d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Replace the ill-named ZERO_COPY_SOCKET kernel option with two

Browse Source 
more appropriate named kernel options for the very distinct
send and receive path.

"options SOCKET_SEND_COW" enables VM page copy-on-write based
sending of data on an outbound socket.

NB: The COW based send mechanism is not safe and may result
in kernel crashes.

"options SOCKET_RECV_PFLIP" enables VM kernel/userspace page
flipping for special disposable pages attached as external
storage to mbufs.

Only the naming of the kernel options is changed and their
corresponding #ifdef sections are adjusted.  No functionality
is added or removed.

Discussed with:	alc (mechanism and limitations of send side COW)
This commit is contained in:
Andre Oppermann 2012-10-23 14:19:44 +00:00
parent 0de9e2cbbd
commit e37e60c379
4 changed files with 40 additions and 28 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
sys/conf/NOTES
View File

@ -964,12 +964,20 @@ options TCP_SIGNATURE #include support for RFC 2385
# a smooth scheduling of the traffic.
options DUMMYNET
# Zero copy sockets support. This enables "zero copy" for sending and
# receiving data via a socket. The send side works for any type of NIC,
# the receive side only works for NICs that support MTUs greater than the
# page size of your architecture and that support header splitting. See
# zero_copy(9) for more details.
options ZERO_COPY_SOCKETS
# "Zero copy" sockets support is split into the send and receive path
# which operate very differently.
# For the send path the VM page with the data is wired into the kernel
# and marked as COW (copy-on-write). If the application touches the
# data while it is still in the send socket buffer the page is copied
# and divorced from its kernel wiring (no longer zero copy).
# The receive side requires explicit NIC driver support to create
# disposable pages which are flipped from kernel to user-space VM.
# See zero_copy(9) for more details.
# XXX: The COW based send mechanism is not safe and may result in
# kernel crashes.
# XXX: None of the current NIC drivers support disposeable pages.
options SOCKET_SEND_COW
options SOCKET_RECV_PFLIP
#####################################################################
# FILESYSTEM OPTIONS

3
sys/conf/options
View File

@ -520,7 +520,8 @@ NGATM_CCATM opt_netgraph.h
# DRM options
DRM_DEBUG opt_drm.h
ZERO_COPY_SOCKETS opt_zero.h
SOCKET_SEND_COW opt_zero.h
SOCKET_RECV_PFLIP opt_zero.h
TI_SF_BUF_JUMBO opt_ti.h
TI_JUMBO_HDRSPLIT opt_ti.h
BCE_JUMBO_HDRSPLIT opt_bce.h

10
sys/kern/subr_uio.c
View File

@ -57,7 +57,7 @@ __FBSDID("$FreeBSD$");
#include <vm/vm_extern.h>
#include <vm/vm_page.h>
#include <vm/vm_map.h>
#ifdef ZERO_COPY_SOCKETS
#ifdef SOCKET_SEND_COW
#include <vm/vm_object.h>
#endif
@ -66,7 +66,7 @@ SYSCTL_INT(_kern, KERN_IOV_MAX, iov_max, CTLFLAG_RD, NULL, UIO_MAXIOV,
static int uiomove_faultflag(void *cp, int n, struct uio *uio, int nofault);
#ifdef ZERO_COPY_SOCKETS
#ifdef SOCKET_SEND_COW
/* Declared in uipc_socket.c */
extern int so_zero_copy_receive;
@ -128,7 +128,7 @@ retry:
vm_map_lookup_done(map, entry);
return(KERN_SUCCESS);
}
#endif /* ZERO_COPY_SOCKETS */
#endif /* SOCKET_SEND_COW */
int
copyin_nofault(const void *udaddr, void *kaddr, size_t len)
@ -261,7 +261,7 @@ uiomove_frombuf(void *buf, int buflen, struct uio *uio)
return (uiomove((char *)buf + offset, n, uio));
}
#ifdef ZERO_COPY_SOCKETS
#ifdef SOCKET_RECV_PFLIP
/*
* Experimental support for zero-copy I/O
*/
@ -356,7 +356,7 @@ uiomoveco(void *cp, int n, struct uio *uio, int disposable)
}
return (0);
}
#endif /* ZERO_COPY_SOCKETS */
#endif /* SOCKET_RECV_PFLIP */
/*
* Give next character to user as result of read.

35
sys/kern/uipc_socket.c
View File

@ -219,17 +219,20 @@ static int numopensockets;
SYSCTL_INT(_kern_ipc, OID_AUTO, numopensockets, CTLFLAG_RD,
&numopensockets, 0, "Number of open sockets");
#ifdef ZERO_COPY_SOCKETS
/* These aren't static because they're used in other files. */
int so_zero_copy_send = 1;
int so_zero_copy_receive = 1;
#if defined(SOCKET_SEND_COW) || defined(SOCKET_RECV_PFLIP)
SYSCTL_NODE(_kern_ipc, OID_AUTO, zero_copy, CTLFLAG_RD, 0,
"Zero copy controls");
#ifdef SOCKET_RECV_PFLIP
int so_zero_copy_receive = 1;
SYSCTL_INT(_kern_ipc_zero_copy, OID_AUTO, receive, CTLFLAG_RW,
&so_zero_copy_receive, 0, "Enable zero copy receive");
#endif
#ifdef SOCKET_SEND_COW
int so_zero_copy_send = 1;
SYSCTL_INT(_kern_ipc_zero_copy, OID_AUTO, send, CTLFLAG_RW,
&so_zero_copy_send, 0, "Enable zero copy send");
#endif /* ZERO_COPY_SOCKETS */
#endif /* SOCKET_SEND_COW */
#endif /* SOCKET_SEND_COW || SOCKET_RECV_PFLIP */
/*
* accept_mtx locks down per-socket fields relating to accept queues. See
@ -903,7 +906,7 @@ sodisconnect(struct socket *so)
return (error);
}
#ifdef ZERO_COPY_SOCKETS
#ifdef SOCKET_SEND_COW
struct so_zerocopy_stats{
int size_ok;
int align_ok;
@ -1008,7 +1011,7 @@ out:
*retmp = top;
return (error);
}
#endif /* ZERO_COPY_SOCKETS */
#endif /* SOCKET_SEND_COW */
#define SBLOCKWAIT(f) (((f) & MSG_DONTWAIT) ? 0 : SBL_WAIT)
@ -1019,7 +1022,7 @@ sosend_dgram(struct socket *so, struct sockaddr *addr, struct uio *uio,
long space;
ssize_t resid;
int clen = 0, error, dontroute;
#ifdef ZERO_COPY_SOCKETS
#ifdef SOCKET_SEND_COW
int atomic = sosendallatonce(so) || top;
#endif
@ -1104,7 +1107,7 @@ sosend_dgram(struct socket *so, struct sockaddr *addr, struct uio *uio,
if (flags & MSG_EOR)
top->m_flags |= M_EOR;
} else {
#ifdef ZERO_COPY_SOCKETS
#ifdef SOCKET_SEND_COW
error = sosend_copyin(uio, &top, atomic, &space, flags);
if (error)
goto out;
@ -1121,7 +1124,7 @@ sosend_dgram(struct socket *so, struct sockaddr *addr, struct uio *uio,
goto out;
}
space -= resid - uio->uio_resid;
#endif
#endif /* SOCKET_SEND_COW */
resid = uio->uio_resid;
}
KASSERT(resid == 0, ("sosend_dgram: resid != 0"));
@ -1293,7 +1296,7 @@ restart:
if (flags & MSG_EOR)
top->m_flags |= M_EOR;
} else {
#ifdef ZERO_COPY_SOCKETS
#ifdef SOCKET_SEND_COW
error = sosend_copyin(uio, &top, atomic,
&space, flags);
if (error != 0)
@ -1313,7 +1316,7 @@ restart:
goto release;
}
space -= resid - uio->uio_resid;
#endif
#endif /* SOCKET_SEND_COW */
resid = uio->uio_resid;
}
if (dontroute) {
@ -1405,7 +1408,7 @@ soreceive_rcvoob(struct socket *so, struct uio *uio, int flags)
if (error)
goto bad;
do {
#ifdef ZERO_COPY_SOCKETS
#ifdef SOCKET_RECV_PFLIP
if (so_zero_copy_receive) {
int disposable;
@ -1419,7 +1422,7 @@ soreceive_rcvoob(struct socket *so, struct uio *uio, int flags)
min(uio->uio_resid, m->m_len),
uio, disposable);
} else
#endif /* ZERO_COPY_SOCKETS */
#endif /* SOCKET_RECV_PFLIP */
error = uiomove(mtod(m, void *),
(int) min(uio->uio_resid, m->m_len), uio);
m = m_free(m);
@ -1743,7 +1746,7 @@ dontblock:
SBLASTRECORDCHK(&so->so_rcv);
SBLASTMBUFCHK(&so->so_rcv);
SOCKBUF_UNLOCK(&so->so_rcv);
#ifdef ZERO_COPY_SOCKETS
#ifdef SOCKET_RECV_PFLIP
if (so_zero_copy_receive) {
int disposable;
@ -1757,7 +1760,7 @@ dontblock:
(int)len, uio,
disposable);
} else
#endif /* ZERO_COPY_SOCKETS */
#endif /* SOCKET_RECV_PFLIP */
error = uiomove(mtod(m, char *) + moff, (int)len, uio);
SOCKBUF_LOCK(&so->so_rcv);
if (error) {