mirror of
https://github.com/hickory-dns/hickory-dns.git
synced 2026-04-25 03:05:51 +03:00
[GH-ISSUE #980] thread 'trust-dns-server-runtime' has overflowed its stack #582
Labels
No labels
blocked
breaking-change
bug
bug:critical
bug:tests
cleanup
compliance
compliance
compliance
crate:all
crate:client
crate:native-tls
crate:proto
crate:recursor
crate:resolver
crate:resolver
crate:rustls
crate:server
crate:util
dependencies
docs
duplicate
easy
easy
enhance
enhance
enhance
feature:dns-over-https
feature:dns-over-quic
feature:dns-over-tls
feature:dnsssec
feature:global_lb
feature:mdns
feature:tsig
features:edns
has workaround
ops
perf
platform:WASM
platform:android
platform:fuchsia
platform:linux
platform:macos
platform:windows
pull-request
question
test
tools
tools
trust
unclear
wontfix
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
starred/hickory-dns#582
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @Darkspirit on GitHub (Jan 4, 2020).
Original GitHub issue: https://github.com/hickory-dns/hickory-dns/issues/980
Describe the bug
Sometimes the server crashes after few seconds or after a few hours:
To ReproduceSteps to the behavior. Sorry, it doesn't seem reliably reproducible at the moment. I'll add more info, as soon as I get more.
Expected behavior
No stack overflow.
System:
Version:
Crate: trust-dns
Version: 0.18.0
Additional context
@bluejekyll commented on GitHub (Jan 4, 2020):
Do you think you could increase the log level and perhaps set the RUST_BACKTRACE=full environment variable?
That might help us see where the program is dying.
@Darkspirit commented on GitHub (Jan 4, 2020):
With
I don't get any backtrace or further error log. There was just a request, then it died.
But it seems I have found STR and could narrow it down to the regressing commit, it will take some time.
Even 0.16 dies with:
@Darkspirit commented on GitHub (Jan 4, 2020):
searching InMemoryAuthority is the last action before the crash.
@bluejekyll commented on GitHub (Jan 4, 2020):
Ok, that's feasible. I think that function is recursive, so it might e doing something incorrectly and thus very possible to overflow the stack.
@Darkspirit commented on GitHub (Jan 4, 2020):
This domain usually just redirects to the GitHub project. I registered it last month before some idiot could take it away from us all.
@Darkspirit commented on GitHub (Jan 4, 2020):
RUST_BACKTRACE=full RUST_LOG=debug gdb --args named --config /home/trustdns/config.toml --zonedir /home/trustdns/zones --debuggdb.txt
(Please instruct me what to do, sorry.)
@bluejekyll commented on GitHub (Jan 4, 2020):
Ok, this is strange. It looks like a wildcard lookup is being triggered, but there is no wildcard in your domain.
@Darkspirit commented on GitHub (Jan 4, 2020):
I've mailed you the full log, including the startup.
@bluejekyll commented on GitHub (Jan 4, 2020):
I'm looking at this, and we have decent test coverage here. I'm hoping to reproduce it with
authority_battery/basic.rstest case, though, it doesn't look simple for some reason.@bluejekyll commented on GitHub (Jan 4, 2020):
Ok, I've reproduced, it looks like it's the
@ 600 IN MX 0 .record.Can you explain what the intention of that is?
@Darkspirit commented on GitHub (Jan 4, 2020):
https://www.hardenize.com/report/rustls.com#email
@Darkspirit commented on GitHub (Jan 4, 2020):
darkspirit.eu has such a record for months, but it seems some mail servers are interested in sending spam to rustls.com, so its MX record was actually requested by someone.
@bluejekyll commented on GitHub (Jan 4, 2020):
Ok, it looks like we aren't properly processing the
NULLrecord in regards to.as the target in the zone file. So this support will need to be added.@bluejekyll commented on GitHub (Jan 4, 2020):
Ok, it turns out that with MX records, when we're looking up additional records for the response, it we ended up continuing to search for ever, bad base case in the recursion.