 byte boundary), now that you've made this structure packed.
It shouldn't be misaligned because the __u16 ssn field is wrapped on a union
with __u32 mid, causing the next field to be aligned on the largest member,
which is __u32.
FWIW, before:
struct sctp_ulpevent {
struct sctp_association * asoc; /* 0 8 */
struct sctp_chunk * chunk; /* 8 8 */
unsigned int rmem_len; /* 16 4 */
__u32 ppid; /* 20 4 */
__u32 tsn; /* 24 4 */
__u32 cumtsn; /* 28 4 */
__u16 stream; /* 32 2 */
__u16 ssn; /* 34 2 */
__u16 flags; /* 36 2 */
__u16 msg_flags; /* 38 2 */
/* size: 40, cachelines: 1, members: 10 */
/* last cacheline: 40 bytes */
};
and after:
struct sctp_ulpevent {
struct sctp_association * asoc; /* 0 8 */
struct sctp_chunk * chunk; /* 8 8 */
unsigned int rmem_len; /* 16 4 */
union {
__u32 mid; /* 4 */
__u16 ssn; /* 2 */
}; /* 20 4 */
union {
__u32 ppid; /* 4 */
__u32 fsn; /* 4 */
}; /* 24 4 */
__u32 tsn; /* 28 4 */
__u32 cumtsn; /* 32 4 */
__u16 stream; /* 36 2 */
__u16 flags; /* 38 2 */
__u16 msg_flags; /* 40 2 */
/* size: 42, cachelines: 1, members: 10 */
/* last cacheline: 42 bytes */
};
Marcelo
--
To unsubscribe from this list: send the line "unsubscribe linux-sctp" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
