You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I've just updated my flake inputs to the latest rev of dsc on this repo (a873d9b) and it looks like at that revision, the postInstall step for the built binary is failing (this is on my M1 macbook and a rosetta builder, but I've verified it on a real linux machine too):
:; nix build .#packages.x86_64-linux.dsc[2/635/638 built, 373 copied (3362.8/3362.9 MiB), 737.4/737.4 MiB DL] building dsc-deps-0.12.0-pre (buildPhase): Compiling serde_derive v1.0.217^Rerror: build of '/nix/store/f4vbjfa5vvcravd2lv0kyz99jr4nwxnl-dsc-0.12.0-pre.drv' on 'ssh-ng://rosetta-builder' failed: builder for '/nix/store/f4vbjfa5vvcravd2lv0kyz99jr4nwxnl-dsc-0.12.0-pre.drv' failed with exit code 2; last 25 log lines: > Running phase: updateAutotoolsGnuConfigScriptsPhase > Running phase: configurePhase > [naersk] cargo_version (read): 1.83.0 (5ffbef321 2024-10-29) > [naersk] cargo_message_format (set): json-diagnostic-rendered-ansi > [naersk] cargo_release: --release > [naersk] cargo_options: > [naersk] cargo_build_options: $cargo_release -j "$NIX_BUILD_CORES" --message-format=$cargo_message_format > [naersk] cargo_test_options: $cargo_release -j "$NIX_BUILD_CORES" > [naersk] RUST_TEST_THREADS: 8 > [naersk] cargo_bins_jq_filter: select(.reason == "compiler-artifact" and .executable != null and .profile.test == false) > [naersk] cargo_build_output_json (created): /build/tmp.9Q7OdtLnBH > [naersk] RUSTFLAGS: > [naersk] CARGO_BUILD_RUSTFLAGS: > [naersk] CARGO_BUILD_RUSTFLAGS (updated): --remap-path-prefix /nix/store/1p8sbpjv32yhhgw6qkrga2rfvpzs77fs-crates-io-dependencies=/sources --remap-path-prefix /nix/store/gx2rmn4950qmmq7ja38p5fkmp4a48544-git-dependencies=/sources > [naersk] pre-installing dep /nix/store/bssbj2dln4x9cr9az3s5pw42g6w68qqc-dsc-deps-0.12.0-pre > Running phase: buildPhase > cargo build $cargo_release -j "$NIX_BUILD_CORES" --message-format=$cargo_message_format > warning: `/build/source/.cargo-home/config` is deprecated in favor of `config.toml` > note: if you need to support cargo 1.38 or earlier, you can symlink `config` to `config.toml` > Compiling dsc v0.12.0-pre (/build/source) > Finished `release` profile [optimized] target(s) in 20.86s > Running phase: installPhase > [naersk] Using file /build/tmp.9Q7OdtLnBH to retrieve build (executable) products > [naersk] found executable dsc -> /nix/store/43kv2abdmqjlgpkwd0cmkwjjc5rc84w6-dsc-0.12.0-pre/bin/dsc > ContextCreate - An error occurred creating the http client: builder error For full logs, run 'nix log /nix/store/f4vbjfa5vvcravd2lv0kyz99jr4nwxnl-dsc-0.12.0-pre.drv'.error: builder for '/nix/store/f4vbjfa5vvcravd2lv0kyz99jr4nwxnl-dsc-0.12.0-pre.drv' failed with exit code 1
With a .overrideAttrs (old: {postInstall = "";}) this error goes away, and the derivation can be built again.
I'm not sure why it's only failing on linux (or what that error even indicates), but looks like something is off there.
The text was updated successfully, but these errors were encountered:
antifuchs
changed the title
flake.nix: installPhase on linux fails with "ContextCreate - An error occurred creating the http client: builder error"
flake.nix: postInstall on linux fails with "ContextCreate - An error occurred creating the http client: builder error"
Jan 27, 2025
Thanks for reporting. This seems to be a problem creating the http client. Unfortunately, builder error is not very helpful :-/ I suspect that creating the client does some impure stuff now and this fails in the nix sandbox. outside, the post-install seems to work fine for me. In code it would be:
I've just updated my flake inputs to the latest rev of dsc on this repo (a873d9b) and it looks like at that revision, the postInstall step for the built binary is failing (this is on my M1 macbook and a rosetta builder, but I've verified it on a real linux machine too):
With a
.overrideAttrs (old: {postInstall = "";})
this error goes away, and the derivation can be built again.I'm not sure why it's only failing on linux (or what that error even indicates), but looks like something is off there.
The text was updated successfully, but these errors were encountered: