freebsd-skq/usr.sbin/ppp/tty.c

757 lines
20 KiB
C
Raw Normal View History

/*-
* Copyright (c) 1999 Brian Somers <brian@Awfulhak.org>
* 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.
*
1999-08-28 01:35:59 +00:00
* $FreeBSD$
*/
#include <sys/param.h>
#include <sys/un.h>
#if defined(__OpenBSD__) || defined(__NetBSD__)
#include <sys/ioctl.h>
#endif
#include <errno.h>
#include <fcntl.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sysexits.h>
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
#include <sys/uio.h>
#include <termios.h>
#include <ttyent.h>
#include <unistd.h>
2002-03-31 01:56:34 +00:00
#ifndef NONETGRAPH
#include <netgraph.h>
#include <netgraph/ng_async.h>
#include <netgraph/ng_message.h>
#include <netgraph/ng_ppp.h>
#include <netgraph/ng_tty.h>
2002-03-31 01:56:34 +00:00
#endif
#include "layer.h"
#include "defs.h"
#include "mbuf.h"
#include "log.h"
#include "timer.h"
#include "lqr.h"
#include "hdlc.h"
#include "throughput.h"
#include "fsm.h"
#include "lcp.h"
#include "ccp.h"
#include "link.h"
#include "async.h"
#include "descriptor.h"
#include "physical.h"
#include "mp.h"
#include "chat.h"
#include "auth.h"
#include "chap.h"
#include "cbcp.h"
#include "datalink.h"
#include "main.h"
#include "id.h"
#include "tty.h"
#if defined(__mac68k__) || defined(__macppc__)
#undef CRTS_IFLOW
#undef CCTS_OFLOW
#define CRTS_IFLOW CDTRCTS
#define CCTS_OFLOW CDTRCTS
#endif
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
#define Online(dev) ((dev)->mbits & TIOCM_CD)
struct ttydevice {
struct device dev; /* What struct physical knows about */
struct pppTimer Timer; /* CD checks */
int mbits; /* Current DCD status */
int carrier_seconds; /* seconds before CD is *required* */
#ifndef NONETGRAPH
struct {
int speed; /* Pre-line-discipline speed */
int fd; /* Pre-line-discipline fd */
int disc; /* Old line-discipline */
} real;
char hook[sizeof NG_ASYNC_HOOK_SYNC]; /* our ng_socket hook */
int cs; /* A netgraph control socket (maybe) */
#endif
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
struct termios ios; /* To be able to reset from raw mode */
};
#define device2tty(d) ((d)->type == TTY_DEVICE ? (struct ttydevice *)d : NULL)
int
tty_DeviceSize(void)
{
return sizeof(struct ttydevice);
}
/*
* tty_Timeout() watches the DCD signal and mentions it if it's status
* changes.
*/
static void
tty_Timeout(void *data)
{
struct physical *p = data;
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
struct ttydevice *dev = device2tty(p->handler);
int ombits, change;
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
timer_Stop(&dev->Timer);
dev->Timer.load = SECTICKS; /* Once a second please */
timer_Start(&dev->Timer);
ombits = dev->mbits;
if (p->fd >= 0) {
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
if (ioctl(p->fd, TIOCMGET, &dev->mbits) < 0) {
/* we must be a pty ? */
if (p->cfg.cd.necessity != CD_DEFAULT)
log_Printf(LogWARN, "%s: Carrier ioctl not supported, "
"using ``set cd off''\n", p->link.name);
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
timer_Stop(&dev->Timer);
1999-12-13 18:45:21 +00:00
dev->mbits = TIOCM_CD;
return;
}
} else
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
dev->mbits = 0;
if (ombits == -1) {
/* First time looking for carrier */
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
if (Online(dev))
log_Printf(LogPHASE, "%s: %s: CD detected\n", p->link.name, p->name.full);
else if (++dev->carrier_seconds >= dev->dev.cd.delay) {
if (dev->dev.cd.necessity == CD_REQUIRED)
log_Printf(LogPHASE, "%s: %s: Required CD not detected\n",
p->link.name, p->name.full);
else {
log_Printf(LogPHASE, "%s: %s doesn't support CD\n",
p->link.name, p->name.full);
dev->mbits = TIOCM_CD; /* Dodgy null-modem cable ? */
}
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
timer_Stop(&dev->Timer);
/* tty_AwaitCarrier() will notice */
} else {
/* Keep waiting */
log_Printf(LogDEBUG, "%s: %s: Still no carrier (%d/%d)\n",
p->link.name, p->name.full, dev->carrier_seconds,
dev->dev.cd.delay);
dev->mbits = -1;
}
} else {
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
change = ombits ^ dev->mbits;
if (change & TIOCM_CD) {
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
if (dev->mbits & TIOCM_CD)
log_Printf(LogDEBUG, "%s: offline -> online\n", p->link.name);
else {
log_Printf(LogDEBUG, "%s: online -> offline\n", p->link.name);
log_Printf(LogPHASE, "%s: Carrier lost\n", p->link.name);
datalink_Down(p->dl, CLOSE_NORMAL);
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
timer_Stop(&dev->Timer);
}
} else
log_Printf(LogDEBUG, "%s: Still %sline\n", p->link.name,
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
Online(dev) ? "on" : "off");
}
}
static void
tty_StartTimer(struct physical *p)
{
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
struct ttydevice *dev = device2tty(p->handler);
timer_Stop(&dev->Timer);
dev->Timer.load = SECTICKS;
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
dev->Timer.func = tty_Timeout;
dev->Timer.name = "tty CD";
dev->Timer.arg = p;
log_Printf(LogDEBUG, "%s: Using tty_Timeout [%p]\n",
p->link.name, tty_Timeout);
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
timer_Start(&dev->Timer);
}
static int
tty_AwaitCarrier(struct physical *p)
{
struct ttydevice *dev = device2tty(p->handler);
if (dev->dev.cd.necessity == CD_NOTREQUIRED || physical_IsSync(p))
return CARRIER_OK;
if (dev->mbits == -1) {
if (dev->Timer.state == TIMER_STOPPED) {
dev->carrier_seconds = 0;
tty_StartTimer(p);
}
return CARRIER_PENDING; /* Not yet ! */
}
1999-12-13 18:45:21 +00:00
return Online(dev) ? CARRIER_OK : CARRIER_LOST;
}
#ifdef NONETGRAPH
#define tty_SetAsyncParams NULL
#define tty_Write NULL
#define tty_Read NULL
#else
static int
isngtty(struct ttydevice *dev)
{
return dev->real.fd != -1;
}
static void
tty_SetAsyncParams(struct physical *p, u_int32_t mymap, u_int32_t hismap)
{
struct ttydevice *dev = device2tty(p->handler);
char asyncpath[NG_PATHSIZ];
struct ng_async_cfg cfg;
if (isngtty(dev)) {
/* Configure the async converter node */
snprintf(asyncpath, sizeof asyncpath, ".:%s", dev->hook);
memset(&cfg, 0, sizeof cfg);
cfg.enabled = 1;
cfg.accm = mymap | hismap;
cfg.amru = MAX_MTU;
cfg.smru = MAX_MRU;
log_Printf(LogDEBUG, "Configure async node at %s\n", asyncpath);
if (NgSendMsg(dev->cs, asyncpath, NGM_ASYNC_COOKIE,
NGM_ASYNC_CMD_SET_CONFIG, &cfg, sizeof cfg) < 0)
log_Printf(LogWARN, "%s: Can't configure async node at %s\n",
p->link.name, asyncpath);
} else
/* No netgraph node, just config the async layer */
async_SetLinkParams(&p->async, mymap, hismap);
}
static int
LoadLineDiscipline(struct physical *p)
{
struct ttydevice *dev = device2tty(p->handler);
u_char rbuf[sizeof(struct ng_mesg) + sizeof(struct nodeinfo)];
struct ng_mesg *reply;
struct nodeinfo *info;
char ttypath[NG_NODESIZ];
struct ngm_mkpeer ngm;
struct ngm_connect ngc;
int ldisc, cs, ds, hot, speed;
/*
* Don't use the netgraph line discipline for now. Using it works, but
* carrier cannot be detected via TIOCMGET and the device doesn't become
* selectable with 0 bytes to read when carrier is lost :(
*/
return 0;
reply = (struct ng_mesg *)rbuf;
info = (struct nodeinfo *)reply->data;
loadmodules(LOAD_VERBOSLY, "netgraph", "ng_tty", "ng_async", "ng_socket",
NULL);
/* Get the speed before loading the line discipline */
speed = physical_GetSpeed(p);
if (ioctl(p->fd, TIOCGETD, &dev->real.disc) < 0) {
log_Printf(LogDEBUG, "%s: Couldn't get tty line discipline\n",
p->link.name);
return 0;
}
ldisc = NETGRAPHDISC;
if (ID0ioctl(p->fd, TIOCSETD, &ldisc) < 0) {
log_Printf(LogDEBUG, "%s: Couldn't set NETGRAPHDISC line discipline\n",
p->link.name);
return 0;
}
/* Get the name of the tty node */
if (ioctl(p->fd, NGIOCGINFO, info) < 0) {
log_Printf(LogWARN, "%s: ioctl(NGIOCGINFO): %s\n", p->link.name,
strerror(errno));
ID0ioctl(p->fd, TIOCSETD, &dev->real.disc);
return 0;
}
snprintf(ttypath, sizeof ttypath, "%s:", info->name);
/* Create a socket node for our endpoint (and to send messages via) */
if (ID0NgMkSockNode(NULL, &cs, &ds) == -1) {
log_Printf(LogWARN, "%s: NgMkSockNode: %s\n", p->link.name,
strerror(errno));
ID0ioctl(p->fd, TIOCSETD, &dev->real.disc);
return 0;
}
/* Set the ``hot char'' on the TTY node */
hot = HDLC_SYN;
log_Printf(LogDEBUG, "%s: Set tty hotchar to 0x%02x\n", p->link.name, hot);
if (NgSendMsg(cs, ttypath, NGM_TTY_COOKIE,
NGM_TTY_SET_HOTCHAR, &hot, sizeof hot) < 0) {
log_Printf(LogWARN, "%s: Can't set hot char\n", p->link.name);
goto failed;
}
/* Attach an async converter node */
snprintf(ngm.type, sizeof ngm.type, "%s", NG_ASYNC_NODE_TYPE);
snprintf(ngm.ourhook, sizeof ngm.ourhook, "%s", NG_TTY_HOOK);
snprintf(ngm.peerhook, sizeof ngm.peerhook, "%s", NG_ASYNC_HOOK_ASYNC);
log_Printf(LogDEBUG, "%s: Send mkpeer async:%s to %s:%s\n", p->link.name,
ngm.peerhook, ttypath, ngm.ourhook);
if (NgSendMsg(cs, ttypath, NGM_GENERIC_COOKIE,
NGM_MKPEER, &ngm, sizeof ngm) < 0) {
log_Printf(LogWARN, "%s: Can't create %s node\n", p->link.name,
NG_ASYNC_NODE_TYPE);
goto failed;
}
/* Connect the async node to our socket */
snprintf(ngc.path, sizeof ngc.path, "%s%s", ttypath, NG_TTY_HOOK);
snprintf(ngc.peerhook, sizeof ngc.peerhook, "%s", NG_ASYNC_HOOK_SYNC);
memcpy(ngc.ourhook, ngc.peerhook, sizeof ngc.ourhook);
log_Printf(LogDEBUG, "%s: Send connect %s:%s to .:%s\n", p->link.name,
ngc.path, ngc.peerhook, ngc.ourhook);
if (NgSendMsg(cs, ".:", NGM_GENERIC_COOKIE, NGM_CONNECT,
&ngc, sizeof ngc) < 0) {
log_Printf(LogWARN, "%s: Can't connect .:%s -> %s.%s: %s\n",
p->link.name, ngc.ourhook, ngc.path, ngc.peerhook,
strerror(errno));
goto failed;
}
/* Get the async node id */
if (NgSendMsg(cs, ngc.path, NGM_GENERIC_COOKIE, NGM_NODEINFO, NULL, 0) < 0) {
log_Printf(LogWARN, "%s: Can't request async node info at %s: %s\n",
p->link.name, ngc.path, strerror(errno));
goto failed;
}
if (NgRecvMsg(cs, reply, sizeof rbuf, NULL) < 0) {
log_Printf(LogWARN, "%s: Can't obtain async node info at %s: %s\n",
p->link.name, ngc.path, strerror(errno));
goto failed;
}
/* All done, set up our device state */
snprintf(dev->hook, sizeof dev->hook, "%s", ngc.ourhook);
dev->cs = cs;
dev->real.fd = p->fd;
p->fd = ds;
dev->real.speed = speed;
physical_SetSync(p);
tty_SetAsyncParams(p, 0xffffffff, 0xffffffff);
physical_SetupStack(p, dev->dev.name, PHYSICAL_NOFORCE);
log_Printf(LogPHASE, "%s: Loaded netgraph tty line discipline\n",
p->link.name);
return 1;
failed:
ID0ioctl(p->fd, TIOCSETD, &dev->real.disc);
close(ds);
close(cs);
return 0;
}
static void
UnloadLineDiscipline(struct physical *p)
{
struct ttydevice *dev = device2tty(p->handler);
if (isngtty(dev)) {
log_Printf(LogPHASE, "back to speed %d\n", dev->real.speed);
if (!physical_SetSpeed(p, dev->real.speed))
log_Printf(LogWARN, "Couldn't reset tty speed to %d\n", dev->real.speed);
dev->real.speed = 0;
close(p->fd);
p->fd = dev->real.fd;
dev->real.fd = -1;
close(dev->cs);
dev->cs = -1;
*dev->hook = '\0';
if (ID0ioctl(p->fd, TIOCSETD, &dev->real.disc) == 0) {
physical_SetupStack(p, dev->dev.name, PHYSICAL_NOFORCE);
log_Printf(LogPHASE, "%s: Unloaded netgraph tty line discipline\n",
p->link.name);
} else
log_Printf(LogWARN, "%s: Failed to unload netgraph tty line discipline\n",
p->link.name);
}
}
static ssize_t
tty_Write(struct physical *p, const void *v, size_t n)
{
struct ttydevice *dev = device2tty(p->handler);
if (isngtty(dev))
return NgSendData(p->fd, dev->hook, v, n) == -1 ? -1 : n;
else
return write(p->fd, v, n);
}
static ssize_t
tty_Read(struct physical *p, void *v, size_t n)
{
struct ttydevice *dev = device2tty(p->handler);
char hook[sizeof NG_ASYNC_HOOK_SYNC];
if (isngtty(dev))
return NgRecvData(p->fd, v, n, hook);
else
return read(p->fd, v, n);
}
#endif /* NETGRAPH */
static int
tty_Raw(struct physical *p)
{
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
struct ttydevice *dev = device2tty(p->handler);
struct termios ios;
int oldflag;
log_Printf(LogDEBUG, "%s: Entering tty_Raw\n", p->link.name);
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
if (p->type != PHYS_DIRECT && p->fd >= 0 && !Online(dev))
log_Printf(LogDEBUG, "%s: Raw: descriptor = %d, mbits = %x\n",
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
p->link.name, p->fd, dev->mbits);
if (!physical_IsSync(p)) {
#ifndef NONETGRAPH
if (!LoadLineDiscipline(p))
#endif
{
tcgetattr(p->fd, &ios);
cfmakeraw(&ios);
if (p->cfg.rts_cts)
ios.c_cflag |= CLOCAL | CCTS_OFLOW | CRTS_IFLOW;
else
ios.c_cflag |= CLOCAL;
if (p->type != PHYS_DEDICATED)
ios.c_cflag |= HUPCL;
if (tcsetattr(p->fd, TCSANOW, &ios) == -1)
log_Printf(LogWARN, "%s: tcsetattr: Failed configuring device\n",
p->link.name);
}
}
oldflag = fcntl(p->fd, F_GETFL, 0);
if (oldflag < 0)
return 0;
fcntl(p->fd, F_SETFL, oldflag | O_NONBLOCK);
return 1;
}
static void
tty_Offline(struct physical *p)
{
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
struct ttydevice *dev = device2tty(p->handler);
if (p->fd >= 0) {
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
timer_Stop(&dev->Timer);
dev->mbits &= ~TIOCM_DTR; /* XXX: Hmm, what's this supposed to do ? */
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
if (Online(dev)) {
struct termios tio;
tcgetattr(p->fd, &tio);
if (cfsetspeed(&tio, B0) == -1 || tcsetattr(p->fd, TCSANOW, &tio) == -1)
log_Printf(LogWARN, "%s: Unable to set physical to speed 0\n",
p->link.name);
}
}
}
static void
tty_Cooked(struct physical *p)
{
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
struct ttydevice *dev = device2tty(p->handler);
int oldflag;
tty_Offline(p); /* In case of emergency close()s */
tcflush(p->fd, TCIOFLUSH);
if (!physical_IsSync(p) && tcsetattr(p->fd, TCSAFLUSH, &dev->ios) == -1)
log_Printf(LogWARN, "%s: tcsetattr: Unable to restore device settings\n",
p->link.name);
#ifndef NONETGRAPH
UnloadLineDiscipline(p);
#endif
if ((oldflag = fcntl(p->fd, F_GETFL, 0)) != -1)
fcntl(p->fd, F_SETFL, oldflag & ~O_NONBLOCK);
}
static void
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
tty_StopTimer(struct physical *p)
{
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
struct ttydevice *dev = device2tty(p->handler);
timer_Stop(&dev->Timer);
}
static void
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
tty_Free(struct physical *p)
{
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
struct ttydevice *dev = device2tty(p->handler);
tty_Offline(p); /* In case of emergency close()s */
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
free(dev);
}
static int
tty_Speed(struct physical *p)
{
struct termios ios;
if (tcgetattr(p->fd, &ios) == -1)
return 0;
return SpeedToInt(cfgetispeed(&ios));
}
static const char *
tty_OpenInfo(struct physical *p)
{
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
struct ttydevice *dev = device2tty(p->handler);
static char buf[13];
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
if (Online(dev))
strcpy(buf, "with");
else
strcpy(buf, "no");
strcat(buf, " carrier");
return buf;
}
static int
tty_Slot(struct physical *p)
{
struct ttyent *ttyp;
int slot;
setttyent();
for (slot = 1; (ttyp = getttyent()); ++slot)
if (!strcmp(ttyp->ty_name, p->name.base)) {
endttyent();
return slot;
}
endttyent();
return -1;
}
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
static void
tty_device2iov(struct device *d, struct iovec *iov, int *niov,
Rewrite the link descriptor transfer code in MP mode. Previously, ppp attempted to bind() to a local domain tcp socket based on the peer authname & enddisc. If it succeeded, it listen()ed and became MP server. If it failed, it connect()ed and became MP client. The server then select()ed on the descriptor, accept()ed it and wrote its pid to it then read the link data & link file descriptor, and finally sent an ack (``!''). The client would read() the server pid, transfer the link lock to that pid, send the link data & descriptor and read the ack. It would then close the descriptor and clean up. There was a race between the bind() and listen() where someone could attempt to connect() and fail. This change removes the race. Now ppp makes the RCVBUF big enough on a socket descriptor and attempts to bind() to a local domain *udp* socket (same name as before). If it succeeds, it becomes MP server. If it fails, it sets the SNDBUF and connect()s, becoming MP client. The server select()s on the descriptor and recvmsg()s the message, insisting on at least two descriptors (plus the link data). It uses the second descriptor to write() its pid then read()s an ack (``!''). The client creates a socketpair() and sendmsg()s the link data, link descriptor and one of the socketpair descriptors. It then read()s the server pid from the other socketpair descriptor, transfers any locks and write()s an ack. Now, there can be no race, and a connect() failure indicates a stale socket file. This also fixes MP ppp over ethernet, where the struct msghdr was being misconstructed when transferring the control socket descriptor. Also, if we fail to send the link, don't hang around in a ``session owner'' state, just do the setsid() and fork() if it's required to disown a tty. UDP idea suggested by: Chris Bennet from Mindspring at FreeBSDCon
1999-11-25 02:47:04 +00:00
int maxiov, int *auxfd, int *nauxfd)
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
{
struct ttydevice *dev = device2tty(d);
int sz = physical_MaxDeviceSize();
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
iov[*niov].iov_base = realloc(d, sz);
if (iov[*niov].iov_base == NULL) {
log_Printf(LogALERT, "Failed to allocate memory: %d\n", sz);
AbortProgram(EX_OSERR);
}
iov[*niov].iov_len = sz;
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
(*niov)++;
#ifndef NONETGRAPH
if (dev->cs >= 0) {
*auxfd = dev->cs;
(*nauxfd)++;
}
#endif
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
if (dev->Timer.state != TIMER_STOPPED) {
timer_Stop(&dev->Timer);
dev->Timer.state = TIMER_RUNNING;
}
}
static struct device basettydevice = {
TTY_DEVICE,
"tty",
0,
{ CD_VARIABLE, DEF_TTYCDDELAY },
tty_AwaitCarrier,
NULL,
tty_Raw,
tty_Offline,
tty_Cooked,
tty_SetAsyncParams,
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
tty_StopTimer,
tty_Free,
tty_Read,
tty_Write,
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
tty_device2iov,
tty_Speed,
tty_OpenInfo,
tty_Slot
};
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
struct device *
tty_iov2device(int type, struct physical *p, struct iovec *iov, int *niov,
int maxiov, int *auxfd, int *nauxfd)
{
if (type == TTY_DEVICE) {
struct ttydevice *dev = (struct ttydevice *)iov[(*niov)++].iov_base;
dev = realloc(dev, sizeof *dev); /* Reduce to the correct size */
if (dev == NULL) {
log_Printf(LogALERT, "Failed to allocate memory: %d\n",
(int)(sizeof *dev));
AbortProgram(EX_OSERR);
}
#ifndef NONETGRAPH
if (*nauxfd) {
dev->cs = *auxfd;
(*nauxfd)--;
} else
dev->cs = -1;
#endif
/* Refresh function pointers etc */
memcpy(&dev->dev, &basettydevice, sizeof dev->dev);
physical_SetupStack(p, dev->dev.name, PHYSICAL_NOFORCE);
if (dev->Timer.state != TIMER_STOPPED) {
dev->Timer.state = TIMER_STOPPED;
p->handler = &dev->dev; /* For the benefit of StartTimer */
tty_StartTimer(p);
}
return &dev->dev;
}
return NULL;
}
struct device *
tty_Create(struct physical *p)
{
struct ttydevice *dev;
struct termios ios;
int oldflag;
if (p->fd < 0 || !isatty(p->fd))
/* Don't want this */
return NULL;
if (*p->name.full == '\0') {
physical_SetDevice(p, ttyname(p->fd));
log_Printf(LogDEBUG, "%s: Input is a tty (%s)\n",
p->link.name, p->name.full);
} else
log_Printf(LogDEBUG, "%s: Opened %s\n", p->link.name, p->name.full);
/* We're gonna return a ttydevice (unless something goes horribly wrong) */
if ((dev = malloc(sizeof *dev)) == NULL) {
/* Complete failure - parent doesn't continue trying to ``create'' */
close(p->fd);
p->fd = -1;
return NULL;
}
memcpy(&dev->dev, &basettydevice, sizeof dev->dev);
1999-05-16 11:58:48 +00:00
memset(&dev->Timer, '\0', sizeof dev->Timer);
dev->mbits = -1;
#ifndef NONETGRAPH
dev->real.speed = 0;
dev->real.fd = -1;
dev->real.disc = -1;
*dev->hook = '\0';
#endif
tcgetattr(p->fd, &ios);
dev->ios = ios;
if (p->cfg.cd.necessity != CD_DEFAULT)
/* Any override is ok for the tty device */
dev->dev.cd = p->cfg.cd;
log_Printf(LogDEBUG, "%s: tty_Create: physical (get): fd = %d,"
" iflag = %lx, oflag = %lx, cflag = %lx\n", p->link.name, p->fd,
(u_long)ios.c_iflag, (u_long)ios.c_oflag, (u_long)ios.c_cflag);
cfmakeraw(&ios);
if (p->cfg.rts_cts)
ios.c_cflag |= CLOCAL | CCTS_OFLOW | CRTS_IFLOW;
else {
ios.c_cflag |= CLOCAL;
ios.c_iflag |= IXOFF;
}
ios.c_iflag |= IXON;
if (p->type != PHYS_DEDICATED)
ios.c_cflag |= HUPCL;
if (p->type != PHYS_DIRECT) {
/* Change tty speed when we're not in -direct mode */
ios.c_cflag &= ~(CSIZE | PARODD | PARENB);
ios.c_cflag |= p->cfg.parity;
if (cfsetspeed(&ios, IntToSpeed(p->cfg.speed)) == -1)
log_Printf(LogWARN, "%s: %s: Unable to set speed to %d\n",
p->link.name, p->name.full, p->cfg.speed);
}
if (tcsetattr(p->fd, TCSADRAIN, &ios) == -1) {
log_Printf(LogWARN, "%s: tcsetattr: Failed configuring device\n",
p->link.name);
if (p->type != PHYS_DIRECT && p->cfg.speed > 115200)
log_Printf(LogWARN, "%.*s Perhaps the speed is unsupported\n",
(int)strlen(p->link.name), "");
}
log_Printf(LogDEBUG, "%s: physical (put): iflag = %lx, oflag = %lx, "
"cflag = %lx\n", p->link.name, (u_long)ios.c_iflag,
(u_long)ios.c_oflag, (u_long)ios.c_cflag);
oldflag = fcntl(p->fd, F_GETFL, 0);
if (oldflag < 0) {
/* Complete failure - parent doesn't continue trying to ``create'' */
log_Printf(LogWARN, "%s: Open: Cannot get physical flags: %s\n",
p->link.name, strerror(errno));
tty_Cooked(p);
close(p->fd);
p->fd = -1;
free(dev);
return NULL;
} else
fcntl(p->fd, F_SETFL, oldflag & ~O_NONBLOCK);
physical_SetupStack(p, dev->dev.name, PHYSICAL_NOFORCE);
return &dev->dev;
}