Struct bundle will have its own struct ccp in the future
too.
o The ``set stopped'' command now requires context and doesn't
work on the IPCP FSM.
o Check if it's time to break out of our top level loop before
doing a select - otherwise, we'll select forever :-(
o Remove `struct link'::ccp (a temporary hack). It turns out
that IpStartOutput() calls link_Output() and link_Output()
incorrectly calls StartOutput() (really modem_StartOutput)
requiring the ccp knowledge so that it can call
IpStartOutput()... The end result is that the whole IP
output queue gets dumped into the modem output queue
and a pile of physical writes are done prematurely. This
makes the (original) code in main() actually work in that
it would not bother selecting() on the tun descriptor when
our modem queue length was 20 or greater. Instead, we now
make that decision based on the overall queue length.
This will need improvement later.
o Move our LCP report timer into struct hdlc - it's really
a hdlc timer (fcs errors etc).
o Make `show hdlc' require context and make the output more
friendly.
o Remove all non-const globals from hdlc.c
o Output peer-rejected protocols by name - not just ones
that we reject.
This is a type of physical link that can chat and talk
LCP & CCP. A bundle contains a list of these (only one
in the list for the moment).
The datalink is a type of descriptor, and dials, enters
LCP (& does CCP), kicks the bundle when its FSMs do
something interesting and does the hangup chat script
on the way down. It also handles redials and reconnects.
There are lots of loose ends, and probably lots of bugs,
but the data structures are getting there !
This will ultimately be a member of a list of descriptors and
their handler functions on which we need to select() in the
main loop.
o Make struct physical into a `sort' of struct descriptor.
o The FSM layering is now more sane.
o Move a lot of the NCP stuff into our ipcpstate rather than having it
in the bundle, including control of the configured IP addresses. We
don't need hacks like the global `linkup' variable any more as the
FSM decides when our ppp.link* files get run. This is going to eventually
be configurable based on FSM events anyway.
o Fix a few inconsistencies when both sides require authentication.
o We now have "Ppp..." and "PPp" prompts, reflecting authentication
and network phase. We don't print loads of spurious prompts as we
change phases any more.
o Our phase is part of the bundle now.
o Fix a bug where the FSM wasn't calling LayerFinish.
o Close the FSM down correctly with a signal rather than slamming it
down as if the line was dropped (the undocumented ``down'' command
is still available though).
o Remove the forgotten `tunno' variable and fix references to it.
This structure contains the asynchronous state of the physical
link.
Unfortunately, just about every .h file is included in every .c
file now. Fixing this can be one of the last jobs.
This structure will eventually contain a list of NCPs (currently
only IPCP is supported) and a list of physical `struct link's.
It will also derive from a struct link itself.
Make ModemTimeout() static - it's way to dangerous to be called
from outside !
Bump version to 1.9. Our first MP release should be 2.0.
IPCP, CCP and LCP are now just derived FSMs.
Comment each of the FSM implementations so that we can
tell what's going on.
Revise the state transitions so that CCP and IPCP actually
send terminate REQs when appropriate.
The OS & IPCP layers are still like spagetti (next job).
The delay defaults to 1 sec (as it always has) unless we've done
a ~p in interactive mode or we've actually detected a HDLC frame.
This is now cleanly implemented (via async timers) so that it is
possible for LCP to come up despite the delay if an LCP REQ is
received.
This will hopefully solve situations with slow servers or slirp
scenarios (where ECHO is left on the port for a second or so before
the peer enters packet mode).
Also, ~p in interactive mode no longer changes the value of the default
openmode delay and -dedicated mode enters packet mode in the right state
according to the value of openmode.
(I completely mis-read the rfc last time 'round!)
This means:
o Better CCP/WARN Reset diagnostics.
o After we've sent a REQ and before we've received an ACK, we drop
incoming compressed data and send another REQ.
o Before sending an ACK, re-sequence all pending PRI_NORMAL data in
the modem queue so that pending packets won't get to the peer
*after* the ResetAck.
o Send ACKs with the `identifier' from the REQ frame.
o After we've received a correct ACK, duplicate ACKs are ok (and will
reset our history).
o Incorrect ACKs (not matching the last REQ) are moaned about and dropped.
Also,
o Calculate the correct FCS after compressing a packet. DEFLATE
*may* produce an mbuf with more than a single link in the chain,
but HdlcOutput didn't know how to calculate the FCS :-(
o Make `struct fsm'::reqid a u_char, not an int.
This fix will prevent us from sending id `255' 2,000,000,000 times
before wrapping to `0' for another 2,000,000,000 sends :-/
o Bump the version number a little.
The end result: DEFLATE now works over an unreliable link layer.
I can txfr a 1.5Mb kernel over a (rather bad) null-modem
cable at an average of 21679 bytes per second using rcp.
Repeat after me: Don't test compression using a loopback ppp/tcp setup as
we never lose packets and therefore never have to reset!
Validate the peers suggested IP by attempting to make a routing table
entry.
Give up IPCP negotiation if the peer NAKs us with an unusable IP.
Always SIOCDIFADDR then SIOCAIFADDR when configuring the tun device.
Using SIOCSIFDSTADDR allows duplicate dst addresses (which we don't
want)!!!
Allow up to 200 interface names (was 50) (now that ppp can play server
properly).
Up the version number (1.5 -> 1.6).
Cosmetic:
Log unexpected CCP packets in the CCP log rather than the ERROR log.
Log unexpected Config Reqs in the appropriate LCP/IPCP/CCP log rather
than the ERROR log.
Log failed route additions and deletions with WARN, not TCPIP.
Log the option id and length for unrecognised IPCP options.
Change some .Sq to .Ar in the man page.
Remove extraneous decls.
Add ``const'' to several places.
Allow ``make NOALIAS=1'' to remove IP aliasing.
Merge with OpenBSD - only the Makefiles vary.
We can now survive a compile with
-Wall -Wbad-function-cast -Wcast-align -Wcast-qual
-Winline -Wmissing-declarations -Wmissing-prototypes
-Wnested-externs -Wpointer-arith -Wredundant-decls
-Wshadow -Wstrict-prototypes -Wwrite-strings -Wchar-subscripts
(although the Makefile just contains -Wall).
o Add missing $Id$s
o Move extern decls from .c -> .h files
o Staticize
o Remove #includes from .h files
o style(9)ify includes
o bcopy -> memcpy
bzero -> memset
bcmp -> memcmp
index -> strchr
rindex -> strrchr
o Move timeout.h -> timer.h (making it consistent w/ timer.c)
o Add -Wmissing-prototypes
receive the ResetAck, NOT when we send the
ResetReq (as per the rfc).
o SILENTLY ignore CCPs that arrive *before*
the network phase (as per the rfc).
o Check that we've actually negotiated PRED1
before sending PRED1 output.
This bug has been around for a *VERY* long
time ! We shouldn't need to explicitly disable
PRED1 now :-)
idependently time out any of the FSMs.
Split LCP logging into LCP, IPCP and CCP logging,
and make room in "struct fsm" for the log level
that the state machine should use.
"set stopped" directive. If the timeout occurs
it will cause a "Down" event, hanging up the line
if it's still up. This *isn't* part of the FSM
diagram, but I consider it ok as a "higher level
implementation specific timeout" as specified in
the rfc ;-}
Discussed briefly with: joerg
o Use syslog
o Remove references to stdout/stderr (incl perror())
o Introduce VarTerm - the interactive terminal or zero
o Allow "set timeout" to affect current session
o Change "set debug" to "set log"
o Allow "set log [+|-]flag"
o Make MSEXT and PASSWDAUTH stuff the default
o Move all #ifdef DEBUG stuff into the code - this
shouldn't be too much overhead. It's now controlled
with "set log +debug"
o Add "set log command, debug, tun, warn, error, alert"
o Remove cdefs.h, and assume an ansi compiler.
o Improve all diagnostic output
o Don't trap SIGSEGV
o SIGHUP now terminates again (log files are controlled
by syslog)
o Call CloseModem() when changing devices
o Fix parsing of third arg of "delete"
I think this fixes the "magic is same" problems that some
people have been experiencing.
The man page is being rewritten. It'll follow soon.
received and after the TerminateAck is sent (as
per rfc1661) rather than to ST_STOPPING. Going
to ST_STOPPING will leave us in a state where
we're waiting for the other side to do something -
not a good idea, especially as the client side sends
a TerminateReq then exits on idle timeout.
This will make a number of things easier in the future, as well as (finally!)
avoiding the Id-smashing problem which has plagued developers for so long.
Boy, I'm glad we're not using sup anymore. This update would have been
insane otherwise.
add some logging functionality which I find very useful.
'set debug link' will record just link up/down and address assignments.
'set debug connect' will record the entire chat dialog
'set debug carrier' will record just chat lines including 'CARRIER'
(so that I can be sure I'm getting a 28.8 line).
There was a global change required to permit LogPrintf to take a bit
mask instead of a bit position value (to permit logging some events
on either of two flags, so that no change in 'set debug lcp' would
result from the code supporting 'link'. Thus the diffs are rather
long for such a small change. The man page is also touched.
Oh, and there was a slight syntax problem in route.c
Reviewed by: phk
Submitted by: Tony Kimball <alk@Think.COM>
ppp based on these patches for about 3 weeks with no downtime.
The original submitters comments:
Two features iijppp has over kernel ppp that I like are predictor1
compression and demand dialing. Here are a few bug fixes.
I expanded the priority queueing scheme and discovered it was broken
due to the assignment at ip.c line 300. All packets were being
queued at the same priority.
Fixing priority queueing broke predictor1 compression. Packets
were compressed before being queued and predictor1 worked as long
as the packets were popped off the queue in the same order they
were pushed onto the queue.
There were a few byte order problems in IP header tests also.
There is a recursion problem in SendLqrReport(). LcpClose() is
called when "Too many echo packets are lost" which winds up in
SendLqrReport() again. I believe the original intention was to
just stop the LQR timer with the call to StopLqr() but the side
effects hurt.
Submitted by: John Capo <jc@irbs.com>
one) as a valid answer to an echo request. This makes the log less
noisy when connecting to Trumpet Winsock or FreeBSD 2.0.5's pppd. :)
Submitted by: melvin@zytek.com (Stephen Melvin)
o Supporting SYNC SIO device (But need a device driver)
- add "set speed sync"
o Fixing bug for Predictor-1 function.
o Add new parameter that re-sent interval for set timeout commands.
o Improving RTT (Round Trip Time) and reducing processor time.
- Previous Timer service was using polling, and now using
SIGALRM ;-)
- A 0.94beta2 will not work correctly....
-- Follows are additinal feature not including 0.94beta2
o Support Proxy ARP
- add "enable/disable proxy" commands
o Marging common routine in CHAP/PAP.
o Enhancing LCP/IPCP log information.
o Support local Authfication connection on port 300x and tty.
- You can set up pair of your "hostname -s" and
password in ppp.secret. if either ppp.secret file nor
your hostname line don't exist, It will notify a message
and working as same as previous version.(Backword compatibility)
- If you did set up them, It's allow connection but nothing to do
except help and passwd command.
- add "passwd yourpasswd" commands
o Support afilter - keep Alive filter that a packet can send/receiving
according to ifilter/ofilter but doesn't count it as preventing idle
timer expires.
- Same syntax of other filters.
o Fixing bugs reported by current user for previous one. Thanks !!
Reviewed by: Atsushi Murai (amurai@spec.co.jp)
