mirror/freebsd
1
0
Fork 0
mirror of https://git.FreeBSD.org/src.git synced 2024-12-27 11:55:06 +00:00
Code Issues Releases Activity
00be90b7c5
freebsd/sys/netinet/accf_http.c

350 lines
8.4 KiB
C
Raw Normal View History

/* -> /*- for license, minor formatting changes
2005-01-07 01:45:51 +00:00
/*-
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
* Copyright (c) 2000 Paycounter, Inc.
* Author: Alfred Perlstein <alfred@paycounter.com>, <alfred@FreeBSD.org>
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
Add FBSDID to all files in netinet so that people can more easily include file version information in bug reports. Approved by: re (kensmith)
2007-10-07 20:44:24 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
#define ACCEPT_FILTER_MOD
#include <sys/param.h>
#include <sys/kernel.h>
Revert the change of #includes in sys/filedesc.h and sys/socketvar.h. Requested by: bde Since locking sigio_lock is usually followed by calling pgsigio(), move the declaration of sigio_lock and the definitions of SIGIO_*() to sys/signalvar.h. While I am here, sort include files alphabetically, where possible.
2002-04-30 01:54:54 +00:00
#include <sys/mbuf.h>
add missing #include <sys/module.h>
2004-05-30 20:27:19 +00:00
#include <sys/module.h>
Revert the change of #includes in sys/filedesc.h and sys/socketvar.h. Requested by: bde Since locking sigio_lock is usually followed by calling pgsigio(), move the declaration of sigio_lock and the definitions of SIGIO_*() to sys/signalvar.h. While I am here, sort include files alphabetically, where possible.
2002-04-30 01:54:54 +00:00
#include <sys/signalvar.h>
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
#include <sys/sysctl.h>
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
#include <sys/socketvar.h>
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
/* check for GET/HEAD */
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
2009-06-01 21:17:03 +00:00
static int sohashttpget(struct socket *so, void *arg, int waitflag);
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
/* check for HTTP/1.0 or HTTP/1.1 */
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
2009-06-01 21:17:03 +00:00
static int soparsehttpvers(struct socket *so, void *arg, int waitflag);
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
/* check for end of HTTP/1.x request */
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
2009-06-01 21:17:03 +00:00
static int soishttpconnected(struct socket *so, void *arg, int waitflag);
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
/* strcmp on an mbuf chain */
static int mbufstrcmp(struct mbuf *m, struct mbuf *npkt, int offset, char *cmp);
/* strncmp on an mbuf chain */
static int mbufstrncmp(struct mbuf *m, struct mbuf *npkt, int offset,
int max, char *cmp);
/* socketbuffer is full */
static int sbfull(struct sockbuf *sb);
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
static struct accept_filter accf_http_filter = {
"httpready",
sohashttpget,
NULL,
NULL
};
static moduledata_t accf_http_mod = {
"accf_http",
accept_filt_generic_mod_event,
&accf_http_filter
};
DECLARE_MODULE(accf_http, accf_http_mod, SI_SUB_DRIVERS, SI_ORDER_MIDDLE);
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
static int parse_http_version = 1;
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
Mark all SYSCTL_NODEs static that have no corresponding SYSCTL_DECLs. The SYSCTL_NODE macro defines a list that stores all child-elements of that node. If there's no SYSCTL_DECL macro anywhere else, there's no reason why it shouldn't be static.
2011-11-07 15:43:11 +00:00
static SYSCTL_NODE(_net_inet_accf, OID_AUTO, http, CTLFLAG_RW, 0,
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
"HTTP accept filter");
SYSCTL_INT(_net_inet_accf_http, OID_AUTO, parsehttpversion, CTLFLAG_RW,
&parse_http_version, 1,
"Parse http version so that non 1.x requests work");
#ifdef ACCF_HTTP_DEBUG
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
#define DPRINT(fmt, args...) \
do { \
printf("%s:%d: " fmt "\n", __func__, __LINE__, ##args); \
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
} while (0)
#else
#define DPRINT(fmt, args...)
#endif
static int
sbfull(struct sockbuf *sb)
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
{
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
DPRINT("sbfull, cc(%ld) >= hiwat(%ld): %d, "
"mbcnt(%ld) >= mbmax(%ld): %d",
sb->sb_cc, sb->sb_hiwat, sb->sb_cc >= sb->sb_hiwat,
sb->sb_mbcnt, sb->sb_mbmax, sb->sb_mbcnt >= sb->sb_mbmax);
return (sb->sb_cc >= sb->sb_hiwat || sb->sb_mbcnt >= sb->sb_mbmax);
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
}
/*
* start at mbuf m, (must provide npkt if exists)
* starting at offset in m compare characters in mbuf chain for 'cmp'
*/
static int
mbufstrcmp(struct mbuf *m, struct mbuf *npkt, int offset, char *cmp)
{
struct mbuf *n;
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
for (; m != NULL; m = n) {
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
n = npkt;
if (npkt)
npkt = npkt->m_nextpkt;
for (; m; m = m->m_next) {
for (; offset < m->m_len; offset++, cmp++) {
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
if (*cmp == '\0')
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
return (1);
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
else if (*cmp != *(mtod(m, char *) + offset))
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
return (0);
}
Fix some edge cases where bad string handling could occur. Submitted by: ps
2002-05-01 08:29:41 +00:00
if (*cmp == '\0')
return (1);
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
offset = 0;
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
}
}
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
return (0);
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
}
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
/*
* start at mbuf m, (must provide npkt if exists)
* starting at offset in m compare characters in mbuf chain for 'cmp'
* stop at 'max' characters
*/
static int
mbufstrncmp(struct mbuf *m, struct mbuf *npkt, int offset, int max, char *cmp)
{
struct mbuf *n;
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
for (; m != NULL; m = n) {
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
n = npkt;
if (npkt)
npkt = npkt->m_nextpkt;
for (; m; m = m->m_next) {
for (; offset < m->m_len; offset++, cmp++, max--) {
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
if (max == 0 || *cmp == '\0')
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
return (1);
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
else if (*cmp != *(mtod(m, char *) + offset))
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
return (0);
}
Fix some edge cases where bad string handling could occur. Submitted by: ps
2002-05-01 08:29:41 +00:00
if (max == 0 || *cmp == '\0')
return (1);
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
offset = 0;
}
}
return (0);
}
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
#define STRSETUP(sptr, slen, str) \
do { \
sptr = str; \
slen = sizeof(str) - 1; \
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
} while(0)
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
2009-06-01 21:17:03 +00:00
static int
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
sohashttpget(struct socket *so, void *arg, int waitflag)
{
The socket field so_state is used to hold a variety of socket related flags relating to several aspects of socket functionality. This change breaks out several bits relating to send and receive operation into a new per-socket buffer field, sb_state, in order to facilitate locking. This is required because, in order to provide more granular locking of sockets, different state fields have different locking properties. The following fields are moved to sb_state: SS_CANTRCVMORE (so_state) SS_CANTSENDMORE (so_state) SS_RCVATMARK (so_state) Rename respectively to: SBS_CANTRCVMORE (so_rcv.sb_state) SBS_CANTSENDMORE (so_snd.sb_state) SBS_RCVATMARK (so_rcv.sb_state) This facilitates locking by isolating fields to be located with other identically locked fields, and permits greater granularity in socket locking by avoiding storing fields with different locking semantics in the same short (avoiding locking conflicts). In the future, we may wish to coallesce sb_state and sb_flags; for the time being I leave them separate and there is no additional memory overhead due to the packing/alignment of shorts in the socket buffer structure.
2004-06-14 18:16:22 +00:00
if ((so->so_rcv.sb_state & SBS_CANTRCVMORE) == 0 && !sbfull(&so->so_rcv)) {
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
struct mbuf *m;
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
char *cmp;
int cmplen, cc;
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
m = so->so_rcv.sb_mb;
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
cc = so->so_rcv.sb_cc - 1;
if (cc < 1)
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
2009-06-01 21:17:03 +00:00
return (SU_OK);
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
switch (*mtod(m, char *)) {
case 'G':
STRSETUP(cmp, cmplen, "ET ");
break;
case 'H':
STRSETUP(cmp, cmplen, "EAD ");
break;
default:
goto fallout;
}
if (cc < cmplen) {
if (mbufstrncmp(m, m->m_nextpkt, 1, cc, cmp) == 1) {
DPRINT("short cc (%d) but mbufstrncmp ok", cc);
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
2009-06-01 21:17:03 +00:00
return (SU_OK);
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
} else {
DPRINT("short cc (%d) mbufstrncmp failed", cc);
goto fallout;
}
}
if (mbufstrcmp(m, m->m_nextpkt, 1, cmp) == 1) {
DPRINT("mbufstrcmp ok");
if (parse_http_version == 0)
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
2009-06-01 21:17:03 +00:00
return (soishttpconnected(so, arg, waitflag));
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
else
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
2009-06-01 21:17:03 +00:00
return (soparsehttpvers(so, arg, waitflag));
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
}
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
DPRINT("mbufstrcmp bad");
Back out my lats commit of locking down a socket, it conflicts with hsu's work. Requested by: hsu
2002-05-31 11:52:35 +00:00
}
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
fallout:
DPRINT("fallout");
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
2009-06-01 21:17:03 +00:00
return (SU_ISCONNECTED);
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
}
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
2009-06-01 21:17:03 +00:00
static int
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
soparsehttpvers(struct socket *so, void *arg, int waitflag)
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
{
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
struct mbuf *m, *n;
int i, cc, spaces, inspaces;
The socket field so_state is used to hold a variety of socket related flags relating to several aspects of socket functionality. This change breaks out several bits relating to send and receive operation into a new per-socket buffer field, sb_state, in order to facilitate locking. This is required because, in order to provide more granular locking of sockets, different state fields have different locking properties. The following fields are moved to sb_state: SS_CANTRCVMORE (so_state) SS_CANTSENDMORE (so_state) SS_RCVATMARK (so_state) Rename respectively to: SBS_CANTRCVMORE (so_rcv.sb_state) SBS_CANTSENDMORE (so_snd.sb_state) SBS_RCVATMARK (so_rcv.sb_state) This facilitates locking by isolating fields to be located with other identically locked fields, and permits greater granularity in socket locking by avoiding storing fields with different locking semantics in the same short (avoiding locking conflicts). In the future, we may wish to coallesce sb_state and sb_flags; for the time being I leave them separate and there is no additional memory overhead due to the packing/alignment of shorts in the socket buffer structure.
2004-06-14 18:16:22 +00:00
if ((so->so_rcv.sb_state & SBS_CANTRCVMORE) != 0 || sbfull(&so->so_rcv))
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
goto fallout;
m = so->so_rcv.sb_mb;
cc = so->so_rcv.sb_cc;
inspaces = spaces = 0;
for (m = so->so_rcv.sb_mb; m; m = n) {
n = m->m_nextpkt;
for (; m; m = m->m_next) {
for (i = 0; i < m->m_len; i++, cc--) {
switch (*(mtod(m, char *) + i)) {
case ' ':
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
/* tabs? '\t' */
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
if (!inspaces) {
spaces++;
inspaces = 1;
}
break;
case '\r':
case '\n':
DPRINT("newline");
goto fallout;
default:
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
if (spaces != 2) {
inspaces = 0;
break;
}
/*
* if we don't have enough characters
* left (cc < sizeof("HTTP/1.0") - 1)
* then see if the remaining ones
* are a request we can parse.
*/
if (cc < sizeof("HTTP/1.0") - 1) {
if (mbufstrncmp(m, n, i, cc,
"HTTP/1.") == 1) {
DPRINT("ok");
goto readmore;
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
} else {
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
DPRINT("bad");
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
goto fallout;
}
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
} else if (
mbufstrcmp(m, n, i, "HTTP/1.0") ||
mbufstrcmp(m, n, i, "HTTP/1.1")) {
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
2009-06-01 21:17:03 +00:00
DPRINT("ok");
return (soishttpconnected(so,
arg, waitflag));
cleanup: fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
} else {
DPRINT("bad");
goto fallout;
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
}
}
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
}
}
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
}
readmore:
DPRINT("readmore");
/*
* if we hit here we haven't hit something
* we don't understand or a newline, so try again
*/
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
2009-06-01 21:17:03 +00:00
soupcall_set(so, SO_RCV, soparsehttpvers, arg);
return (SU_OK);
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
fallout:
DPRINT("fallout");
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
2009-06-01 21:17:03 +00:00
return (SU_ISCONNECTED);
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
}
#define NCHRS 3
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
2009-06-01 21:17:03 +00:00
static int
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
soishttpconnected(struct socket *so, void *arg, int waitflag)
{
char a, b, c;
struct mbuf *m, *n;
int ccleft, copied;
DPRINT("start");
The socket field so_state is used to hold a variety of socket related flags relating to several aspects of socket functionality. This change breaks out several bits relating to send and receive operation into a new per-socket buffer field, sb_state, in order to facilitate locking. This is required because, in order to provide more granular locking of sockets, different state fields have different locking properties. The following fields are moved to sb_state: SS_CANTRCVMORE (so_state) SS_CANTSENDMORE (so_state) SS_RCVATMARK (so_state) Rename respectively to: SBS_CANTRCVMORE (so_rcv.sb_state) SBS_CANTSENDMORE (so_snd.sb_state) SBS_RCVATMARK (so_rcv.sb_state) This facilitates locking by isolating fields to be located with other identically locked fields, and permits greater granularity in socket locking by avoiding storing fields with different locking semantics in the same short (avoiding locking conflicts). In the future, we may wish to coallesce sb_state and sb_flags; for the time being I leave them separate and there is no additional memory overhead due to the packing/alignment of shorts in the socket buffer structure.
2004-06-14 18:16:22 +00:00
if ((so->so_rcv.sb_state & SBS_CANTRCVMORE) != 0 || sbfull(&so->so_rcv))
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
goto gotit;
/*
* Walk the socketbuffer and copy the last NCHRS (3) into a, b, and c
* copied - how much we've copied so far
* ccleft - how many bytes remaining in the socketbuffer
* just loop over the mbufs subtracting from 'ccleft' until we only
* have NCHRS left
*/
copied = 0;
ccleft = so->so_rcv.sb_cc;
if (ccleft < NCHRS)
goto readmore;
a = b = c = '\0';
for (m = so->so_rcv.sb_mb; m; m = n) {
n = m->m_nextpkt;
for (; m; m = m->m_next) {
ccleft -= m->m_len;
if (ccleft <= NCHRS) {
char *src;
int tocopy;
tocopy = (NCHRS - ccleft) - copied;
src = mtod(m, char *) + (m->m_len - tocopy);
while (tocopy--) {
switch (copied++) {
case 0:
a = *src++;
break;
case 1:
b = *src++;
break;
case 2:
c = *src++;
break;
}
}
}
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
}
}
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
if (c == '\n' && (b == '\n' || (b == '\r' && a == '\n'))) {
/* we have all request headers */
goto gotit;
}
readmore:
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
2009-06-01 21:17:03 +00:00
soupcall_set(so, SO_RCV, soishttpconnected, arg);
return (SU_OK);
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
Accept filter maintainance Update copyrights. Introduce a new sysctl node: net.inet.accf Although acceptfilters need refcounting to be properly (safely) unloaded as a temporary hack allow them to be unloaded if the sysctl net.inet.accf.unloadable is set, this is really for developers who want to work on thier own filters. A near complete re-write of the accf_http filter: 1) Parse check if the request is HTTP/1.0 or HTTP/1.1 if not dump to the application. Because of the performance implications of this there is a sysctl 'net.inet.accf.http.parsehttpversion' that when set to non-zero parses the HTTP version. The default is to parse the version. 2) Check if a socket has filled and dump to the listener 3) optimize the way that mbuf boundries are handled using some voodoo 4) even though you'd expect accept filters to only be used on TCP connections that don't use m_nextpkt I've fixed the accept filter for socket connections that use this. This rewrite of accf_http should allow someone to use them and maintain full HTTP compliance as long as net.inet.accf.http.parsehttpversion is set.
2000-09-06 18:49:13 +00:00
gotit:
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
2009-06-01 21:17:03 +00:00
return (SU_ISCONNECTED);
return of the accept filter part II accept filters are now loadable as well as able to be compiled into the kernel. two accept filters are provided, one that returns sockets when data arrives the other when an http request is completed (doesn't work with 0.9 requests) Reviewed by: jmg
2000-06-20 01:09:23 +00:00
}
Reference in New Issue Copy Permalink