sysmocom
/
ofono
mirror of git://git.sysmocom.de/ofono
11
0
Fork 0
Code Issues Releases Activity
3,551 Commits 12 Branches 0 Tags 8.9 MiB
Commit Graph

463 Commits

Author SHA1 Message Date
Denis Kenzior 3946d7e07c ppp: Keep these defines private 2010-05-10 16:28:23 -05:00
Kristen Carlson Accardi 125fc62526 ppp: set address and control field 
Before sending to hdlc, set the address and control field.
Fix hardcode of ppp header size.
 2010-05-10 16:27:15 -05:00
Kristen Carlson Accardi f13d5e2bad gsmdial: use g_at_ppp_new_from_io() 2010-05-10 15:41:39 -05:00
Kristen Carlson Accardi 72c38f7adb gatchat: implement g_at_chat_get_io() 2010-05-10 15:41:39 -05:00
Kristen Carlson Accardi cb36130b13 gsmdial: use g_at_chat_suspend 2010-05-10 15:41:39 -05:00
Kristen Carlson Accardi 2d8a1462db ppp: transition to dead when lcp finished 
Transition to DEAD when lcp is finished and ignore any io
disconnects if we are already dead.
 2010-05-10 15:29:44 -05:00
Denis Kenzior 02c2cdaf21 gsmdial: Port to new PPP api 2010-04-30 15:32:45 -05:00
Denis Kenzior 6d20194e75 ppp: Refactor connect / disconnect callbacks 
Right now it is very hard to figure out whether we should be calling the
connect callback or the disconnect callback.  So refactor as follows:
	- Connect callback is only called once the net is actually up
	- Disconnect callback is called once ppp is down, with a reason
	  for why it is so.
 2010-04-30 15:31:26 -05:00
Denis Kenzior 9ae0dcb47d ppp: get rid of ppp_enter_phase 
This function simply didn't have the context of why the phase was being
entered.  Instead have each protocol notify GAtPPP as to what is
happening.  We already had this more or less for IPCP and AUTH events,
this just now formalizes it for LCP as well.
 2010-04-30 14:37:17 -05:00
Denis Kenzior c7ef06f91e ppp: Add _from_io constructor 2010-04-30 10:52:16 -05:00
Kristen Carlson Accardi 5ef90f934b ppp: fix return of drop_packet for NETWORK phase 2010-04-29 18:49:16 -05:00
Denis Kenzior 95e0a15010 gathdlc: Add from_io constructor 2010-04-29 18:49:16 -05:00
Kristen Carlson Accardi 7d59d7a377 ppp: silently drop invalid packets 2010-04-29 14:40:20 -05:00
Kristen Carlson Accardi 9d5eb5110f ppp: transition to dead when read io destroyed 
remove call to signal_close for lcp object, it is redundent.  If
our io is destroyed we are dead, so transition the ppp phase to
DEAD so our disconnect callback can be called.
 2010-04-29 14:24:47 -05:00
Denis Kenzior c9edf56fc6 ppp: Hang set_recording off the GAtPPP object 
Marcel: recording right now only works for PPP, so we'd need some sort
of multi-protocol support.

So for now expose set_recording to be used through the main PPP object.
HDLC object recording support needs to be extended.
 2010-04-28 20:25:10 -05:00
Denis Kenzior fb8ca665dd ppp: Re-add capability to detect io down events 2010-04-28 17:27:36 -05:00
Denis Kenzior 378f3c61e3 gathdlc: Add accessor for GAtIO 2010-04-28 17:27:36 -05:00
Denis Kenzior 3b98ed067a gathdlc: Port GAtHDLC to use GAtIO 2010-04-28 17:27:36 -05:00
Denis Kenzior d271216112 ppp: Don't crash if the interface wasn't created 2010-04-28 17:27:36 -05:00
Denis Kenzior db0b7c8a98 ppp: Accept 1500 byte MTUs 2010-04-28 17:27:36 -05:00
Denis Kenzior 7d5a3730d1 gsmdial: Use set recording on the HDLC object 2010-04-28 17:27:36 -05:00
Denis Kenzior ed15727b98 ppp: Add getter for HDLC object 2010-04-28 17:27:36 -05:00
Denis Kenzior 4cdf49f0de ppp: Initial port of PPP to use GAtHDLC 2010-04-28 17:27:35 -05:00
Denis Kenzior fa8614cc78 ppp: Make pppcp functions const correct 2010-04-28 17:27:35 -05:00
Denis Kenzior 87340d64e2 ppp: Make ppp_chap_process_packet const correct 2010-04-28 17:27:35 -05:00
Denis Kenzior 955673573b ppp: Make ppp_net_process_packet const correct 2010-04-28 17:27:35 -05:00
Denis Kenzior 5acb633a1d gathdlc: Add recording support, ported from PPP 
For protocol examination using wireshark
 2010-04-28 17:27:35 -05:00
Denis Kenzior 41ce6212a1 gathdlc: Add receive ACCM support 2010-04-28 17:27:35 -05:00
Denis Kenzior 7c6c72f558 gathdlc: Support transmit ACCM 2010-04-28 17:27:35 -05:00
Denis Kenzior e088f3c2bb gatchat: Don't try to cleanup twice 
In case we get unrefed from the disconnect callback
 2010-04-28 17:27:35 -05:00
Denis Kenzior 0845dc3b5d gatchat: Use GAtIO for write watchers 2010-04-28 17:27:34 -05:00
Denis Kenzior 6a66505e46 gatio: support for blocking / nonblocking writes 2010-04-28 17:27:34 -05:00
Denis Kenzior 85448aac8c gatchat: get_channel should return one from GAtIO 2010-04-28 17:27:34 -05:00
Denis Kenzior 19115e0a15 gatchat: Channel is now setup by GAtIO 2010-04-28 17:27:34 -05:00
Denis Kenzior 033b576593 gatchat: Better re-entrancy handling 2010-04-28 17:27:34 -05:00
Denis Kenzior 9c3ae4746e gsmdial: Get rid of g_at_chat_shutdown usage 2010-04-28 17:27:34 -05:00
Denis Kenzior d875298404 gsmdial: Unref the chat instead of using shutdown 2010-04-28 17:27:34 -05:00
Denis Kenzior 8645b05c4f gatmux: Unref the chat instead of using shutdown 2010-04-28 17:27:34 -05:00
Denis Kenzior f9b2e5cf5a gatchat: Only set debug on IO if we have it 2010-04-28 17:27:33 -05:00
Denis Kenzior 3aa1faddef gatchat: Replace shutdown by suspend/resume 
Shutdown was actually pointless and according to Marcel he only had it
in mind for PPP useage.
 2010-04-28 17:27:33 -05:00
Denis Kenzior 7b8f4f5dc8 gatchat: port gatchat to use gatio 2010-04-28 17:27:33 -05:00
Denis Kenzior 43c30e2292 gatio: Remove g_at_io_shutdown 2010-04-28 17:27:33 -05:00
Denis Kenzior 204814da05 gatio: Get re-entrancy correct 2010-04-28 17:27:32 -05:00
Denis Kenzior c614254ae9 gatio: Style fix for gsource id comparisons 2010-04-28 17:27:32 -05:00
Denis Kenzior 2d2bc8b5b3 gatio: Kill the reader when no more space 2010-04-28 17:27:32 -05:00
Denis Kenzior f09ce13ad7 gatchat: Remove g_at_chat_set_syntax 
Never used and too unsafe in its current form
 2010-04-28 17:27:32 -05:00
Denis Kenzior 5fc7c1e1d0 gatchat: Add initial GAtIO implementation 2010-04-28 17:27:32 -05:00
Kristen Carlson Accardi a9c0be7339 ppp: use default ACCM when sending LCP codes 1-7 
According to the spec, you must transmit all Link Configuration,
Termination, and Code-Reject packets as if no options had been
negotiated.  This requires that when encoding we use the
default ACCM of 0xffffffff when sending these types of packets.
 2010-04-26 15:31:17 -05:00
Kristen Carlson Accardi fce78feb54 ppp: discard non-LCP packets in ESTABLISHMENT phase 2010-04-26 12:38:31 -05:00
Marcel Holtmann cc1975270c Fix broken HDLC flag handling 2010-04-24 19:12:30 +02:00
Marcel Holtmann 057bd9f0b3 Fix wrong escape character check in HDLC transmit 2010-04-24 19:02:04 +02:00
Marcel Holtmann 46839fbd15 Allow real wrapping around of HDLC transmit ring buffer 2010-04-24 19:00:54 +02:00
Marcel Holtmann 8f05535bb2 Add offset parameter to ring_buffer_write_ptr() function 2010-04-24 18:59:44 +02:00
Marcel Holtmann 9fd7d841c8 Add support for wrapping of HDLC transmit ring buffer 2010-04-24 18:54:34 +02:00
Marcel Holtmann d2c8f124e6 Use HDLC_FCS helper where possible 2010-04-23 17:22:56 +02:00
Marcel Holtmann 2a609f9054 Use HDLC constants instead of magic numbers 2010-04-23 17:21:49 +02:00
Marcel Holtmann a31870f542 Add support for wrapping of HDLC receive ring buffer 2010-04-23 17:18:30 +02:00
Denis Kenzior fa5e8cf56b ppp: Fix trivial style error 2010-04-22 17:58:29 -05:00
Kristen Carlson Accardi e98fb9a6bd ppp: change MTU on TUN device when MRU option received 2010-04-22 17:56:31 -05:00
Denis Kenzior fde23f113a ppp: Request MRU if we ever get NAKed with it 
Huawei E160G hardware seems to NAK our configure request and suggest
that it will never send packets bigger than 1440 bytes.  Since we don't
particularly care (our receive ring buffer is 4K, so it can handle 2048
byte packets), we just re-send the Configure Request with the preferred
value.
 2010-04-22 15:08:05 -05:00
Denis Kenzior 07280e1bff ppp: Fix trivial copy paste error 2010-04-22 15:07:42 -05:00
Denis Kenzior 226865272c ppp: Shut off IPCP when entering TERMINATING phase 2010-04-22 15:05:18 -05:00
Kristen Carlson Accardi 71775550fe ppp: implement MRU option 
If the peer requests a MRU option, set the mtu for the network
phase.  When we are in link establishment phase, we should
continue to behave as if no option has been set and the peer
should use the default MRU.

This option is required for the Huawei E160G modem.
 2010-04-22 07:00:40 +02:00
Marcel Holtmann 123b87de2c No to check for NULL pointers with g_free 2010-04-15 04:42:03 -07:00
Kristen Carlson Accardi 60643ee287 ppp: nak unknown auth protocol 
If we are sent a Config-Request for an auth proto other than
CHAP with MD5, send a NAK.
 2010-04-14 17:16:03 -05:00
Denis Kenzior 15d93ad0b9 Fix: GAtMux channels should return EAGAIN 
GAtChat uses non-blocking semantics, so the GAtMux channels should
return the EAGAIN status to make GAtChat work properly.
 2010-04-14 10:39:47 -05:00
Denis Kenzior c8aa645973 ppp: Refing the channel is actually not necessary 
The read_watch and write_watches both take a ref
 2010-04-13 15:57:07 -05:00
Denis Kenzior 8e15816ac1 Fix: Don't try to set channel flags if they're 0 2010-04-13 15:37:05 -05:00
Denis Kenzior c74e2ab189 ppp: TUN channel should not be opened NONBLOCKING 
We don't handle it as a nonblocking channel and getting it right would
be a nightmare.
 2010-04-13 15:35:55 -05:00
Denis Kenzior 7085818f23 ppp: Cleanup function definitions in ppp.h 2010-04-13 15:32:35 -05:00
Kristen Carlson Accardi b74a9342d2 ppp: remove pfc and acfc 
We will not support pfc or acfc
 2010-04-13 15:32:35 -05:00
Kristen Carlson Accardi 9c3631f24e ppp: remove references to magic number 
We will not support loopback detection.
 2010-04-13 15:32:35 -05:00
Denis Kenzior 0aaf98a354 ppp: Get rid of net_open and net_close 
There really isn't a need for these now
 2010-04-13 15:14:40 -05:00
Denis Kenzior 1d939052fc ppp: Introduce ppp_net_down_notify 2010-04-13 14:59:13 -05:00
Denis Kenzior d1d73a75fc ppp: Introduce ppp_net_up_notify and use it 
This is slightly cleaner way than defining a weird callback function.
 2010-04-13 14:55:10 -05:00
Denis Kenzior 6525826455 ppp: Hide ppp_net definition in ppp_net.c 2010-04-13 14:42:59 -05:00
Denis Kenzior 71ddfca269 ppp: Use chap functions directly, instead of auth 2010-04-13 14:21:16 -05:00
Denis Kenzior 18e5e26acb ppp: Remove auth_ functions from ppp_auth.c 
These really serve no purpose right now as we use only CHAP.  So they
only take up space and make the code harder to read.  If we implement
1-3 auth protocols, then they're easier handled inside gatppp.c.  If we
have more, then a proper auth driver framework is required.
 2010-04-13 14:21:16 -05:00
Denis Kenzior 42f6e8ec6e ppp: introduce ppp_auth_notify 
This function will be notified whenever authentication has succeeded /
failed.  This can happen in the authentication phase or during the
network phase.  If auth fails, then we should proceed to the terminate
phase.
 2010-04-13 14:21:16 -05:00
Denis Kenzior e21bb72df0 ppp: Add set / get password & username 
Also refactor the set credentials function, we will be removing the auth
object shortly
 2010-04-13 14:21:16 -05:00
Denis Kenzior fed99b9265 ppp: Transition the phase directly 
Use of the generate event function, while more 'pure' with regard to how
the spec views transitions, actually makes code more difficult to read.

Instead use phase transitions directly inside gatppp.  This still bleeds
through a little into lcp code, and probably should be fixed in a better
way eventually.
 2010-04-13 14:21:12 -05:00
Denis Kenzior 552db428ef ppp: Refactor how tls/tlu/tld/tlf are used 
With the upper layer driving these, the special handling is no longer
required.
 2010-04-13 14:00:01 -05:00
Denis Kenzior 9116b7e8f8 ppp: Let the upper layer handle open / up events 
This removes the need for the layer_started functions in lcp and ipcp.
For LCP the link is always up unless the socket has been closed, and for
IPCP the link should be opened as soon as LCP is ready anyway.
 2010-04-13 13:58:12 -05:00
Denis Kenzior 87939c2fcf Merge: make link 2010-04-13 13:57:03 -05:00
Denis Kenzior dbbaa3c416 ppp: Reset the options whenever the layer is down 
So we can re-negotiate the options if the layer is opened again.
 2010-04-13 13:34:12 -05:00
Denis Kenzior 8c1677713f ppp: Rename data to pppcp 2010-04-13 13:27:23 -05:00
Denis Kenzior 843db6f9a5 ppp: unref should mean a hard shutdown 
This can happen when e.g. the modem is physically removed from the
system and it is not feasible to wait for the nice shutdown state to be
reached.
 2010-04-13 12:03:32 -05:00
Denis Kenzior b916f189f5 ppp: Signal Down & Close when socket is closed 2010-04-13 12:03:32 -05:00
Denis Kenzior a661dd1583 ppp: Add pppcp_signal_down 
For use when the link is abruptly terminated by the remote side
 2010-04-13 12:03:32 -05:00
Denis Kenzior f5270cc9bd ppp: Use pppcp_signal functions directly 
No real need to wrap them behind lcp_ functions
 2010-04-13 12:03:32 -05:00
Denis Kenzior 36d15b4366 ppp: Keep track when read watcher dies 2010-04-13 12:03:32 -05:00
Denis Kenzior 79d81af431 ppp: Remove unused define 2010-04-13 12:03:32 -05:00
Denis Kenzior 6ac0025767 Fix: busy loop in atserver 2010-04-12 22:33:42 -05:00
Marcel Holtmann 7a27cc42fc Reindent the handler registration inside AT server 2010-04-12 18:13:02 -07:00
Marcel Holtmann f319e65b99 Use ppp_debug() for PPP CP event debug statements 2010-04-12 18:02:11 -07:00
Marcel Holtmann 143d6db7c3 Remove useless debug in PPP transmit destroy callback 2010-04-12 17:53:29 -07:00
Marcel Holtmann a977ecf260 Remove various GDestroyNotify function casting 2010-04-12 17:48:20 -07:00
Marcel Holtmann 5b229e60c9 The exit() function should always use positive exit values 2010-04-12 17:37:50 -07:00
Marcel Holtmann 0e178517ce Fix file descriptor leakage when closing PPP recording 2010-04-12 17:34:24 -07:00
Marcel Holtmann 313eb4f285 The AT+CGDATA command should return CONNECT on success 2010-04-12 16:36:06 -07:00