success. I increased `:datasize-cur:` and `:datasize-max:` in /etc/login.conf for the `default` and `staff` classes so it applied to my user, which I verified using `ulimit -a`, and then that specific cpp file succeeded in getting built

the default value was 1536M, and I changed it to 5536M

So it *was* a heap issue, and compiling that file required more than 1.5G of heap space

.merges

.merge+ 9079 9080 9069 9072 9073 9076 9077

Added

awww yes, the build succeeded!

it feels so good to succeed at building the repo after so much effort

So the issue was just openbsd restricting core usage ?

core usage? I think it was heap space, given my previous messages

Yes. That.

I see. Well yeah, it was openbsd having low default limits for the heap size

You may want to update the OpenBSD section in the README.md with your changes then.

and I increased it as described above, by changing `:datasize-cur:` (important!) in `/etc/login.conf` such that my `ulimit -a` specified I had 5G instead of 1G of heap size

ooh sweet

Actually, it might do the same.,,

yeah... erm. I think I succeeded in building without reading this section

I'm sorry

<0​xfffc:matrix.org> mesaoptimizer

<o​frnxmr:monero.social> > No. Try the stack size limit increase. That should work, if it didn’t post the failure log.

<o​frnxmr:monero.social> .

the unit tests fail for me on the debug build

one failed test: node_server.bind_same_p2p_port

Which line ?

How do I figure out which line it is? I just ran `ctest` in the `build/debug/tests/unit_tests

I thought it'd mention it.

Nope, it didn't -- the last few lines are mesaoptimizer.com/test_failure.txt

Did you look at the test output above ?

unfortunately my terminal trashed the output that would have contained this detail

okay got it

line 324 and 327 based on mesaoptimizer.com/LastTest.log

seems like this may be an issue on my end, given that what is failing is an init() call

okay, I ran `doas ifconfig lo0 alias 127.0.0.2` and reran the failing test and it succeeded

I should probably add a relevant line to the file

here's a patch to document this fix: mesaoptimizer.com/0001-OpenBSD-node…-fix-IP-localhost-by-aliasing.patch

mesaoptimizer: do you have a github account?

otherwise I can submit the patch

selsta: I do, I use the `mesaoptimizer0` github account when I need to use github

I submitted the patch since I thought it would be easier than doing a fork + commit + pull request

hash-slow-4 failed for `gmake release-test` due to segfault. Will investigate further tomorrow