| 14 Aug 2021 |
 lukegb (he/him) | if we want to eke all the life we can from wendy then we should remove the big-parallel required feature and add features for processor extensions to all the builders that do have them | 14:40:36 |
| lukegb (he/him) | but at the moment: the work wendy is doing could be done faster and better by one of the other workers | 14:41:14 |
 Vladimír Čunát | My understanding is that wendy got switched to big-parallel jobs exactly in order to avoid pytorch. | 15:35:53 |
| lukegb (he/him) | Yeah, but then I made pytorch big-parallel because the build takes a long time and is parallelisable xD | 15:43:18 |
| Vladimír Čunát | Oh, I didn't notice the reply and found out independently. | 15:51:25 |
| Vladimír Čunát | From what I've seen, wendy really is more trouble than worth. | 15:53:03 |
| 15 Aug 2021 |
| lukegb (he/him) | grahamc (he/him): can I get a new hydra jobset ala staging (with... whatever shares are appropriate) pointed at https://github.com/lukegb/nixpkgs.git xcrypt? This is for https://github.com/NixOS/nixpkgs/issues/112371 | 02:16:50 |
| Vladimír Čunát | In reply to @lukegb:zxcvbnm.ninja
grahamc (he/him): can I get a new hydra jobset ala staging (with... whatever shares are appropriate) pointed at https://github.com/lukegb/nixpkgs.git xcrypt? This is for https://github.com/NixOS/nixpkgs/issues/112371 I wonder if building all individual packages is an overkill. Only relatively few packages use the library, I think; Arch list: https://archlinux.org/todo/libxcrypt-rebuild/
It might make sense to start just locally. Or do a -small jobset and then (using that from cache) try those packages locally.
A potential risk: I wonder if some build scripts might just silently turn off some features when failing to find libcrypt (or even try to only find it during runtime). But I'd hope the impact won't be significant here.
| 08:25:53 |
| Vladimír Čunát |
It might make sense to start just locally.
Anyway, gcc won't build, so there's not much to do yet :-)
../../../../gcc-10.3.0/libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.cpp:143:10:
crypt.h: No such file or directory
| 09:35:52 |
| Vladimír Čunát | *
It might make sense to start just locally.
Anyway, gcc won't build, so there's not much to rebuild yet :-)
../../../../gcc-10.3.0/libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.cpp:143:10:
crypt.h: No such file or directory
| 09:36:33 |
| lukegb (he/him) | I guess my point was: it's going to be a full rebuild anyway, even of unrelated packages | 10:50:29 |
| lukegb (he/him) | For the same reason as glibc version bumps are full rebuilds | 10:50:41 |
| lukegb (he/him) | I guess maybe I'll just merge it into staging and then we can fix it there rather than have a new jobset if we're reluctant to do that | 11:07:27 |
| lukegb (he/him) | In reply to @vcunat:matrix.org
It might make sense to start just locally.
Anyway, gcc won't build, so there's not much to rebuild yet :-)
../../../../gcc-10.3.0/libsanitizer/sanitizer_common/sanitizer_platform_limits_posix.cpp:143:10:
crypt.h: No such file or directory
Oh right, that's fixed locally, I haven't pushed it yet | 11:09:58 |
| Vladimír Čunát | Yes, it will be a full rebuild either way, but the extra jobset would serve to find breakages, not to utilize those binaries (because staging/next is somewhere else or would move in the meantime anyway). | 11:44:47 |
| Vladimír Čunát | Right now Hydra is fully utilized by staging-next - and I believe it should take precedence even if I created your jobset immediately. (the current iteration has been stabilizing for more than two weeks already) | 11:50:04 |
| lukegb (he/him) | One of the problems we're going to run into though is that bootstrapping it is hard, because they don't ship tarball releases | 12:52:45 |
| 16 Aug 2021 |
| Vladimír Čunát | lukegb (he/him): if you push it, I can build e.g. that list of packages from Arch to test. | 16:15:33 |
| Vladimír Čunát | ... and after those pass, I suspect it might be ready to merge to staging anyway. | 16:16:27 |
| lukegb (he/him) | I got stuck untangling the build with gcc in a satisfactory way | 16:46:28 |
| lukegb (he/him) | I might try and ask upstream to start making "real" releases | 16:46:44 |
| lukegb (he/him) | oh, there's already a bug from the gentoo folks | 17:26:53 |
|  Finn Behrens (he/him) changed their display name from Finn Behrens to Finn Behrens (he/him). | 20:40:50 |
 hexa | https://github.com/NixOS/nixpkgs/pull/133375#issuecomment-899806058 | 23:32:37 |
| hexa | fails to build because it cannot resolve … anything?
building '/nix/store/a6vp8nwp14wl6vsvlbw7124pba93n58l-c-ares-1.17.2.tar.gz.drv'...
warning: unable to download 'http://tarballs.nixos.org/sha256/4803c844ce20ce510ef0eb83f8ea41fa24ecaae9d280c468c582d2bb25b3913d': Couldn't resolve host name (6); retrying in 256 ms
warning: unable to download 'http://tarballs.nixos.org/sha256/4803c844ce20ce510ef0eb83f8ea41fa24ecaae9d280c468c582d2bb25b3913d': Couldn't resolve host name (6); retrying in 600 ms
warning: unable to download 'http://tarballs.nixos.org/sha256/4803c844ce20ce510ef0eb83f8ea41fa24ecaae9d280c468c582d2bb25b3913d': Couldn't resolve host name (6); retrying in 1285 ms
warning: unable to download 'http://tarballs.nixos.org/sha256/4803c844ce20ce510ef0eb83f8ea41fa24ecaae9d280c468c582d2bb25b3913d': Couldn't resolve host name (6); retrying in 2350 ms
warning: unable to download 'https://c-ares.haxx.se/download/c-ares-1.17.2.tar.gz': Couldn't resolve host name (6); retrying in 287 ms
warning: unable to download 'https://c-ares.haxx.se/download/c-ares-1.17.2.tar.gz': Couldn't resolve host name (6); retrying in 707 ms
warning: unable to download 'https://c-ares.haxx.se/download/c-ares-1.17.2.tar.gz': Couldn't resolve host name (6); retrying in 1237 ms
warning: unable to download 'https://c-ares.haxx.se/download/c-ares-1.17.2.tar.gz': Couldn't resolve host name (6); retrying in 2032 ms
error: unable to download 'https://c-ares.haxx.se/download/c-ares-1.17.2.tar.gz': Couldn't resolve host name (6)
builder for '/nix/store/a6vp8nwp14wl6vsvlbw7124pba93n58l-c-ares-1.17.2.tar.gz.drv' failed with exit code 1
cannot build derivation '/nix/store/dzwb1hs1wmmhvw5xxiwklf62ba183xg7-c-ares-1.17.2.drv': 1 dependencies couldn't be built
cannot build derivation '/nix/store/4x03gxybhr6n8k29n021al2zkpjgy37k-nghttp2-1.43.0.drv': 1 dependencies couldn't be built
cannot build derivation '/nix/store/l121a81z0p248f4rmrjd9hxdnlfmnw69-curl-7.76.1.drv': 1 dependencies couldn't be built
cannot build derivation '/nix/store/76ffqb7vv76kambacap9ayy5dz35p9cl-c-ares-1.17.2.tar.gz.drv': 1 dependencies couldn't be built
cannot build derivation '/nix/store/7yizrfya9qn4gcn3by89vqzq0gd33hzg-c-ares-1.17.2.drv': 1 dependencies couldn't be built
error: build of '/nix/store/7yizrfya9qn4gcn3by89vqzq0gd33hzg-c-ares-1.17.2.drv' failed
| 23:43:29 |
| 17 Aug 2021 |
| hexa | works if I prefetch the tarball | 00:33:10 |
 andi- | Perhaps the same as https://github.com/NixOS/nix/issues/5089 | 10:37:42 |
| hexa | Also works If i revert the c-ares bump | 11:53:57 |
| andi- | Was that failure on darwin? | 14:16:28 |
| hexa | I reproduced that on my x86_64 desktop | 14:21:54 |
