.merges

.merge+ 8677 8678 8675 8676

Added

sech1: I added both your PRs

we have quite a lot of changes which would make sense to tag v0.18.2.0 and also do some testing on testnet / stagenet before release

tag it as v0.18.2.0 instead of v0.18.1.3

selsta so 2-3 weeks for testing and then release? Can the release wait until February? I'm planning a P2Pool fork: SChernykh/p2pool #224 so it would be nice to sync releases of GUI v0.18.2.0 and P2Pool v3.0.

sech1: yes the release can wait until February and we can do testing in the meantime

I'll be doing P2Pool testing too with a few other people, and I'll ask them to use the current monero release branch :)

we also have Dandelion++ changes

hope luigi can do merges soon™ :D

was there some progress also on community nodes?

rbrunner: Regarding your second point. Supposing we move to Seraphis or similar down the line. Would the approach be to convert an "old" wallet2 style multisig wallet, into a new Seraphis compatible wallet. Or... would they simply create a new multisig wallet with Seraphis, and then transfer the funds over to it

#no-wallet-left-behind:monero.social john_r365:

Domt kidding me please

So converting 24 words to 25 words will my 25 words match or not if someone else creates new monero wallet? For example, if it does match, what is the probability?

I want to answer

I love you monero

Answer me ofrnxmr

Ofrnxmr if you don't answer me, I will f*ck your wife as5 h0le. I will f1sting your wife as5 h0le.

Shes gonna fist you 😀

If you have no wife, I make your mom

Dont threaten me with a good time

Do you still want those pics

Foot pict?

So converting 24 words to 25 words will my 25 words match or not if someone else creates new monero wallet? For example, if it does match, what is the probability?

Negligible

Answer mee

So converting 24 words to 25 words will my 25 words match or not if someone else creates new monero wallet? For example, if it does match, what is the probability?

sorry, ignore maratov^

So converting 24 words to 25 words will my 25 words match or not if someone else creates new monero wallet? For example, if it does match, what is the probability?

I answered, negligible is the correct term

I suggest not answering flaming assholes (at least after they prove to be flaming assholes).

.merges

sech1: kayabanerve[m]: can you reply here so we can get this merged? monero-project/monero #8675#issuecomment-1379186484

I think the comment was in regards to a different proposed solution, not the PR itself

yes, that's my understanding. My question is how does the current solution (cache in this pr) handle it?

.merges

I answered

thanks. All merged

ty

sech1: I'm getting "Couldn't allocate RandomX light VM using large pages" and "Couldn't allocate RandomX cache using large pages" on start

macOS

You need to enable huge pages

or reboot if they're already enabled

does that even exist on macOS? and we can't expect all monerod users to enable this?

according xmrig.com/docs/miner/hugepages they're called super pages on macOS

this is just a warning and it monerod can work fine if it allocates RandomX without large pages (it will hash a bit slower)

also, this warning doesn't appear in logs by default

you need to enable "randomx" log catergory to see it

hmm ok, wait

I think I know the issue

sech1: paste.debian.net/hidden/8ab71a75

this is with log level 0

can you check bitmonero.log? It shows log categories at startup

New log categories: *:WARNING,net:FATAL,net.http:FATAL,net.ssl:FATAL,net.p2p:FATAL,net.cn:FATAL,daemon.rpc:FATAL,global:INFO,verify:FATAL,serialization:FATAL,daemon.rpc.payment:ERROR,stacktrace:INFO,logging:INFO,msgwriter:INFO

ah right

because it's a warning

but you shouldn't get these warnings all the time. It tries to allocate VM only once per thread

selsta so from what I can see, when it's syncing, it's checking a lot of RandomX hashes in worker threads at the same time, so you have 16 worker threads there and get 16 warnings

it keeps printing this message 4x

it printed it hundreds of times now during sync

everytime there is a "Synced 2782981/2797403 (99%, 14422 left)" message there are also 4x warning messages

that's not good

so it creates new threads in groups of 4 every time

and then allocates VMs in these threads and then these VMs leak

slow_hash_free_state() should be called when a thread that calculated RandomX hashes before, exits

it is probably the case here, but I need to double check

yes, it's Blockchain::block_longhash_worker() responsible for warning spam, but it cleans up everything correctly, so no memory leak

and the warning spam is _because_ it cleans up everything after every batch, and then tries to allocate VMs again

it should be an easy fix - just print a warning once

ok will test once you have a patch

selsta monero-project/monero #8705

sech1: works now but is it a good idea to print this at log level 0? maybe better to move it to log level 1?

it's not at log level 0, it's a warning

and default logger settings show all warnings

we will get lots of reports about what this is and if randomx is broken lol

an opportunity to educate people about large pages :P

anyway, people can discuss it in 8705 now - either leave it as is, or demote it from warning to info message

It seems like a pointless warning to issue repeatedly. The patch seems fine.