Reaching the tip needed two historical facts encoded, and the second one was
not in the original hypothesis.
First, the disarmed window: 36 anchor heights (13,267,824 to 13,268,944) whose
blocks carry an attached certificate but no digest, and whose signer indices are
not strictly increasing, produced while anchoring was disarmed fleet-wide to
unblock the chain after an incident. Encoded as a named list of historical
windows with fixed bounds in code, after the precedent of the base-fee lapse
windows: inside, the header is accepted as written; outside, nothing changes.
Second, the recovery, found by measuring rather than guessing. After the
emergency re-arm the fleet ran for a stretch with a lowered seal threshold, so
proposers legitimately wrote shorter certificates. Its extent was measured by
reading 25,252 anchor heights one at a time, with NO binary search, because the
property is not monotone: only about 15 percent of heights are affected and the
largest gap between two affected heights is 44, so a binary search would have
returned an answer that looks exactly like a good one. Result: 13,268,976 to
13,890,544, thresholds measured at 1 or 2, never 0.
The second window relaxes the seal COUNT only. Digest binding and index
ordering stay enforced and every seal is still verified. Widening the disarmed
window to cover both would have been easier and would have thrown away
certificate binding on 19,425 anchor heights, which is the one property the
anchor exists for.
Negative control in four directions, because an exception can fail both ways:
the window predicate forced always-false turns 10 tests red; forced always-true
turns 39 red, of which 22 are pre-existing strictness tests, making the
exception swallowing the chain visible; the historical threshold emptied turns 7
red; pinned at 1 turns 8 red. Restored: 641 tests, 0 failures, counted from XML.
Patch verified on a pristine upstream checkout, alone and in series, and the
resulting tree compiles.
The published patch carried a one-block boundary defect: the seal rule asks the
index-to-address map at the parent height (H-1) while the registry schedule
binds inclusively from H, so at the very first anchor the schedule was empty and
the lookup fell back to a head map that is empty on a node synced from genesis.
A valid certificate was rejected. Fixed in both lookup paths: when the schedule
has an entry at exactly blockNumber+1, answer from that entry's verified bound
registry. The boundary stays exactly one block.
Proven by import, not by assertion: the package-built node was stuck at
13,013,999 with 6,646 rejections; with the fix it crossed 13,014,000 in 42
seconds and imported 253,729 further blocks through the anchor era with zero
rejections. Tests: new PqParentHeightAlignmentTest 3/3, consensus:common 391/0,
consensus:qbft 217/0. Negative control measured: disarm the condition and the
measuring test goes red; restore it and it is green.
Patch verified to apply cleanly on a pristine upstream checkout, alone and in
series with 0001/0004/0005; file list is the previous 75 plus the new test, and
every untouched section is byte-identical to the old patch. The anchor/ source
copy in this package was brought to the same state so the two cannot diverge.
Honest limit, recorded in IMPORT-PROOF-STARE: further along, at 13,267,729, the
node stops again for a DIFFERENT reason. Blocks in a band there carry an
attached certificate whose vanityData is the ordinary client string rather than
the digest, at heights a single static interval treats as anchor heights. The
best-supported reading is that the fleet ran a different anchoring configuration
in that window (these values are not consensus-bound), so the package needs a
HISTORICAL anchoring schedule, not one value. Stated rather than implied.
We went and read the field instead of asserting into it, and two things in the
README were saying more than the literature supports.
The attack has a name we were not using. It is posterior corruption, also called
the founders' attack, long-range attack, history revision or costless simulation.
Tas, Tse, Gai, Kannan, Maddah-Ali and Yu (arXiv 2207.08392) survey the four known
mitigations: social consensus, key-evolving signatures, verifiable delay
functions, and timestamping onto a proof-of-work chain.
That paper proves an impossibility, and the previous wording here would have read
as if we had stepped around it. We had not, because it says something narrower
than it first appears: slashable safety, which is safety plus the ability to
punish the attacker, needs an external trusted source. Plain safety against
posterior corruption does not, and key-evolving signatures are shown to give it
under an honest majority. This anchor does not claim slashable safety and the
README now says so.
What it does claim is now stated where it is actually strong. Key-evolving
signatures rest on an old key being gone once deleted. A quantum adversary does
not need it to have been kept: QBFT headers carry addresses, but every seal is an
ECDSA signature from which the public key is recoverable, and from that the
private one. Deleting the key buys nothing when the chain reconstructs it. The
certificate under the block hash still has to be forged, and that needs a secret
no amount of computation derives from the chain.
So this is the post-quantum analogue of what key-evolving signatures give
classically, in the setting where their central assumption is void. Narrower than
checkpointing, and borrowing nothing from another chain. Both true at once.
Sources opened and checked, not recalled: arXiv 2207.08392, eprint 2019/1440,
arXiv 2208.05408.
We cite Azouvi, Danezis and Nikolaenko's "Winkle" for the threat this anchor
addresses. Today we opened the paper instead of repeating the citation, and two
things need saying, because a reader who opens it will notice both.
Winkle does not mention quantum adversaries anywhere. It treats old validator
keys becoming compromised, by any means. The quantum framing is ours.
And Winkle's defence is not this one. It adds a secondary layer of client-based
validation in which clients sign a hash of the previously sequenced block. The
same authors' later work, Pikachu, checkpoints into Bitcoin instead. Both are
established answers to this threat, and a reader weighing our design deserves to
be pointed at them rather than left to find them.
So the README now says it plainly: the threat has peer-reviewed grounding, the
defence in this repository is ours, and the honest difference from checkpointing
is that checkpointing borrows security from another chain while this borrows
nothing.
Citations verified at source, not from memory: eprint.iacr.org/2019/1440, AFT
2020 pp. 189-201; arXiv 2208.05408, 2022.
Caught by our own licence gate the minute the anchor went up, which is the only
reason this is a same-day correction rather than something a reader finds first.
Two separate requirements, both real:
- section 4(b): the twenty upstream files this overlay modifies must carry a
prominent notice that we changed them. They did not. They do now, placed after
the upstream copyright header rather than over it, because 4(c) requires that
header to survive untouched. It does: eighteen still read "Copyright ConsenSys
AG.", two "Copyright contributors to Besu."
- section 4(d): NOTICE must carry the attribution notices of the work this
derives from. It named Hyperledger Besu only, while the files themselves carry
three distinct notices. All three are now reproduced. Naming one of three was a
smaller truth than the files tell.
The patch is regenerated from the corrected files and re-verified end to end, not
assumed: git apply --check and git apply both 0 on a pristine d2032017 checkout,
the resulting tree byte-identical to anchor/ (75 files compared, 0 differences),
and 605 tests with 0 failures across consensus:common and consensus:qbft.
This repository argues, in its own README, that a patch against a named upstream
commit is the honest way to publish a contribution to a million-line project: the
diff is the contribution, the reader fetches the rest from upstream, and a patch
fails loudly when upstream moves while a whole file overwrites in silence.
The anchor was published as whole files only, which contradicted that argument.
It is now both: patches/0003 to apply, anchor/ to read.
Verified on a pristine d2032017 checkout rather than assumed:
- git apply --check and git apply both returned 0
- the resulting tree is byte-identical to anchor/: 75 files compared, 0 differences
- :consensus:common:test and :consensus:qbft:test returned 605 tests, 0 failures,
the same count class by class as the same tree built from the files
The README also said "no build file changes are required" in a place that now
reads as if it covered all three patches. It covered the first two. Patch 0003
changes one build file by one line, and that line is named.
This is the code that puts a post-quantum validator certificate under the block
hash. It is the thing this project exists to do, and it is published so that the
claim can be checked rather than believed.
What it is. In QBFT the block hash is computed over a re-encoding of the decoded
extraData with the seals removed, so anything the decoder does not know about is
dropped before hashing. Appending a certificate as a new element gives you a
certificate that is stored, gossiped, and entirely absent from the hash. The
design that works instead puts a 32-byte digest of the certificate into
vanityData, which is already under keccak. anchor/README.md sets out the four
designs that died before this one and why.
Scope, stated in the README and repeated here because it matters: consensus on
chain 2800 is classical secp256k1 ECDSA. This binds a post-quantum certificate to
the block hash. It does not make consensus post-quantum and is never described as
such.
What is here: the anchor, the validation rules, the wiring, and the tests,
including the negative controls. Applied to upstream d2032017bb, the pinned base
named in anchor/BASE.txt. One build file changes, by one line, and the README says
which and why. No cryptography is implemented here; Falcon verification calls
Bouncy Castle.
What is not here: no keys, no fleet configuration, and nothing about what is armed
on any running network.
Measured before publishing, on upstream d2032017bb with this overlay applied:
consensus:common and consensus:qbft, 605 tests, 0 failures, identical to the same
tree before this work, class by class.
Three things were found while preparing it, and all three are fixed here:
- the code spoke Romanian in 134 comment lines and 43 strings, 37 of them on
production paths, which is to say in the messages a node prints when it refuses
to start. An auditor given the code to check the guards could not read the
guards.
- ten test classes carried internal issue numbers in their names. They now say
what they test.
- the suite was green partly by ordering luck. One class cleared its system
properties but not the configuration PqAnchorProducer remembers, so it left the
anchor armed for whichever class ran next. Renaming the classes changed the
order and four tests began failing on a guard that was firing correctly. Fixed
where it leaks, with the negative control measured: remove the line and the
pair goes red, restore it and it goes green.
