[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <1452271040.8255.226.camel@edumazet-glaptop2.roam.corp.google.com>
Date: Fri, 08 Jan 2016 08:37:20 -0800
From: Eric Dumazet <eric.dumazet@...il.com>
To: David Miller <davem@...emloft.net>
Cc: netdev <netdev@...r.kernel.org>,
Willem de Bruijn <willemb@...gle.com>
Subject: [PATCH net-next] net: add scheduling point in recvmmsg/sendmmsg
From: Eric Dumazet <edumazet@...gle.com>
Applications often have to reduce number of datagrams
they receive or send per system call to avoid starvation problems.
Really the kernel should take care of this by using cond_resched(),
so that applications can experiment bigger batch sizes.
Signed-off-by: Eric Dumazet <edumazet@...gle.com>
---
net/socket.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/net/socket.c b/net/socket.c
index d730ef9dfbf0..91c2de6f5020 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -2041,6 +2041,7 @@ int __sys_sendmmsg(int fd, struct mmsghdr __user *mmsg, unsigned int vlen,
if (err)
break;
++datagrams;
+ cond_resched();
}
fput_light(sock->file, fput_needed);
@@ -2236,6 +2237,7 @@ int __sys_recvmmsg(int fd, struct mmsghdr __user *mmsg, unsigned int vlen,
/* Out of band data, return right away */
if (msg_sys.msg_flags & MSG_OOB)
break;
+ cond_resched();
}
out_put:
Powered by blists - more mailing lists