On Mon, 8 Jan 2024 04:16:20 -0500 Jon Maloy wrote:

When reading received messages from a socket with MSG_PEEK, we may want to read the contents with an offset, like we can do with pread/preadv() when reading files. Currently, it is not possible to do that.

In this commit, we allow the user to set iovec.iov_base in the first vector entry to NULL. This tells the socket to skip the first entry, hence letting the iov_len field of that entry indicate the offset value. This way, there is no need to add any new arguments or flags.

In the iperf3 logs examples shown below, we can observe a throughput improvement of ~15 % in the direction host->namespace when using the protocol splicer 'pasta' (https://passt.top). This is a consistent result.

'passt' is a tool used to support VMs in containers, such as KubeVirt, and is also generally supported in libvirt/QEMU since release 9.2 / 7.2. 'pasta' is the pure namespace variety of the same tool.

It's not necessarily VMs in containers, and "support" doesn't quite explain we're doing user-mode networking, or why we need this. Perhaps, instead of this paragraph: pasta(1) and passt(1) implement user-mode networking for network namespaces (containers) and virtual machines by means of a translation layer between Layer-2 network interface and native Layer-4 sockets (TCP, UDP, ICMP/ICMPv6 echo). Received, pending TCP data to the container/guest is kept in kernel buffers until acknowledged, so the tool routinely needs to fetch new data from socket, skipping data that was already sent. At the moment this is implemented using a dummy buffer passed to recvmsg(). With this change, we don't need a dummy buffer and the related buffer copy (copy_to_user()) anymore.

jmaloy@lubu:~/passt/passt$ ../passt//net/tools/perf/perf record -g ./pasta --config-net -f MSG_PEEK with offset not supported by kernel.

root@lubu:~/passt/passt# iperf3 -s ----------------------------------------------------------- Server listening on 5201 (test #1) ----------------------------------------------------------- Accepted connection from 192.168.122.1, port 44822 [ 5] local 192.168.122.180 port 5201 connected to 192.168.122.1 port 44832 [ ID] Interval Transfer Bitrate [ 5] 0.00-1.00 sec 1.02 GBytes 8.78 Gbits/sec [ 5] 1.00-2.00 sec 1.06 GBytes 9.08 Gbits/sec [ 5] 2.00-3.00 sec 1.07 GBytes 9.15 Gbits/sec [ 5] 3.00-4.00 sec 1.10 GBytes 9.46 Gbits/sec [ 5] 4.00-5.00 sec 1.03 GBytes 8.85 Gbits/sec [ 5] 5.00-6.00 sec 1.10 GBytes 9.44 Gbits/sec [ 5] 6.00-7.00 sec 1.11 GBytes 9.56 Gbits/sec [ 5] 7.00-8.00 sec 1.07 GBytes 9.20 Gbits/sec [ 5] 8.00-9.00 sec 667 MBytes 5.59 Gbits/sec [ 5] 9.00-10.00 sec 1.03 GBytes 8.83 Gbits/sec [ 5] 10.00-10.04 sec 30.1 MBytes 6.36 Gbits/sec - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate [ 5] 0.00-10.04 sec 10.3 GBytes 8.78 Gbits/sec receiver ----------------------------------------------------------- Server listening on 5201 (test #2) ----------------------------------------------------------- ^Ciperf3: interrupt - the server has terminated root@lubu:~/passt/passt# logout [ perf record: Woken up 23 times to write data ] [ perf record: Captured and wrote 5.696 MB perf.data (35580 samples) ] jmaloy@lubu:~/passt/passt$

jmaloy@lubu:~/passt/passt$ /home/jmaloy/passt//net/tools/perf/perf record -g ./pasta --config-net -f MSG_PEEK with offset supported by kernel.

root@lubu:~/passt/passt# iperf3 -s ----------------------------------------------------------- Server listening on 5201 (test #1) ----------------------------------------------------------- Accepted connection from 192.168.122.1, port 40854 [ 5] local 192.168.122.180 port 5201 connected to 192.168.122.1 port 40862 [ ID] Interval Transfer Bitrate [ 5] 0.00-1.00 sec 1.22 GBytes 10.5 Gbits/sec [ 5] 1.00-2.00 sec 1.19 GBytes 10.2 Gbits/sec [ 5] 2.00-3.00 sec 1.22 GBytes 10.5 Gbits/sec [ 5] 3.00-4.00 sec 1.11 GBytes 9.56 Gbits/sec [ 5] 4.00-5.00 sec 1.20 GBytes 10.3 Gbits/sec [ 5] 5.00-6.00 sec 1.14 GBytes 9.80 Gbits/sec [ 5] 6.00-7.00 sec 1.17 GBytes 10.0 Gbits/sec [ 5] 7.00-8.00 sec 1.12 GBytes 9.61 Gbits/sec [ 5] 8.00-9.00 sec 1.13 GBytes 9.74 Gbits/sec [ 5] 9.00-10.00 sec 1.26 GBytes 10.8 Gbits/sec - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate [ 5] 0.00-10.04 sec 11.8 GBytes 10.1 Gbits/sec receiver ----------------------------------------------------------- Server listening on 5201 (test #2) ----------------------------------------------------------- ^Ciperf3: interrupt - the server has terminated logout [ perf record: Woken up 20 times to write data ] [ perf record: Captured and wrote 5.040 MB perf.data (33411 samples) ] jmaloy@lubu:~/passt/passt$

The perf record confirms this result. Below, we can observe that the CPU spends significantly less time in the function ____sys_recvmsg() when we have offset support.

Without offset support: ---------------------- jmaloy@lubu:~/passt/net/tools/perf$ ./perf report -q --symbol-filter=do_syscall_64 -p ____sys_recvmsg -x --stdio -i /home/jmaloy/passt/passt/perf.data | head -1 46.32% 0.00% passt.avx2 [kernel.vmlinux] [k] do_syscall_64 ____sys_recvmsg jmaloy@lubu:~$

With offset support: ---------------------- jmaloy@lubu:~/passt/net/tools/perf$ ./perf report -q --symbol-filter=do_syscall_64 -p ____sys_recvmsg -x --stdio -i /home/jmaloy/passt/passt/perf.data | head -1 27.24% 0.00% passt.avx2 [kernel.vmlinux] [k] do_syscall_64 ____sys_recvmsg jmaloy@lubu:~$

Signed-off-by: Jon Maloy

--- v3: Made changes suggested by Stefano Brivio: - Added perf result to commit log - Separated parameter sanity tests from code logics

Signed-off-by: Jon Maloy --- net/ipv4/tcp.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)

diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index ff6838ca2e58..007b56dfc9e0 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -2310,6 +2310,7 @@ static int tcp_recvmsg_locked(struct sock *sk, struct msghdr *msg, size_t len, int *cmsg_flags) { struct tcp_sock *tp = tcp_sk(sk); + size_t peek_offset;

I still think that the declaration of 'peek_offset' should be moved below, to the block you're adding. Nowadays even sparse or checkpatch.pl (I don't remember which one) report unnecessarily broad scopes.

int copied = 0; u32 peek_seq; u32 *seq; @@ -2353,6 +2354,18 @@ static int tcp_recvmsg_locked(struct sock *sk, struct msghdr *msg, size_t len, if (flags & MSG_PEEK) { peek_seq = tp->copied_seq; seq = &peek_seq; + if (!msg->msg_iter.__iov[0].iov_base) { + peek_offset = msg->msg_iter.__iov[0].iov_len; + if (peek_offset >= len || msg->msg_iter.nr_segs <= 1) { + err = -EINVAL; + goto out; + } + msg->msg_iter.__iov = &msg->msg_iter.__iov[1]; + msg->msg_iter.nr_segs -= 1; + msg->msg_iter.count -= peek_offset;

I spent a while on this but I couldn't find an obvious place in the path (from the syscall entry point, all the way down to copy_to_iter()) where the first iov_len (peek_offset) is checked against msg->msg_iter.count, like in the test you had in v2. I couldn't even find out where msg_iter.count is calculated, actually, so I'm not sure how to test this. If you are confident it's fine, I guess it is.

+ len -= peek_offset; + *seq += peek_offset; + } }

target = sock_rcvlowat(sk, flags & MSG_WAITALL, len);

Everything else looks good to me. -- Stefano