Commit Graph

239 Commits

Author SHA1 Message Date
Julian Elischer
079b7badea Pre-KSE/M3 commit.
this is a low-functionality change that changes the kernel to access the main
thread of a process via the linked list of threads rather than
assuming that it is embedded in the process. It IS still embeded there
but remove all teh code that assumes that in preparation for the next commit
which will actually move it out.

Reviewed by: peter@freebsd.org, gallatin@cs.duke.edu, benno rice,
2002-02-07 20:58:47 +00:00
Archie Cobbs
a3e232d65b The hardware checksum code makes the assumption that a packet routed out
a particular Ethernet interface will actually be delivered by (only) that
device driver. This is not necessarily true when ng_ether(4) is used.

To word around this, while a ng_ether(4)'s "upper" hook is connected,
turn off all hardware checksum, fragmentation, etc., features for that
interface.

PR:		kern/31586
MFC after:	1 week
2002-02-05 18:27:30 +00:00
Matthew Dillon
ecde8f7c29 Get rid of the twisted MFREE() macro entirely.
Reviewed by:	dg, bmilekic
MFC after:	3 days
2002-02-05 02:00:56 +00:00
Archie Cobbs
d9bfecab53 Some netgraph parse types (such as for the 'value' field in ng_ksocket's
'struct ng_ksocket_sockopt') like to peek into the ng_mesg header for
information. Make sure when generating default values that we provide
a valid header to peek into.

MFC after:	1 week
2002-02-01 02:21:41 +00:00
Alfred Perlstein
a4db49537b Replace ffind_* with fget calls.
Make fget MPsafe.

Make fgetvp and fgetsock use the fget subsystem to reduce code bloat.

Push giant down in fpathconf().
2002-01-14 00:13:45 +00:00
Alfred Perlstein
426da3bcfb SMP Lock struct file, filedesc and the global file list.
Seigo Tanimura (tanimura) posted the initial delta.

I've polished it quite a bit reducing the need for locking and
adapting it for KSE.

Locks:

1 mutex in each filedesc
   protects all the fields.
   protects "struct file" initialization, while a struct file
     is being changed from &badfileops -> &pipeops or something
     the filedesc should be locked.

1 mutex in each struct file
   protects the refcount fields.
   doesn't protect anything else.
   the flags used for garbage collection have been moved to
     f_gcflag which was the FILLER short, this doesn't need
     locking because the garbage collection is a single threaded
     container.
  could likely be made to use a pool mutex.

1 sx lock for the global filelist.

struct file *	fhold(struct file *fp);
        /* increments reference count on a file */

struct file *	fhold_locked(struct file *fp);
        /* like fhold but expects file to locked */

struct file *	ffind_hold(struct thread *, int fd);
        /* finds the struct file in thread, adds one reference and
                returns it unlocked */

struct file *	ffind_lock(struct thread *, int fd);
        /* ffind_hold, but returns file locked */

I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
Mike Smith
b78929a257 Staticise the socket list. 2002-01-08 10:30:34 +00:00
Archie Cobbs
dc9c2e0149 Avoid reentrantly sending on the same socket, which causes a kernel panic. 2002-01-06 01:08:30 +00:00
Robert Watson
9c4d63da6d o Make the credential used by socreate() an explicit argument to
socreate(), rather than getting it implicitly from the thread
  argument.

o Make NFS cache the credential provided at mount-time, and use
  the cached credential (nfsmount->nm_cred) when making calls to
  socreate() on initially connecting, or reconnecting the socket.

This fixes bugs involving NFS over TCP and ipfw uid/gid rules, as well
as bugs involving NFS and mandatory access control implementations.

Reviewed by:	freebsd-arch
2001-12-31 17:45:16 +00:00
Archie Cobbs
b9bc94b33f Typo. 2001-12-15 20:53:15 +00:00
Archie Cobbs
6c12c2b195 Don't free a structure containing a 'struct callout' structure while that
callout is still pending.

MFC after:	3 days
2001-12-15 20:48:53 +00:00
Archie Cobbs
34fd23818a Add support for 56 bit MPPE encryption.
MFC after:	3 days
2001-12-15 02:07:32 +00:00
David E. O'Brien
6e551fb628 Update to C99, s/__FUNCTION__/__func__/,
also don't use ANSI string concatenation.
2001-12-10 08:09:49 +00:00
Archie Cobbs
19ff9e5f3e When a socket is not connected, allow the peer "struct sockaddr"
to be included in the meta information that is associated with
incoming and outgoing packets.

Reviewed by:	julian
MFC after:	1 week
2001-11-28 19:39:58 +00:00
Julian Elischer
10d7ccab88 A node that allows ethernet type packets to be filtered to different
hooks depending on ethertype. Great for prototyping protocols.
connects to the lower and upper hooks of an ethernet type of node.

Obtained from: Monzoon Networks.
	Thanks to Andre Oppermann, May 2001.
2001-10-30 07:28:17 +00:00
Matthew Dillon
e74e9032da log routine called w/ %ld and int argument, cast argument to long 2001-10-29 02:22:49 +00:00
Archie Cobbs
6628011155 Fix bug that cheated hook names out of the last 2 bytes.
MFC after:      3 days
2001-10-10 19:58:11 +00:00
Archie Cobbs
129bc89568 Let "raw" mean IPPROTO_RAW instead of IPPROTO_IP.
Noticed by:	jdp
MFC after:	3 days
2001-10-10 19:51:13 +00:00
David Malone
2bc21ed985 Hopefully improve control message passing over Unix domain sockets.
1) Allow the sending of more than one control message at a time
over a unix domain socket. This should cover the PR 29499.

2) This requires that unp_{ex,in}ternalize and unp_scan understand
mbufs with more than one control message at a time.

3) Internalize and externalize used to work on the mbuf in-place.
This made life quite complicated and the code for sizeof(int) <
sizeof(file *) could end up doing the wrong thing. The patch always
create a new mbuf/cluster now. This resulted in the change of the
prototype for the domain externalise function.

4) You can now send SCM_TIMESTAMP messages.

5) Always use CMSG_DATA(cm) to determine the start where the data
in unp_{ex,in}ternalize. It was using ((struct cmsghdr *)cm + 1)
in some places, which gives the wrong alignment on the alpha.
(NetBSD made this fix some time ago).

This results in an ABI change for discriptor passing and creds
passing on the alpha. (Probably on the IA64 and Spare ports too).

6) Fix userland programs to use CMSG_* macros too.

7) Be more careful about freeing mbufs containing (file *)s.
This is made possible by the prototype change of externalise.

PR:		29499
MFC after:	6 weeks
2001-10-04 13:11:48 +00:00
Bruce Evans
4c6bb41ef2 Fixed pedantic syntax error (trailing semicolon in enum). 2001-10-04 07:51:42 +00:00
Brooks Davis
7360079ab3 Remove a couple unintentional mentions of Ethernet that crept in from
ng_ether.c.
2001-09-28 00:02:50 +00:00
Brooks Davis
c2eed10556 Add ng_ip_input. A new netgraph node for queuing IP packets into the
main IP input processing code.
2001-09-27 21:54:27 +00:00
Brooks Davis
cf776d8152 The initial commit contained an error in the license, this is the
correct one.
2001-09-27 00:04:29 +00:00
Brooks Davis
94408d94c3 /home/brooks/ng_gif.message 2001-09-26 23:50:17 +00:00
Julian Elischer
b40ce4165d KSE Milestone 2
Note ALL MODULES MUST BE RECOMPILED
make the kernel aware that there are smaller units of scheduling than the
process. (but only allow one thread per process at this time).
This is functionally equivalent to teh previousl -current except
that there is a thread associated with each process.

Sorry john! (your next MFC will be a doosie!)

Reviewed by: peter@freebsd.org, dillon@freebsd.org

X-MFC after:    ha ha ha ha
2001-09-12 08:38:13 +00:00
Julian Elischer
f97e0a0719 First pass at porting John's "accept" changes to
allow an in-kernel webserver (or similar) to accept
and handle incoming connections using netgraph without ever leaving the
kernel. (allows incoming tunnel requests to be
handled totally within the kernel for example)

Needs work, but shouldn't break existing functionality.

Submitted by:	John Polstra <jdp@polstra.com>
MFC after:	2 weeks
2001-09-07 07:12:51 +00:00
Julian Elischer
94142c49dc MFS: change name of sysctl to something more diplomatic. 2001-09-04 06:29:35 +00:00
Archie Cobbs
cf2010b81f Fix kernel crash when VLANs are combined with ng_ether(4), by attaching
interfaces of type IFT_L2VLAN as well as IFT_ETHER during module load.

Submitted by:	yar
2001-08-30 19:09:10 +00:00
Brian Somers
376958b412 Fix two typos 2001-08-21 13:20:02 +00:00
Brian Somers
2b5dcd2ffe Pack struct uniqtag declarations to stop our data field from being pushed
4 bytes to the right on the alpha.

Tested by:	Thomas Pornin <Thomas.Pornin@ens.fr>
MFC after:	1 week
2001-08-02 09:28:31 +00:00
Brian Somers
9088fa05ae If an attempt is made to LISTEN for a service tag that's already being
LISTENed for, return EEXISTS.

Only match the magic "*" service tag if no other LISTEN service tags
match.

Require an explicit LISTEN for an empty service tag in order to match
empty service requests.

Approved by:	julian
MFC after:	3 days
2001-07-25 03:34:07 +00:00
Brooks Davis
bb5c977fbd General cleanup providing better style(9) conformance and generally
improved readability.  The two real functional changes are that
netgraph now sees this as the "split" node type rather then the
"ng_split" node type and that meta data is passed through without
processing rather then being dropped.

Reviewed by:	jhb, julian
MFC after:	7 weeks
2001-07-24 23:33:06 +00:00
Julian Elischer
c31b4a5381 Add an external function to unlink a netgraph type from the types list. 2001-07-23 21:14:57 +00:00
Julian Elischer
8b68f82f72 remember to set the return address in a message when
sending it along a hook.

PR: 27906
Submitted by: Harti Brandt <brandt@fokus.gmd.de>

no MFC as code is different in 4.x
2001-07-14 05:51:33 +00:00
Peter Wemm
bc66c1fd1f Fix warning: (int/long mixup in printf)
475: warning: unsigned int format, long unsigned int arg (arg 3)
2001-06-15 07:35:25 +00:00
John Polstra
a514569e9a Fix a range checking bug in ng_int32_parse which affected 64-bit
machines.  The code formerly read:

    long val;
    if (val < (long)-0x80000000 || ...)
            return EINVAL;

The constant 0x80000000 has type unsigned int.  The unary `-'
operator does not change the type (or the value, in this case).
Therefore the promotion to long is done by 0-extension, giving
0x0000000080000000 instead of the desired 0xffffffff80000000.  I
got rid of the `-' and changed the cast to (int32_t) to give proper
sign-extension on all architectures and to better reflect the fact
that we are range-checking a 32-bit value.

This commit also makes the analogous changes to ng_int{8,16}_parse
for consistency.

MFC after:	3 days
2001-05-19 19:36:32 +00:00
John Baldwin
f5d325c599 Remove unneeded includes in the i386 case. 2001-05-15 23:16:18 +00:00
Archie Cobbs
422c727634 Don't reference a node after we dropped a reference to it
(same as in previous checkin, but in a different function).
2001-04-11 22:04:47 +00:00
John Baldwin
f34fa851e0 Catch up to header include changes:
- <sys/mutex.h> now requires <sys/systm.h>
- <sys/mutex.h> and <sys/sx.h> now require <sys/lock.h>
2001-03-28 09:17:56 +00:00
Julian Elischer
5951069a87 netgraph.h:
Change a prototype.
  Add a function version of ng_ref_node() when debugging so
  a breakpoint can be set on it.
ng_base.c:
  add 'node' as an argument to ng_apply_item so that it is up
  to the caller to take over and release the item's reference on
  the node. If the release reports back that the node went away
  due to the reference going to 0, the caller should cease referencing
  the now defunct node. (e.g. the item was a 'kill node' message).
  Alter ng_unref_node to report back the residual references as a result.
ng_pptpgre.c:
  Don't reference a node after we dropped a reference to it.
  (What if it was the last?)
Fixes a node leak reported by Harti Brandt <brandt@fokus.gmd.de>
 which was due to an incorrect earlier attempt to fix the
 "accessing node after dropping the last reference" problem.
2001-03-10 16:31:00 +00:00
Archie Cobbs
22dfb9bdb7 Fix potential crash caused by packets with bogus ACK's.
Reported by:	Fabien THOMAS <fabient@netasq.com>
2001-03-08 20:10:02 +00:00
Julian Elischer
8714210340 Cleanups to Macros for sending data between netgraph nodes. 2001-03-03 05:52:49 +00:00
Julian Elischer
e7af56e53e Add parenthesis to a macro.
This took me 2 whole days to track down. (bleah)
2001-03-03 05:50:47 +00:00
Julian Elischer
a96dcd84d2 Shuffle netgraph mutexes a bit and hold a reference on a node
from the function that is calling the destructor.
2001-02-28 18:49:09 +00:00
Julian Elischer
e08d3e3c33 Allow a changed MAC address to show up in ifconfig by changing it
in the ifaddr list as well. Also change an error return in the base system.
2001-02-26 09:31:54 +00:00
Julian Elischer
950809d7ac slight cleanups during testing. 2001-02-25 16:49:04 +00:00
Julian Elischer
2b2c95c996 Add a node that looks to all the word like an ethernet but delivers its
ehternet frames to a netgraph  hook.

Submitted by: "Vitaly V. Belekhov" <vitaly@riss-telecom.ru>
translated to 5.0 by me. man page not yet written.

This node still needs a little work.. don't use yet. Not yet linked into
the build.
2001-02-25 05:46:52 +00:00
Julian Elischer
e519ede348 Make the sample netgraph node compileable again.
Makes it easier for people if they can start with something
that actually compiles.
2001-02-25 05:36:25 +00:00
Julian Elischer
33338e7370 Add knowledge of the netgraph spinlocks into the Witness code.
Well, at least I think that's how it's done.
2001-02-24 14:29:47 +00:00
Julian Elischer
bfa7e882d1 Shuffle sysctls a bit (thankyou whoever made them dynamic for modules)
and add a sysctl to pppoe to activate non standard ethertypes
so that idiot ISPs (apparently in France) who use
equipment from idiot suppliers (rumour says 3com)
who use nonstandard ethertypes can still connect.

 "yep, sure we do pppoe, we use a different identifier to that dictated in
 the standard, but sure it's pppoe!"

sysctl -w net.graph.stupid_isp=1 enables the changeover.
2001-02-23 16:34:22 +00:00