| 6 Nov 2024 |
Artturin |
I sent to someone
Literally SomeoneSerge (utc+3)
| 23:06:43 |
Artturin | haha | 23:06:48 |
frontear | appreciated looks like need to write down some notes as i read these thank you | 23:10:18 |
Artturin | https://github.com/NixOS/nixpkgs/blob/master/pkgs/stdenv/linux/default.nix | 23:11:19 |
Artturin | vs cross https://github.com/NixOS/nixpkgs/blob/master/pkgs/stdenv/cross/default.nix | 23:11:31 |
Artturin | 699 lines vs 107 | 23:11:41 |
frontear | honestly i expected the complete opposite | 23:12:00 |
frontear | id expected cross setup to be more difficult | 23:12:11 |
Artturin | The hard work of bootstrapping is done in the native stages | 23:12:37 |
Artturin | So cross is mostly just compiling packages like normal | 23:13:28 |
| 7 Nov 2024 |
trofi | Yeah. It still needs to build cross-gcc twice (and arguably does it slightly incorrectly WRT libgcc.so), but it's not as involved as native tools are already there (and use native bootstrap). | 16:56:46 |
| 8 Nov 2024 |
| moved to @amadaluzia:tchncs.de changed their display name from (artur 'manuel) to moved to @amadaluzia:tchncs.de. | 14:29:49 |
| 9 Nov 2024 |
emily | is doing stdenv.override { whateverPlatform = …; } meant to work? | 08:43:56 |
emily | it seems to not regenerate wrappers or anything | 08:44:00 |
emily | (looking for an LLVM stdenv that targets a different platform. I assume this should not require rebuilding all of LLVM since it's multi-target) | 08:49:12 |
Artturin | In reply to @emilazy:matrix.org it seems to not regenerate wrappers or anything CC not overriden automatically https://github.com/NixOS/nixpkgs/blob/master/pkgs/stdenv/generic/default.nix you'll have to pass CC yourself https://github.com/NixOS/nixpkgs/blob/a31f2a7b37f2c0cc463768c2637413838da0f35b/pkgs/stdenv/cross/default.nix#L83 | 14:48:04 |
emily | right | 14:48:25 |
emily | it seemed like I'd need multiple layers of nested overrides and I resorted to pkgsCross in the end | 14:48:43 |
Tristan Ross | In reply to @emilazy:matrix.org is doing stdenv.override { whateverPlatform = …; } meant to work? Oh, that might explain why I see the Flutter engine update script not working after the changes with the platform attributes and using the ones from stdenv. | 15:46:11 |
emily | hm? | 15:46:27 |
emily | what changes? | 15:46:30 |
Tristan Ross | The ones that changes the consumption of `*Platform` -> `stdenv.*Platform` | 15:47:12 |
Tristan Ross | I think Artturin did that PR | 15:47:26 |
Tristan Ross | It seems to have messed up the source fetching since Flutter engine sources are based on host and target. | 15:48:03 |
Tristan Ross | I tried fixing it with overriding stdenv but that didn't help | 15:48:33 |
emily | that seems really unlikely | 15:49:18 |
emily | there's no semantic difference | 15:49:22 |
Artturin | In reply to @rosscomputerguy:matrix.org It seems to have messed up the source fetching since Flutter engine sources are based on host and target. The pr only unalised the usages | 15:49:26 |
Artturin | * The pr only unaliased the usages | 15:49:40 |
Tristan Ross | In reply to @emilazy:matrix.org there's no semantic difference Then why does the update script now fetch aarch64-linux for x86_64-linux while host is aarch64-linux. | 15:50:42 |