Commit graph

1111 commits

Author SHA1 Message Date
John Barboza
31b9b6bc1c clear onchaind subd in channel on error
When the subdaemon closes unexpectedly, clear the reference and
update the channel billboard with the description
2018-03-04 16:29:44 +01:00
ZmnSCPxj
1f6008689d payalgo: Throttle pay command if failure is due to blockheight disagreement. 2018-03-03 18:19:53 +01:00
John Barboza
b50912d8b1 Print node alias when listing peers 2018-03-02 02:25:30 +00:00
Rusty Russell
c5d41a23d7 short_channel_id: just use structeq.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-03-01 23:33:56 +01:00
Rusty Russell
042d5d13f5 short_channel_id: don't use bitfields.
I leave all the now-unnecessary accessors in place to avoid churn, but
the use of bitfields has been more pain than help.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-03-01 23:33:56 +01:00
Igor Cota
cb6820d445 Do preimage, hash and consequentially bolt11 generation before we store the invoice in db.
This way we store bolt11 to the table as well, in case a client needs it later
2018-03-01 03:55:48 +00:00
Christian Decker
a90502d917 jsonrpc: Refuse to forget a channel with an open HTLC
... instead provide useful suggestions as to how to close it.

Signed-off-by: Christian Decker <decker.christian@gmail.com>
2018-03-01 03:50:50 +00:00
Christian Decker
aef5780f36 htlc: Extracted htlc detection from the channel destructor
Let's have a simple function that allows us to check whether a channel
still has an HTLC open.

Signed-off-by: Christian Decker <decker.christian@gmail.com>
2018-03-01 03:50:50 +00:00
Corné Plooy
f32ebb7dca Indentation fix. 2018-02-28 21:21:25 +01:00
Corné Plooy
1e879de89e json-rpc: connection_complete_ok and json_command_malformed should never receive NULL arguments. Pass "null" instead, where needed. 2018-02-28 21:21:25 +01:00
Corné Plooy
d86ad15b53 json-rpc: translate NULL into "null" instead of "(null)". 2018-02-28 21:21:25 +01:00
practicalswift
dbb7ea1e41 Adjust log level for logging disconnected RPC users ("Command failed after jcon close")
This is a follow-up to #1121
2018-02-28 11:38:22 +01:00
ZmnSCPxj
978e5c67d8 invoices: Remove persistent in-memory invoice structures. 2018-02-28 11:17:08 +01:00
ZmnSCPxj
f05c86618c invoices: Change iteration interface to be more abstract. 2018-02-28 11:17:08 +01:00
ZmnSCPxj
9b4c6699f9 invoices: Semantically separate invoice details from invoice.
In preparation for removing in-memory invoice structures.
Invoice details are requested rarely anyway.
2018-02-28 11:17:08 +01:00
practicalswift
a20cf5463d Adjust log level for logging disconnected RPC users ("Abandoning command" + "Command returned result after jcon close")
The following command can be used to trigger these messages:

```
$ timeout 0.01 cli/lightning-cli connect [insert-syntactically-valid-peer-id-here] 123.123.123.123 # where 123.123.123.123 is unreachable
```
2018-02-28 01:53:52 +00:00
ZmnSCPxj
73cda2f2ae payalgo: Report route, and result of trying route. 2018-02-27 13:38:32 +01:00
ZmnSCPxj
21cfec816d payalgo: Report number of tries on failure. 2018-02-27 13:38:32 +01:00
ZmnSCPxj
6c9d81ef42 payalgo: Remove reporting PAY_TRY_ANOTHER_ROUTE and PAY_UNPARSEABLE_ONION.
These error codes will cause `pay` to retry, so `pay` will never
actually report those error codes.
Those error codes will only get reported at the `sendpay` level.
2018-02-27 13:38:32 +01:00
Rusty Russell
b7ed5670d5 lightningd: close and reopen db across fork for daemonize
Fixes: #1092
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-27 00:09:34 +01:00
Rusty Russell
cf3f19524e gossip: formalize passing of siphash_seed.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-26 06:35:02 +00:00
ZmnSCPxj
f7f55edcdb wireaddr: Do hostname resolution in parse_wireaddr.
Fixes: #834
2018-02-26 03:13:55 +00:00
ZmnSCPxj
bd07a91782 jsonrpc: Let Base58 '2..' address work on both regtest and testnet. 2018-02-26 03:09:15 +00:00
John Barboza
b028a0a439 invoice: add fallback address to invoice command
* Modifies invoice command to have the following format
  invoice <msatoshi> <label> <desc> <?expiry> <?fallbackaddr>
* Adds support for Segwit bcrt1 addresses for withdraw
* Add test case for fallback address in invoice creation
* Create a common json_tok_address_scriptpubkey to be used
  by invoice and withdraw commands.
2018-02-26 03:09:15 +00:00
ZmnSCPxj
d23650d2ed Explain fuzz initial values. 2018-02-26 02:36:27 +00:00
ZmnSCPxj
2db6ff9186 gossip_control: Change defult fuzz of getroute to 75%. 2018-02-26 02:36:27 +00:00
ZmnSCPxj
6767434ea9 routing: Use siphash24 for route randomization
Primary idea by @rustyrussell
2018-02-26 02:36:27 +00:00
ZmnSCPxj
61d0b50c54 payalgo: Randomize paths as long as we respect maxfeepercent. 2018-02-26 02:36:27 +00:00
ZmnSCPxj
9693843141 payalgo: Change sendpay_parent to try_parent.
Now try_parent is used as the parent for all allocations needed
  for a try.
2018-02-26 02:36:27 +00:00
ZmnSCPxj
216c52940d gossipd: Add fuzz and seed to getroute request. 2018-02-26 02:36:27 +00:00
ZmnSCPxj
2b782ee436 gossip_control: Add fuzz and seed parameters.
In preparation for improved route randomization.
2018-02-26 02:36:27 +00:00
ZmnSCPxj
e34be575aa lightning/peer_control.h: Minor cleanup of nonexistent functions. 2018-02-25 20:41:50 +00:00
Rusty Russell
50171d3e1a lightningd: add --bitcoin-cli arg for testing.
And remove unused bitcoin_datadir and BITCOIN_CLI.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-24 10:37:28 +01:00
Rusty Russell
f0a12c5c23 bitcoind: retry after one second if a call fails.
There are two recurring calls: the estimatefee call and the
getblockcount call.  Currently we simply discard them on error, the
timer isn't rearmed.

This should fix a number of cases where bitcoind has an intermittant
failure and lightningd simply stops collecting blocks.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-24 10:37:28 +01:00
Rusty Russell
0204c44243 bitcoind: allow processing callbacks to indicate that a failure is spurious.
In particular, process_getblockhash() exits with status 8 when the block
number is out of range, which is expected.  Any other exit status should
be treated as a spurious error.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-24 10:37:28 +01:00
Rusty Russell
cccdb53bec channel_states: fold all the ONCHAIND states into one.
The billboard is now far more useful to tell what's going on, and this
gets us closer to a state == owner mapping.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-23 18:02:00 +01:00
Rusty Russell
8aa7a48138 listpeers: display billboard as 'status' array.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-23 18:02:00 +01:00
Rusty Russell
cca0a5412e subd: clear transient billboard on start and shutdown.
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>
2018-02-23 18:02:00 +01:00
Rusty Russell
26b004e5af subd: handle status_peer_billboard messages from subdaemons.
We use a callback which updates the appropriate slot.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-23 18:02:00 +01:00
Rusty Russell
db4de95033 lightningd: single transient billboard for opening peers.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-23 18:02:00 +01:00
Rusty Russell
86a04c59d4 lightningd: maintain a status billboard for each channel.
Each state (effectively, each daemon) has two slots: a permanent slot
if something permanent happens (usually, a failure), and a transient
slot which summarizes what's happening right now.

Uncommitted channels only have a transient slot, by their very nature.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-23 18:02:00 +01:00
Christian Decker
6f6176a91e options: Add --offline to disable listening and auto-reconnection 2018-02-23 06:30:15 +00:00
Rusty Russell
85f1a9b825 lightningd: append (as much as we can) version into default name if DEVELOPER=1
This helps for bug reporting, where we may need to diagnose the other
end.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2018-02-22 12:24:26 +01:00
practicalswift
1b51b5ae0f Remove unused parameter topo in txw_fire(struct chain_topology *topo, ...) 2018-02-22 10:46:30 +01:00
practicalswift
0046ace318 Remove unused parameter topo in txowatch_fire(struct chain_topology *topo, ...) 2018-02-22 10:46:30 +01:00
practicalswift
74841ef567 Remove unused parameter payment_hash in forward_htlc(..., const struct sha256 *payment_hash, ...) 2018-02-22 10:46:30 +01:00
practicalswift
d873bf60bf Remove unused parameter fieldname in json_add_log(..., const char *fieldname, ...) 2018-02-22 10:46:30 +01:00
practicalswift
2ab274595e Remove unused parameter ptr in corrupt(const void *ptr, ...) 2018-02-22 10:46:30 +01:00
practicalswift
f24eae60b7 Remove unused parameter why in delete_channel(..., const char *why) 2018-02-22 10:46:30 +01:00
practicalswift
f15868a94d Remove unused function json_dev_broadcast(...) 2018-02-22 09:45:22 +00:00