I screwed up the rotation logic in an earlier varient of this PR, and
it lead me to discover why test_mpp_interference_2 was flaky.
Really, we should keep a fuzzy estimator of how much payment is
outstanding, but in practice rotation is probably good enough.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
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>
Some additional detail was warranted in several places, though the
listconfigs is better off simply referring to lightningd-config.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Changelog-Fixed: doc: Many missing manual pages were completed
It's now only needed by devtools/mkfunding, so include a reduced one
there, and this also means we remove tx_spending_utxos().
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This removes the reservation cleanup at startup, too, now they're all
using 'reserved_til'.
This changes test_withdraw, since it asserted that outputs were marked
spent as soon as we broadcast a transaction: now they're reserved until
it's mined. Similarly, test_addfunds_from_block assumed we'd see funds
as soon as we broadcast the tx.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Changelog-Changed: JSON-RPC: `withdraw` now randomizes input and output order, not BIP69.
These tests appear in both test_misc.py and test_wallet.py.
I checked, and they're the only duplcates; they were moved in
0543149b89 then re-added (merge error?)
in 67fc8ffbcf.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
First, simplify: amount is set to 1000000, but then we deposit 1000000 + 0.01btc
(i.e. 2000000), and we always use 2 * amount. Just use a single constant to
make it clear.
Secondly, we assume that the wallet considers outputs spent as soon as
we created the tx: this will not be true once withdraw uses sendpsbt.
So, we generate blocks, but now sometimes withdraw will pick up change
txs, so we need to reserve them to avoid that messing our coinmovements.
Finally, we assumed the withdrawl order was BIP69, which becomes
variable.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This is a little lazy, but simpler than extracting the common parts
or making withdraw a plugin which calls txprepare (which should be
deprecated soon in favor of fundpsbt etc).
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Marking spent means if the transaction doesn't confirm for some
reason, the user will need to force a rescan to find the funds. Now
we have timed reservations, reserving for (an additional) 12 hours
should be sufficient.
We also take this opportunity (now we have our own callback path)
to record the tx in the wallet only on success.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Some minor phrasing differences cause test changes.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Changelog-Changed: txprepare reservations stay across restarts: use fundpsbt/reservepsbt/unreservepsbt
Changelog-Removed: txprepare `destination` `satoshi` argument form removed (deprecated v0.7.3)
This uses `fundpsbt` and similar to simulate the txprepare command.
It has one difference (when complete), in that it those reservations
are now timed and don't get reset on restart.
It also doesn't have the restriction that `all` can only be used with
no other output, as I didn't realize that when I implemented it!
Note that change is now inserted in a random position, not sorted
into BIP69 order.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
If all is set, diff may not be:
==1230518== Use of uninitialised value of size 8
==1230518== at 0x4C5781B: _itoa_word (_itoa.c:179)
==1230518== by 0x4C736F4: __vfprintf_internal (vfprintf-internal.c:1687)
==1230518== by 0x4C88119: __vsnprintf_internal (vsnprintf.c:114)
==1230518== by 0x1D44B6: do_vfmt (str.c:66)
==1230518== by 0x1D45A0: tal_vfmt_ (str.c:92)
==1230518== by 0x1D4401: tal_fmt_ (str.c:44)
==1230518== by 0x15D30F: fmt_amount_sat (amount.c:60)
==1230518== by 0x15D338: fmt_amount_sat_ (amount.c:62)
==1230518== by 0x178C45: type_to_string_ (type_to_string.c:35)
==1230518== by 0x1B8F75: json_fundpsbt (reservation.c:394)
==1230518== by 0x12D0EC: command_exec (jsonrpc.c:602)
==1230518== by 0x12D598: rpc_command_hook_callback (jsonrpc.c:712)
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This avoids overwriting the ones in git, and generally makes things neater.
We have convenience headers wire/peer_wire.h and wire/onion_wire.h to
avoid most #ifdefs: simply include those.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We're going to make experimental versions of these completely separate files.
Also remove the dependency on the Makefile itself: it simply causes
unnecessary churn. We can always force-rebuild when we change a rule.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
This will make the `openchannel_hook` chainable. Logic is as follows:
- The first plugin that rejects terminates the chain
- If more than one plugin uses the `close_to` parameter, take the first
value and log_unusual for the others.
Changelog-Added: openchannel_hook is now chainable
It is not part of the default documentation generation, thus will show up in
the git diff every time a maintainer-clean is done. We could also add it to
the default documentation generation, however it requires some rather
heavy-weight latex dependencies, and the file hasn't changed in several years.
Since we started committing generated files as well we can mark them as
derived so that they appear collapsed in the diff, reducing noise while
reviewing.
Generating these requires python tooling which may not be present, so only
clean them in maintainer mode. Also remove `wallet/statements.po` which is no
longer used or generated at all.
PR #3957 improved performance considerably, however we still look over the
entire message for the message separator. If instead we just look in the
incrementally read data, we remove the quadratic behavior for large messages.
This is safe since we then loop over the messages which would drain any
message separator from the buffer before we attempt the next read.
Changelog-Fixed: bcli: Significant speedups for block synchronization
We're not going to mutate transactions in a block, so computing the txids
every time we need them is a waste, let's compute them upfront and use them
afterwards.
We were using `tal_fmt` to truncate off the last byte of the
response (newline), which required an allocation, a call to `vsnprintf` and a
copy of the block contents. This being >2MB in size on mainnet was rather
expensive.
We now just signal the end of the string by overwriting the trailing newline,
and stealing directly onto the result.
Rusty Russell
Vincenzo Palazzo
Michael Schmoock
Christian Decker