* [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets
@ 2010-10-04 15:38 Alban Crequy
2010-10-04 16:41 ` Eric Dumazet
0 siblings, 1 reply; 9+ messages in thread
From: Alban Crequy @ 2010-10-04 15:38 UTC (permalink / raw)
Cc: Lennart Poettering, Alban Crequy, David S. Miller,
Stephen Hemminger, Eric Dumazet, Cyrill Gorcunov, Alexey Dobriyan,
linux-kernel, netdev
Userspace applications can already request to receive timestamps with:
setsockopt(sockfd, SOL_SOCKET, SO_TIMESTAMP, ...)
Although setsockopt() returns zero (success), timestamps are not added to the
ancillary data. This patch fixes that.
Signed-off-by: Alban Crequy <alban.crequy@collabora.co.uk>
---
net/unix/af_unix.c | 4 ++++
1 files changed, 4 insertions(+), 0 deletions(-)
diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index 617bea4..142ccea 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -1697,6 +1697,8 @@ static int unix_dgram_recvmsg(struct kiocb *iocb, struct socket *sock,
if (err)
goto out_free;
+ sock_recv_timestamp(msg, sk, skb);
+
if (!siocb->scm) {
siocb->scm = &tmp_scm;
memset(&tmp_scm, 0, sizeof(tmp_scm));
@@ -1877,6 +1879,8 @@ static int unix_stream_recvmsg(struct kiocb *iocb, struct socket *sock,
copied += chunk;
size -= chunk;
+ sock_recv_timestamp(msg, sk, skb);
+
/* Mark read part of skb as used */
if (!(flags & MSG_PEEK)) {
skb_pull(skb, chunk);
--
1.7.1
^ permalink raw reply related [flat|nested] 9+ messages in thread* Re: [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets
2010-10-04 15:38 [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets Alban Crequy
@ 2010-10-04 16:41 ` Eric Dumazet
2010-10-04 18:00 ` Alban Crequy
0 siblings, 1 reply; 9+ messages in thread
From: Eric Dumazet @ 2010-10-04 16:41 UTC (permalink / raw)
To: Alban Crequy
Cc: Lennart Poettering, David S. Miller, Stephen Hemminger,
Cyrill Gorcunov, Alexey Dobriyan, linux-kernel, netdev
Le lundi 04 octobre 2010 à 16:38 +0100, Alban Crequy a écrit :
> Userspace applications can already request to receive timestamps with:
> setsockopt(sockfd, SOL_SOCKET, SO_TIMESTAMP, ...)
>
> Although setsockopt() returns zero (success), timestamps are not added to the
> ancillary data. This patch fixes that.
>
> Signed-off-by: Alban Crequy <alban.crequy@collabora.co.uk>
> ---
> net/unix/af_unix.c | 4 ++++
> 1 files changed, 4 insertions(+), 0 deletions(-)
>
> diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
> index 617bea4..142ccea 100644
> --- a/net/unix/af_unix.c
> +++ b/net/unix/af_unix.c
> @@ -1697,6 +1697,8 @@ static int unix_dgram_recvmsg(struct kiocb *iocb, struct socket *sock,
> if (err)
> goto out_free;
>
> + sock_recv_timestamp(msg, sk, skb);
> +
> if (!siocb->scm) {
> siocb->scm = &tmp_scm;
> memset(&tmp_scm, 0, sizeof(tmp_scm));
> @@ -1877,6 +1879,8 @@ static int unix_stream_recvmsg(struct kiocb *iocb, struct socket *sock,
> copied += chunk;
> size -= chunk;
>
> + sock_recv_timestamp(msg, sk, skb);
> +
> /* Mark read part of skb as used */
> if (!(flags & MSG_PEEK)) {
> skb_pull(skb, chunk);
Are you sure its needed for unix_stream case ?
We dont do this for TCP for example, only for datagrams.
As shown in the past, sock_recv_timestamp() is a bit expensive because
it takes care of many possible options.
It would be better to use in AF_UNIX case (only software timestamps) :
Solution 1)
if (sock_flag(sk, SOCK_RCVTSTAMP))
__sock_recv_timestamp(msg, sk, skb);
Solution 2)
Or something already used elsewhere since 2.6.35 and commit
767dd03369ac1 (net: speedup sock_recv_ts_and_drops()) :
sock_recv_ts_and_drops(msg, sk, skb);
I would vote for the 1) solution
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets
2010-10-04 16:41 ` Eric Dumazet
@ 2010-10-04 18:00 ` Alban Crequy
2010-10-04 18:01 ` Alban Crequy
0 siblings, 1 reply; 9+ messages in thread
From: Alban Crequy @ 2010-10-04 18:00 UTC (permalink / raw)
To: Eric Dumazet
Cc: Lennart Poettering, David S. Miller, Stephen Hemminger,
Cyrill Gorcunov, Alexey Dobriyan, linux-kernel, netdev,
Alban Crequy
Le Mon, 04 Oct 2010 18:41:44 +0200,
Eric Dumazet <eric.dumazet@gmail.com> a écrit :
> Le lundi 04 octobre 2010 à 16:38 +0100, Alban Crequy a écrit :
> > Userspace applications can already request to receive timestamps
> > with: setsockopt(sockfd, SOL_SOCKET, SO_TIMESTAMP, ...)
> >
> > Although setsockopt() returns zero (success), timestamps are not
> > added to the ancillary data. This patch fixes that.
> >
> > Signed-off-by: Alban Crequy <alban.crequy@collabora.co.uk>
> > ---
> > net/unix/af_unix.c | 4 ++++
> > 1 files changed, 4 insertions(+), 0 deletions(-)
> >
> > diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
> > index 617bea4..142ccea 100644
> > --- a/net/unix/af_unix.c
> > +++ b/net/unix/af_unix.c
> > @@ -1697,6 +1697,8 @@ static int unix_dgram_recvmsg(struct kiocb
> > *iocb, struct socket *sock, if (err)
> > goto out_free;
> >
> > + sock_recv_timestamp(msg, sk, skb);
> > +
> > if (!siocb->scm) {
> > siocb->scm = &tmp_scm;
> > memset(&tmp_scm, 0, sizeof(tmp_scm));
> > @@ -1877,6 +1879,8 @@ static int unix_stream_recvmsg(struct kiocb
> > *iocb, struct socket *sock, copied += chunk;
> > size -= chunk;
> >
> > + sock_recv_timestamp(msg, sk, skb);
> > +
> > /* Mark read part of skb as used */
> > if (!(flags & MSG_PEEK)) {
> > skb_pull(skb, chunk);
>
> Are you sure its needed for unix_stream case ?
I was about to say it does not cost anything to add it...
> We dont do this for TCP for example, only for datagrams.
>
> As shown in the past, sock_recv_timestamp() is a bit expensive because
> it takes care of many possible options.
Is it really expensive? It looks like a few flag checks in an inline
function to me.
> It would be better to use in AF_UNIX case (only software timestamps) :
>
> Solution 1)
> if (sock_flag(sk, SOCK_RCVTSTAMP))
> __sock_recv_timestamp(msg, sk, skb);
>
> Solution 2)
> Or something already used elsewhere since 2.6.35 and commit
> 767dd03369ac1 (net: speedup sock_recv_ts_and_drops()) :
>
> sock_recv_ts_and_drops(msg, sk, skb);
>
>
> I would vote for the 1) solution
The patch in the next email implements the 1) solution on SOCK_DGRAM and
SOCK_SEQPACKET (without SOCK_STREAM).
--
Alban
^ permalink raw reply [flat|nested] 9+ messages in thread* [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets
2010-10-04 18:00 ` Alban Crequy
@ 2010-10-04 18:01 ` Alban Crequy
2010-10-04 18:09 ` Eric Dumazet
0 siblings, 1 reply; 9+ messages in thread
From: Alban Crequy @ 2010-10-04 18:01 UTC (permalink / raw)
Cc: Lennart Poettering, Alban Crequy, David S. Miller,
Stephen Hemminger, Eric Dumazet, Cyrill Gorcunov, Alexey Dobriyan,
linux-kernel, netdev
Userspace applications can already request to receive timestamps with:
setsockopt(sockfd, SOL_SOCKET, SO_TIMESTAMP, ...)
Although setsockopt() returns zero (success), timestamps are not added to the
ancillary data. This patch fixes that on SOCK_DGRAM and SOCK_SEQPACKET Unix
sockets.
Signed-off-by: Alban Crequy <alban.crequy@collabora.co.uk>
---
net/unix/af_unix.c | 3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index 617bea4..bca0c1e 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -1697,6 +1697,9 @@ static int unix_dgram_recvmsg(struct kiocb *iocb, struct socket *sock,
if (err)
goto out_free;
+ if (sock_flag(sk, SOCK_RCVTSTAMP))
+ __sock_recv_timestamp(msg, sk, skb);
+
if (!siocb->scm) {
siocb->scm = &tmp_scm;
memset(&tmp_scm, 0, sizeof(tmp_scm));
--
1.7.1
^ permalink raw reply related [flat|nested] 9+ messages in thread* Re: [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets
2010-10-04 18:01 ` Alban Crequy
@ 2010-10-04 18:09 ` Eric Dumazet
2010-10-04 18:47 ` Alban Crequy
0 siblings, 1 reply; 9+ messages in thread
From: Eric Dumazet @ 2010-10-04 18:09 UTC (permalink / raw)
To: Alban Crequy
Cc: Lennart Poettering, David S. Miller, Stephen Hemminger,
Cyrill Gorcunov, Alexey Dobriyan, linux-kernel, netdev
Le lundi 04 octobre 2010 à 19:01 +0100, Alban Crequy a écrit :
> Userspace applications can already request to receive timestamps with:
> setsockopt(sockfd, SOL_SOCKET, SO_TIMESTAMP, ...)
>
> Although setsockopt() returns zero (success), timestamps are not added to the
> ancillary data. This patch fixes that on SOCK_DGRAM and SOCK_SEQPACKET Unix
> sockets.
>
> Signed-off-by: Alban Crequy <alban.crequy@collabora.co.uk>
> ---
> net/unix/af_unix.c | 3 +++
> 1 files changed, 3 insertions(+), 0 deletions(-)
>
> diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
> index 617bea4..bca0c1e 100644
> --- a/net/unix/af_unix.c
> +++ b/net/unix/af_unix.c
> @@ -1697,6 +1697,9 @@ static int unix_dgram_recvmsg(struct kiocb *iocb, struct socket *sock,
> if (err)
> goto out_free;
>
> + if (sock_flag(sk, SOCK_RCVTSTAMP))
> + __sock_recv_timestamp(msg, sk, skb);
> +
> if (!siocb->scm) {
> siocb->scm = &tmp_scm;
> memset(&tmp_scm, 0, sizeof(tmp_scm));
Hmm, but how is set skb->tstamp ?
I think its zero at this point, so __sock_recv_timestamp() use current
time, not time of send().
gettimeofday() might be better/cheaper :)
^ permalink raw reply [flat|nested] 9+ messages in thread* Re: [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets
2010-10-04 18:09 ` Eric Dumazet
@ 2010-10-04 18:47 ` Alban Crequy
2010-10-04 18:48 ` Alban Crequy
0 siblings, 1 reply; 9+ messages in thread
From: Alban Crequy @ 2010-10-04 18:47 UTC (permalink / raw)
To: Eric Dumazet
Cc: Lennart Poettering, David S. Miller, Stephen Hemminger,
Cyrill Gorcunov, Alexey Dobriyan, linux-kernel, netdev,
Alban Crequy
Le Mon, 04 Oct 2010 20:09:49 +0200,
Eric Dumazet <eric.dumazet@gmail.com> a écrit :
> Le lundi 04 octobre 2010 à 19:01 +0100, Alban Crequy a écrit :
> > Userspace applications can already request to receive timestamps
> > with: setsockopt(sockfd, SOL_SOCKET, SO_TIMESTAMP, ...)
> >
> > Although setsockopt() returns zero (success), timestamps are not
> > added to the ancillary data. This patch fixes that on SOCK_DGRAM
> > and SOCK_SEQPACKET Unix sockets.
> >
> > Signed-off-by: Alban Crequy <alban.crequy@collabora.co.uk>
> > ---
> > net/unix/af_unix.c | 3 +++
> > 1 files changed, 3 insertions(+), 0 deletions(-)
> >
> > diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
> > index 617bea4..bca0c1e 100644
> > --- a/net/unix/af_unix.c
> > +++ b/net/unix/af_unix.c
> > @@ -1697,6 +1697,9 @@ static int unix_dgram_recvmsg(struct kiocb
> > *iocb, struct socket *sock, if (err)
> > goto out_free;
> >
> > + if (sock_flag(sk, SOCK_RCVTSTAMP))
> > + __sock_recv_timestamp(msg, sk, skb);
> > +
> > if (!siocb->scm) {
> > siocb->scm = &tmp_scm;
> > memset(&tmp_scm, 0, sizeof(tmp_scm));
>
> Hmm, but how is set skb->tstamp ?
>
> I think its zero at this point, so __sock_recv_timestamp() use current
> time, not time of send().
>
> gettimeofday() might be better/cheaper :)
You're right, I just tested it and it was the time of reception. Thanks
for the review.
^ permalink raw reply [flat|nested] 9+ messages in thread* [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets
2010-10-04 18:47 ` Alban Crequy
@ 2010-10-04 18:48 ` Alban Crequy
2010-10-04 19:20 ` Eric Dumazet
0 siblings, 1 reply; 9+ messages in thread
From: Alban Crequy @ 2010-10-04 18:48 UTC (permalink / raw)
Cc: Lennart Poettering, Alban Crequy, David S. Miller,
Stephen Hemminger, Eric Dumazet, Cyrill Gorcunov, Alexey Dobriyan,
linux-kernel, netdev
Userspace applications can already request to receive timestamps with:
setsockopt(sockfd, SOL_SOCKET, SO_TIMESTAMP, ...)
Although setsockopt() returns zero (success), timestamps are not added to the
ancillary data. This patch fixes that on SOCK_DGRAM and SOCK_SEQPACKET Unix
sockets.
Signed-off-by: Alban Crequy <alban.crequy@collabora.co.uk>
---
net/unix/af_unix.c | 5 +++++
1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index 617bea4..b03baa8 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -1484,6 +1484,8 @@ restart:
goto restart;
}
+ if (sock_flag(other, SOCK_RCVTSTAMP))
+ __net_timestamp(skb);
skb_queue_tail(&other->sk_receive_queue, skb);
unix_state_unlock(other);
other->sk_data_ready(other, len);
@@ -1697,6 +1699,9 @@ static int unix_dgram_recvmsg(struct kiocb *iocb, struct socket *sock,
if (err)
goto out_free;
+ if (sock_flag(sk, SOCK_RCVTSTAMP))
+ __sock_recv_timestamp(msg, sk, skb);
+
if (!siocb->scm) {
siocb->scm = &tmp_scm;
memset(&tmp_scm, 0, sizeof(tmp_scm));
--
1.7.1
^ permalink raw reply related [flat|nested] 9+ messages in thread* Re: [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets
2010-10-04 18:48 ` Alban Crequy
@ 2010-10-04 19:20 ` Eric Dumazet
2010-10-05 21:55 ` David Miller
0 siblings, 1 reply; 9+ messages in thread
From: Eric Dumazet @ 2010-10-04 19:20 UTC (permalink / raw)
To: Alban Crequy
Cc: Lennart Poettering, David S. Miller, Stephen Hemminger,
Cyrill Gorcunov, Alexey Dobriyan, linux-kernel, netdev
Le lundi 04 octobre 2010 à 19:48 +0100, Alban Crequy a écrit :
> Userspace applications can already request to receive timestamps with:
> setsockopt(sockfd, SOL_SOCKET, SO_TIMESTAMP, ...)
>
> Although setsockopt() returns zero (success), timestamps are not added to the
> ancillary data. This patch fixes that on SOCK_DGRAM and SOCK_SEQPACKET Unix
> sockets.
>
> Signed-off-by: Alban Crequy <alban.crequy@collabora.co.uk>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Note: tstamp is sampled at the time of skb queueing, _after_ eventual
sleep in sock_alloc_send_skb() or memcpy_fromiovec()
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets
2010-10-04 19:20 ` Eric Dumazet
@ 2010-10-05 21:55 ` David Miller
0 siblings, 0 replies; 9+ messages in thread
From: David Miller @ 2010-10-05 21:55 UTC (permalink / raw)
To: eric.dumazet
Cc: alban.crequy, lennart, shemminger, gorcunov, adobriyan,
linux-kernel, netdev
From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Mon, 04 Oct 2010 21:20:52 +0200
> Le lundi 04 octobre 2010 à 19:48 +0100, Alban Crequy a écrit :
>> Userspace applications can already request to receive timestamps with:
>> setsockopt(sockfd, SOL_SOCKET, SO_TIMESTAMP, ...)
>>
>> Although setsockopt() returns zero (success), timestamps are not added to the
>> ancillary data. This patch fixes that on SOCK_DGRAM and SOCK_SEQPACKET Unix
>> sockets.
>>
>> Signed-off-by: Alban Crequy <alban.crequy@collabora.co.uk>
>
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Applied.
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2010-10-05 21:55 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-10-04 15:38 [PATCH] AF_UNIX: Implement SO_TIMESTAMP and SO_TIMETAMPNS on Unix sockets Alban Crequy
2010-10-04 16:41 ` Eric Dumazet
2010-10-04 18:00 ` Alban Crequy
2010-10-04 18:01 ` Alban Crequy
2010-10-04 18:09 ` Eric Dumazet
2010-10-04 18:47 ` Alban Crequy
2010-10-04 18:48 ` Alban Crequy
2010-10-04 19:20 ` Eric Dumazet
2010-10-05 21:55 ` David Miller
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox