We currently just ignore them. This is one reason the hsm (in some places)
explicitly calls log_broken so we get some idea.
This was the only subdaemon which had a NULL msgcb and msgname, so eliminate
those checks in subd.c.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Code changes:
1. Expose daemon_poll() so lightningd can call it directly, which avoids us
having store a global and document it.
2. Remove the (undocumented, unused, forgotten) --rpc-file="" option to disable
JSON RPC.
3. Move the ickiness of finding the executable path into subd.c, so it doesn't
distract from lightningd.c overview.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We're going to call out to subds for memleak detection, and the disabler
looks like a memleak if we're inside a callback.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
I crashed the HSMD, and it gave no output at all. That's because we
were only reading the status fd when we were waiting for a reply.
Fix this by using a separate request fd and status fd, which also means
that hsm_sync_read() is no longer required.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
(This was sitting in my gossip-enchancement patch queue, but it simplifies
this set too, so I moved it here).
In 94711969f we added an explicit gossip_index so when gossipd gets
peers back from other daemons, it knows what gossip it has sent (since
gossipd can send gossip after the other daemon is already complete).
This solution is insufficient for the more general case where gossipd
wants to send other messages reliably, so replace it with the other
solution: have gossipd drain the "gossip fd" which the daemon returns.
This turns out to be quite simple, and is probably how I should have
done it originally :(
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Use NULL on the callback to mean "clear the slot", and call it.
We have do this in two places: the old daemon might die, or the new
daemon might start first.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We always hand in "NULL" (which means use tal_len on the msg), except
for two places which do that manually for no good reason.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Several daemons (onchaind, hsm) want to use the status messages, but
don't communicate with peers. The coming changes made them drag in
more code they didn't need, so instead we have a different
non-overlapping type.
We combine the status_received_errmsg and status_sent_errmsg
into a single status_peer_error, with the presence or not of the
'error_for_them' field indicating direction.
We also rename status_fatal_connection_lost() to
peer_failed_connection_lost() to fit in.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This means the caller needs to supply an explicit log to base the
subd log on, and also a callback for error handling.
The callback is kind of ugly, but it gets reworked towards the end
of this series.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We usually did this, but sometimes they were named after what they did,
rather than what they cleaned up.
There are still a few exceptions:
1. I didn't bother creating destroy_xxx wrappers for htable routines
which already existed.
2. Sometimes destructors really are used for side-effects (eg. to simply
mark that something was freed): these are clearer with boutique names.
3. Generally destructors are static, but they don't need to be: in some
cases we attach a destructor then remove it later, or only attach
to *some* cases. These are best with qualifiers in the destroy_<type>
name.
Suggested-by: @ZmnSCPxj
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Much like the database; peer contains id, address, channel contains
per-channel information. Where we create a channel, we always create
the peer too.
For the moment, peer->log and channel->log coexist side-by-side, to
reduce some of the churn.
Note that this changes the API to dev-forget-channel: if we have more
than one channel, we insist they specify the short-channel-id.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Now we have wirestring, this is much more natural. And with the
24M length limit, we needn't be so concerned about dumping 64k peer
messages in hex.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This makes much more sense when you ask for a specific peer's log.
Also, we put the peerid rather than pid ().
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This means we print out the correct path with --debugger, which
can be vital if there are multiple binaries (eg. compiled vs installed).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
As demonstrated in the test at the end of this series, openingd dying
spontaneously causes the conn to be freed which causes the subd to be
destroyed, which fails the peer, which hits the db.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This is a bit messier than I'd like, but we want to clearly remove all
dev code (not just have it uncalled), so we remove fields and functions
altogether rather than stub them out. This means we put #ifdefs in callers
in some places, but at least it's explicit.
We still run tests, but only a subset, and we run with NO_VALGRIND under
Travis to avoid increasing test times too much.
See-also: #176
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
There are now only two kinds of subdaemons: global ones (hsmd, gossipd) and
per-peer ones. We can handle many callbacks internally now.
We can have a handler to set a new peer owner, and automatically do
the cleanup of the old one if necessary, since we now know which ones
are per-peer.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We have to do a dance when we get a reconnect in openingd, because we
don't normally expect to free both owner and peer. It's a layering
violation: freeing a peer should clean up the owner's pointer to it,
to avoid a double free, and we can eliminate this dance.
The free order is now different, and the test_reconnect_openingd was
overprecise.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We were sending a channeld message to onchaind, which was v. confusing
due to overlap. We make all the numbers distinct, which means we can
also add an assert() that it's valid for that daemon, which catches
such errors immediately.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Either when it exits with a signal, or sends an error status message.
Then we make test_lightningd.py use it.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This change is really to allow us to have a --dev-fail-on-subdaemon-fail option
so we can handle failures from subdaemons generically.
It also neatens handling so we can have an explicit callback for "peer
did something wrong" (which matters if we want to close the channel in
that case).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Some fields were redundant, some are simply moved into 'struct lightningd'.
All routines updated to hand 'struct lightningd *ld' now.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Also, we split the more sophisticated json_add helpers to avoid pulling in
everything into lightning-cli, and unify the routines to print struct
short_channel_id (it's ':', not '/' too).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We had a terrible hack in gossip when a peer didn't exist. Formalize
a pattern when code+200 is a failure (with no fds passed), and use it
here.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This matters in one case: channeld receiving a bad message is a
permenant failure, whereas losing a connection is transient.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We use a file descriptor, so when we consume an entry, we move past it
(and everyone shares a file offset, so this works).
The file contains packet names prefixed by - (treat fd as closed when
we try to write this packet), + (write the packet then ensure the file
descriptor fails), or @ ("lose" the packet then ensure the file
descriptor fails).
The sync and async peer-write functions hook this in automatically.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Header from folded patch 'test-run-cryptomsg__fix_compilation.patch':
test/run-cryptomsg: fix compilation.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
If a peer dies, and then we get a reply, that can cause access after free.
The usual way to handle this is to make the request a child of the peer,
but in fact we still want to catch (and disard) it, so it's a little
more complex internally.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
The STDOUT fd being reused as communication sockets with other daemons
was causing some unexpected crashes if the sub-daemon wrote something,
e.g., using `log_*`. Not closing it should avoid that conflict.
Instead of indicating where to place the fd, you say how many: the
fd array gets passed into the callback.
This is also clearer for the users.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
subd_req() needs to get the type before it calls subd_send_msg, because
if it's take() then msg_enqueue() may reallocate.
Which also made me realize that subd_send_message() should not try to dup,
since msg_enqueue() handles that itself.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We have some duplication in handling queues, so this is an attempt at
deduplicating some of that work. `daemon_conn` now uses the
`msg_queue` and `channeld` was also migrated to `msg_queue`. At the
same time I made `msg_queue` create a copy of the messages or takes
over messages marked with `take()`. This should make cleaning up
messages easier.
This uses a single fd for both status and control.
To make this work, we enforce the convention that replies are the same
as requests + 100, and that their name ends in "_REPLY".
This also means that various daemons can simply exit when done; there's
no race between reading request and closing status fds.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
