v29.0 Testing #32052

1. 
   fanquake commented at 5:57 am on March 13, 2025: member

   Umbrella issue for 29.0 testing. Please help testing on a wide variety of supported platforms, as well as interaction with different software.

   Let us know which version you tested on which operating system.

   If you find an issue, please search Github for known issues first and then open a new Github issue.

   This meta issue should not be used to report bugs, as a single thread makes it impossible to track more than one topic.

   See 29.0 Release Notes Draft for a list of changes.

   See here for the testing guide: https://github.com/bitcoin-core/bitcoin-devwiki/wiki/29.0-Release-Candidate-Testing-Guide.

2. fanquake added this to the milestone 29.0 on Mar 13, 2025
3. fanquake pinned this on Mar 13, 2025
4. 
   achow101 commented at 0:32 am on March 14, 2025: member

   #31844 Broke MacOS codesigning and notarization as it adds the manpage to the app, which signapple currently cannot cover in its signature. Since we did not have the manpage there before, and I don’t think MacOS can even use that manpage, I think it would be reasonable to drop that from the app. I think we can skip codesigning for 29.0rc1.
5. fanquake closed this on Mar 14, 2025

   ---

6. fanquake referenced this in commit f4b3a5858a on Mar 14, 2025
7. achow101 reopened this on Mar 14, 2025

   ---

8. 
   Sjors commented at 4:18 pm on March 14, 2025: member

   This was fixed in #32064. So I guess we can’t do the guix-sign step until the next release candidate?
9. 
   hebasto commented at 4:08 pm on March 24, 2025: member

   I’ve completed testing the v29.0.0rc2 release candidate on Windows 11 Pro 24H2, using the default Edge browser for downloading.

   All signatures look good. The browser flagged bitcoin-29.0rc2-win64-setup-unsigned.exe, which is expected:

   

   The binaries bitcoind.exe, bitcoin-cli.exe, bitcoin-qt.exe and test-bitcoin.exe function as expected.

   I was unable to test the -natpmp option, unfortunately.

10. 
    laanwj commented at 10:16 am on March 25, 2025: member

    ✅ i tested that the new NATPMP implementation works with protonvpn’s automatic port forwarding (v29.0rc1 tag built from git):

     0# ip netns exec bitcoin-test bin/bitcoind  -regtest -natpmp=1 -debug=net
     12025-03-25T10:08:03Z Bitcoin Core version v29.0rc1 (release build)
     2⋮
     32025-03-25T10:09:20Z mapport thread start
     42025-03-25T10:09:20Z [net] portmap: gateway [IPv4]: 10.2.0.1
     52025-03-25T10:09:20Z [net] pcp: Requesting port mapping for addr 0.0.0.0 port 18444 from gateway 10.2.0.1
     62025-03-25T10:09:20Z Bound to 127.0.0.1:18445
     72025-03-25T10:09:20Z [net] pcp: Internal address after connect: 10.2.0.2
     8⋮
     92025-03-25T10:09:20Z [net] pcp: Received response of 8 bytes: 008000010058139c
    102025-03-25T10:09:20Z [net] portmap: Got unsupported PCP version response, falling back to NAT-PMP
    112025-03-25T10:09:20Z [net] natpmp: Requesting port mapping port 18444 from gateway 10.2.0.1
    122025-03-25T10:09:20Z [net] natpmp: Received response of 12 bytes: 008000000058139c4f7f83dd
    132025-03-25T10:09:20Z [net] natpmp: Received response of 16 bytes: 008200000058139c480ca7720000003c
    142025-03-25T10:09:20Z [net:info] portmap: Added mapping natpmp:79.127.131.221:42866 -> 10.2.0.2:18444 (for 60s)
    152025-03-25T10:09:20Z AddLocal(79.127.131.221:42866,3)
    162025-03-25T10:10:34Z [net] portmap: Could not determine IPv6 default gateway
    17⋮
    182025-03-25T10:09:57Z [net] portmap: gateway [IPv4]: 10.2.0.1
    192025-03-25T10:09:57Z [net] pcp: Requesting port mapping for addr 0.0.0.0 port 18444 from gateway 10.2.0.1
    202025-03-25T10:09:57Z [net] pcp: Internal address after connect: 10.2.0.2
    

    It works successfully, and it re-requests the lease in time, even though the expiry time they give is non-characteristically short (60s). It also adds the correct external address/port for advertising even though it doesn’t match the internal one.

11. 
    darosior commented at 8:22 pm on March 26, 2025: member

    I’ve tested to run a downstream project’s (Liana’s) functional test suite against v29.0rc2. It did not surface any bug nor RPC interface regression. I also ran it with a bitcoind under Valgrind.

    I’ve also been running a (non-listening) v29.0rc2 mainnet node under Valgrind for a few days, sanity checking various operations but nothing really involved.

    I’ve also tried to get end users to test the new port mapping implementation on whatever router they use at home. Work in progress but a few of them have been sharing results in #31663.

12. 
    Lagrang3 commented at 3:09 pm on March 28, 2025: none

    I’ve upgraded my mainnet node to v29.0rc2 with no issues running the following services:

    • electrs
    • datum
    • and core-lightning

Contributors
fanquake achow101 Sjors hebasto laanwj darosior Lagrang3

Milestone
29.0