#archlinux32 | Logs for 2024-03-01
Back
[02:57:58] -!- bdju has quit [Ping timeout: 246 seconds]
[03:00:13] -!- bdju has joined #archlinux32
[03:11:56] <KitsuWhooa> abaumann: they seem to be running now
[03:47:23] mavicaway is now known as mavica
[04:14:34] <KitsuWhooa> Is there an easy way to inject variables into builders? /etc/environment doesn't seem to do it
[04:14:43] <KitsuWhooa> I want to try CARGO_REGISTRIES_CRATES_IO_PROTOCOL=sparse
[05:16:20] -!- ssserpent has joined #archlinux32
[06:45:19] -!- titus_livius has joined #archlinux32
[06:46:49] -!- abaumann has joined #archlinux32
[06:48:30] <abaumann> KitsuWhooa: nice, build machines up again :-)
[06:48:45] <abaumann> environment variables, I would probably have added them to the build scripts directly
[06:51:46] <buildmaster> Hi abaumann!
[06:51:46] <buildmaster> !rq abaumann
[06:51:47] <phrik> buildmaster: <abaumann> that's what happens if fixing a bug is faster than checking out the git repo of the linux kernel.
[06:52:15] <KitsuWhooa> that was delayed
[06:52:22] <abaumann> yep.
[06:52:27] <abaumann> I rebooted some stuff
[06:52:38] <abaumann> though, not the buildmaster, mmh..
[06:52:39] <KitsuWhooa> 08:48:45 <abaumann> environment variables, I would probably have added them to the build scripts directly <-- build scripts, as in?
[06:52:54] <abaumann> the core-staging-xxx ones
[06:53:06] <KitsuWhooa> woudln't it need to be added to every single one?
[06:53:18] <KitsuWhooa> seems a bit inefficient
[06:53:21] <abaumann> yes, but those scripts get generated when building devtools32
[06:53:43] <abaumann> there is a Makefile there which adds all kind of things to some templates and then generates the whole bunch
[06:53:47] <KitsuWhooa> ah
[06:54:01] <KitsuWhooa> https://git.archlinux32.org
[06:54:02] <phrik> Title: makechrootpkg.in « src - devtools32 - Archlinux32 fork of devtools (at git.archlinux32.org)
[06:54:04] <KitsuWhooa> guess it'd be this one, right?
[06:55:10] <KitsuWhooa> https://git.archlinux32.org
[06:55:10] <phrik> Title: makechrootpkg.in « src - devtools32 - Archlinux32 fork of devtools (at git.archlinux32.org)
[06:55:12] <KitsuWhooa> I guess I can add it there
[06:55:41] <abaumann> try that. I fail to find the scripts just now..
[06:55:58] <KitsuWhooa> hm
[06:56:05] <KitsuWhooa> would it be possible to add it in makepkg.conf?
[06:56:27] <abaumann> makepkg I think is just taking the variables from the user starting makepkg
[06:56:41] <abaumann> so I doubt there is a special mechanism to add more environment variables in makepkg.conf itself
[06:56:45] <KitsuWhooa> there's a MAKEFLAGS in makepkg.conf that I use to do -j32
[06:56:49] <KitsuWhooa> and that works fine
[06:56:58] <abaumann> yes, but those are flags for 'make'
[06:56:59] <KitsuWhooa> but I don't know if that's treated as an env var or not
[06:57:06] <abaumann> which look like environment variables, but are not.
[06:57:10] <KitsuWhooa> yeah, okay
[06:57:15] <KitsuWhooa> (MAKEFLAGS is an env var though)
[06:57:19] <abaumann> A=B make, A is an enviroment variable
[06:57:28] <abaumann> make A=B, A is a make variable
[06:57:32] <KitsuWhooa> yup
[06:58:50] <abaumann> ah, core-testing-i486-build and friends are links to archbuild
[06:59:12] <abaumann> which calls exec makechrootpkg
[07:00:23] <abaumann> AAA=a exec env | grep AAA
[07:00:30] <abaumann> yields AAA=a
[07:00:36] <KitsuWhooa> yes
[07:00:42] <abaumann> ok, so exec leaves the envoriment as is
[07:00:42] <KitsuWhooa> in this case though, sudo won't preserve the env var
[07:00:53] <abaumann> sudo might have options for that
[07:00:56] <KitsuWhooa> it does
[07:01:01] <KitsuWhooa> --preserve-env=ENV_VAR
[07:01:32] <KitsuWhooa> also, something funny
[07:01:35] <KitsuWhooa> namcap is broken :p
[07:01:49] <abaumann> guess in which language it is written.. ;-)
[07:01:52] <KitsuWhooa> I know :p
[07:01:59] <abaumann> no worries
[07:02:08] <abaumann> namcap runs upstream already on the package
[07:02:24] <abaumann> and we never added arch32-specific rules - though we had the plan to do so
[07:02:38] <KitsuWhooa> There's another dependency loop in two python packages, and I need to rebuild one of them so that it gets the correct dependency on python-jaraco.functools so that everything else works
[07:02:53] <KitsuWhooa> I wonder if I should just copy the existing package, modify the dependencies and inject it
[07:03:42] <abaumann> you mean, copy PKGBUILD?
[07:03:50] <KitsuWhooa> no, the actual built package
[07:03:52] <abaumann> ah.
[07:03:57] <abaumann> yeah, why not.
[07:03:58] <KitsuWhooa> the PKGBUILD is correct
[07:04:58] <abaumann> mmh. eurobuild6 is quite silent..
[07:05:01] <abaumann> ..let's investigate
[07:05:12] <abaumann> I might have the same issue as you had with your build machines..
[07:06:41] <KitsuWhooa> The reason I wanted to inject the env var is because of https://blog.rust-lang.org
[07:06:42] <phrik> Title: Help test Cargo's new index protocol | Inside Rust Blog (at blog.rust-lang.org)
[07:06:47] <KitsuWhooa> it should make things faster
[07:07:02] <abaumann> yes, but I think rust is broken anyway and needs a bootstrap.
[07:07:05] <KitsuWhooa> oh I know
[07:07:12] <KitsuWhooa> but it'll make builds fail faster in the meantime
[07:07:21] <abaumann> and make them fail faster.
[07:07:24] <KitsuWhooa> yes
[07:07:27] <abaumann> yes, that's a good thing. :-)
[07:07:40] <KitsuWhooa> and use less bandwidth too
[07:07:44] <KitsuWhooa> or at least I hope
[07:25:14] -!- abouvier has quit [Quit: kthxbye]
[07:32:36] -!- abouvier has joined #archlinux32
[07:41:47] -!- abouvier has quit [Quit: kthxbye]
[07:42:26] -!- abouvier has joined #archlinux32
[08:34:37] -!- mvchtz has quit [Ping timeout: 264 seconds]
[08:36:36] mavica is now known as mavicaway
[08:38:21] -!- mvchtz has joined #archlinux32
[11:47:46] -!- lithiumpt has quit [Quit: ZNC 1.8.2 - https://znc.in]
[11:48:06] -!- lithiumpt has joined #archlinux32
[13:41:45] <KitsuWhooa> oh wow this is so much faster
[13:41:54] <KitsuWhooa> (CARGO_REGISTRIES_CRATES_IO_PROTOCOL=sparse)
[14:12:33] -!- ssserpent has quit [Quit: WeeChat 4.2.1]
[14:46:16] <abaumann> good :-)
[16:16:08] -!- abaumann has quit [Quit: leaving]
[16:44:12] <KitsuWhooa> Maybe I'm wrong
[16:44:18] <KitsuWhooa> I think it's faster but I can't tell
[17:36:11] mavicaway is now known as mavica
[18:08:08] -!- buildmaster has quit [Read error: Connection reset by peer]
[18:10:26] -!- buildmaster has joined #archlinux32
[21:18:27] T`aBoWRiS is now known as T`aZeL
[21:20:51] -!- GNUtoo has quit [Remote host closed the connection]
[21:21:14] -!- GNUtoo has joined #archlinux32
[22:00:49] T`aZeL is now known as B`aZeL