diff options
author | jhb <jhb@FreeBSD.org> | 2009-06-01 21:17:03 +0000 |
---|---|---|
committer | jhb <jhb@FreeBSD.org> | 2009-06-01 21:17:03 +0000 |
commit | a1af9ecca44f362b24fe3a8342ca6ed8676a399c (patch) | |
tree | 13628b6be10af95db7dc7d8ef88b3291d48583ab /sys/kern/uipc_socket.c | |
parent | 9956d85f164d16d3c1db67cc01f521c1c09d5fdb (diff) | |
download | FreeBSD-src-a1af9ecca44f362b24fe3a8342ca6ed8676a399c.zip FreeBSD-src-a1af9ecca44f362b24fe3a8342ca6ed8676a399c.tar.gz |
Rework socket upcalls to close some races with setup/teardown of upcalls.
- Each socket upcall is now invoked with the appropriate socket buffer
locked. It is not permissible to call soisconnected() with this lock
held; however, so socket upcalls now return an integer value. The two
possible values are SU_OK and SU_ISCONNECTED. If an upcall returns
SU_ISCONNECTED, then the soisconnected() will be invoked on the
socket after the socket buffer lock is dropped.
- A new API is provided for setting and clearing socket upcalls. The
API consists of soupcall_set() and soupcall_clear().
- To simplify locking, each socket buffer now has a separate upcall.
- When a socket upcall returns SU_ISCONNECTED, the upcall is cleared from
the receive socket buffer automatically. Note that a SO_SND upcall
should never return SU_ISCONNECTED.
- All this means that accept filters should now return SU_ISCONNECTED
instead of calling soisconnected() directly. They also no longer need
to explicitly clear the upcall on the new socket.
- The HTTP accept filter still uses soupcall_set() to manage its internal
state machine, but other accept filters no longer have any explicit
knowlege of socket upcall internals aside from their return value.
- The various RPC client upcalls currently drop the socket buffer lock
while invoking soreceive() as a temporary band-aid. The plan for
the future is to add a new flag to allow soreceive() to be called with
the socket buffer locked.
- The AIO callback for socket I/O is now also invoked with the socket
buffer locked. Previously sowakeup() would drop the socket buffer
lock only to call aio_swake() which immediately re-acquired the socket
buffer lock for the duration of the function call.
Discussed with: rwatson, rmacklem
Diffstat (limited to 'sys/kern/uipc_socket.c')
-rw-r--r-- | sys/kern/uipc_socket.c | 69 |
1 files changed, 63 insertions, 6 deletions
diff --git a/sys/kern/uipc_socket.c b/sys/kern/uipc_socket.c index 0992b3f..80f9a55 100644 --- a/sys/kern/uipc_socket.c +++ b/sys/kern/uipc_socket.c @@ -3054,8 +3054,10 @@ soisconnecting(struct socket *so) void soisconnected(struct socket *so) { - struct socket *head; + struct socket *head; + int ret; +restart: ACCEPT_LOCK(); SOCK_LOCK(so); so->so_state &= ~(SS_ISCONNECTING|SS_ISDISCONNECTING|SS_ISCONFIRMING); @@ -3075,13 +3077,17 @@ soisconnected(struct socket *so) wakeup_one(&head->so_timeo); } else { ACCEPT_UNLOCK(); - so->so_upcall = - head->so_accf->so_accept_filter->accf_callback; - so->so_upcallarg = head->so_accf->so_accept_filter_arg; - so->so_rcv.sb_flags |= SB_UPCALL; + soupcall_set(so, SO_RCV, + head->so_accf->so_accept_filter->accf_callback, + head->so_accf->so_accept_filter_arg); so->so_options &= ~SO_ACCEPTFILTER; + ret = head->so_accf->so_accept_filter->accf_callback(so, + head->so_accf->so_accept_filter_arg, M_DONTWAIT); + if (ret == SU_ISCONNECTED) + soupcall_clear(so, SO_RCV); SOCK_UNLOCK(so); - so->so_upcall(so, so->so_upcallarg, M_DONTWAIT); + if (ret == SU_ISCONNECTED) + goto restart; } return; } @@ -3146,6 +3152,57 @@ sodupsockaddr(const struct sockaddr *sa, int mflags) } /* + * Register per-socket buffer upcalls. + */ +void +soupcall_set(struct socket *so, int which, + int (*func)(struct socket *, void *, int), void *arg) +{ + struct sockbuf *sb; + + switch (which) { + case SO_RCV: + sb = &so->so_rcv; + break; + case SO_SND: + sb = &so->so_snd; + break; + default: + panic("soupcall_set: bad which"); + } + SOCKBUF_LOCK_ASSERT(sb); +#if 0 + /* XXX: accf_http actually wants to do this on purpose. */ + KASSERT(sb->sb_upcall == NULL, ("soupcall_set: overwriting upcall")); +#endif + sb->sb_upcall = func; + sb->sb_upcallarg = arg; + sb->sb_flags |= SB_UPCALL; +} + +void +soupcall_clear(struct socket *so, int which) +{ + struct sockbuf *sb; + + switch (which) { + case SO_RCV: + sb = &so->so_rcv; + break; + case SO_SND: + sb = &so->so_snd; + break; + default: + panic("soupcall_clear: bad which"); + } + SOCKBUF_LOCK_ASSERT(sb); + KASSERT(sb->sb_upcall != NULL, ("soupcall_clear: no upcall to clear")); + sb->sb_upcall = NULL; + sb->sb_upcallarg = NULL; + sb->sb_flags &= ~SB_UPCALL; +} + +/* * Create an external-format (``xsocket'') structure using the information in * the kernel-format socket structure pointed to by so. This is done to * reduce the spew of irrelevant information over this interface, to isolate |