Hi! I've seen this self-check error only randomly, and only ever saw it on s390x and riscv64. On most builds it works fine.
Failing build log:
https://buildd.debian.org/status/fetch.php?pkg=golang-filippo-sunlight&arch=s390x&ver=0.8.0-1&stamp=1776364990&raw=0
Working build log:
https://buildd.debian.org/status/fetch.php?pkg=golang-filippo-sunlight&arch=s390x&ver=0.8.0-1&stamp=1776369096&raw=0
It is identical code and (should be) identical host and software environment, only two different runs of the build.
The failing log ends like this:
testlog_test.go:90: time=2026-04-16T18:42:03.156Z level=ERROR source=ctlog.go:880 msg="failed to trim entry" err="failed to parse certificate: asn1: struc
ture error: tags don't match (16 vs {class:0 tag:17 length:52 isCompound:true}) {optional:false explicit:false application:false private:false defaultValue:<nil>
tag:<nil> stringType:0 timeType:0 set:false omitEmpty:false lax:true name:} certificate @2"
testlog_test.go:90: time=2026-04-16T18:42:03.156Z level=ERROR source=ctlog.go:880 msg="failed to trim entry" err="failed to parse certificate: asn1: struc
ture error: tags don't match (16 vs {class:0 tag:17 length:52 isCompound:true}) {optional:false explicit:false application:false private:false defaultValue:<nil>
tag:<nil> stringType:0 timeType:0 set:false omitEmpty:false lax:true name:} certificate @2"
ctlog_test.go:87: expected size 1500005, got 1497005
--- FAIL: TestSequenceLargeLog (255.24s)
The many ERROR lines does not seem like the trigger, they occur in the passing self tests too. Are they intentional or not? They cause the build log to be quite big.
Thanks,
Simon
Hi! I've seen this self-check error only randomly, and only ever saw it on s390x and riscv64. On most builds it works fine.
Failing build log:
https://buildd.debian.org/status/fetch.php?pkg=golang-filippo-sunlight&arch=s390x&ver=0.8.0-1&stamp=1776364990&raw=0
Working build log:
https://buildd.debian.org/status/fetch.php?pkg=golang-filippo-sunlight&arch=s390x&ver=0.8.0-1&stamp=1776369096&raw=0
It is identical code and (should be) identical host and software environment, only two different runs of the build.
The failing log ends like this:
The many
ERRORlines does not seem like the trigger, they occur in the passing self tests too. Are they intentional or not? They cause the build log to be quite big.Thanks,
Simon