bug: bls12: failed to verify (aggregated) message

[meling]

I have not investigated this, but it only occurred for two out of 2880 benchmark runs. It occurred for both old and new runs, and the issue is not consistent across a specific set of parameters. This is all the information I have, as it occurred during benchmark runs without any other logging:

(old version)

--- FAIL: BenchmarkVerifyAggregateQC/Crypto=bls12/Cache=0/Participants=800/QCsPerParticipant=4-12
    cert_test.go:460: VerifyAggregateQC failed: bls12: failed to verify message

--- FAIL: BenchmarkVerifyAggregateQC/Crypto=bls12/Cache=10/Participants=800/QCsPerParticipant=1-12
    cert_test.go:460: VerifyAggregateQC failed: bls12: failed to verify aggregated message

(new version)

--- FAIL: BenchmarkVerifyAggregateQC/Crypto=bls12/Cache=10/Participants=400/QCsPerParticipant=4-12
    cert_test.go:460: VerifyAggregateQC failed: bls12: failed to verify aggregated message

--- FAIL: BenchmarkVerifyAggregateQC/Crypto=bls12/Cache=10/Participants=800/QCsPerParticipant=40-12
    cert_test.go:460: VerifyAggregateQC failed: bls12: failed to verify message

[meling]

Trying to debug this I added a panic and more debug output and ran with high -count value on the test:

=== RUN   TestVerifyAggregateQCReproduceBLS12FailVerify/Crypto=bls12/Cache=10/Participants=800/QCsPerParticipant=40
Failed to verify message: pubKey=9110a12631844394471459978e7d85cb1f92b5a0eb4963e229f1ade6e9ff8d31b300096370071db40360a53e4ab98554, message=9110a12631844394471459978e7d85cb1f92b5a0eb4963e229f1ade6e9ff8d31b300096370071db40360a53e4ab98554, signature=a046d72d0e7c1b6910ececc0e9efc82ed7174bf08ee5f00038865330e8e98a81569435c7d32963426194251d4a50121e0b7212567b12a2f9c514afe7d8b313f78340521b454f0addb18bf6d3692f58b6cd94a065b714dab69b984e5eacdac9bf
engine: &{0x14004fd3090 0x14004fd30e0 0x14004f210e0 0x14007029640 {[0x1400502f8c0 0x1400502f920 0x1400502f980 0x1400502f9e0 0x1400502fa40 0x1400502faa0 0x1400502fb00 0x1400502fb60 0x1400502fbc0] [[[[[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]]] [[[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]]]] [[[[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]]] [[[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]]]] [[[[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]]] [[[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]]]]]} []}
engine result: &[[[[8505329371266088957 17002214543764226050 6865905132761471162 8632934651105793861 6631298214892334189 1582556514881692819] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]]] [[[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0]]]]
signature: &[[[3963448189575496513 12883331095396856690 12512881503907965865 7964920898647163380 13571063699431002408 608149944225842178] [18242113633994204789 4664896052734048334 11428417160730411639 14043890458199195797 16966916673528871487 37718794275284723]] [[14959396524692122192 15371270268078706150 3561745321422278379 3686834865569669584 13249600719543493187 1073968602886936036] [4194557025160364193 13755251249120667681 2845399486965878670 16223990394999992877 16878611295472031775 206659576077864173]] [[8505329371266088957 17002214543764226050 6865905132761471162 8632934651105793861 6631298214892334189 1582556514881692819] [0 0 0 0 0 0]]]
Failed to verify aggregated message: signature=a046d72d0e7c1b6910ececc0e9efc82ed7174bf08ee5f00038865330e8e98a81569435c7d32963426194251d4a50121e0b7212567b12a2f9c514afe7d8b313f78340521b454f0addb18bf6d3692f58b6cd94a065b714dab69b984e5eacdac9bf
--- FAIL: TestVerifyAggregateQCReproduceBLS12FailVerify (35.57s)
    --- PASS: TestVerifyAggregateQCReproduceBLS12FailVerify/Crypto=bls12/Cache=0/Participants=800/QCsPerParticipant=4 (5.93s)
    --- PASS: TestVerifyAggregateQCReproduceBLS12FailVerify/Crypto=bls12/Cache=10/Participants=800/QCsPerParticipant=1 (4.49s)
    --- PASS: TestVerifyAggregateQCReproduceBLS12FailVerify/Crypto=bls12/Cache=10/Participants=400/QCsPerParticipant=4 (2.97s)
    --- FAIL: TestVerifyAggregateQCReproduceBLS12FailVerify/Crypto=bls12/Cache=10/Participants=800/QCsPerParticipant=40 (22.18s)
panic: bls12: verification failed to verify message [recovered, repanicked]

goroutine 569 [running]:
testing.tRunner.func1.2({0x104ffba40, 0x1051079f0})
	/opt/homebrew/Cellar/go/1.25.3/libexec/src/testing/testing.go:1872 +0x190
testing.tRunner.func1()
	/opt/homebrew/Cellar/go/1.25.3/libexec/src/testing/testing.go:1875 +0x31c
panic({0x104ffba40?, 0x1051079f0?})
	/opt/homebrew/Cellar/go/1.25.3/libexec/src/runtime/panic.go:783 +0x120
github.com/relab/hotstuff/security/crypto.(*bls12Base).coreVerify(0x0?, 0x140004938d0, {0x1400501f1d0, 0x30, 0x30}, 0x1400502b200, {0x1054eee00, 0x2b, 0x2b})
	/Users/meling/work/hotstuff/security/crypto/bls12.go:228 +0x324
github.com/relab/hotstuff/security/crypto.(*bls12Base).popVerify(0x14001caa240, 0x140004938d0, 0x1400502b200)
	/Users/meling/work/hotstuff/security/crypto/bls12.go:244 +0x5c
github.com/relab/hotstuff/security/crypto.(*bls12Base).checkPop(0x14001caa240, 0x140007d1c40)
	/Users/meling/work/hotstuff/security/crypto/bls12.go:269 +0x308
github.com/relab/hotstuff/security/crypto.(*bls12Base).publicKey(0x14001caa240, 0x1e8)
	/Users/meling/work/hotstuff/security/crypto/bls12.go:182 +0x80
github.com/relab/hotstuff/security/crypto.(*bls12Base).BatchVerify(0x14001caa240, {0x10510e128?, 0x14006774b40?}, 0x1400839f620)
	/Users/meling/work/hotstuff/security/crypto/bls12.go:433 +0x3d0
github.com/relab/hotstuff/security/cert.(*Cache).BatchVerify(0x14008500240, {0x10510e128, 0x14006774b40}, 0x1400839f620)
	/Users/meling/work/hotstuff/security/cert/cache.go:108 +0x22c
github.com/relab/hotstuff/security/cert.(*Authority).VerifyAggregateQC(0x14006fb58e0, {0x1400839f5f0?, {0x10510e128?, 0x14006774b40?}, 0x6ef?})
	/Users/meling/work/hotstuff/security/cert/auth.go:168 +0x340
github.com/relab/hotstuff/security/cert_test.TestVerifyAggregateQCReproduceBLS12FailVerify.func1(0x14000103c00)
	/Users/meling/work/hotstuff/security/cert/cert_test.go:478 +0x50
testing.tRunner(0x14000103c00, 0x14001caa060)
	/opt/homebrew/Cellar/go/1.25.3/libexec/src/testing/testing.go:1934 +0xc8
created by testing.(*T).Run in goroutine 578
	/opt/homebrew/Cellar/go/1.25.3/libexec/src/testing/testing.go:1997 +0x364
exit status 2
FAIL	github.com/relab/hotstuff/security/cert	3267.427s

I suspect there is a bug in the bls12 library that we use, which is no longer maintained; see issue #242. Therefore, I will not debug this more until we have replaced the bls12 dependency with a maintained one.