mirrors/core-lightning
1
0
Fork 0
mirror of https://github.com/ElementsProject/lightning.git synced 2024-11-19 09:54:16 +01:00
Code Issues Projects Releases Wiki Activity
67d667db90
core-lightning/lightningd/routehint.c

261 lines
8.2 KiB
C
Raw Normal View History

check: make sure all files outside contrib/ include "config.h" first. And turn "" includes into full-path (which makes it easier to put config.h first, and finds some cases check-includes.sh missed previously). config.h sets _GNU_SOURCE which really needs to be done before any '#includes': we mainly got away with it with glibc, but other platforms like Alpine may have stricter requirements. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2021-12-04 12:23:56 +01:00
#include "config.h"
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
#include <common/bolt11.h>
common: clean up json routine locations. We have them split over common/param.c, common/json.c, common/json_helpers.c, common/json_tok.c and common/json_stream.c. Change that to: * common/json_parse (all the json_to_xxx routines) * common/json_parse_simple (simplest the json parsing routines, for cli too) * common/json_stream (all the json_add_xxx routines) * common/json_param (all the param and param_xxx routines) Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-07-04 05:49:38 +02:00
#include <common/json_parse.h>
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
#include <gossipd/gossipd_wiregen.h>
cleanup: remove unneeded includes in header files. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2021-09-16 07:00:42 +02:00
#include <lightningd/channel.h>
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
#include <lightningd/lightningd.h>
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
#include <lightningd/peer_control.h>
#include <lightningd/routehint.h>
static bool scid_in_arr(const struct short_channel_id *scidarr,
bitcoin/short_channel_id: pass by copy everywhere. It's a u64, we should pass by copy. This is a big sweeping change, but mainly mechanical (change one, compile, fix breakage, repeat). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 02:59:51 +01:00
struct short_channel_id scid)
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
{
for (size_t i = 0; i < tal_count(scidarr); i++)
bitcoin/short_channel_id: pass by copy everywhere. It's a u64, we should pass by copy. This is a big sweeping change, but mainly mechanical (change one, compile, fix breakage, repeat). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 02:59:51 +01:00
if (short_channel_id_eq(scidarr[i], scid))
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
return true;
return false;
}
struct routehint_candidate *
routehint_candidates(const tal_t *ctx,
struct lightningd *ld,
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
const char *buf,
const jsmntok_t *toks,
const bool *expose_all_private,
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
const struct short_channel_id *hints,
bool *none_public,
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
struct amount_msat *avail_capacity,
struct amount_msat *private_capacity,
struct amount_msat *deadend_capacity,
struct amount_msat *offline_capacity)
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
{
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
struct routehint_candidate *candidates, *privcandidates;
const jsmntok_t *t, *arr;
size_t i;
log_debug(ld->log, "routehint: %.*s",
json_tok_full_len(toks),
json_tok_full(buf, toks));
/* We get the full JSON, including result. */
t = json_get_member(buf, toks, "result");
if (!t)
fatal("Missing result from listincoming: %.*s",
json_tok_full_len(toks),
json_tok_full(buf, toks));
arr = json_get_member(buf, t, "incoming");
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
candidates = tal_arr(ctx, struct routehint_candidate, 0);
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
privcandidates = tal_arr(tmpctx, struct routehint_candidate, 0);
*none_public = true;
*deadend_capacity = AMOUNT_MSAT(0);
*offline_capacity = AMOUNT_MSAT(0);
*avail_capacity = AMOUNT_MSAT(0);
*private_capacity = AMOUNT_MSAT(0);
/* We combine the JSON output which knows the peers' details,
* with our internal information */
json_for_each_arr(i, t, arr) {
struct amount_msat capacity;
const char *err;
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
struct routehint_candidate candidate;
lightningd: check htlc_maximum_msat of channels for routehints. We still use the channel hint here (as it's the only option), we just warn about lack of capacity. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-03-21 01:58:28 +01:00
struct amount_msat fee_base, htlc_max;
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
struct route_info *r;
bool is_public;
r = tal(tmpctx, struct route_info);
err = json_scan(tmpctx, buf, t,
"{id:%"
",short_channel_id:%"
",fee_base_msat:%"
",fee_proportional_millionths:%"
",cltv_expiry_delta:%"
lightningd: check htlc_maximum_msat of channels for routehints. We still use the channel hint here (as it's the only option), we just warn about lack of capacity. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-03-21 01:58:28 +01:00
",htlc_max_msat:%"
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
",incoming_capacity_msat:%"
"}",
JSON_SCAN(json_to_node_id, &r->pubkey),
JSON_SCAN(json_to_short_channel_id,
&r->short_channel_id),
JSON_SCAN(json_to_msat, &fee_base),
JSON_SCAN(json_to_u32,
&r->fee_proportional_millionths),
JSON_SCAN(json_to_u16, &r->cltv_expiry_delta),
lightningd: check htlc_maximum_msat of channels for routehints. We still use the channel hint here (as it's the only option), we just warn about lack of capacity. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-03-21 01:58:28 +01:00
JSON_SCAN(json_to_msat, &htlc_max),
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
JSON_SCAN(json_to_msat, &capacity));
if (err) {
fatal("Invalid return from listincoming (%s): %.*s",
err,
json_tok_full_len(toks),
json_tok_full(buf, toks));
}
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
lightningd: make listincoming use gossmods_from_listpeerchannels(). We temporarily use a second gossmap so we can just switch private info off for listincoming and not listchannels. Note that listchannels now uses the local alias (if no scid), so we have to change that in the routehint caller. Since we now *always* use a channel alias in hints if one exists, a test broke, so fix that. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-12-13 06:35:43 +01:00
/* Note: listincoming returns real scid or local alias if no real scid. */
bitcoin/short_channel_id: pass by copy everywhere. It's a u64, we should pass by copy. This is a big sweeping change, but mainly mechanical (change one, compile, fix breakage, repeat). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 02:59:51 +01:00
candidate.c = any_channel_by_scid(ld, r->short_channel_id, true);
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
if (!candidate.c) {
lightningd: don't assume a single channel per peer. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-03-22 23:59:20 +01:00
log_debug(ld->log, "%s: channel not found in peer %s",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_short_channel_id(tmpctx, r->short_channel_id),
fmt_node_id(tmpctx, &r->pubkey));
lightningd: don't assume a single channel per peer. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-03-22 23:59:20 +01:00
continue;
}
lightningd: make listincoming use gossmods_from_listpeerchannels(). We temporarily use a second gossmap so we can just switch private info off for listincoming and not listchannels. Note that listchannels now uses the local alias (if no scid), so we have to change that in the routehint caller. Since we now *always* use a channel alias in hints if one exists, a test broke, so fix that. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-12-13 06:35:43 +01:00
/* Check channel is in CHANNELD_NORMAL or CHANNELD_AWAITING_SPLICE */
lightningd: make channel-query functions all take state. It has the information we need, now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-10-02 00:59:51 +02:00
if (!channel_state_can_add_htlc(candidate.c->state)) {
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
log_debug(ld->log, "%s: abnormal channel",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_short_channel_id(tmpctx,
r->short_channel_id));
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
continue;
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
}
setchannel: add minhtlc Suggested by @m-schmook, I realized that if we append it later I'll never get it right: I expect parameters min and max, not max and min! Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: Protocol: you can now alter the `htlc_minimum_msat` and `htlc_maximum_msat` your node advertizes.
2022-03-21 01:58:54 +01:00
/* FIXME: we don't actually check htlc_minimum_msat! */
lightningd: check htlc_maximum_msat of channels for routehints. We still use the channel hint here (as it's the only option), we just warn about lack of capacity. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-03-21 01:58:28 +01:00
/* If they set an htlc_maximum_msat, consider that the
* capacity ceiling. We *could* do multiple HTLCs,
* but presumably that would defeat the spirit of the
* limit anyway */
lightningd: don't assume a single channel per peer. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-03-22 23:59:20 +01:00
/* FIXME: Present max capacity of multiple channels? */
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
candidate.capacity = channel_amount_receivable(candidate.c);
lightningd: check htlc_maximum_msat of channels for routehints. We still use the channel hint here (as it's the only option), we just warn about lack of capacity. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2022-03-21 01:58:28 +01:00
if (amount_msat_greater(candidate.capacity, htlc_max))
candidate.capacity = htlc_max;
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
/* Now we can tell if it's public. If so (even if it's otherwise
* unusable), we *don't* expose private channels! */
is_public = (candidate.c->channel_flags
& CHANNEL_FLAGS_ANNOUNCE_CHANNEL);
if (is_public)
*none_public = false;
/* If they explicitly say to expose all private ones, consider
* it public. */
if (expose_all_private != NULL && *expose_all_private)
is_public = true;
r->fee_base_msat = fee_base.millisatoshis; /* Raw: route_info */
/* Could wrap: if so ignore */
if (!amount_msat_eq(amount_msat(r->fee_base_msat), fee_base)) {
log_debug(ld->log,
"Peer charging insane fee %.*s; ignoring",
json_tok_full_len(t),
json_tok_full(buf, t));
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
continue;
}
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
/* Consider only hints they gave */
if (hints) {
log_debug(ld->log, "We have hints!");
lightnind: use aliases in routehints for private channels. We *should* remember the channel type, since this is only required if they set the channel_type to include option_scid_alias. However, since we support channel upgrade, channel_type really needs a new table. I have a patch for that, from my abandoned original "fastopen" branch for aliases, but it's too big a chance for rc2 IMHO. Meanwhile, we allow exposeprivatechannels's scids to be either real or the aliases. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: Protocol: invoice routehints will use fake short-channel-ids for private channels if channel opened with option_scid_alias-supporting peer.
2022-08-08 21:58:07 +02:00
/* Allow specification by alias, too */
bitcoin/short_channel_id: pass by copy everywhere. It's a u64, we should pass by copy. This is a big sweeping change, but mainly mechanical (change one, compile, fix breakage, repeat). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 02:59:51 +01:00
if (!scid_in_arr(hints, r->short_channel_id)
lightnind: use aliases in routehints for private channels. We *should* remember the channel type, since this is only required if they set the channel_type to include option_scid_alias. However, since we support channel upgrade, channel_type really needs a new table. I have a patch for that, from my abandoned original "fastopen" branch for aliases, but it's too big a chance for rc2 IMHO. Meanwhile, we allow exposeprivatechannels's scids to be either real or the aliases. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: Protocol: invoice routehints will use fake short-channel-ids for private channels if channel opened with option_scid_alias-supporting peer.
2022-08-08 21:58:07 +02:00
&& (!candidate.c->alias[REMOTE]
bitcoin/short_channel_id: pass by copy everywhere. It's a u64, we should pass by copy. This is a big sweeping change, but mainly mechanical (change one, compile, fix breakage, repeat). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 02:59:51 +01:00
|| !scid_in_arr(hints, *candidate.c->alias[REMOTE]))) {
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
log_debug(ld->log, "scid %s not in hints",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_short_channel_id(tmpctx,
r->short_channel_id));
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
continue;
}
/* If they give us a hint, we use even if capacity 0 */
common/amount: rename amount_sat_zero/amount_msat_zerp -> amount_sat_is_zero/amount_msat_is_zero. I used `amount_msat_eq(x, AMOUNT_MSAT(0))` because I forgot this function existed. I probably missed it because the name is surprising, so add "is" in there to make it clear it's a boolean function. You'll note almost all the places which did use it are Eduardo's and Lisa's code, so maybe it's just me. Fix up a few places which I could use it, too. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-09-18 09:34:27 +02:00
} else if (amount_msat_is_zero(capacity)) {
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
log_debug(ld->log, "%s: deadend",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_short_channel_id(tmpctx,
r->short_channel_id));
common/amount: add amount_msat_accumulate() Saves some typing, and is clearer than checking if both args really are the same! Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-09-18 09:33:27 +02:00
if (!amount_msat_accumulate(deadend_capacity,
candidate.capacity))
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
fatal("Overflow summing deadend capacity!");
continue;
}
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
/* Is it offline? */
if (candidate.c->owner == NULL) {
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
log_debug(ld->log, "%s: offline",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_short_channel_id(tmpctx,
r->short_channel_id));
common/amount: add amount_msat_accumulate() Saves some typing, and is clearer than checking if both args really are the same! Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-09-18 09:33:27 +02:00
if (!amount_msat_accumulate(offline_capacity,
candidate.capacity))
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
fatal("Overflow summing offline capacity!");
continue;
}
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
global: fix up quotes which are now in the master branch. 1. onion-message 2. blinded-payments 3. route-blinding 4. channel-type 5. warnings. Now they'll be checked correctly, and if the spec changes, we'll know to reexamine this code. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-18 04:01:06 +01:00
/* BOLT #2:
lightnind: use aliases in routehints for private channels. We *should* remember the channel type, since this is only required if they set the channel_type to include option_scid_alias. However, since we support channel upgrade, channel_type really needs a new table. I have a patch for that, from my abandoned original "fastopen" branch for aliases, but it's too big a chance for rc2 IMHO. Meanwhile, we allow exposeprivatechannels's scids to be either real or the aliases. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: Protocol: invoice routehints will use fake short-channel-ids for private channels if channel opened with option_scid_alias-supporting peer.
2022-08-08 21:58:07 +02:00
* - if `channel_type` has `option_scid_alias` set:
* - MUST NOT use the real `short_channel_id` in
* BOLT 11 `r` fields.
*/
lightningd: make listincoming use gossmods_from_listpeerchannels(). We temporarily use a second gossmap so we can just switch private info off for listincoming and not listchannels. Note that listchannels now uses the local alias (if no scid), so we have to change that in the routehint caller. Since we now *always* use a channel alias in hints if one exists, a test broke, so fix that. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-12-13 06:35:43 +01:00
if (channel_has(candidate.c, OPT_SCID_ALIAS)) {
if (!candidate.c->alias[REMOTE]) {
log_debug(ld->log, "%s: no remote alias (yet?)",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_short_channel_id(tmpctx,
r->short_channel_id));
lightningd: make listincoming use gossmods_from_listpeerchannels(). We temporarily use a second gossmap so we can just switch private info off for listincoming and not listchannels. Note that listchannels now uses the local alias (if no scid), so we have to change that in the routehint caller. Since we now *always* use a channel alias in hints if one exists, a test broke, so fix that. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-12-13 06:35:43 +01:00
continue;
}
lightnind: use aliases in routehints for private channels. We *should* remember the channel type, since this is only required if they set the channel_type to include option_scid_alias. However, since we support channel upgrade, channel_type really needs a new table. I have a patch for that, from my abandoned original "fastopen" branch for aliases, but it's too big a chance for rc2 IMHO. Meanwhile, we allow exposeprivatechannels's scids to be either real or the aliases. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: Protocol: invoice routehints will use fake short-channel-ids for private channels if channel opened with option_scid_alias-supporting peer.
2022-08-08 21:58:07 +02:00
r->short_channel_id = *candidate.c->alias[REMOTE];
lightningd: make listincoming use gossmods_from_listpeerchannels(). We temporarily use a second gossmap so we can just switch private info off for listincoming and not listchannels. Note that listchannels now uses the local alias (if no scid), so we have to change that in the routehint caller. Since we now *always* use a channel alias in hints if one exists, a test broke, so fix that. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-12-13 06:35:43 +01:00
} else {
/* Older code didn't remember type correctly,
* or negotiate properly, so best effort here */
/* Note explicit flag test here: if we're told to expose all
* private channels, then "is_public" is forced true */
if (candidate.c->alias[REMOTE]
/* Use remote alias if it's all we've got, or unannounced channel */
&& (!(candidate.c->channel_flags & CHANNEL_FLAGS_ANNOUNCE_CHANNEL)
|| !candidate.c->scid)) {
r->short_channel_id = *candidate.c->alias[REMOTE];
} else if (candidate.c->scid) {
r->short_channel_id = *candidate.c->scid;
} else {
/* Haven't got remote alias yet? Can't use ut. */
log_debug(ld->log, "%s: no alias",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_short_channel_id(tmpctx,
r->short_channel_id));
lightningd: make listincoming use gossmods_from_listpeerchannels(). We temporarily use a second gossmap so we can just switch private info off for listincoming and not listchannels. Note that listchannels now uses the local alias (if no scid), so we have to change that in the routehint caller. Since we now *always* use a channel alias in hints if one exists, a test broke, so fix that. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2023-12-13 06:35:43 +01:00
continue;
}
lightnind: use aliases in routehints for private channels. We *should* remember the channel type, since this is only required if they set the channel_type to include option_scid_alias. However, since we support channel upgrade, channel_type really needs a new table. I have a patch for that, from my abandoned original "fastopen" branch for aliases, but it's too big a chance for rc2 IMHO. Meanwhile, we allow exposeprivatechannels's scids to be either real or the aliases. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: Protocol: invoice routehints will use fake short-channel-ids for private channels if channel opened with option_scid_alias-supporting peer.
2022-08-08 21:58:07 +02:00
}
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
/* OK, finish it and append to one of the arrays. */
if (is_public) {
log_debug(ld->log, "%s: added to public",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_short_channel_id(tmpctx,
r->short_channel_id));
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
candidate.r = tal_steal(candidates, r);
tal_arr_expand(&candidates, candidate);
common/amount: add amount_msat_accumulate() Saves some typing, and is clearer than checking if both args really are the same! Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-09-18 09:33:27 +02:00
if (!amount_msat_accumulate(avail_capacity,
candidate.capacity)) {
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
fatal("Overflow summing pub capacities %s + %s",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_amount_msat(tmpctx,
*avail_capacity),
fmt_amount_msat(tmpctx,
candidate.capacity));
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
}
} else {
log_debug(ld->log, "%s: added to private",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_short_channel_id(tmpctx,
r->short_channel_id));
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
candidate.r = tal_steal(privcandidates, r);
tal_arr_expand(&privcandidates, candidate);
common/amount: add amount_msat_accumulate() Saves some typing, and is clearer than checking if both args really are the same! Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-09-18 09:33:27 +02:00
if (!amount_msat_accumulate(private_capacity,
candidate.capacity)) {
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
fatal("Overflow summing priv capacities %s + %s",
global: change all type_to_string to fmt_X. This has the benefit of being shorter, as well as more reliable (you will get a link error if we can't print it, not a runtime one!). Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2024-03-20 01:47:52 +01:00
fmt_amount_msat(tmpctx,
*private_capacity),
fmt_amount_msat(tmpctx,
candidate.capacity));
invoice: overhaul routehints to use topology.listincoming, cleanup. This turned into a more extensive cleanup than intended. The previous warnings were overlapping and confusing, especially now MPP is the norm. *warning_capacity* is now the "even under best circumstances, we don't have enough incoming capacity", which is really what warning_mpp_capacity was trying to say (no longer printed). *warning_offline* and *warning_deadends* are only given if adding such peers would have helped give capacity (i.e. not if *warning_capacity* is set). The new *warning_private_unused* tells you that we would have sufficient capacity, but we refused to expose private channels. The test cases have been enhanced to cover the new warnings. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Changelog-Added: JSON-RPC: `invoice` now gives `warning_private_unused` if unused unannounced channels could have provided sufficient capacity. Changelog-Changed: JSON-RPC: `invoice` warnings are now better defined, and `warning_mpp_capacity` is no longer included (since `warning_capacity` covers that).
2021-06-14 23:07:38 +02:00
}
}
}
/* By default, only consider private channels if there are
* no public channels *at all* */
if (expose_all_private == NULL
&& tal_count(candidates) == 0 && *none_public) {
log_debug(ld->log, "No publics: using private channels");
tal_free(candidates);
candidates = tal_steal(ctx, privcandidates);
*avail_capacity = *private_capacity;
/* This reflects *unused* private capacity. */
*private_capacity = AMOUNT_MSAT(0);
lightningd: extract routehint selection code. We're going to want this for bolt13 formation as well. As a result of reworking the logic into "candidate selection" then "route hint selection", we need to change the way round-robin works. We use a simple incrementing index now. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
2020-09-08 06:57:53 +02:00
}
return candidates;
}
Reference in New Issue Copy Permalink