08:03:33 <m-relay> <j​effro256:monero.social> Is anyone else's compiler seg faulting very often on the core codebase? Mine started doing it a few weeks ago with g++ 12.3.0. Switching to g++ 13 didn't help AFAICT. It's infuriating. I have to rerun `make` like 30% of the time. I don't think it's from not having enough memory since peak RAM usage is 75% and swap space (8GB) is empty
08:04:57 <moneromoooo> After a git pull ?
08:05:52 <moneromoooo> You could try checking out the old source and seeing if that helps. Could be a compiler bug. For OOM, it should log those in dmesg.
08:06:16 <m-relay> <0​xfffc:matrix.org> Can you run make in verbose mode and find which command is seg faulting
08:12:49 <m-relay> <j​effro256:monero.social> moneromoooo: i'll try reproducing after a git pull. seems almost entirely random at this point. The file the trace log returns as the culprit also seems random
08:13:23 <m-relay> <j​effro256:monero.social> 0xFFFC: are you saying to run g++ itself in verbose mode ?
08:17:31 <m-relay> <0​xfffc:matrix.org> That is next step. For ninja, I use `-v`. For make I don’t remember of the top of my head since haven’t used it for few years. I believe it had `-n` which only prints the command (not run). Try with `-d`
08:21:33 <m-relay> <j​effro256:monero.social> moneromooo: doesn't have to be after a git pull
08:37:07 <moneromoooo> make VERBOSE=1 will show what make rnus, but not what gcc runs (ie, linker, cpp, etc).
08:37:42 <moneromoooo> I did suggest a git pull, I asked whether the problem started (shortly) after a git pull.
08:37:52 <moneromoooo> I did *not* suggest a git pull, I asked whether the problem started (shortly) after a git pull.
08:38:14 <moneromoooo> A compiler bug often triggers on particular code.
08:38:53 <moneromoooo> You then said random location though, so that looks more like OOM...
08:39:07 <moneromoooo> Or broken hardware.
08:49:06 <m-relay> <0​xfffc:matrix.org> Yes. Once we find out which gcc command fails. Next step is to find out which component fails. And fire up gdb
08:53:15 <sech1> you can edit CMakeLists.txt - find "set(CMAKE_CXX_FLAGS_RELEASE "-DNDEBUG ${OPT_FLAGS_RELEASE}")" line and add "-v" flag. It will make g++ print a lot of stuff for each file, including command line and search paths for includes
09:00:42 <moneromoooo> What I do when I want to run custom commands is: make target VERBOSE=1 
09:01:01 <moneromoooo> Then ^C when I see what I want, I copy/paste/alter that command. 
09:01:25 <moneromoooo> Changing CMakeLists.txt will try to rebiuld everything.
09:01:50 <moneromoooo> Low tech, but works when I want to build just one file with -g -O0 :)
09:15:30 <sech1> Rebuild everything is more likely to trigger the bug though, if the bug is random
09:15:56 <moneromoooo> Ah, good point...
15:30:10 <m-relay> <h​ushrom:matrix.org> Do you guys know the algorithm used by Monero wallets such as monerujo and featherwallet to create derived mnemonic seeds out of offset passphrases?
15:31:20 <hyc> pretty sure it's just a standard libwallet feature, no?
15:32:00 <hyc> otherwise go to their github repo and ask them...
15:50:26 <wasmutable> https://wasmutable.wpenginepowered.com/technology/white-paper-implementing-quantum-syntax-in-monero/
22:55:34 <m-relay> <t​obtoht:monero.social> cryptonote::decrypt_key
23:06:21 <selsta> jeffro256: please also open against release branch
23:55:35 <m-relay> <r​4v3r23:monero.social> jeffro256thanks for your review of 8617
23:55:55 <m-relay> <r​4v3r23:monero.social> selsta what else is needed to get this merged
23:58:03 <selsta> will try to get it merged after we put out 0.18.3.2