| 17 Jan 2026 |
asa | In reply to @k900:0upti.me And even -C opt-level=3 is different from -O3 never thought about this but I guess it's a good thing rust release builds don't use fast-math by default... | 22:44:11 |
asa | i looked into this more and it seems thin local lto is redundant with codegen-units=1 anyway (and rustc disables thin local lto if codegen-units=1 and the lto level is unspecified)
buildRustCrate has defaultCodegenUnits ? 1 since the derivation doesn't allow for rebuilds on a higher granularity than per-crate anyway
| 23:28:51 |
| 18 Jan 2026 |
debtquity | dynamic derivations | 00:21:58 |
debtquity | ? | 00:21:59 |
| isabel changed their profile picture. | 20:43:40 |
debtquity | i get flashbacks to random segfaults in c/c++ when this happens:
> thread 'main' (13304) panicked at cargo-auditable/src/rustc_wrapper.rs:35:10:
> rustc was terminated by a deadly signal
> stack backtrace:
> 0: __rustc::rust_begin_unwind
> 1: core::panicking::panic_fmt
> 2: core::option::expect_failed
> 3: cargo_auditable::rustc_wrapper::main
> 4: cargo_auditable::main
> note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
💀
| 23:43:35 |
debtquity | beginning to understand why the maintainer abandoned trying to update this package. somehow it compiles fine on darwin, but trying with aarch64 it throws this junk | 23:44:43 |
Winter | what are you trying to compile | 23:46:03 |
debtquity | this one: https://github.com/NixOS/nixpkgs/blob/49213403e38ca893cafc83571d05bebbb782ff89/pkgs/by-name/st/stalwart-mail/package.nix | 23:48:12 |
debtquity | think it might just be a local issue with vm, probably wont see this issue on hydra | 23:56:07 |
Winter | what signal was rustc terminated by | 23:56:56 |
debtquity | not sure, only got this in the output:
Caused by:
Caused by:
process didn't exit successfully: `/nix/store/y7xffr129r4iqjwnrzsylp59837lkfbq-cargo-auditable-0.7.2/bin/cargo-auditable rustc ... (exit status: 101)
| 23:58:51 |
| 19 Jan 2026 |
rosssmyth | It does default to thin local LTO between codegen units within a single crate aka between modules | 01:31:31 |
| Sapphire changed their profile picture. | 20:16:20 |
| 20 Jan 2026 |
| cheako joined the room. | 05:54:14 |
cheako | I must be doing something wrong because I get proc-macro server version check failed from rust-analyzer. I'm on a steamdeck and it's all setup to run most everything from podman and nix making assumptions about a "pristine" system breaks things. Like keeping the provided LD_LIBRARY_PATH, this made the installer(nix-store) segfault. | 06:09:47 |
cheako | I actually came to ask about developing for android, from amd64. I think I should be using cargo-ndk and there is a package for that. | 06:11:20 |
cheako | /nix/store/n234xvnsajgvn5lym30idk66397s86sq-rust-nightly-complete-2026-01-19/bin/rust-analyzer and...
/nix/store/n234xvnsajgvn5lym30idk66397s86sq-rust-nightly-complete-2026-01-19/libexec/rust-analyzer-proc-macro-srv | 06:16:40 |
cheako | (deck@steamdeck-oled ~)$ head -n30 .local/bin/{.codium.nix,codium}
==> .local/bin/.codium.nix <==
with (import <nixpkgs>) { };
mkShell {
nativeBuildInputs = [
pkg-config
clang_18
lld_18
];
buildInputs = [
vscodium
neovim
nixfmt
(let
fenix = import (fetchTarball "https://github.com/nix-community/fenix/archive/main.tar.gz") { };
in
fenix.complete.toolchain)
pre-commit
# rustfmt
# rustPackages.clippy
alsa-lib
udev
vulkan-loader
vulkan-headers
vulkan-tools
xorg.libX11
xorg.libXrandr
xorg.libXcursor
xorg.libXi
(android-studio.withSdk (androidenv.composeAndroidPackages { includeNDK = true; }).androidsdk)
cargo-ndk
];
==> .local/bin/codium <==
#!/bin/sh
exec nix-shell --run codium "$(dirname "$(realpath $0)")"/.codium.nix
| 06:18:30 |
cheako | shellHook = ''export LD_LIBRARY_PATH="''${LD_LIBRARY_PATH:+$LD_LIBRARY_PATH:}${
pkgs.lib.makeLibraryPath [
alsa-lib
udev
vulkan-loader
libxkbcommon
]
}"'';
RUST_SRC_PATH = rustPlatform.rustLibSrc;
}
| 06:20:47 |
cheako | Also I added [ "$TEMP" ] && mkdir -p "$TEMP" to my .profile, but still I need to run it manually. | 06:22:50 |
cheako | I'll be back in 8hrs. | 06:24:39 |
| innocentzer0 joined the room. | 14:40:21 |
cheako | (deck@steamdeck-oled ~)$ nix-shell -p nix-info --run "nix-info -m"
- system: `"x86_64-linux"`
- host os: `Linux 6.11.11-valve27-1-neptune-611-g60ef8556a811, SteamOS, noversion, 20260108.1`
- multi-user?: `yes`
- sandbox: `yes`
- version: `nix-env (Nix) 2.31.2+2`
- channels(root): `"nixpkgs"`
- nixpkgs: `/nix/store/ss16zjzngmqn3idqk8a8csh7x1xjkqf7-nixpkgs/nixpkgs`
| 17:47:36 |
| faye joined the room. | 17:52:10 |
Marie | have you tried rust from nixpkgs instead of fenix? | 19:13:21 |
cheako | Yes, I moved to fenix because rust-analyzer was unable to process macros. | 19:24:03 |
Pol | Hello,
I'm trying to build fresh-editor 0.1.86 and I got this issue: https://github.com/NixOS/nixpkgs/pull/482054 | 19:56:36 |
Epic Eric | I attached a breakpointHook and got the following error while trying to compile it manually: | 21:30:12 |
Epic Eric | --- stderr
thread 'main' (7836) panicked at /build/fresh-0.1.86-vendor/bindgen-0.72.1/lib.rs:616:27:
Unable to find libclang: "couldn't find any valid shared libraries matching: ['libclang.so', 'libclang-*.so', 'libclang.so.*', 'libclang-*.so.*'], set the `LIBCLANG_PATH` environment variable to a path where one of these files can be found (invalid: [])"
stack backtrace:
0: __rustc::rust_begin_unwind
1: core::panicking::panic_fmt
2: core::result::unwrap_failed
3: core::result::Result<T,E>::expect
4: bindgen::ensure_libclang_is_loaded::{{closure}}
5: std::sync::once_lock::OnceLock<T>::get_or_init::{{closure}}
6: std::sync::once_lock::OnceLock<T>::initialize::{{closure}}
7: std::sync::poison::once::Once::call_once_force::{{closure}}
8: std::sys::sync::once::futex::Once::call
9: std::sync::poison::once::Once::call_once_force
10: std::sync::once_lock::OnceLock<T>::initialize
11: std::sync::once_lock::OnceLock<T>::get_or_try_init
12: std::sync::once_lock::OnceLock<T>::get_or_init
13: bindgen::ensure_libclang_is_loaded
14: bindgen::Bindings::generate
15: bindgen::Builder::generate
16: build_script_build::bindgen
17: build_script_build::main
18: core::ops::function::FnOnce::call_once
| 21:30:37 |