Aussie Dollar is tanking, wish I had bought at the dip last friday but was at work =/

let it die

gov fiat

hmm, monerod dies with a segfault after running start_mining in the cli wallet

is that a known problem?

it happened with 15.1 and 15.5 as well

running in gdb but it's a release build

here's the bt

Get us a stack trace (bt in gdb).

ty

randomx_calculate_hash(

)

hmm

randomx is said to be exercising CPUs hard and some are buggy. Seems a bit of a copout, but... :)

try: disas

Wait:

thread 41

disas

No function contains program counter for selected frame.

Try up, then disas again.

ok, will recurse up until i get something

And after done: disas 0x00007fe9e5f90983

ok

(gdb) disas 0x00007fe9e5f90983

No function contains specified address.

hmm, call stack garbled?

disas 0x00007fe9e5f90900-0x00007fe9e5f90a00

(gdb) disas 0x00007fe9e5f90900-0x00007fe9e5f90a00

No function contains specified address.

Why does it care about this -_-

gdb you mean? no idea.. maybe there's a flag to make it not care

let's see

Anyway, you'll want to wait for sech1 or tevador to show up, they'll know much better what info they need to work out whether it's a real bug.

oh syntax

it's start,end, not start-end

there's a disassembly for you

Hmm. No memory access there, so SIGSEGV looks unlikely.

from the first location in the bt i get:

0x00007fe9e5df3800: imul $0x42,%ebx,%esp

my assembly fu is old. is that an indirect adressing there?

The SIGSEGV seems to come from 983.

right. one off error :)

It weirds me out to have esp be patr of an imul.

Didn't even know they could encode that.

hmm, lemme run the damn thing in valgrind. maybe i get something interesting by tomorrow ;)

it doesn't like being run in valgrind at all it seems

fps what CPU do you run it on?

a ryzen something something

hold on

model name : AMD Ryzen 7 1700X Eight-Core Processor

Some first gen Ryzens are known to crash on RandomX.

XMRig has a workaround for it, but it's not in monerod

it's a hardware bug

is there possibly a microcode fix?

You can try to turn off opcache in BIOS

search for "opcache" there

ok, got it. thanks

Finally, after weeks: "SYNCHRONIZED OK"

The RaspberryPi has synced :)

\o/

Would anyone here send me a quarter of a cent (or smaller) so I can experiment with the setup a bit?

Feuermagier: sure

<B

binaryFate: why are the newest cli binaries not signed with fluffypony's key, but instead with yours?

and why did fluffypony not sign your key?

fluffypony wanted out of this, and binaryFate volunteered to take over. binaryFate's key should be in the git repo, and pony should have signed the merge commit. Is this not the case (double check first) ?

he has signed my key, refresh from servers maybe

as did others

see web.getmonero.org/2019/12/16/technical-responsibilities-update.html for more info

ewaf ^

binaryFate: i imported your key from the git repo (aswell as fluffypony's key) and gpg --list-sig [key id] just outputs user id not found

(well i didn't import fluffypony's key, i had it already)

the repo contains keys only, not signatures

import my key from a key server using its ID and you should get signatures along with it

but doesn't [user id not found] say that there are signatures attached?

mmm not sure about the error message sorry

you can import from hkps:/peegeepee.com or hkp://pgp.mit.edu with ID F0AF4D462A0BDF92

it should keep my key (as you have it already) but download sigs as well

but why don't you attach fluffypony's signature to the key in the repo, so it's easier for users who haven't used monero in a while and don't want to use keyservers because it might break their gpg installation through key poisoning?

because nobody asked and we didn't consider :)

I'll check how to do that

ok, i guess i'm the only one who's verifying signatures but not keeping up with blog posts

ewaf: I can do that, but didn't I merge binaryFate's key into the repo anyway?

iirc

i just imported the key from a keyserver and the sig is listed now

ok cool

can i still use wallet v 14

no

i try to open monerod and it tells me to delete blockchain

jeez fluffypony

i never have this many problems with bitcoin wallet bro

it's telling you to delete blockchain.info

it's helping you

i gotta bypass content filters just to download it

gah

i'm trying .15 now

.15 works i like it better with monerod running in background

fluffypony hows the coronavirus situation in your area

not great - total lack of testing

USA is closed pretty much

they've only really been testing people who have an international travel history

ic

so I think there are a lot of people infected who don't know yet

give it a week or two and it's going to be a mess

my company is based out of germany and everything has been shut down

no meetings more than 10 people

no travel

heavy

stay safe fluffypony

Hello! I apologize if this would be better answered within the price chat but I was wondering if you guys are familiar with software that works with the terminal for linux to get crypto prices? I found a few but they all rely on nodejs and I really want to avoid dealing with that and the npm package management hell.

npnx: you can use curl and jq to call the api of a site like coinpaprika.com or coinmarketcap.com

npnx: `curl rate.sx`

Thanks guys. Probably the simplest and most cross platform. Thank you. How is everyone? Staying healthy?

npnx: if you want something scriptable: curl -s api.coinpaprika.com/v1/tickers/xmr-monero | jq '.quotes.USD.price'

selsta: that is honestly amazing! Thank you and same to you asymptotically.

there is also curl rate.sx

curl rate.sx/xmr and so on

Yeah. That is the type of simple solution I needed. Very nice.

the latter version has nice color tty charts :)

curl rate.sx/:help for more goodies

it's pretty featureful

Yeah. It is very neat.

unrelated to crypto, same author has wttr.in which is pretty terminal weather forecasts

It kills me that no one has ported the xmr wallets to openBSD. Not even the cli one. That is also really cool. Thank you.

the daemon and CLI wallet both work on openbsd

Really?! I have had no luck with it before. Let me try again.

hello

after inserting trace paste.debian.net/1135645 i had not smth new paste.debian.net/1135646

You get the line numbers in the log file.

there are no line numbers

just the output that u c

Did you actually look in the log file ?

yes nothing new

I see: src/cryptonote_core/blockchain.cpp:1129coinbase transaction in the block has no inputs

this was before too but the first error is src/cryptonote_basic/cryptonote_format_utils.cpp:204 Failed to parse transaction from blob

if i remove from basic.h major_version 12 daemon start

Ah, true.

Then go to that line 204, and adds logs to whatever this calls, etc, till you find what exactly is the low level test that fails.

when i add to 204 cant compile

only after BEGIN_SERIALIZE() mginfo("trace"0 works

u cant add everywhere

<i​dunno> its 5 ppl now

i cant add mginfo everywhere what to do?

Learn C. It has to be synctactically correct. MGINFO() is a statement, it goes where statements can go.

BEGIN_SERIALIZE is a macro that's a function definition header IIRC. Behaves like one anyway.

back

moneromooo where to add that mginfo

Hey ndorf. I am getting the dependencies for building monero on OpenBSD. No package for openpgm but I want to say OpenBSD has a native pgm setup. Is that what you used?

Or is that what you are aware of for those that wre able to use monero-cli on openbsd?

IIRC i built libzmq and openpgm myself

I know zeromq exists as a package for OpenBSD. I will keep that in mind. Thank you.

ah yeah, i can't remember why i built it manually at the moment

this is on one of the laptops i haven't seen in a couple of months, but if you get really stuck hit me up again and i'll dig it out

Means a lot. Thank you.

ndorf: did you build FreeBSD or Linux?

build monero you mean? yes

i don't think i had to build any dependencies on FreeBSD (or Linux), just packages