Elias Rohrer [Thu, 1 Aug 2024 16:51:53 +0000 (11:51 -0500)]
Pin `tokio` in CI to fix MSRV
The recently released `tokio` 1.39 bumped their MSRV to rustc 1.70.
Here, we pin the `tokio` version to 1.38 for users that require to
maintain our MSRV of rustc 1.63.
Matt Corallo [Mon, 22 Jul 2024 20:03:41 +0000 (20:03 +0000)]
Enforce segwit inputs for all "safe" funding transactions
8403755a2a524beb9f6c8951f51dd60f7c54c912 introduced a separate path
for funding a channel without a full funding transaction, relying
on users to manually broadcast the funding tx. One of the major
things that makes this path less safe is that for other paths we're
supposed to validate that all inputs have witnesses, making the
funding transaction (likely) txid-non-malleable.
However, in one of several rewrites of that commit the funding tx
tests ended up getting elided in some call paths, which is fixed
here.
InvoiceRequest and Refund have payer_metadata which consists of an
encrypted payment id and a nonce used to derive its signing keys and
authenticate any corresponding invoices. Now that the blinded paths
include this data in their OffersContext, remove the nonce as it is now
redundant. Keep the encrypted payment id as some data is needed in the
payer metadata according to the spec. This saves space and prevents
de-anonymization attacks as along as the nonce isn't revealed.
When authenticating that an invoice is for a valid invoice request, the
payer metadata is needed. Some of this data will be removed in the next
commit and instead be included in the message context of the request's
reply path. Add this data to Event::InvoiceReceived so that asynchronous
invoice handling can verify properly.
When a Bolt12Invoice is handled with an OfferContext, use both the
containing payment_id and nonce to verify that it is for a pending
outbound payment. Previously, the nonce the payment_id were taken from
the payer_metadata and the latter was compared against the payment_id
in the OfferContext. The payer_metadata thus no longer needs to include
either when a blinded path is used. However, some payer_metadata will
still be needed as per the spec.
To authenticate that a Bolt12Invoice is for a valid InvoiceRequest or
Refund, include the nonce from the payer_metadata in the InvoiceRequest
reply path or Refund::paths, respectively. This can be used to prevent
de-anonymization attacks where an attacker sends invoices using
self-constructed paths to nodes near the blinded paths' introduction
nodes.
Invoices are authenticated by checking the payer metadata in the
corresponding invoice request or refund. For all invoices requests and
for refunds using blinded paths, this will be the encrypted payment id
and a 128-bit nonce. Allows checking the unencrypted payment id and
nonce explicitly instead of the payer metadata. This will be used by an
upcoming change that includes the payment id and nonce in the invoice
request's reply path and the refund's blinded paths instead of
completely in the payer metadata, which mitigates de-anonymization
attacks.
When using RefundBuilder::deriving_payer_id, the nonce generated needs
to be the same one included in any RefundBuilder::paths. This is because
the nonce is used along with the refund TLVs to derive a payer id and
will soon be used to authenticate any invoices.
When using InvoiceRequestBuilder::deriving_payer_id, the nonce generated
needs to be the same one included in any reply path. This is because the
nonce is used along with the invoice request TLVs to derive a payer id.
While this data is also included in the payer_metadata, including it in
the blinded path would allow reducing the amount of data needed there to
just enough to provide entropy (i.e., 16 bytes). This is more important
for Refund because it can be transmitted via a QR code. But using the
same payer_metadata structure for both InvoiceRequest and Refund would
be beneficial to avoid more code.
When an invoice or invoice request cannot be authenticated from an
OffersContext, simply do not respond instead of sending an InvoiceError
message. According to BOLT4, messages sent over a blinded path not
intended for its use MUST be ignored.
When a Bolt12Invoice is handled with an OfferContext, use the
containing payment_id to verify that it is for a pending outbound
payment. Only invoices for refunds without any blinded paths can be
verified without an OfferContext.
Jeffrey Czyz [Thu, 20 Jun 2024 16:43:57 +0000 (11:43 -0500)]
Elide metadata from Offer with derived keys
When an Offer uses blinded paths, its metadata consists of a nonce used
to derive its signing keys. Now that the blinded paths contain this
nonce, elide the metadata as it is now redundant. This saves space and
also makes it impossible to derive the signing keys if an invoice
request is received with the incorrect nonce. The nonce shouldn't be
revealed in this case either to prevent de-anonymization attacks.
When an InvoiceRequest is handled with an OfferContext, use the
containing nonce to verify that it is for a valid Offer. Otherwise, fall
back to using Offer::metadata, which also contains the nonce. The latter
is useful for supporting offers without blinded paths or those created
prior to including an OffersContext in their blinded paths.
To authenticate that an InvoiceRequest is for a valid Offer, include the
nonce from the Offer::metadata in the Offer::paths. This can be used to
prevent de-anonymization attacks where an attacker sends requests using
self-constructed paths to nodes near the Offer::paths' introduction
nodes.
Metadata is an internal type used within Offer messages. For any
constructed message, Metadata::Bytes is always used. The other variants
are used during construction or verification time. Document this and
debug_assert!(false) accordingly.
Jeffrey Czyz [Thu, 20 Jun 2024 00:18:26 +0000 (19:18 -0500)]
Add InvoiceRequest::verify_using_recipient_data
Invoice requests are authenticated by checking the metadata in the
corresponding offer. For offers using blinded paths, this will simply be
a 128-bit nonce. Allows checking this nonce explicitly instead of the
metadata. This will be used by an upcoming change that includes the
nonce in the offer's blinded paths instead of the metadata, which
mitigate de-anonymization attacks.
Jeffrey Czyz [Wed, 19 Jun 2024 17:42:16 +0000 (12:42 -0500)]
Pass Nonce directly to OfferBuilder
When using OfferBuilder::deriving_signing_pubkey, the nonce generated
needs to be the same one included in any OfferBuilder::paths. This is
because the nonce is used along with the offer TLVs to derive a signing
pubkey and will soon be elided from the metadata entirely.
Jeffrey Czyz [Wed, 19 Jun 2024 16:51:53 +0000 (11:51 -0500)]
Move Nonce to a separate offers sub-module
Nonce is used when constructing Offer::metadata and will soon be need
when constructing BlindedPath for use in authentication. Move it to
separate module now that it is public and will be more widely used.
Jeffrey Czyz [Wed, 19 Jun 2024 16:29:20 +0000 (11:29 -0500)]
Change Nonce visibility to pub
A nonce is generated in OfferBuilder::deriving_signing_pubkey from an
EntropySource for use in Offer::metadata. The same nonce will need to be
included as recipient data in any blinded paths in the Offer. Increase
the visibility to allow for this.
Matt Corallo [Sun, 21 Jul 2024 23:29:19 +0000 (23:29 +0000)]
[block-sync] Don't hold client-cache lock while connecting
`lightning-block-sync`'s REST and RPC clients both hold a cache for
a connected client to avoid the extra connection round-trip on each
request. Because only one client can be using a connection at once,
the connection is `take()`n out of an `Option` behind a `Mutex` and
if there isn't one present, we call `HttpClient::connect` to build
a new one.
However, this full logic is completed in one statement, causing a
client-cache lock to be held during `HttpClient::connect`. This
can turn into quite a bit of contention when using these clients as
gossip verifiers as we can create many requests back-to-back during
startup.
I noticed this as my node during startup only seemed to be
saturating one core and managed to get a backtrace that showed
several threads being blocked on this mutex when hitting a Bitcoin
Core node over REST that is on the same LAN, but not the same
machine.
Matt Corallo [Sat, 20 Jul 2024 22:57:09 +0000 (22:57 +0000)]
Ensure `ChannelMonitorUpdate`s are ordered with full monitor writes
When we update a channel, then while connecting a block persist a
full `ChannelMonitor` prior to persisting the
`ChannelMonitorUpdate`, users can end up seeing a full
`ChannelMonitor` with a given `latest_update_id` prior to seeing
the `ChannelMonitorUpdate` with the same `update_id`. This
could cause users to have a full `ChannelMonitor` on disk as well
as a `ChannelMonitorUpdate` which was already applied. While this
isn't an issue for the LDK-provided update-based `Persist`, its
somewhat surprising for users so we avoid it.
The BOLT11 and BOLT12 outbound payment initiation code differ in that
the latter re-uses the retry path (i.e., find_route_and_send_payment).
The drawback of this is that Ok is returned even if there is an error
finding a route. Refactor send_payment_for_bolt12_invoice such that it
re-uses find_initial_route instead so that errors can be returned.
Elias Rohrer [Tue, 2 Jul 2024 10:04:33 +0000 (12:04 +0200)]
Handle fallible events in `OnionMessenger`
Previously, we would just fire-and-forget in `OnionMessenger`'s event
handling. Since we now introduced the possibility of event handling
failures, we here adapt the event handling logic to retain any
events which we failed to handle to have them replayed upon the next
invocation of `process_pending_events`/`process_pending_events_async`.
Elias Rohrer [Mon, 15 Apr 2024 08:35:52 +0000 (10:35 +0200)]
Make event handling fallible
Previously, we would require our users to handle all events
successfully inline or panic will trying to do so. If they would exit
the `EventHandler` any other way we'd forget about the event and
wouldn't replay them after restart.
Here, we implement fallible event handling, allowing the user to return
`Err(())` which signals to our event providers they should abort event
processing and replay any unhandled events later (i.e., in the next
invocation).
Elias Rohrer [Tue, 2 Jul 2024 09:11:40 +0000 (11:11 +0200)]
Hold sep. Mutexes for pending `intercepted_msgs`/`peer_connected` events
This is a minor refactor that will allow us to access the individual
event queue Mutexes separately, allowing us to drop the locks earlier
when processing them individually.
Matt Corallo [Mon, 8 Jul 2024 19:21:19 +0000 (19:21 +0000)]
Make `impl_writeable_tlv_based_enum*` actually upgradable
In cc78b77c715d6ef62693d4c1bc7190da990ec0fa it was discovered that
`impl_writeable_tlv_based_enum_upgradable` wasn't actually
upgradable - tuple variants weren't written with length-prefixes,
causing downgrades with new tuple variants to be unreadable by
older clients as they wouldn't know where to stop reading.
This was fixed by simply assuming that any new variants will be
non-tuple variants with a length prefix, but no code write-side
changes were made, allowing new code to freely continue to use the
broken tuple-variant serialization.
Here we address this be defining yet more serialization macros
which aren't broken, and convert existing usage of the existing
macros using non-length-prefixed tuple variants to renamed
`*_legacy` macros.
Note that this changes the serialization format of
`impl_writeable_tlv_based_enum[_upgradable]` when tuple fields are
written, and as such deliberately changes the call semantics for
such tuples.
Only the serialization format of `MessageContext` is changed here
which is fine as it has not yet reached a release of LDK.
Elias Rohrer [Tue, 16 Jul 2024 08:28:58 +0000 (10:28 +0200)]
Clarify maximum length of an invoice description
We previously stated in the docs that the invoice description can be at most `1023`
bytes long, which is wrong. According to BOLT 11 it's at most 1023*5 bits (639 bytes) long.
Alec Chen [Sun, 30 Jun 2024 23:05:48 +0000 (16:05 -0700)]
Test async get per commitment point for revoke_and_ack
Note: this does not test the CS -> RAA resend ordering, because this
requires handling async get_per_commitment_point for channel
reestablishment, which will be addressed in a follow up PR.
Alec Chen [Wed, 5 Jun 2024 22:40:38 +0000 (17:40 -0500)]
Change get_per_commitment_point to return result type
Includes simple changes to test util signers and tests, as well as
handling the error case for get_per_commitment_point in
HolderCommitmentPoint. This leaves a couple `.expect`s in places
that will be addressed in a separate PR for handling funding.
Matt Corallo [Mon, 15 Jul 2024 18:34:04 +0000 (18:34 +0000)]
Drop unnecessary `strict` feature from `lightning-invoice`
99aa6e27f616c96dda2b49d09bafbc0b982251e0 detected that we had an
undefined feature in `lightning-invoice` called `strict`, which was
used to turn on `deny(warnings)`. It resolved that by adding the
feature to the `Cargo.toml`, but we actually don't need it - our CI
already builds with `-Dwarnings`, so any warnings should be
rejected during CI and there's not a lot of value in having a
(public) feature to do the same.