2018-09-19 20:56:44 +00:00
|
|
|
/*
|
|
|
|
* Supporting routines used in common by all the various components of
|
|
|
|
* the SSH system.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <assert.h>
|
2018-09-19 19:36:40 +00:00
|
|
|
#include <stdlib.h>
|
2018-09-19 20:56:44 +00:00
|
|
|
|
2018-09-19 19:36:40 +00:00
|
|
|
#include "putty.h"
|
2018-09-19 20:56:44 +00:00
|
|
|
#include "ssh.h"
|
2018-09-24 12:45:10 +00:00
|
|
|
#include "sshbpp.h"
|
Move most of ssh.c out into separate source files.
I've tried to separate out as many individually coherent changes from
this work as I could into their own commits, but here's where I run
out and have to commit the rest of this major refactoring as a
big-bang change.
Most of ssh.c is now no longer in ssh.c: all five of the main
coroutines that handle layers of the SSH-1 and SSH-2 protocols now
each have their own source file to live in, and a lot of the
supporting functions have moved into the appropriate one of those too.
The new abstraction is a vtable called 'PacketProtocolLayer', which
has an input and output packet queue. Each layer's main coroutine is
invoked from the method ssh_ppl_process_queue(), which is usually
(though not exclusively) triggered automatically when things are
pushed on the input queue. In SSH-2, the base layer is the transport
protocol, and it contains a pair of subsidiary queues by which it
passes some of its packets to the higher SSH-2 layers - first userauth
and then connection, which are peers at the same level, with the
former abdicating in favour of the latter at the appropriate moment.
SSH-1 is simpler: the whole login phase of the protocol (crypto setup
and authentication) is all in one module, and since SSH-1 has no
repeat key exchange, that setup layer abdicates in favour of the
connection phase when it's done.
ssh.c itself is now about a tenth of its old size (which all by itself
is cause for celebration!). Its main job is to set up all the layers,
hook them up to each other and to the BPP, and to funnel data back and
forth between that collection of modules and external things such as
the network and the terminal. Once it's set up a collection of packet
protocol layers, it communicates with them partly by calling methods
of the base layer (and if that's ssh2transport then it will delegate
some functionality to the corresponding methods of its higher layer),
and partly by talking directly to the connection layer no matter where
it is in the stack by means of the separate ConnectionLayer vtable
which I introduced in commit 8001dd4cb, and to which I've now added
quite a few extra methods replacing services that used to be internal
function calls within ssh.c.
(One effect of this is that the SSH-1 and SSH-2 channel storage is now
no longer shared - there are distinct struct types ssh1_channel and
ssh2_channel. That means a bit more code duplication, but on the plus
side, a lot fewer confusing conditionals in the middle of half-shared
functions, and less risk of a piece of SSH-1 escaping into SSH-2 or
vice versa, which I remember has happened at least once in the past.)
The bulk of this commit introduces the five new source files, their
common header sshppl.h and some shared supporting routines in
sshcommon.c, and rewrites nearly all of ssh.c itself. But it also
includes a couple of other changes that I couldn't separate easily
enough:
Firstly, there's a new handling for socket EOF, in which ssh.c sets an
'input_eof' flag in the BPP, and that responds by checking a flag that
tells it whether to report the EOF as an error or not. (This is the
main reason for those new BPP_READ / BPP_WAITFOR macros - they can
check the EOF flag every time the coroutine is resumed.)
Secondly, the error reporting itself is changed around again. I'd
expected to put some data fields in the public PacketProtocolLayer
structure that it could set to report errors in the same way as the
BPPs have been doing, but in the end, I decided propagating all those
data fields around was a pain and that even the BPPs shouldn't have
been doing it that way. So I've reverted to a system where everything
calls back to functions in ssh.c itself to report any connection-
ending condition. But there's a new family of those functions,
categorising the possible such conditions by semantics, and each one
has a different set of detailed effects (e.g. how rudely to close the
network connection, what exit status should be passed back to the
whole application, whether to send a disconnect message and/or display
a GUI error box).
I don't expect this to be immediately perfect: of course, the code has
been through a big upheaval, new bugs are expected, and I haven't been
able to do a full job of testing (e.g. I haven't tested every auth or
kex method). But I've checked that it _basically_ works - both SSH
protocols, all the different kinds of forwarding channel, more than
one auth method, Windows and Linux, connection sharing - and I think
it's now at the point where the easiest way to find further bugs is to
let it out into the wild and see what users can spot.
2018-09-24 17:28:16 +00:00
|
|
|
#include "sshppl.h"
|
2018-09-19 20:56:44 +00:00
|
|
|
#include "sshchan.h"
|
|
|
|
|
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Implementation of PacketQueue.
|
|
|
|
*/
|
|
|
|
|
2018-09-23 15:35:29 +00:00
|
|
|
static void pq_ensure_unlinked(PacketQueueNode *node)
|
|
|
|
{
|
|
|
|
if (node->on_free_queue) {
|
|
|
|
node->next->prev = node->prev;
|
|
|
|
node->prev->next = node->next;
|
|
|
|
} else {
|
|
|
|
assert(!node->next);
|
|
|
|
assert(!node->prev);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-09-19 20:56:44 +00:00
|
|
|
void pq_base_push(PacketQueueBase *pqb, PacketQueueNode *node)
|
|
|
|
{
|
2018-09-23 15:35:29 +00:00
|
|
|
pq_ensure_unlinked(node);
|
2018-09-19 20:56:44 +00:00
|
|
|
node->next = &pqb->end;
|
|
|
|
node->prev = pqb->end.prev;
|
|
|
|
node->next->prev = node;
|
|
|
|
node->prev->next = node;
|
2018-09-21 12:16:38 +00:00
|
|
|
|
|
|
|
if (pqb->ic)
|
|
|
|
queue_idempotent_callback(pqb->ic);
|
2018-09-19 20:56:44 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void pq_base_push_front(PacketQueueBase *pqb, PacketQueueNode *node)
|
|
|
|
{
|
2018-09-23 15:35:29 +00:00
|
|
|
pq_ensure_unlinked(node);
|
2018-09-19 20:56:44 +00:00
|
|
|
node->prev = &pqb->end;
|
|
|
|
node->next = pqb->end.next;
|
|
|
|
node->next->prev = node;
|
|
|
|
node->prev->next = node;
|
2018-09-21 12:16:38 +00:00
|
|
|
|
|
|
|
if (pqb->ic)
|
|
|
|
queue_idempotent_callback(pqb->ic);
|
2018-09-19 20:56:44 +00:00
|
|
|
}
|
|
|
|
|
2018-09-23 15:35:29 +00:00
|
|
|
static PacketQueueNode pktin_freeq_head = {
|
|
|
|
&pktin_freeq_head, &pktin_freeq_head, TRUE
|
|
|
|
};
|
|
|
|
|
|
|
|
static void pktin_free_queue_callback(void *vctx)
|
|
|
|
{
|
|
|
|
while (pktin_freeq_head.next != &pktin_freeq_head) {
|
|
|
|
PacketQueueNode *node = pktin_freeq_head.next;
|
2018-10-05 22:49:08 +00:00
|
|
|
PktIn *pktin = container_of(node, PktIn, qnode);
|
2018-09-23 15:35:29 +00:00
|
|
|
pktin_freeq_head.next = node->next;
|
|
|
|
sfree(pktin);
|
|
|
|
}
|
|
|
|
|
|
|
|
pktin_freeq_head.prev = &pktin_freeq_head;
|
|
|
|
}
|
|
|
|
|
|
|
|
static IdempotentCallback ic_pktin_free = {
|
|
|
|
pktin_free_queue_callback, NULL, FALSE
|
|
|
|
};
|
|
|
|
|
2018-10-06 17:42:08 +00:00
|
|
|
static PktIn *pq_in_after(PacketQueueBase *pqb,
|
|
|
|
PacketQueueNode *prev, int pop)
|
2018-09-19 20:56:44 +00:00
|
|
|
{
|
2018-10-06 17:42:08 +00:00
|
|
|
PacketQueueNode *node = prev->next;
|
2018-09-19 20:56:44 +00:00
|
|
|
if (node == &pqb->end)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
if (pop) {
|
|
|
|
node->next->prev = node->prev;
|
|
|
|
node->prev->next = node->next;
|
2018-09-23 15:35:29 +00:00
|
|
|
|
|
|
|
node->prev = pktin_freeq_head.prev;
|
|
|
|
node->next = &pktin_freeq_head;
|
|
|
|
node->next->prev = node;
|
|
|
|
node->prev->next = node;
|
|
|
|
node->on_free_queue = TRUE;
|
|
|
|
queue_idempotent_callback(&ic_pktin_free);
|
2018-09-19 20:56:44 +00:00
|
|
|
}
|
|
|
|
|
2018-10-05 22:49:08 +00:00
|
|
|
return container_of(node, PktIn, qnode);
|
2018-09-19 20:56:44 +00:00
|
|
|
}
|
|
|
|
|
2018-10-06 17:42:08 +00:00
|
|
|
static PktOut *pq_out_after(PacketQueueBase *pqb,
|
|
|
|
PacketQueueNode *prev, int pop)
|
2018-09-19 20:56:44 +00:00
|
|
|
{
|
2018-10-06 17:42:08 +00:00
|
|
|
PacketQueueNode *node = prev->next;
|
2018-09-19 20:56:44 +00:00
|
|
|
if (node == &pqb->end)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
if (pop) {
|
|
|
|
node->next->prev = node->prev;
|
|
|
|
node->prev->next = node->next;
|
|
|
|
node->prev = node->next = NULL;
|
|
|
|
}
|
|
|
|
|
2018-10-05 22:49:08 +00:00
|
|
|
return container_of(node, PktOut, qnode);
|
2018-09-19 20:56:44 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void pq_in_init(PktInQueue *pq)
|
|
|
|
{
|
2018-09-21 12:16:38 +00:00
|
|
|
pq->pqb.ic = NULL;
|
2018-09-19 20:56:44 +00:00
|
|
|
pq->pqb.end.next = pq->pqb.end.prev = &pq->pqb.end;
|
2018-10-06 17:42:08 +00:00
|
|
|
pq->after = pq_in_after;
|
2018-09-19 20:56:44 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void pq_out_init(PktOutQueue *pq)
|
|
|
|
{
|
2018-09-21 12:16:38 +00:00
|
|
|
pq->pqb.ic = NULL;
|
2018-09-19 20:56:44 +00:00
|
|
|
pq->pqb.end.next = pq->pqb.end.prev = &pq->pqb.end;
|
2018-10-06 17:42:08 +00:00
|
|
|
pq->after = pq_out_after;
|
2018-09-19 20:56:44 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void pq_in_clear(PktInQueue *pq)
|
|
|
|
{
|
|
|
|
PktIn *pkt;
|
2018-09-21 12:16:38 +00:00
|
|
|
pq->pqb.ic = NULL;
|
2018-09-23 15:35:29 +00:00
|
|
|
while ((pkt = pq_pop(pq)) != NULL) {
|
|
|
|
/* No need to actually free these packets: pq_pop on a
|
|
|
|
* PktInQueue will automatically move them to the free
|
|
|
|
* queue. */
|
|
|
|
}
|
2018-09-19 20:56:44 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void pq_out_clear(PktOutQueue *pq)
|
|
|
|
{
|
|
|
|
PktOut *pkt;
|
2018-09-21 12:16:38 +00:00
|
|
|
pq->pqb.ic = NULL;
|
2018-09-19 20:56:44 +00:00
|
|
|
while ((pkt = pq_pop(pq)) != NULL)
|
|
|
|
ssh_free_pktout(pkt);
|
|
|
|
}
|
|
|
|
|
2018-09-19 19:30:40 +00:00
|
|
|
/*
|
|
|
|
* Concatenate the contents of the two queues q1 and q2, and leave the
|
|
|
|
* result in qdest. qdest must be either empty, or one of the input
|
|
|
|
* queues.
|
|
|
|
*/
|
|
|
|
void pq_base_concatenate(PacketQueueBase *qdest,
|
|
|
|
PacketQueueBase *q1, PacketQueueBase *q2)
|
2018-09-19 20:56:44 +00:00
|
|
|
{
|
2018-09-19 19:30:40 +00:00
|
|
|
struct PacketQueueNode *head1, *tail1, *head2, *tail2;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Extract the contents from both input queues, and empty them.
|
|
|
|
*/
|
|
|
|
|
|
|
|
head1 = (q1->end.next == &q1->end ? NULL : q1->end.next);
|
|
|
|
tail1 = (q1->end.prev == &q1->end ? NULL : q1->end.prev);
|
|
|
|
head2 = (q2->end.next == &q2->end ? NULL : q2->end.next);
|
|
|
|
tail2 = (q2->end.prev == &q2->end ? NULL : q2->end.prev);
|
|
|
|
|
|
|
|
q1->end.next = q1->end.prev = &q1->end;
|
|
|
|
q2->end.next = q2->end.prev = &q2->end;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Link the two lists together, handling the case where one or
|
|
|
|
* both is empty.
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (tail1)
|
|
|
|
tail1->next = head2;
|
|
|
|
else
|
|
|
|
head1 = head2;
|
|
|
|
|
|
|
|
if (head2)
|
|
|
|
head2->prev = tail1;
|
|
|
|
else
|
2018-09-22 08:32:08 +00:00
|
|
|
tail2 = tail1;
|
2018-09-19 19:30:40 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Check the destination queue is currently empty. (If it was one
|
|
|
|
* of the input queues, then it will be, because we emptied both
|
|
|
|
* of those just a moment ago.)
|
|
|
|
*/
|
|
|
|
|
|
|
|
assert(qdest->end.next == &qdest->end);
|
|
|
|
assert(qdest->end.prev == &qdest->end);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If our concatenated list has anything in it, then put it in
|
|
|
|
* dest.
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (!head1) {
|
|
|
|
assert(!tail2);
|
|
|
|
} else {
|
|
|
|
assert(tail2);
|
|
|
|
qdest->end.next = head1;
|
|
|
|
qdest->end.prev = tail2;
|
|
|
|
head1->prev = &qdest->end;
|
|
|
|
tail2->next = &qdest->end;
|
2018-09-21 12:16:38 +00:00
|
|
|
|
|
|
|
if (qdest->ic)
|
|
|
|
queue_idempotent_callback(qdest->ic);
|
2018-09-19 19:30:40 +00:00
|
|
|
}
|
2018-09-19 20:56:44 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Low-level functions for the packet structures themselves.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static void ssh_pkt_BinarySink_write(BinarySink *bs,
|
|
|
|
const void *data, size_t len);
|
|
|
|
PktOut *ssh_new_packet(void)
|
|
|
|
{
|
|
|
|
PktOut *pkt = snew(PktOut);
|
|
|
|
|
|
|
|
BinarySink_INIT(pkt, ssh_pkt_BinarySink_write);
|
|
|
|
pkt->data = NULL;
|
|
|
|
pkt->length = 0;
|
|
|
|
pkt->maxlen = 0;
|
|
|
|
pkt->downstream_id = 0;
|
|
|
|
pkt->additional_log_text = NULL;
|
|
|
|
pkt->qnode.next = pkt->qnode.prev = NULL;
|
2018-09-23 15:35:29 +00:00
|
|
|
pkt->qnode.on_free_queue = FALSE;
|
2018-09-19 20:56:44 +00:00
|
|
|
|
|
|
|
return pkt;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void ssh_pkt_ensure(PktOut *pkt, int length)
|
|
|
|
{
|
|
|
|
if (pkt->maxlen < length) {
|
|
|
|
pkt->maxlen = length + 256;
|
|
|
|
pkt->data = sresize(pkt->data, pkt->maxlen, unsigned char);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
static void ssh_pkt_adddata(PktOut *pkt, const void *data, int len)
|
|
|
|
{
|
|
|
|
pkt->length += len;
|
|
|
|
ssh_pkt_ensure(pkt, pkt->length);
|
|
|
|
memcpy(pkt->data + pkt->length - len, data, len);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void ssh_pkt_BinarySink_write(BinarySink *bs,
|
|
|
|
const void *data, size_t len)
|
|
|
|
{
|
|
|
|
PktOut *pkt = BinarySink_DOWNCAST(bs, PktOut);
|
|
|
|
ssh_pkt_adddata(pkt, data, len);
|
|
|
|
}
|
|
|
|
|
|
|
|
void ssh_free_pktout(PktOut *pkt)
|
|
|
|
{
|
|
|
|
sfree(pkt->data);
|
|
|
|
sfree(pkt);
|
|
|
|
}
|
2018-09-24 17:08:09 +00:00
|
|
|
|
2018-09-19 19:31:19 +00:00
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Implement zombiechan_new() and its trivial vtable.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static void zombiechan_free(Channel *chan);
|
|
|
|
static int zombiechan_send(Channel *chan, int is_stderr, const void *, int);
|
|
|
|
static void zombiechan_set_input_wanted(Channel *chan, int wanted);
|
|
|
|
static void zombiechan_do_nothing(Channel *chan);
|
|
|
|
static void zombiechan_open_failure(Channel *chan, const char *);
|
|
|
|
static int zombiechan_want_close(Channel *chan, int sent_eof, int rcvd_eof);
|
|
|
|
static char *zombiechan_log_close_msg(Channel *chan) { return NULL; }
|
|
|
|
|
|
|
|
static const struct ChannelVtable zombiechan_channelvt = {
|
|
|
|
zombiechan_free,
|
|
|
|
zombiechan_do_nothing, /* open_confirmation */
|
|
|
|
zombiechan_open_failure,
|
|
|
|
zombiechan_send,
|
|
|
|
zombiechan_do_nothing, /* send_eof */
|
|
|
|
zombiechan_set_input_wanted,
|
|
|
|
zombiechan_log_close_msg,
|
|
|
|
zombiechan_want_close,
|
2018-09-26 16:34:20 +00:00
|
|
|
chan_no_exit_status,
|
|
|
|
chan_no_exit_signal,
|
|
|
|
chan_no_exit_signal_numeric,
|
2018-09-26 17:02:33 +00:00
|
|
|
chan_no_request_response,
|
2018-09-19 19:31:19 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
Channel *zombiechan_new(void)
|
|
|
|
{
|
|
|
|
Channel *chan = snew(Channel);
|
|
|
|
chan->vt = &zombiechan_channelvt;
|
|
|
|
chan->initial_fixed_window_size = 0;
|
|
|
|
return chan;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void zombiechan_free(Channel *chan)
|
|
|
|
{
|
|
|
|
assert(chan->vt == &zombiechan_channelvt);
|
|
|
|
sfree(chan);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void zombiechan_do_nothing(Channel *chan)
|
|
|
|
{
|
|
|
|
assert(chan->vt == &zombiechan_channelvt);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void zombiechan_open_failure(Channel *chan, const char *errtext)
|
|
|
|
{
|
|
|
|
assert(chan->vt == &zombiechan_channelvt);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int zombiechan_send(Channel *chan, int is_stderr,
|
|
|
|
const void *data, int length)
|
|
|
|
{
|
|
|
|
assert(chan->vt == &zombiechan_channelvt);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void zombiechan_set_input_wanted(Channel *chan, int enable)
|
|
|
|
{
|
|
|
|
assert(chan->vt == &zombiechan_channelvt);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int zombiechan_want_close(Channel *chan, int sent_eof, int rcvd_eof)
|
|
|
|
{
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
2018-09-19 19:32:27 +00:00
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Centralised standard methods for other channel implementations to
|
|
|
|
* borrow.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void chan_remotely_opened_confirmation(Channel *chan)
|
|
|
|
{
|
|
|
|
assert(0 && "this channel type should never receive OPEN_CONFIRMATION");
|
|
|
|
}
|
|
|
|
|
|
|
|
void chan_remotely_opened_failure(Channel *chan, const char *errtext)
|
|
|
|
{
|
|
|
|
assert(0 && "this channel type should never receive OPEN_FAILURE");
|
|
|
|
}
|
|
|
|
|
|
|
|
int chan_no_eager_close(Channel *chan, int sent_local_eof, int rcvd_remote_eof)
|
|
|
|
{
|
|
|
|
return FALSE; /* default: never proactively ask for a close */
|
|
|
|
}
|
2018-09-19 19:36:40 +00:00
|
|
|
|
2018-09-26 16:34:20 +00:00
|
|
|
int chan_no_exit_status(Channel *chan, int status)
|
|
|
|
{
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
int chan_no_exit_signal(
|
|
|
|
Channel *chan, ptrlen signame, int core_dumped, ptrlen msg)
|
|
|
|
{
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
int chan_no_exit_signal_numeric(
|
|
|
|
Channel *chan, int signum, int core_dumped, ptrlen msg)
|
|
|
|
{
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
2018-09-26 17:02:33 +00:00
|
|
|
void chan_no_request_response(Channel *chan, int success)
|
|
|
|
{
|
|
|
|
assert(0 && "this channel type should never send a want-reply request");
|
|
|
|
}
|
|
|
|
|
2018-09-19 19:36:40 +00:00
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Common routine to marshal tty modes into an SSH packet.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void write_ttymodes_to_packet_from_conf(
|
New abstraction 'Seat', to pass to backends.
This is a new vtable-based abstraction which is passed to a backend in
place of Frontend, and it implements only the subset of the Frontend
functions needed by a backend. (Many other Frontend functions still
exist, notably the wide range of things called by terminal.c providing
platform-independent operations on the GUI terminal window.)
The purpose of making it a vtable is that this opens up the
possibility of creating a backend as an internal implementation detail
of some other activity, by providing just that one backend with a
custom Seat that implements the methods differently.
For example, this refactoring should make it feasible to directly
implement an SSH proxy type, aka the 'jump host' feature supported by
OpenSSH, aka 'open a secondary SSH session in MAINCHAN_DIRECT_TCP
mode, and then expose the main channel of that as the Socket for the
primary connection'. (Which of course you can already do by spawning
'plink -nc' as a separate proxy process, but this would permit it in
the _same_ process without anything getting confused.)
I've centralised a full set of stub methods in misc.c for the new
abstraction, which allows me to get rid of several annoying stubs in
the previous code. Also, while I'm here, I've moved a lot of
duplicated modalfatalbox() type functions from application main
program files into wincons.c / uxcons.c, which I think saves
duplication overall. (A minor visible effect is that the prefixes on
those console-based fatal error messages will now be more consistent
between applications.)
2018-10-11 18:58:42 +00:00
|
|
|
BinarySink *bs, Seat *seat, Conf *conf,
|
2018-09-19 19:36:40 +00:00
|
|
|
int ssh_version, int ospeed, int ispeed)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Codes for terminal modes.
|
|
|
|
* Most of these are the same in SSH-1 and SSH-2.
|
|
|
|
* This list is derived from RFC 4254 and
|
|
|
|
* SSH-1 RFC-1.2.31.
|
|
|
|
*/
|
|
|
|
static const struct ssh_ttymode {
|
|
|
|
const char *mode;
|
|
|
|
int opcode;
|
|
|
|
enum { TTY_OP_CHAR, TTY_OP_BOOL } type;
|
|
|
|
} ssh_ttymodes[] = {
|
|
|
|
/* "V" prefix discarded for special characters relative to SSH specs */
|
|
|
|
{ "INTR", 1, TTY_OP_CHAR },
|
|
|
|
{ "QUIT", 2, TTY_OP_CHAR },
|
|
|
|
{ "ERASE", 3, TTY_OP_CHAR },
|
|
|
|
{ "KILL", 4, TTY_OP_CHAR },
|
|
|
|
{ "EOF", 5, TTY_OP_CHAR },
|
|
|
|
{ "EOL", 6, TTY_OP_CHAR },
|
|
|
|
{ "EOL2", 7, TTY_OP_CHAR },
|
|
|
|
{ "START", 8, TTY_OP_CHAR },
|
|
|
|
{ "STOP", 9, TTY_OP_CHAR },
|
|
|
|
{ "SUSP", 10, TTY_OP_CHAR },
|
|
|
|
{ "DSUSP", 11, TTY_OP_CHAR },
|
|
|
|
{ "REPRINT", 12, TTY_OP_CHAR },
|
|
|
|
{ "WERASE", 13, TTY_OP_CHAR },
|
|
|
|
{ "LNEXT", 14, TTY_OP_CHAR },
|
|
|
|
{ "FLUSH", 15, TTY_OP_CHAR },
|
|
|
|
{ "SWTCH", 16, TTY_OP_CHAR },
|
|
|
|
{ "STATUS", 17, TTY_OP_CHAR },
|
|
|
|
{ "DISCARD", 18, TTY_OP_CHAR },
|
|
|
|
{ "IGNPAR", 30, TTY_OP_BOOL },
|
|
|
|
{ "PARMRK", 31, TTY_OP_BOOL },
|
|
|
|
{ "INPCK", 32, TTY_OP_BOOL },
|
|
|
|
{ "ISTRIP", 33, TTY_OP_BOOL },
|
|
|
|
{ "INLCR", 34, TTY_OP_BOOL },
|
|
|
|
{ "IGNCR", 35, TTY_OP_BOOL },
|
|
|
|
{ "ICRNL", 36, TTY_OP_BOOL },
|
|
|
|
{ "IUCLC", 37, TTY_OP_BOOL },
|
|
|
|
{ "IXON", 38, TTY_OP_BOOL },
|
|
|
|
{ "IXANY", 39, TTY_OP_BOOL },
|
|
|
|
{ "IXOFF", 40, TTY_OP_BOOL },
|
|
|
|
{ "IMAXBEL", 41, TTY_OP_BOOL },
|
|
|
|
{ "IUTF8", 42, TTY_OP_BOOL },
|
|
|
|
{ "ISIG", 50, TTY_OP_BOOL },
|
|
|
|
{ "ICANON", 51, TTY_OP_BOOL },
|
|
|
|
{ "XCASE", 52, TTY_OP_BOOL },
|
|
|
|
{ "ECHO", 53, TTY_OP_BOOL },
|
|
|
|
{ "ECHOE", 54, TTY_OP_BOOL },
|
|
|
|
{ "ECHOK", 55, TTY_OP_BOOL },
|
|
|
|
{ "ECHONL", 56, TTY_OP_BOOL },
|
|
|
|
{ "NOFLSH", 57, TTY_OP_BOOL },
|
|
|
|
{ "TOSTOP", 58, TTY_OP_BOOL },
|
|
|
|
{ "IEXTEN", 59, TTY_OP_BOOL },
|
|
|
|
{ "ECHOCTL", 60, TTY_OP_BOOL },
|
|
|
|
{ "ECHOKE", 61, TTY_OP_BOOL },
|
|
|
|
{ "PENDIN", 62, TTY_OP_BOOL }, /* XXX is this a real mode? */
|
|
|
|
{ "OPOST", 70, TTY_OP_BOOL },
|
|
|
|
{ "OLCUC", 71, TTY_OP_BOOL },
|
|
|
|
{ "ONLCR", 72, TTY_OP_BOOL },
|
|
|
|
{ "OCRNL", 73, TTY_OP_BOOL },
|
|
|
|
{ "ONOCR", 74, TTY_OP_BOOL },
|
|
|
|
{ "ONLRET", 75, TTY_OP_BOOL },
|
|
|
|
{ "CS7", 90, TTY_OP_BOOL },
|
|
|
|
{ "CS8", 91, TTY_OP_BOOL },
|
|
|
|
{ "PARENB", 92, TTY_OP_BOOL },
|
|
|
|
{ "PARODD", 93, TTY_OP_BOOL }
|
|
|
|
};
|
|
|
|
|
|
|
|
/* Miscellaneous other tty-related constants. */
|
|
|
|
enum {
|
|
|
|
/* The opcodes for ISPEED/OSPEED differ between SSH-1 and SSH-2. */
|
|
|
|
SSH1_TTY_OP_ISPEED = 192,
|
|
|
|
SSH1_TTY_OP_OSPEED = 193,
|
|
|
|
SSH2_TTY_OP_ISPEED = 128,
|
|
|
|
SSH2_TTY_OP_OSPEED = 129,
|
|
|
|
|
|
|
|
SSH_TTY_OP_END = 0
|
|
|
|
};
|
|
|
|
|
|
|
|
for (i = 0; i < lenof(ssh_ttymodes); i++) {
|
|
|
|
const struct ssh_ttymode *mode = ssh_ttymodes + i;
|
|
|
|
const char *sval = conf_get_str_str(conf, CONF_ttymodes, mode->mode);
|
|
|
|
char *to_free = NULL;
|
|
|
|
|
|
|
|
/* Every mode known to the current version of the code should be
|
|
|
|
* mentioned; this was ensured when settings were loaded. */
|
|
|
|
|
|
|
|
/*
|
|
|
|
* sval[0] can be
|
|
|
|
* - 'V', indicating that an explicit value follows it;
|
|
|
|
* - 'A', indicating that we should pass the value through from
|
|
|
|
* the local environment via get_ttymode; or
|
|
|
|
* - 'N', indicating that we should explicitly not send this
|
|
|
|
* mode.
|
|
|
|
*/
|
|
|
|
if (sval[0] == 'A') {
|
New abstraction 'Seat', to pass to backends.
This is a new vtable-based abstraction which is passed to a backend in
place of Frontend, and it implements only the subset of the Frontend
functions needed by a backend. (Many other Frontend functions still
exist, notably the wide range of things called by terminal.c providing
platform-independent operations on the GUI terminal window.)
The purpose of making it a vtable is that this opens up the
possibility of creating a backend as an internal implementation detail
of some other activity, by providing just that one backend with a
custom Seat that implements the methods differently.
For example, this refactoring should make it feasible to directly
implement an SSH proxy type, aka the 'jump host' feature supported by
OpenSSH, aka 'open a secondary SSH session in MAINCHAN_DIRECT_TCP
mode, and then expose the main channel of that as the Socket for the
primary connection'. (Which of course you can already do by spawning
'plink -nc' as a separate proxy process, but this would permit it in
the _same_ process without anything getting confused.)
I've centralised a full set of stub methods in misc.c for the new
abstraction, which allows me to get rid of several annoying stubs in
the previous code. Also, while I'm here, I've moved a lot of
duplicated modalfatalbox() type functions from application main
program files into wincons.c / uxcons.c, which I think saves
duplication overall. (A minor visible effect is that the prefixes on
those console-based fatal error messages will now be more consistent
between applications.)
2018-10-11 18:58:42 +00:00
|
|
|
sval = to_free = seat_get_ttymode(seat, mode->mode);
|
2018-09-19 19:36:40 +00:00
|
|
|
} else if (sval[0] == 'V') {
|
|
|
|
sval++; /* skip the 'V' */
|
|
|
|
} else {
|
|
|
|
/* else 'N', or something from the future we don't understand */
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (sval) {
|
|
|
|
/*
|
|
|
|
* Parse the string representation of the tty mode
|
|
|
|
* into the integer value it will take on the wire.
|
|
|
|
*/
|
|
|
|
unsigned ival = 0;
|
|
|
|
|
|
|
|
switch (mode->type) {
|
|
|
|
case TTY_OP_CHAR:
|
|
|
|
if (*sval) {
|
|
|
|
char *next = NULL;
|
|
|
|
/* We know ctrlparse won't write to the string, so
|
|
|
|
* casting away const is ugly but allowable. */
|
|
|
|
ival = ctrlparse((char *)sval, &next);
|
|
|
|
if (!next)
|
|
|
|
ival = sval[0];
|
|
|
|
} else {
|
|
|
|
ival = 255; /* special value meaning "don't set" */
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case TTY_OP_BOOL:
|
|
|
|
if (stricmp(sval, "yes") == 0 ||
|
|
|
|
stricmp(sval, "on") == 0 ||
|
|
|
|
stricmp(sval, "true") == 0 ||
|
|
|
|
stricmp(sval, "+") == 0)
|
|
|
|
ival = 1; /* true */
|
|
|
|
else if (stricmp(sval, "no") == 0 ||
|
|
|
|
stricmp(sval, "off") == 0 ||
|
|
|
|
stricmp(sval, "false") == 0 ||
|
|
|
|
stricmp(sval, "-") == 0)
|
|
|
|
ival = 0; /* false */
|
|
|
|
else
|
|
|
|
ival = (atoi(sval) != 0);
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
assert(0 && "Bad mode->type");
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* And write it into the output packet. The parameter
|
|
|
|
* value is formatted as a byte in SSH-1, but a uint32
|
|
|
|
* in SSH-2.
|
|
|
|
*/
|
|
|
|
put_byte(bs, mode->opcode);
|
|
|
|
if (ssh_version == 1)
|
|
|
|
put_byte(bs, ival);
|
|
|
|
else
|
|
|
|
put_uint32(bs, ival);
|
|
|
|
}
|
|
|
|
|
|
|
|
sfree(to_free);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Finish off with the terminal speeds (which are formatted as
|
|
|
|
* uint32 in both protocol versions) and the end marker.
|
|
|
|
*/
|
|
|
|
put_byte(bs, ssh_version == 1 ? SSH1_TTY_OP_ISPEED : SSH2_TTY_OP_ISPEED);
|
|
|
|
put_uint32(bs, ispeed);
|
|
|
|
put_byte(bs, ssh_version == 1 ? SSH1_TTY_OP_OSPEED : SSH2_TTY_OP_OSPEED);
|
|
|
|
put_uint32(bs, ospeed);
|
|
|
|
put_byte(bs, SSH_TTY_OP_END);
|
|
|
|
}
|
2018-09-19 19:37:12 +00:00
|
|
|
|
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Routine for allocating a new channel ID, given a means of finding
|
|
|
|
* the index field in a given channel structure.
|
|
|
|
*/
|
|
|
|
|
|
|
|
unsigned alloc_channel_id_general(tree234 *channels, size_t localid_offset)
|
|
|
|
{
|
|
|
|
const unsigned CHANNEL_NUMBER_OFFSET = 256;
|
|
|
|
search234_state ss;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* First-fit allocation of channel numbers: we always pick the
|
|
|
|
* lowest unused one.
|
|
|
|
*
|
|
|
|
* Every channel before that, and no channel after it, has an ID
|
|
|
|
* exactly equal to its tree index plus CHANNEL_NUMBER_OFFSET. So
|
|
|
|
* we can use the search234 system to identify the length of that
|
|
|
|
* initial sequence, in a single log-time pass down the channels
|
|
|
|
* tree.
|
|
|
|
*/
|
|
|
|
search234_start(&ss, channels);
|
|
|
|
while (ss.element) {
|
|
|
|
unsigned localid = *(unsigned *)((char *)ss.element + localid_offset);
|
|
|
|
if (localid == ss.index + CHANNEL_NUMBER_OFFSET)
|
|
|
|
search234_step(&ss, +1);
|
|
|
|
else
|
|
|
|
search234_step(&ss, -1);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Now ss.index gives exactly the number of channels in that
|
|
|
|
* initial sequence. So adding CHANNEL_NUMBER_OFFSET to it must
|
|
|
|
* give precisely the lowest unused channel number.
|
|
|
|
*/
|
|
|
|
return ss.index + CHANNEL_NUMBER_OFFSET;
|
|
|
|
}
|
2018-09-19 19:39:25 +00:00
|
|
|
|
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Functions for handling the comma-separated strings used to store
|
|
|
|
* lists of protocol identifiers in SSH-2.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int first_in_commasep_string(char const *needle, char const *haystack,
|
|
|
|
int haylen)
|
|
|
|
{
|
|
|
|
int needlen;
|
|
|
|
if (!needle || !haystack) /* protect against null pointers */
|
|
|
|
return 0;
|
|
|
|
needlen = strlen(needle);
|
|
|
|
|
|
|
|
if (haylen >= needlen && /* haystack is long enough */
|
|
|
|
!memcmp(needle, haystack, needlen) && /* initial match */
|
|
|
|
(haylen == needlen || haystack[needlen] == ',')
|
|
|
|
/* either , or EOS follows */
|
|
|
|
)
|
|
|
|
return 1;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
int in_commasep_string(char const *needle, char const *haystack, int haylen)
|
|
|
|
{
|
|
|
|
char *p;
|
|
|
|
|
|
|
|
if (!needle || !haystack) /* protect against null pointers */
|
|
|
|
return FALSE;
|
|
|
|
/*
|
|
|
|
* Is it at the start of the string?
|
|
|
|
*/
|
|
|
|
if (first_in_commasep_string(needle, haystack, haylen))
|
|
|
|
return TRUE;
|
|
|
|
/*
|
|
|
|
* If not, search for the next comma and resume after that.
|
|
|
|
* If no comma found, terminate.
|
|
|
|
*/
|
|
|
|
p = memchr(haystack, ',', haylen);
|
|
|
|
if (!p)
|
|
|
|
return FALSE;
|
|
|
|
/* + 1 to skip over comma */
|
|
|
|
return in_commasep_string(needle, p + 1, haylen - (p + 1 - haystack));
|
|
|
|
}
|
|
|
|
|
|
|
|
void add_to_commasep(strbuf *buf, const char *data)
|
|
|
|
{
|
|
|
|
if (buf->len > 0)
|
|
|
|
put_byte(buf, ',');
|
|
|
|
put_data(buf, data, strlen(data));
|
|
|
|
}
|
2018-09-19 19:40:21 +00:00
|
|
|
|
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Functions for translating SSH packet type codes into their symbolic
|
|
|
|
* string names.
|
|
|
|
*/
|
|
|
|
|
2018-09-24 12:29:09 +00:00
|
|
|
#define TRANSLATE_UNIVERSAL(y, name, value) \
|
|
|
|
if (type == value) return #name;
|
|
|
|
#define TRANSLATE_KEX(y, name, value, ctx) \
|
|
|
|
if (type == value && pkt_kctx == ctx) return #name;
|
|
|
|
#define TRANSLATE_AUTH(y, name, value, ctx) \
|
|
|
|
if (type == value && pkt_actx == ctx) return #name;
|
|
|
|
|
2018-09-19 19:40:21 +00:00
|
|
|
const char *ssh1_pkt_type(int type)
|
|
|
|
{
|
2018-09-24 12:29:09 +00:00
|
|
|
SSH1_MESSAGE_TYPES(TRANSLATE_UNIVERSAL, y);
|
2018-09-19 19:40:21 +00:00
|
|
|
return "unknown";
|
|
|
|
}
|
|
|
|
const char *ssh2_pkt_type(Pkt_KCtx pkt_kctx, Pkt_ACtx pkt_actx, int type)
|
|
|
|
{
|
2018-09-24 12:29:09 +00:00
|
|
|
SSH2_MESSAGE_TYPES(TRANSLATE_UNIVERSAL, TRANSLATE_KEX, TRANSLATE_AUTH, y);
|
2018-09-19 19:40:21 +00:00
|
|
|
return "unknown";
|
|
|
|
}
|
2018-09-24 12:29:09 +00:00
|
|
|
|
|
|
|
#undef TRANSLATE_UNIVERSAL
|
|
|
|
#undef TRANSLATE_KEX
|
|
|
|
#undef TRANSLATE_AUTH
|
2018-09-24 12:45:10 +00:00
|
|
|
|
Move most of ssh.c out into separate source files.
I've tried to separate out as many individually coherent changes from
this work as I could into their own commits, but here's where I run
out and have to commit the rest of this major refactoring as a
big-bang change.
Most of ssh.c is now no longer in ssh.c: all five of the main
coroutines that handle layers of the SSH-1 and SSH-2 protocols now
each have their own source file to live in, and a lot of the
supporting functions have moved into the appropriate one of those too.
The new abstraction is a vtable called 'PacketProtocolLayer', which
has an input and output packet queue. Each layer's main coroutine is
invoked from the method ssh_ppl_process_queue(), which is usually
(though not exclusively) triggered automatically when things are
pushed on the input queue. In SSH-2, the base layer is the transport
protocol, and it contains a pair of subsidiary queues by which it
passes some of its packets to the higher SSH-2 layers - first userauth
and then connection, which are peers at the same level, with the
former abdicating in favour of the latter at the appropriate moment.
SSH-1 is simpler: the whole login phase of the protocol (crypto setup
and authentication) is all in one module, and since SSH-1 has no
repeat key exchange, that setup layer abdicates in favour of the
connection phase when it's done.
ssh.c itself is now about a tenth of its old size (which all by itself
is cause for celebration!). Its main job is to set up all the layers,
hook them up to each other and to the BPP, and to funnel data back and
forth between that collection of modules and external things such as
the network and the terminal. Once it's set up a collection of packet
protocol layers, it communicates with them partly by calling methods
of the base layer (and if that's ssh2transport then it will delegate
some functionality to the corresponding methods of its higher layer),
and partly by talking directly to the connection layer no matter where
it is in the stack by means of the separate ConnectionLayer vtable
which I introduced in commit 8001dd4cb, and to which I've now added
quite a few extra methods replacing services that used to be internal
function calls within ssh.c.
(One effect of this is that the SSH-1 and SSH-2 channel storage is now
no longer shared - there are distinct struct types ssh1_channel and
ssh2_channel. That means a bit more code duplication, but on the plus
side, a lot fewer confusing conditionals in the middle of half-shared
functions, and less risk of a piece of SSH-1 escaping into SSH-2 or
vice versa, which I remember has happened at least once in the past.)
The bulk of this commit introduces the five new source files, their
common header sshppl.h and some shared supporting routines in
sshcommon.c, and rewrites nearly all of ssh.c itself. But it also
includes a couple of other changes that I couldn't separate easily
enough:
Firstly, there's a new handling for socket EOF, in which ssh.c sets an
'input_eof' flag in the BPP, and that responds by checking a flag that
tells it whether to report the EOF as an error or not. (This is the
main reason for those new BPP_READ / BPP_WAITFOR macros - they can
check the EOF flag every time the coroutine is resumed.)
Secondly, the error reporting itself is changed around again. I'd
expected to put some data fields in the public PacketProtocolLayer
structure that it could set to report errors in the same way as the
BPPs have been doing, but in the end, I decided propagating all those
data fields around was a pain and that even the BPPs shouldn't have
been doing it that way. So I've reverted to a system where everything
calls back to functions in ssh.c itself to report any connection-
ending condition. But there's a new family of those functions,
categorising the possible such conditions by semantics, and each one
has a different set of detailed effects (e.g. how rudely to close the
network connection, what exit status should be passed back to the
whole application, whether to send a disconnect message and/or display
a GUI error box).
I don't expect this to be immediately perfect: of course, the code has
been through a big upheaval, new bugs are expected, and I haven't been
able to do a full job of testing (e.g. I haven't tested every auth or
kex method). But I've checked that it _basically_ works - both SSH
protocols, all the different kinds of forwarding channel, more than
one auth method, Windows and Linux, connection sharing - and I think
it's now at the point where the easiest way to find further bugs is to
let it out into the wild and see what users can spot.
2018-09-24 17:28:16 +00:00
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Common helper function for clients and implementations of
|
|
|
|
* PacketProtocolLayer.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void ssh_ppl_replace(PacketProtocolLayer *old, PacketProtocolLayer *new)
|
|
|
|
{
|
|
|
|
new->bpp = old->bpp;
|
|
|
|
ssh_ppl_setup_queues(new, old->in_pq, old->out_pq);
|
|
|
|
new->selfptr = old->selfptr;
|
|
|
|
new->user_input = old->user_input;
|
New abstraction 'Seat', to pass to backends.
This is a new vtable-based abstraction which is passed to a backend in
place of Frontend, and it implements only the subset of the Frontend
functions needed by a backend. (Many other Frontend functions still
exist, notably the wide range of things called by terminal.c providing
platform-independent operations on the GUI terminal window.)
The purpose of making it a vtable is that this opens up the
possibility of creating a backend as an internal implementation detail
of some other activity, by providing just that one backend with a
custom Seat that implements the methods differently.
For example, this refactoring should make it feasible to directly
implement an SSH proxy type, aka the 'jump host' feature supported by
OpenSSH, aka 'open a secondary SSH session in MAINCHAN_DIRECT_TCP
mode, and then expose the main channel of that as the Socket for the
primary connection'. (Which of course you can already do by spawning
'plink -nc' as a separate proxy process, but this would permit it in
the _same_ process without anything getting confused.)
I've centralised a full set of stub methods in misc.c for the new
abstraction, which allows me to get rid of several annoying stubs in
the previous code. Also, while I'm here, I've moved a lot of
duplicated modalfatalbox() type functions from application main
program files into wincons.c / uxcons.c, which I think saves
duplication overall. (A minor visible effect is that the prefixes on
those console-based fatal error messages will now be more consistent
between applications.)
2018-10-11 18:58:42 +00:00
|
|
|
new->seat = old->seat;
|
Move most of ssh.c out into separate source files.
I've tried to separate out as many individually coherent changes from
this work as I could into their own commits, but here's where I run
out and have to commit the rest of this major refactoring as a
big-bang change.
Most of ssh.c is now no longer in ssh.c: all five of the main
coroutines that handle layers of the SSH-1 and SSH-2 protocols now
each have their own source file to live in, and a lot of the
supporting functions have moved into the appropriate one of those too.
The new abstraction is a vtable called 'PacketProtocolLayer', which
has an input and output packet queue. Each layer's main coroutine is
invoked from the method ssh_ppl_process_queue(), which is usually
(though not exclusively) triggered automatically when things are
pushed on the input queue. In SSH-2, the base layer is the transport
protocol, and it contains a pair of subsidiary queues by which it
passes some of its packets to the higher SSH-2 layers - first userauth
and then connection, which are peers at the same level, with the
former abdicating in favour of the latter at the appropriate moment.
SSH-1 is simpler: the whole login phase of the protocol (crypto setup
and authentication) is all in one module, and since SSH-1 has no
repeat key exchange, that setup layer abdicates in favour of the
connection phase when it's done.
ssh.c itself is now about a tenth of its old size (which all by itself
is cause for celebration!). Its main job is to set up all the layers,
hook them up to each other and to the BPP, and to funnel data back and
forth between that collection of modules and external things such as
the network and the terminal. Once it's set up a collection of packet
protocol layers, it communicates with them partly by calling methods
of the base layer (and if that's ssh2transport then it will delegate
some functionality to the corresponding methods of its higher layer),
and partly by talking directly to the connection layer no matter where
it is in the stack by means of the separate ConnectionLayer vtable
which I introduced in commit 8001dd4cb, and to which I've now added
quite a few extra methods replacing services that used to be internal
function calls within ssh.c.
(One effect of this is that the SSH-1 and SSH-2 channel storage is now
no longer shared - there are distinct struct types ssh1_channel and
ssh2_channel. That means a bit more code duplication, but on the plus
side, a lot fewer confusing conditionals in the middle of half-shared
functions, and less risk of a piece of SSH-1 escaping into SSH-2 or
vice versa, which I remember has happened at least once in the past.)
The bulk of this commit introduces the five new source files, their
common header sshppl.h and some shared supporting routines in
sshcommon.c, and rewrites nearly all of ssh.c itself. But it also
includes a couple of other changes that I couldn't separate easily
enough:
Firstly, there's a new handling for socket EOF, in which ssh.c sets an
'input_eof' flag in the BPP, and that responds by checking a flag that
tells it whether to report the EOF as an error or not. (This is the
main reason for those new BPP_READ / BPP_WAITFOR macros - they can
check the EOF flag every time the coroutine is resumed.)
Secondly, the error reporting itself is changed around again. I'd
expected to put some data fields in the public PacketProtocolLayer
structure that it could set to report errors in the same way as the
BPPs have been doing, but in the end, I decided propagating all those
data fields around was a pain and that even the BPPs shouldn't have
been doing it that way. So I've reverted to a system where everything
calls back to functions in ssh.c itself to report any connection-
ending condition. But there's a new family of those functions,
categorising the possible such conditions by semantics, and each one
has a different set of detailed effects (e.g. how rudely to close the
network connection, what exit status should be passed back to the
whole application, whether to send a disconnect message and/or display
a GUI error box).
I don't expect this to be immediately perfect: of course, the code has
been through a big upheaval, new bugs are expected, and I haven't been
able to do a full job of testing (e.g. I haven't tested every auth or
kex method). But I've checked that it _basically_ works - both SSH
protocols, all the different kinds of forwarding channel, more than
one auth method, Windows and Linux, connection sharing - and I think
it's now at the point where the easiest way to find further bugs is to
let it out into the wild and see what users can spot.
2018-09-24 17:28:16 +00:00
|
|
|
new->ssh = old->ssh;
|
|
|
|
|
|
|
|
*new->selfptr = new;
|
|
|
|
ssh_ppl_free(old);
|
|
|
|
|
|
|
|
/* The new layer might need to be the first one that sends a
|
|
|
|
* packet, so trigger a call to its main coroutine immediately. If
|
|
|
|
* it doesn't need to go first, the worst that will do is return
|
|
|
|
* straight away. */
|
|
|
|
queue_idempotent_callback(&new->ic_process_queue);
|
|
|
|
}
|
|
|
|
|
|
|
|
void ssh_ppl_free(PacketProtocolLayer *ppl)
|
|
|
|
{
|
|
|
|
delete_callbacks_for_context(ppl);
|
|
|
|
ppl->vt->free(ppl);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void ssh_ppl_ic_process_queue_callback(void *context)
|
|
|
|
{
|
|
|
|
PacketProtocolLayer *ppl = (PacketProtocolLayer *)context;
|
|
|
|
ssh_ppl_process_queue(ppl);
|
|
|
|
}
|
|
|
|
|
|
|
|
void ssh_ppl_setup_queues(PacketProtocolLayer *ppl,
|
|
|
|
PktInQueue *inq, PktOutQueue *outq)
|
|
|
|
{
|
|
|
|
ppl->in_pq = inq;
|
|
|
|
ppl->out_pq = outq;
|
|
|
|
ppl->in_pq->pqb.ic = &ppl->ic_process_queue;
|
|
|
|
ppl->ic_process_queue.fn = ssh_ppl_ic_process_queue_callback;
|
|
|
|
ppl->ic_process_queue.ctx = ppl;
|
|
|
|
|
|
|
|
/* If there's already something on the input queue, it will want
|
|
|
|
* handling immediately. */
|
|
|
|
if (pq_peek(ppl->in_pq))
|
|
|
|
queue_idempotent_callback(&ppl->ic_process_queue);
|
|
|
|
}
|
|
|
|
|
|
|
|
void ssh_ppl_user_output_string_and_free(PacketProtocolLayer *ppl, char *text)
|
|
|
|
{
|
|
|
|
/* Messages sent via this function are from the SSH layer, not
|
|
|
|
* from the server-side process, so they always have the stderr
|
|
|
|
* flag set. */
|
New abstraction 'Seat', to pass to backends.
This is a new vtable-based abstraction which is passed to a backend in
place of Frontend, and it implements only the subset of the Frontend
functions needed by a backend. (Many other Frontend functions still
exist, notably the wide range of things called by terminal.c providing
platform-independent operations on the GUI terminal window.)
The purpose of making it a vtable is that this opens up the
possibility of creating a backend as an internal implementation detail
of some other activity, by providing just that one backend with a
custom Seat that implements the methods differently.
For example, this refactoring should make it feasible to directly
implement an SSH proxy type, aka the 'jump host' feature supported by
OpenSSH, aka 'open a secondary SSH session in MAINCHAN_DIRECT_TCP
mode, and then expose the main channel of that as the Socket for the
primary connection'. (Which of course you can already do by spawning
'plink -nc' as a separate proxy process, but this would permit it in
the _same_ process without anything getting confused.)
I've centralised a full set of stub methods in misc.c for the new
abstraction, which allows me to get rid of several annoying stubs in
the previous code. Also, while I'm here, I've moved a lot of
duplicated modalfatalbox() type functions from application main
program files into wincons.c / uxcons.c, which I think saves
duplication overall. (A minor visible effect is that the prefixes on
those console-based fatal error messages will now be more consistent
between applications.)
2018-10-11 18:58:42 +00:00
|
|
|
seat_stderr(ppl->seat, text, strlen(text));
|
Move most of ssh.c out into separate source files.
I've tried to separate out as many individually coherent changes from
this work as I could into their own commits, but here's where I run
out and have to commit the rest of this major refactoring as a
big-bang change.
Most of ssh.c is now no longer in ssh.c: all five of the main
coroutines that handle layers of the SSH-1 and SSH-2 protocols now
each have their own source file to live in, and a lot of the
supporting functions have moved into the appropriate one of those too.
The new abstraction is a vtable called 'PacketProtocolLayer', which
has an input and output packet queue. Each layer's main coroutine is
invoked from the method ssh_ppl_process_queue(), which is usually
(though not exclusively) triggered automatically when things are
pushed on the input queue. In SSH-2, the base layer is the transport
protocol, and it contains a pair of subsidiary queues by which it
passes some of its packets to the higher SSH-2 layers - first userauth
and then connection, which are peers at the same level, with the
former abdicating in favour of the latter at the appropriate moment.
SSH-1 is simpler: the whole login phase of the protocol (crypto setup
and authentication) is all in one module, and since SSH-1 has no
repeat key exchange, that setup layer abdicates in favour of the
connection phase when it's done.
ssh.c itself is now about a tenth of its old size (which all by itself
is cause for celebration!). Its main job is to set up all the layers,
hook them up to each other and to the BPP, and to funnel data back and
forth between that collection of modules and external things such as
the network and the terminal. Once it's set up a collection of packet
protocol layers, it communicates with them partly by calling methods
of the base layer (and if that's ssh2transport then it will delegate
some functionality to the corresponding methods of its higher layer),
and partly by talking directly to the connection layer no matter where
it is in the stack by means of the separate ConnectionLayer vtable
which I introduced in commit 8001dd4cb, and to which I've now added
quite a few extra methods replacing services that used to be internal
function calls within ssh.c.
(One effect of this is that the SSH-1 and SSH-2 channel storage is now
no longer shared - there are distinct struct types ssh1_channel and
ssh2_channel. That means a bit more code duplication, but on the plus
side, a lot fewer confusing conditionals in the middle of half-shared
functions, and less risk of a piece of SSH-1 escaping into SSH-2 or
vice versa, which I remember has happened at least once in the past.)
The bulk of this commit introduces the five new source files, their
common header sshppl.h and some shared supporting routines in
sshcommon.c, and rewrites nearly all of ssh.c itself. But it also
includes a couple of other changes that I couldn't separate easily
enough:
Firstly, there's a new handling for socket EOF, in which ssh.c sets an
'input_eof' flag in the BPP, and that responds by checking a flag that
tells it whether to report the EOF as an error or not. (This is the
main reason for those new BPP_READ / BPP_WAITFOR macros - they can
check the EOF flag every time the coroutine is resumed.)
Secondly, the error reporting itself is changed around again. I'd
expected to put some data fields in the public PacketProtocolLayer
structure that it could set to report errors in the same way as the
BPPs have been doing, but in the end, I decided propagating all those
data fields around was a pain and that even the BPPs shouldn't have
been doing it that way. So I've reverted to a system where everything
calls back to functions in ssh.c itself to report any connection-
ending condition. But there's a new family of those functions,
categorising the possible such conditions by semantics, and each one
has a different set of detailed effects (e.g. how rudely to close the
network connection, what exit status should be passed back to the
whole application, whether to send a disconnect message and/or display
a GUI error box).
I don't expect this to be immediately perfect: of course, the code has
been through a big upheaval, new bugs are expected, and I haven't been
able to do a full job of testing (e.g. I haven't tested every auth or
kex method). But I've checked that it _basically_ works - both SSH
protocols, all the different kinds of forwarding channel, more than
one auth method, Windows and Linux, connection sharing - and I think
it's now at the point where the easiest way to find further bugs is to
let it out into the wild and see what users can spot.
2018-09-24 17:28:16 +00:00
|
|
|
sfree(text);
|
|
|
|
}
|
|
|
|
|
2018-09-24 12:45:10 +00:00
|
|
|
/* ----------------------------------------------------------------------
|
2018-09-24 17:08:09 +00:00
|
|
|
* Common helper functions for clients and implementations of
|
|
|
|
* BinaryPacketProtocol.
|
2018-09-24 12:45:10 +00:00
|
|
|
*/
|
|
|
|
|
2018-09-24 17:08:09 +00:00
|
|
|
static void ssh_bpp_input_raw_data_callback(void *context)
|
|
|
|
{
|
|
|
|
BinaryPacketProtocol *bpp = (BinaryPacketProtocol *)context;
|
|
|
|
ssh_bpp_handle_input(bpp);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void ssh_bpp_output_packet_callback(void *context)
|
|
|
|
{
|
|
|
|
BinaryPacketProtocol *bpp = (BinaryPacketProtocol *)context;
|
|
|
|
ssh_bpp_handle_output(bpp);
|
|
|
|
}
|
|
|
|
|
|
|
|
void ssh_bpp_common_setup(BinaryPacketProtocol *bpp)
|
|
|
|
{
|
|
|
|
pq_in_init(&bpp->in_pq);
|
|
|
|
pq_out_init(&bpp->out_pq);
|
Move most of ssh.c out into separate source files.
I've tried to separate out as many individually coherent changes from
this work as I could into their own commits, but here's where I run
out and have to commit the rest of this major refactoring as a
big-bang change.
Most of ssh.c is now no longer in ssh.c: all five of the main
coroutines that handle layers of the SSH-1 and SSH-2 protocols now
each have their own source file to live in, and a lot of the
supporting functions have moved into the appropriate one of those too.
The new abstraction is a vtable called 'PacketProtocolLayer', which
has an input and output packet queue. Each layer's main coroutine is
invoked from the method ssh_ppl_process_queue(), which is usually
(though not exclusively) triggered automatically when things are
pushed on the input queue. In SSH-2, the base layer is the transport
protocol, and it contains a pair of subsidiary queues by which it
passes some of its packets to the higher SSH-2 layers - first userauth
and then connection, which are peers at the same level, with the
former abdicating in favour of the latter at the appropriate moment.
SSH-1 is simpler: the whole login phase of the protocol (crypto setup
and authentication) is all in one module, and since SSH-1 has no
repeat key exchange, that setup layer abdicates in favour of the
connection phase when it's done.
ssh.c itself is now about a tenth of its old size (which all by itself
is cause for celebration!). Its main job is to set up all the layers,
hook them up to each other and to the BPP, and to funnel data back and
forth between that collection of modules and external things such as
the network and the terminal. Once it's set up a collection of packet
protocol layers, it communicates with them partly by calling methods
of the base layer (and if that's ssh2transport then it will delegate
some functionality to the corresponding methods of its higher layer),
and partly by talking directly to the connection layer no matter where
it is in the stack by means of the separate ConnectionLayer vtable
which I introduced in commit 8001dd4cb, and to which I've now added
quite a few extra methods replacing services that used to be internal
function calls within ssh.c.
(One effect of this is that the SSH-1 and SSH-2 channel storage is now
no longer shared - there are distinct struct types ssh1_channel and
ssh2_channel. That means a bit more code duplication, but on the plus
side, a lot fewer confusing conditionals in the middle of half-shared
functions, and less risk of a piece of SSH-1 escaping into SSH-2 or
vice versa, which I remember has happened at least once in the past.)
The bulk of this commit introduces the five new source files, their
common header sshppl.h and some shared supporting routines in
sshcommon.c, and rewrites nearly all of ssh.c itself. But it also
includes a couple of other changes that I couldn't separate easily
enough:
Firstly, there's a new handling for socket EOF, in which ssh.c sets an
'input_eof' flag in the BPP, and that responds by checking a flag that
tells it whether to report the EOF as an error or not. (This is the
main reason for those new BPP_READ / BPP_WAITFOR macros - they can
check the EOF flag every time the coroutine is resumed.)
Secondly, the error reporting itself is changed around again. I'd
expected to put some data fields in the public PacketProtocolLayer
structure that it could set to report errors in the same way as the
BPPs have been doing, but in the end, I decided propagating all those
data fields around was a pain and that even the BPPs shouldn't have
been doing it that way. So I've reverted to a system where everything
calls back to functions in ssh.c itself to report any connection-
ending condition. But there's a new family of those functions,
categorising the possible such conditions by semantics, and each one
has a different set of detailed effects (e.g. how rudely to close the
network connection, what exit status should be passed back to the
whole application, whether to send a disconnect message and/or display
a GUI error box).
I don't expect this to be immediately perfect: of course, the code has
been through a big upheaval, new bugs are expected, and I haven't been
able to do a full job of testing (e.g. I haven't tested every auth or
kex method). But I've checked that it _basically_ works - both SSH
protocols, all the different kinds of forwarding channel, more than
one auth method, Windows and Linux, connection sharing - and I think
it's now at the point where the easiest way to find further bugs is to
let it out into the wild and see what users can spot.
2018-09-24 17:28:16 +00:00
|
|
|
bpp->input_eof = FALSE;
|
2018-09-24 17:08:09 +00:00
|
|
|
bpp->ic_in_raw.fn = ssh_bpp_input_raw_data_callback;
|
|
|
|
bpp->ic_in_raw.ctx = bpp;
|
|
|
|
bpp->ic_out_pq.fn = ssh_bpp_output_packet_callback;
|
|
|
|
bpp->ic_out_pq.ctx = bpp;
|
|
|
|
bpp->out_pq.pqb.ic = &bpp->ic_out_pq;
|
|
|
|
}
|
|
|
|
|
|
|
|
void ssh_bpp_free(BinaryPacketProtocol *bpp)
|
|
|
|
{
|
|
|
|
delete_callbacks_for_context(bpp);
|
|
|
|
bpp->vt->free(bpp);
|
|
|
|
}
|
|
|
|
|
2018-09-24 17:14:33 +00:00
|
|
|
void ssh2_bpp_queue_disconnect(BinaryPacketProtocol *bpp,
|
|
|
|
const char *msg, int category)
|
|
|
|
{
|
|
|
|
PktOut *pkt = ssh_bpp_new_pktout(bpp, SSH2_MSG_DISCONNECT);
|
|
|
|
put_uint32(pkt, category);
|
|
|
|
put_stringz(pkt, msg);
|
|
|
|
put_stringz(pkt, "en"); /* language tag */
|
|
|
|
pq_push(&bpp->out_pq, pkt);
|
|
|
|
}
|
|
|
|
|
2018-09-24 12:45:10 +00:00
|
|
|
#define BITMAP_UNIVERSAL(y, name, value) \
|
|
|
|
| (value >= y && value < y+32 ? 1UL << (value-y) : 0)
|
|
|
|
#define BITMAP_CONDITIONAL(y, name, value, ctx) \
|
|
|
|
BITMAP_UNIVERSAL(y, name, value)
|
|
|
|
#define SSH2_BITMAP_WORD(y) \
|
|
|
|
(0 SSH2_MESSAGE_TYPES(BITMAP_UNIVERSAL, BITMAP_CONDITIONAL, \
|
|
|
|
BITMAP_CONDITIONAL, (32*y)))
|
|
|
|
|
|
|
|
int ssh2_bpp_check_unimplemented(BinaryPacketProtocol *bpp, PktIn *pktin)
|
|
|
|
{
|
|
|
|
static const unsigned valid_bitmap[] = {
|
|
|
|
SSH2_BITMAP_WORD(0),
|
|
|
|
SSH2_BITMAP_WORD(1),
|
|
|
|
SSH2_BITMAP_WORD(2),
|
|
|
|
SSH2_BITMAP_WORD(3),
|
|
|
|
SSH2_BITMAP_WORD(4),
|
|
|
|
SSH2_BITMAP_WORD(5),
|
|
|
|
SSH2_BITMAP_WORD(6),
|
|
|
|
SSH2_BITMAP_WORD(7),
|
|
|
|
};
|
|
|
|
|
|
|
|
if (pktin->type < 0x100 &&
|
|
|
|
!((valid_bitmap[pktin->type >> 5] >> (pktin->type & 0x1F)) & 1)) {
|
|
|
|
PktOut *pkt = ssh_bpp_new_pktout(bpp, SSH2_MSG_UNIMPLEMENTED);
|
|
|
|
put_uint32(pkt, pktin->sequence);
|
2018-09-24 17:08:09 +00:00
|
|
|
pq_push(&bpp->out_pq, pkt);
|
2018-09-24 12:45:10 +00:00
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
#undef BITMAP_UNIVERSAL
|
|
|
|
#undef BITMAP_CONDITIONAL
|
|
|
|
#undef SSH1_BITMAP_WORD
|
2018-09-24 13:15:32 +00:00
|
|
|
|
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Function to check a host key against any manually configured in Conf.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int verify_ssh_manual_host_key(
|
|
|
|
Conf *conf, const char *fingerprint, ssh_key *key)
|
|
|
|
{
|
|
|
|
if (!conf_get_str_nthstrkey(conf, CONF_ssh_manual_hostkeys, 0))
|
|
|
|
return -1; /* no manual keys configured */
|
|
|
|
|
|
|
|
if (fingerprint) {
|
|
|
|
/*
|
|
|
|
* The fingerprint string we've been given will have things
|
|
|
|
* like 'ssh-rsa 2048' at the front of it. Strip those off and
|
|
|
|
* narrow down to just the colon-separated hex block at the
|
|
|
|
* end of the string.
|
|
|
|
*/
|
|
|
|
const char *p = strrchr(fingerprint, ' ');
|
|
|
|
fingerprint = p ? p+1 : fingerprint;
|
|
|
|
/* Quick sanity checks, including making sure it's in lowercase */
|
|
|
|
assert(strlen(fingerprint) == 16*3 - 1);
|
|
|
|
assert(fingerprint[2] == ':');
|
|
|
|
assert(fingerprint[strspn(fingerprint, "0123456789abcdef:")] == 0);
|
|
|
|
|
|
|
|
if (conf_get_str_str_opt(conf, CONF_ssh_manual_hostkeys, fingerprint))
|
|
|
|
return 1; /* success */
|
|
|
|
}
|
|
|
|
|
|
|
|
if (key) {
|
|
|
|
/*
|
|
|
|
* Construct the base64-encoded public key blob and see if
|
|
|
|
* that's listed.
|
|
|
|
*/
|
|
|
|
strbuf *binblob;
|
|
|
|
char *base64blob;
|
|
|
|
int atoms, i;
|
|
|
|
binblob = strbuf_new();
|
|
|
|
ssh_key_public_blob(key, BinarySink_UPCAST(binblob));
|
|
|
|
atoms = (binblob->len + 2) / 3;
|
|
|
|
base64blob = snewn(atoms * 4 + 1, char);
|
|
|
|
for (i = 0; i < atoms; i++)
|
|
|
|
base64_encode_atom(binblob->u + 3*i,
|
|
|
|
binblob->len - 3*i, base64blob + 4*i);
|
|
|
|
base64blob[atoms * 4] = '\0';
|
|
|
|
strbuf_free(binblob);
|
|
|
|
if (conf_get_str_str_opt(conf, CONF_ssh_manual_hostkeys, base64blob)) {
|
|
|
|
sfree(base64blob);
|
|
|
|
return 1; /* success */
|
|
|
|
}
|
|
|
|
sfree(base64blob);
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
Move most of ssh.c out into separate source files.
I've tried to separate out as many individually coherent changes from
this work as I could into their own commits, but here's where I run
out and have to commit the rest of this major refactoring as a
big-bang change.
Most of ssh.c is now no longer in ssh.c: all five of the main
coroutines that handle layers of the SSH-1 and SSH-2 protocols now
each have their own source file to live in, and a lot of the
supporting functions have moved into the appropriate one of those too.
The new abstraction is a vtable called 'PacketProtocolLayer', which
has an input and output packet queue. Each layer's main coroutine is
invoked from the method ssh_ppl_process_queue(), which is usually
(though not exclusively) triggered automatically when things are
pushed on the input queue. In SSH-2, the base layer is the transport
protocol, and it contains a pair of subsidiary queues by which it
passes some of its packets to the higher SSH-2 layers - first userauth
and then connection, which are peers at the same level, with the
former abdicating in favour of the latter at the appropriate moment.
SSH-1 is simpler: the whole login phase of the protocol (crypto setup
and authentication) is all in one module, and since SSH-1 has no
repeat key exchange, that setup layer abdicates in favour of the
connection phase when it's done.
ssh.c itself is now about a tenth of its old size (which all by itself
is cause for celebration!). Its main job is to set up all the layers,
hook them up to each other and to the BPP, and to funnel data back and
forth between that collection of modules and external things such as
the network and the terminal. Once it's set up a collection of packet
protocol layers, it communicates with them partly by calling methods
of the base layer (and if that's ssh2transport then it will delegate
some functionality to the corresponding methods of its higher layer),
and partly by talking directly to the connection layer no matter where
it is in the stack by means of the separate ConnectionLayer vtable
which I introduced in commit 8001dd4cb, and to which I've now added
quite a few extra methods replacing services that used to be internal
function calls within ssh.c.
(One effect of this is that the SSH-1 and SSH-2 channel storage is now
no longer shared - there are distinct struct types ssh1_channel and
ssh2_channel. That means a bit more code duplication, but on the plus
side, a lot fewer confusing conditionals in the middle of half-shared
functions, and less risk of a piece of SSH-1 escaping into SSH-2 or
vice versa, which I remember has happened at least once in the past.)
The bulk of this commit introduces the five new source files, their
common header sshppl.h and some shared supporting routines in
sshcommon.c, and rewrites nearly all of ssh.c itself. But it also
includes a couple of other changes that I couldn't separate easily
enough:
Firstly, there's a new handling for socket EOF, in which ssh.c sets an
'input_eof' flag in the BPP, and that responds by checking a flag that
tells it whether to report the EOF as an error or not. (This is the
main reason for those new BPP_READ / BPP_WAITFOR macros - they can
check the EOF flag every time the coroutine is resumed.)
Secondly, the error reporting itself is changed around again. I'd
expected to put some data fields in the public PacketProtocolLayer
structure that it could set to report errors in the same way as the
BPPs have been doing, but in the end, I decided propagating all those
data fields around was a pain and that even the BPPs shouldn't have
been doing it that way. So I've reverted to a system where everything
calls back to functions in ssh.c itself to report any connection-
ending condition. But there's a new family of those functions,
categorising the possible such conditions by semantics, and each one
has a different set of detailed effects (e.g. how rudely to close the
network connection, what exit status should be passed back to the
whole application, whether to send a disconnect message and/or display
a GUI error box).
I don't expect this to be immediately perfect: of course, the code has
been through a big upheaval, new bugs are expected, and I haven't been
able to do a full job of testing (e.g. I haven't tested every auth or
kex method). But I've checked that it _basically_ works - both SSH
protocols, all the different kinds of forwarding channel, more than
one auth method, Windows and Linux, connection sharing - and I think
it's now at the point where the easiest way to find further bugs is to
let it out into the wild and see what users can spot.
2018-09-24 17:28:16 +00:00
|
|
|
|
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Common get_specials function for the two SSH-1 layers.
|
|
|
|
*/
|
|
|
|
|
|
|
|
int ssh1_common_get_specials(
|
|
|
|
PacketProtocolLayer *ppl, add_special_fn_t add_special, void *ctx)
|
|
|
|
{
|
|
|
|
/*
|
|
|
|
* Don't bother offering IGNORE if we've decided the remote
|
|
|
|
* won't cope with it, since we wouldn't bother sending it if
|
|
|
|
* asked anyway.
|
|
|
|
*/
|
|
|
|
if (!(ppl->remote_bugs & BUG_CHOKES_ON_SSH1_IGNORE)) {
|
|
|
|
add_special(ctx, "IGNORE message", SS_NOP, 0);
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* ----------------------------------------------------------------------
|
|
|
|
* Other miscellaneous utility functions.
|
|
|
|
*/
|
|
|
|
|
|
|
|
void free_rportfwd(struct ssh_rportfwd *rpf)
|
|
|
|
{
|
|
|
|
if (rpf) {
|
|
|
|
sfree(rpf->log_description);
|
|
|
|
sfree(rpf->shost);
|
|
|
|
sfree(rpf->dhost);
|
|
|
|
sfree(rpf);
|
|
|
|
}
|
|
|
|
}
|