| 2 May 2023 |
 raitobezarius | ·72445:·00000000015c6d40····32·FUNC····LOCAL··DEFAULT···12·_RINvNtCsdmovq6qZ2Qe_4core3ptr13drop_in_placeINtNtCsiWvv4Beivar_5alloc7raw_vec6RawVecNtNtCs6wlDC2A5WUZ_15rustc_mir_build6errors7OverlapEEB1o_.llvm.16271328637079777826 just disappear | 21:32:28 |
 trofi | In reply to @raboof:matrix.org
anyone know how to easily reproduce the /nix/store/4rll8l0q3ycfpykwn1wqk93x1jqayqkl-x86_64-unknown-linux-musl-stage-static-gcc-12.2.0.drv problem? If you are willing to tweak gcc derivation you can try GNUMAKEFLAGS = "--shuffle"; https://trofi.github.io/posts/238-new-make-shuffle-mode.html (worth double-checking if make-4.4 is used for that derivation first). | 21:33:47 |
| raitobezarius | but is even rustc supposed to be reproducible? | 21:33:51 |
| raitobezarius | I thought that rustc could output different stuff by compiling the same source | 21:34:05 |
| raitobezarius | https://doc.rust-lang.org/nightly/nightly-rustc/rustc_mir_build/errors/struct.Overlap.html | 21:34:11 |
| raitobezarius |
Note: Most layout information is completely unstable and may even differ between compilations. The only exception is types with certain repr(...) attributes. Please see the Rust Reference’s “Type Layout” chapter for details on type layout guarantees.
| 21:34:18 |
 delroth | I think rustc is supposed to be reproducible, yes: https://github.com/rust-lang/rust/issues/34902 | 21:35:37 |
| delroth | (and its output) | 21:35:44 |
| 4 May 2023 |
|  Kashmira Qeel changed their display name from Keel to Kashmira Qeel. | 08:52:22 |
| Kashmira Qeel changed their display name from Kashmira Qeel to KashmiraQeel. | 08:52:40 |
| Kashmira Qeel changed their display name from KashmiraQeel to Kashmira Qeel. | 08:53:07 |
|  @witcher:envs.net joined the room. | 17:35:40 |
| @witcher:envs.net left the room. | 17:37:10 |
 raboof | finally hacked in a way to get a report of just the runtime dependencies, https://reproducible.nixos.org/nixos-iso-minimal-runtime/ - no surprises there, the Linux BPF/BTF problem and Python remaining. Hoping to do the graphical iso later. | 20:41:36 |
| raitobezarius | what does runtime means? | 20:42:08 |
 baloo | dependencies pulled in the image, filtering out the dependencies necessary to build the image but not included in the image. At least that's what I understand. | 20:44:40 |
| raitobezarius | oh makes sense | 20:44:57 |
| raitobezarius | so that would potentially take care of haskell stuff | 20:45:05 |
| baloo | would still be appreciated for haskell to pull their thing together :D | 20:48:05 |
| baloo | for all the talks about purity and what not | 20:48:13 |
| raitobezarius | afaik it's a very complicated and non actionable problem | 20:48:38 |
| raitobezarius | maybe I'm soft for Haskell but I am not really concerned about it, the existing workaround is to build with 1 core | 20:49:02 |
| baloo | last time I've checked, that was fixed in recent version of ghc but not the whole ecosystem was compatible | 20:49:49 |
| baloo | or at least, the few leads I followed were leading to me fixes made in 9.X (don't remember which version that was) | 20:50:38 |
 Artturin | https://github.com/NixOS/nixpkgs/issues/151347 | 20:52:12 |
| baloo | shit my own commit :D | 20:52:36 |
| baloo | * shit my own comment :D | 20:52:40 |
| Artturin | the issue is still open | 20:52:49 |
| Artturin | * the ghc issue is still open | 20:52:55 |
| baloo | https://gitlab.haskell.org/ghc/ghc/-/merge_requests/7353 that was the commit I had in mind | 20:53:28 |
