19 Oct 2024 |
Fabián Heredia | just upgraded a couple of weeks ago because I did got several OOMs during the previous cycle, now at 128gb memory (and was only building qtwebengine since it had failed yesteday in the same way | 19:19:20 |
K900 | Oh god wait | 19:20:11 |
K900 | It's debug info | 19:20:14 |
K900 | Is the debug info for qtwebengine just | 19:20:21 |
K900 | TOO BIG | 19:20:22 |
Fabián Heredia | * just upgraded a couple of weeks ago because I did got several OOMs during the previous cycle, now at 128gb memory (and was only building qtwebengine since it had failed yesteday in the same way) | 19:20:47 |
Fabián Heredia | Will attempt reverting this commit then:
commit 9c47bda6cc20cf976e5174c149997149e8a518ec
Author: K900 <me@0upti.me>
Date: Fri Oct 18 10:15:35 2024 +0300
qt6: enable debug info for the rest of them
| 19:25:45 |
Fabián Heredia | oh wait no | 19:26:20 |
Fabián Heredia | there is a qtbase one and a "rest of modules" one | 19:26:51 |
Fabián Heredia | Should I try only disabling it for qtwebengine K900 ? | 19:27:16 |
K900 | I'm working on it | 19:27:30 |
K900 | Looks like it's producing more than 4GB of debug info | 19:27:49 |
K900 | So it's not 32-bit addressable | 19:27:53 |
K900 | Which | 19:27:55 |
K900 | Yeah we don't want to ship that | 19:27:59 |
Fabián Heredia | https://github.com/NixOS/nixpkgs/pull/349846 | 19:34:03 |
Fabián Heredia | currently building | 19:34:17 |
Fabián Heredia | ah nvm didn't eval | 19:35:01 |
K900 | Pushed a fix directly | 19:35:07 |
Fabián Heredia | ty, building with your fix/commit | 19:36:40 |
Randy Eckenrode | Re-asking a question from #macos:nixos.org: for PRs that want to use the new SDK that would normally target master, should they target staging-next? | 21:17:27 |
20 Oct 2024 |
Fabián Heredia | afaik if a change that would normally target the main branch dependends on a change that is part of the staging-next cycle it is ok to send it there. | 07:55:04 |
Fabián Heredia |
- though being mindful not send too many separate PRs
| 07:56:00 |
Fabián Heredia | (spaking for myself, though quickly reading the docs those specify that only unbreaking is to be targeted into staging-next) | 07:58:15 |
Fabián Heredia | * (speaking for myself, though quickly reading the docs those specify that only unbreaking is to be targeted into staging-next) | 07:58:21 |
emily | I think we were specifically concerned about mass rebuilds going to staging-next | 07:59:42 |
Fabián Heredia | * afaik if a change that would normally target the main branch dependends on a change that is part of the staging-next cycle it is ok to send it to staging-next. | 08:01:01 |
trofi | https://github.com/NixOS/nixpkgs/pull/350074 should fix eval failure caused by a merge that pulls in duplicate attribute for Open python3Packages.dbt-adapters | 18:10:03 |
trofi | * https://github.com/NixOS/nixpkgs/pull/350074 should fix eval failure caused by a merge that pulls in duplicate attribute for python3Packages.dbt-adapters | 18:10:11 |
Artturin | In reply to @trofi:matrix.org https://github.com/NixOS/nixpkgs/pull/350074 should fix eval failure caused by a merge that pulls in duplicate attribute for python3Packages.dbt-adapters Oops I did the merge | 19:25:24 |