--- Log opened Tue Apr 13 00:00:24 2021
03:13 -!- mol [~mol@unaffiliated/molly] has joined #bitcoin-builds
03:15 -!- molz_ [~mol@unaffiliated/molly] has quit [Ping timeout: 265 seconds]
03:20 -!- Newton15Denesik [~Newton15D@static.57.1.216.95.clients.your-server.de] has joined #bitcoin-builds
04:49 -!- Newton15Denesik [~Newton15D@static.57.1.216.95.clients.your-server.de] has quit [Ping timeout: 240 seconds]
06:03 -!- mol_ [~mol@unaffiliated/molly] has joined #bitcoin-builds
06:07 -!- mol [~mol@unaffiliated/molly] has quit [Ping timeout: 252 seconds]
06:29 -!- mol_ [~mol@unaffiliated/molly] has quit [Ping timeout: 268 seconds]
06:33 -!- mol [~mol@unaffiliated/molly] has joined #bitcoin-builds
07:17 -!- belcher_ is now known as belcher
07:55 -!- mol [~mol@unaffiliated/molly] has quit [Ping timeout: 240 seconds]
07:57 < dongcarl> hebasto: Added to my list
08:04 < dongcarl> sipa: I think you should be able to build all HOSTs other than i686 and x86_64 since qt (and depends) makes erroneous assumptions that CC is always multilib
08:06 < hebasto> dongcarl: ty
08:59 -!- sipa [~pw@gateway/tor-sasl/sipa1024] has quit [Ping timeout: 240 seconds]
09:06 -!- sipa [~pw@gateway/tor-sasl/sipa1024] has joined #bitcoin-builds
09:16 -!- mol [~mol@unaffiliated/molly] has joined #bitcoin-builds
09:23 < sipa> dongcarl: others fail also
09:24 < dongcarl> Weird... Lmk if you can send some logs!
09:24 < sipa> which ones you want?
09:25 < sipa> (i don't have the logs anymore, but recreating should be fast)
09:54 -!- mol_ [~mol@unaffiliated/molly] has joined #bitcoin-builds
09:57 -!- mol [~mol@unaffiliated/molly] has quit [Ping timeout: 246 seconds]
12:36 < dongcarl> sipa: Could you send me a combined stdout/err of riscv64-linux-gnu?
12:41 < sipa> dongcarl: running... seems guix has some work with riscv64 first, so may be a few hours
12:42 < dongcarl> sipa: Okie!
12:46 < sipa> this is on an odroid device btw, which seems much easier to get running
12:46 < sipa> it's a 4.9 kernel, and guix seems to work with it
12:59 < dongcarl> Haha... bisecting the kernel my switching devices... nice...
13:00 < dongcarl> I'm sure odroid has their own custom kernel fork for BIG.little or sth
13:00 < dongcarl> s/my/by/
13:06 < sipa> dongcarl: i have no idea how well that's supported really... i think linux just thinks there are 6 threads
13:07 < dongcarl> Right...
13:16 -!- mol_ [~mol@unaffiliated/molly] has quit [Ping timeout: 240 seconds]
13:18 -!- mol [~mol@unaffiliated/molly] has joined #bitcoin-builds
14:50 < sipa> &^!^&%# it seems to be building now
14:50 < sipa> i tried x86_64 and aarch64 before
15:24 < sipa> dongcarl: https://zerobin.net/?6ae3e53f196e3bb6#lvCbbqtAnvAJcO5JlOKtaw5S+uUiEcBuL5B2oHntvcU=
15:25 < sipa> oh, that's maybe a boring OOM
15:57 < emzy> looks like it.
16:02 < emzy> btw. bad idea to run guix --bootstrap --no-substitutes on an old AMD64 thin client. It's the second day and still running. :)
16:04 < sipa> it took many hours (possibly over a day) on my threadripper 2950x as well
16:06 < emzy> oof.
16:06 < sipa> part of the problem is that job/thread management is hard
16:07 < sipa> lots of phases during the process are effectively single-worker limited (because everything before point P is needed to compile python say, and everythibg after it needs python)
16:08 < sipa> so for those you want 1 worker, but with lots of threads
16:08 < emzy> The thin client has 2 cores. No problem there. ;)
16:09 < sipa> but during other times there are tons of tiny jobs that can't effectively use many threads (e.g. because most of its work in a single-threaded configure scriot)
16:11 < emzy> Maybe the packages could be build in parallel, if they are not dependent.
16:12 < sipa> yes, that would be good... but you can't set both the worker count and thread count high... because if ou somehow use both maximally (16 workers, each with 16 threads say... 256 jobs in total) you'll definitely OOM
16:13 < sipa> so ideally there was some way for the build system to distribute cpu slots dynamically to workers
16:13 < sipa> but that sounds very hard
16:14 < emzy> Also after the first run it sould be much faster. I try to use this thin client than only for guix bitcoin builds. That's the plan.
16:23 -!- sipa [~pw@gateway/tor-sasl/sipa1024] has quit [Ping timeout: 240 seconds]
16:30 -!- sipa [~pw@gateway/tor-sasl/sipa1024] has joined #bitcoin-builds
17:27 -!- mol [~mol@unaffiliated/molly] has quit [Ping timeout: 240 seconds]
17:30 -!- mol [~mol@unaffiliated/molly] has joined #bitcoin-builds
17:32 < dongcarl> emzy: Haha yes this might take quite a while for you, lmk if you run into errors
17:33 < dongcarl> sipa: Still running I'm guessing?
17:53 -!- mol_ [~mol@unaffiliated/molly] has joined #bitcoin-builds
17:56 -!- mol [~mol@unaffiliated/molly] has quit [Ping timeout: 240 seconds]
18:10 < sipa> dongcarl: same
18:10 < sipa> this time i used JOBS=2
18:11 < sipa> 6200.53user 223.21system 1:02:02elapsed 172%CPU (0avgtext+0avgdata 1876556maxresident)k
18:11 < sipa> that does not sound like OOM
18:11 < sipa> i'll try JOBS=1...
18:11 < fanquake> not very parallel 😢
18:11 < sipa> system has 4 GiB or RAM
18:12 < fanquake> "Kupdating" is that a typo?
18:12 < fanquake> oh no, I see "[K" is all over the place
19:16 -!- mol [~mol@unaffiliated/molly] has joined #bitcoin-builds
19:20 -!- mol_ [~mol@unaffiliated/molly] has quit [Ping timeout: 240 seconds]
19:22 -!- mol [~mol@unaffiliated/molly] has quit [Ping timeout: 265 seconds]
19:22 < fanquake> utlizing some CPU with a --bootstrap --no-substitutes Guix build
20:23 -!- belcher_ [~belcher@unaffiliated/belcher] has joined #bitcoin-builds
20:26 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 265 seconds]
20:27 < sipa> dongcarl: build for riscv64 completed, but it's not a match
20:27 < sipa> dist-archive does match
20:42 < sipa> trying powerpc64le now (which iirc failed before)
21:26 -!- mol [~mol@unaffiliated/molly] has joined #bitcoin-builds
22:30 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Ping timeout: 246 seconds]
22:31 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #bitcoin-builds
22:34 -!- lukedashjr [~luke-jr@unaffiliated/luke-jr] has joined #bitcoin-builds
22:35 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Ping timeout: 252 seconds]
22:39 -!- lukedashjr is now known as luke-jr
23:34 -!- elichai2 [sid212594@gateway/web/irccloud.com/x-iknxywtqwegwtadk] has quit [Ping timeout: 245 seconds]
23:34 -!- felixweis [sid154231@gateway/web/irccloud.com/x-hizsttpdglmlqbhs] has quit [Read error: Connection reset by peer]
23:35 -!- jakesyl [sid56879@gateway/web/irccloud.com/x-aqlohovkfuqlzkiw] has quit [Read error: Connection reset by peer]
23:36 -!- felixweis [sid154231@gateway/web/irccloud.com/x-rjfaljpefwtmhbry] has joined #bitcoin-builds
23:36 -!- elichai2 [sid212594@gateway/web/irccloud.com/x-pgjznmqnvkisbftf] has joined #bitcoin-builds
23:37 -!- jakesyl [sid56879@gateway/web/irccloud.com/x-xnxvcvtpceedkggt] has joined #bitcoin-builds
--- Log closed Wed Apr 14 00:00:25 2021