Commit graph

15570 commits

Author SHA1 Message Date
eldritch horrors 07ef9f9f5d docs: skip do custom include processing for mdbook
mdbook already does include processing of its own, and the custom
processing code has always admitted as much. we don't need it for
the mdbook build at this point if we run our preprocessors in the
right order, and maybe we can even have mdbook to return complete
pages to us that we only have to pass to lowdown without any more
preprocessing of our own.

Change-Id: Icd978acbc3b1e215fee8f062c53ab2cb2a222ab1
2024-04-09 02:09:36 +00:00
eldritch horrors 14088e1ec5 docs: remove many link-to-self
for some reason these three were anchors, not links, but had they been
links they wouldn't've worked because they're not defined anywhere but
here. in the print version of the manual they're duplicated many times
over (creating id collisions), so we should better remove them anyway.

Change-Id: I8988a7c32c812dee0f0b6d4953faa7cd1255228d
2024-04-09 02:09:36 +00:00
Rebecca Turner 727b43478c Add repl-overlays
Adds a `repl-overlays` option, which specifies files that can overlay
and modify the top-level bindings in `nix repl`. For example, with the
following contents in `~/.config/nix/repl.nix`:

    info: final: prev: let
      optionalAttrs = predicate: attrs:
        if predicate
        then attrs
        else {};
    in
      optionalAttrs (prev ? legacyPackages && prev.legacyPackages ? ${info.currentSystem})
      {
        pkgs = prev.legacyPackages.${info.currentSystem};
      }

We can run `nix repl` and use `pkgs` to refer to `legacyPackages.${currentSystem}`:

    $ nix repl --repl-overlays ~/.config/nix/repl.nix nixpkgs
    Lix 2.90.0
    Type :? for help.
    Loading installable 'flake:nixpkgs#'...
    Added 5 variables.
    Loading 'repl-overlays'...
    Added 6 variables.
    nix-repl> pkgs.bash
    «derivation /nix/store/g08b5vkwwh0j8ic9rkmd8mpj878rk62z-bash-5.2p26.drv»

Change-Id: Ic12e0f2f210b2f46e920c33088dfe1083f42391a
2024-04-08 17:11:47 -07:00
jade 1e74bffd5c pre-commit check for pragma once and ///@file
This is in our style guide, we can cheaply enforce it, let's do it.

```
$ pre-commit
check-case-conflicts.....................................................Passed
check-executables-have-shebangs..........................................Passed
check-headers............................................................Failed
- hook id: check-headers
- exit code: 1

Missing pattern @file in file src/libexpr/value.hh

We found some header files that don't conform to the style guide.

The Lix style guide requests that header files:
- Begin with `#pragma once` so they only get parsed once
- Contain a doxygen comment (`/**` or `///`) containing `@file`, for
  example, `///@file`, which will make doxygen generate docs for them.

  When adding that, consider also adding a `@brief` with a sentence
  explaining what the header is for.

For more details: https://wiki.lix.systems/link/3#bkmrk-header-files

check-merge-conflicts....................................................Passed
check-shebang-scripts-are-executable.....................................Passed
check-symlinks.......................................(no files to check)Skipped
end-of-file-fixer........................................................Passed
mixed-line-endings.......................................................Passed
no-commit-to-branch......................................................Passed
release-notes........................................(no files to check)Skipped
treefmt..................................................................Passed
trim-trailing-whitespace.................................................Passed
```

Fixes: lix-project/lix#233
Change-Id: I77150b9298c844ffedd0f85cc5250ae9208502e3
2024-04-08 16:10:57 -07:00
jade c58e3f826e pragma once and ///@file everything missing it
Change-Id: Ia1a72348336794b5fb9f2694dd750266089b904e
2024-04-08 15:40:12 -07:00
jade 54bc1f1b98 "but doctor, I AM the untrusted store": nix doctor had wrong trustedness
This probably snuck in in a refactor using truthiness or so. The
trustedness flag was having the optional fullness checked, rather than
the actual contained trust level.

Also adds some tests.

```
m1@6876551b-255d-4cb0-af02-8a4f17b27e2e ~ % nix store ping
warning: 'nix store ping' is a deprecated alias for 'nix store info'
Store URL: daemon
Version: 2.20.4
Trusted: 0
m1@6876551b-255d-4cb0-af02-8a4f17b27e2e ~ % nix doctor
warning: 'doctor' is a deprecated alias for 'config check'
[PASS] PATH contains only one nix version.
[PASS] All profiles are gcroots.
[PASS] Client protocol matches store protocol.
[INFO] You are trusted by store uri: daemon
```

Fixes: lix-project/lix#232
Change-Id: I21576e2a0a755036edf8814133345987617ba3d0
2024-04-08 15:40:12 -07:00
jade 0d37b99a15 build: enable libstdc++ assertions
Benchmarks say that it does not regress performance by more than 1%
(which is where it gets really hard to measure accurately anyhow).

Meson appears to be planning to do this for us without asking us in a
release we will get in the future, and it seems good enough to ship
today:
https://mesonbuild.com/Release-notes-for-1-4-0.html#ndebug-setting-now-controls-c-stdlib-assertions

Benchmarks:

| Command | Mean [ms] | Min [ms] | Max [ms] | Relative |
|:---|---:|---:|---:|---:|
| `result-asserts/bin/nix --extra-experimental-features 'nix-command flakes' eval -f ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix` | 418.4 ± 25.0 | 396.9 | 451.2 | 1.01 ± 0.08 |
| `result/bin/nix --extra-experimental-features 'nix-command flakes' eval -f ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix` | 416.1 ± 23.9 | 397.1 | 445.4 | 1.00 |

| Command | Mean [s] | Min [s] | Max [s] | Relative |
|:---|---:|---:|---:|---:|
| `GC_INITIAL_HEAP_SIZE=10g result-asserts/bin/nix eval --extra-experimental-features 'nix-command flakes' --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'` | 4.147 ± 0.021 | 4.123 | 4.195 | 1.00 |
| `GC_INITIAL_HEAP_SIZE=10g result/bin/nix eval --extra-experimental-features 'nix-command flakes' --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'` | 4.149 ± 0.027 | 4.126 | 4.215 | 1.00 ± 0.01 |

| Command | Mean [s] | Min [s] | Max [s] | Relative |
|:---|---:|---:|---:|---:|
| `result-asserts/bin/nix --extra-experimental-features 'nix-command flakes' eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'` | 5.838 ± 0.023 | 5.799 | 5.867 | 1.01 ± 0.01 |
| `result/bin/nix --extra-experimental-features 'nix-command flakes' eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'` | 5.788 ± 0.044 | 5.715 | 5.876 | 1.00 |

| Command | Mean [s] | Min [s] | Max [s] | Relative |
|:---|---:|---:|---:|---:|
| `result-asserts/bin/nix --extra-experimental-features 'nix-command flakes' search --no-eval-cache github:nixos/nixpkgs/e1fa12d4f6c6fe19ccb59cac54b5b3f25e160870 hello` | 15.993 ± 0.081 | 15.829 | 16.096 | 1.01 ± 0.01 |
| `result/bin/nix --extra-experimental-features 'nix-command flakes' search --no-eval-cache github:nixos/nixpkgs/e1fa12d4f6c6fe19ccb59cac54b5b3f25e160870 hello` | 15.897 ± 0.075 | 15.807 | 16.047 | 1.00 |

Fixes: lix-project/lix#4

Change-Id: Id3a6f38274ba94d5d10b09edd19dfd96bc3e7d5f
2024-04-08 15:40:12 -07:00
jade 06f17a5c78 release-notes: check with pre-commit
This required making the build-release-notes script understand how to
check multiple directories.

Change-Id: I057f5f636155ab6c6fb5755da5217b7e72249ece
2024-04-08 15:40:12 -07:00
jade 6fcab7ee95 pre-commit: stop using the flake
The flake for pre-commit-checks is rather questionable. We ignored
it so it uses our own nixpkgs and doesn't reimport nixpkgs. This should
save a couple of seconds of eval time!

Change-Id: I4584982beb32e0122f791fa29f6a544bdbb9e201
2024-04-08 15:29:23 -07:00
jade 860b34cf07 build-release-notes: support multiple PRs/CLs
Change-Id: I923e4af099d02d2324777cc9f7e6d38b130aac10
2024-04-08 15:29:19 -07:00
Rebecca Turner 5ec256625e Merge "Don't run libstore unit tests in project root" into main 2024-04-08 22:08:34 +00:00
Rebecca Turner 3910d4aa68 Merge "Fix REPL test parser nits" into main 2024-04-08 21:52:06 +00:00
Rebecca Turner cfbcf12276 Format Nix code with nixfmt
Change-Id: I61efeb666ff7481c05fcb247168290e86a250151
2024-04-08 13:00:00 -07:00
Rebecca Turner 12f2565f91 Don't run libstore unit tests in project root
This keeps the libstore unit tests from writing
`libstore-unit-tests.xml` to the project root.

Change-Id: I0d9909aabf9f3574cc1e72a5ae81daefba9a394b
2024-04-08 12:43:35 -07:00
Rebecca Turner 4a49a497eb Fix REPL test parser nits
Follow-up to https://gerrit.lix.systems/c/lix/+/546

Change-Id: Ie603f01e5520329bf879e061cea9e3fba45213fd
2024-04-08 11:14:34 -07:00
Rebecca Turner b323340538 Add nixfmt
Change-Id: I7f21695e3971cfd02b2cce0dd016ff6eb3389905
2024-04-08 09:42:34 -07:00
jade e55fc5af71 Merge "nix-doc -> lix-doc, make self-contained in package.nix" into main 2024-04-08 07:45:00 +00:00
Rebecca Turner 4600eabdae Merge "Rename some parse failure tests to parse-fail-*" into main 2024-04-08 06:33:51 +00:00
Rebecca Turner 8ed8b9aef0 Merge "Add PathsSetting" into main 2024-04-08 06:31:11 +00:00
jade 0145d45806 nix-doc -> lix-doc, make self-contained in package.nix
package.nix previously needed this callPackage'd externally, which
didn't make a lot of sense to us since this is an internal dependency.
Thus we changed it to make it more self contained.

Change-Id: I4935bc0bc80e1a132bc9b1519e917791da95037c
2024-04-08 04:05:13 +00:00
Robert Hensing b995c17f0e Merge pull request #10208 from 9999years/print-strings-directly
`:print` strings directly in `nix repl`

(cherry picked from commit 3539172fd2f7cee639ce46423c58beca4231f2db)
Change-Id: I1972f3bf3b56312851f38288509d371d37f21677
Upstream-PR: https://github.com/NixOS/nix/pull/10208
2024-04-07 19:10:43 -07:00
Rebecca Turner 6f863e8ccd Add PathsSetting
Change-Id: I1165f6ef033a5f757ca3716d3f8008ba36b01fd0
2024-04-07 18:22:15 -07:00
jade de20392c37 flake: remove dead code
Some of this code existed for installer tests, and indeed its removal is
an indication that our daemon cross-compatibility tests were removed.
Although these are not like, super critical tests, we would like to
restore them.

See: lix-project/lix#33
Change-Id: I75c733b25c00eca3a9676d498703bbfc1d6ec21b
2024-04-07 17:30:22 -07:00
Qyriad 84599abd4c Merge "flake: fix rl-next and rl-next-dev checks after 32eaa8a29 (fix nix flake check)" into main 2024-04-07 23:26:19 +00:00
Qyriad 0c5e2cfb33 package: just use fileset from lib
The following command is now sufficient to build Lix from outside of the
flake:

nix-build -E 'let pkgs = import <nixpkgs> { }; in pkgs.callPackage
./package.nix { build-release-notes = false; nix-doc = pkgs.callPackage
./nix-doc/package.nix { }; }'

Change-Id: Ie6b14b446480ac07c7266d4fba20042b04cc35b9
2024-04-07 23:10:05 +00:00
Qyriad 55e0c73695 flake: fix rl-next and rl-next-dev checks after 32eaa8a29 (fix nix flake check)
follow-up to 32eaa8a29[1] "flake: move release note checks to hydraJobs",
this commit fixes a load-bearing typo for`checks.rl-next` and
`checks.rl-next-dev`.

[1]: 32eaa8a2910793538deab31f85534faf7e722ef7

Change-Id: I9383ed21f7eccc337c0c2f65525418b735a94a1d
2024-04-07 17:05:51 -06:00
Qyriad 3ac2dd0613 package: put boehmgc patch logic in package.nix
In our view it really doesn't make sense to not have this in in
package.nix in some way. These patches aren't just for performance or
something -- Lix flat out doesn't build without these patches.

(Arguably that makes them a buildsystem responsibility as well, but that
can wait for when we're ready to start adding subproject fallback
dependency resolution to Meson.)

This is a step towards making `package.nix` more self-sufficient and
`callPackage`able without excessive external logic.

With this change the following command is enough to build Lix from out
of the flake:

nix-build -E 'let pkgs = import <nixpkgs> { }; in pkgs.callPackage
./package.nix { build-release-notes = false; inherit (pkgs.lib) fileset;
nix-doc = pkgs.callPackage ./nix-doc/package.nix { }; }'

Change-Id: Ia37fe8171f87d3293033de8be07d9bab12716f1d
2024-04-07 22:44:04 +00:00
Rebecca Turner 06e11778b5 Merge "Rewrite REPL test parser" into main 2024-04-06 21:36:26 +00:00
Qyriad 5bac308c7c meson: fix warm nix3 CLI manual generation
`nix eval --write-to` refuses to write to a directory that exists at
all, so now we generate in a temporary directory, and copy the generated
tree to the build directory. This is equivalent to what the Make
buildsystem did, actually, but hopefully more robust.

Future work: documenting the doc generation architecture in the
top-level meson.build outline comment.

Change-Id: Ic3eb6d26e3cc249a1c042fd3ced22d637ac66a69
2024-04-06 14:43:14 -06:00
eldritch horrors 1b5b9de04e flake: move release note checks to hydraJobs
having them in checks only does not run them in CI, which can cause
broken release notes entries to pass.

fixes #228

Change-Id: If0ba7b1be0b6525fc884a27e941cbc84b5a160f9
2024-04-06 15:51:52 +00:00
terru - a33c95be5b avoid markdown which the repl's :doc cannot handle
code blocks, if not surrounded by empty lines, have the language
tags (in these cases, always `nix`) show up in the output of :doc.

for example:

  nix-repl> :doc builtins.parseFlakeRef
  Synopsis: builtins.parseFlakeRef flake-ref

    Parse a flake reference, and return its exploded form.

    For example: nix builtins.parseFlakeRef
    "github:NixOS/nixpkgs/23.05?dir=lib"  evaluates to: nix { dir =
    "lib"; owner = "NixOS"; ref = "23.05"; repo = "nixpkgs"; type =
    "github"; }

is now instead:
  nix-repl> :doc builtins.parseFlakeRef
  Synopsis: builtins.parseFlakeRef flake-ref

    Parse a flake reference, and return its exploded form.

    For example:

      | builtins.parseFlakeRef "github:NixOS/nixpkgs/23.05?dir=lib"

    evaluates to:

      | { dir = "lib"; owner = "NixOS"; ref = "23.05"; repo = "nixpkgs"; type = "github"; }

(closes #225)

Change-Id: I0741aeb1006a5376bb2f663d202c7a4da7e38cce
2024-04-06 13:14:14 +00:00
Qyriad fae48884c5 Merge "package: add python to nativeBuildInputs" into main 2024-04-06 05:29:34 +00:00
Qyriad 9bd52cb952 Merge "always re-eval cached failures" into main 2024-04-06 05:20:14 +00:00
jade 9e166f3456 Merge "Create clang-tidy check to rename all our includes" into main 2024-04-06 05:01:21 +00:00
jade 43cf487c25 Create clang-tidy check to rename all our includes
It is a little bit scuffed, but it seems to produce correct results. We
can run it at a later date when we want to explode every in-flight
commit in existence and then need to filter-branch them.

Fixes: lix-project/lix#188

Change-Id: Id97e4651f78804a941d941df02c7c1b21ce453b6
2024-04-06 04:40:19 +00:00
Qyriad a4ef195a9f always re-eval cached failures
This is terrible UX, and frankly an eval failure should be a cache
invalidation anyway.

This removes the CachedEvalError type entirely.

Fixes #223.

Change-Id: I91f8003eabd0ea45003024e96d1de3c7ae8e49d8
2024-04-06 04:35:25 +00:00
Qyriad 11ad383327 package: add python to nativeBuildInputs
Commit c21d11ac0 "docs: replace sed invocation with an mdbook
preprocessor for @docroot@" added a direct build dependency on Python.
This has been accidentally working so far because Python is already a
*transitive* dependency of Lix's derivation.

Change-Id: I32d6b4f2665dbbfad7014613457dd58aa4ec73da
2024-04-05 22:28:27 -06:00
Qyriad b47c159aec Merge "docs: replace sed invocation with an mdbook preprocessor for @docroot@" into main 2024-04-06 01:05:32 +00:00
Qyriad 4e11b0d04d Merge "meson: add missing tests: ca, dyn-drv, plugins, libstoreconsumer" into main 2024-04-06 01:05:09 +00:00
eldritch horrors e9e1b6963c Merge changes I1fa30114,I3ca208b6,Ide4c6e00,I74c46b9f,I05fa6a9d, ... into main
* changes:
  Revert "libutil: drop Pool resources on exceptional free"
  Revert "libutil: remove Pool::Handle::bad"
  Revert "libstore: remove one Resource::good flag"
  Revert "libstore: using throwing finally in withFramedSink"
  Revert "libutil: allow graceful dropping of Pool::Handle"
  Revert "libutil: drop Fs{Source,Sink}::good"
  libutil: guard Finally against invalid exception throws
2024-04-05 23:17:18 +00:00
Qyriad 405e41e288 Merge "meson: run repl characterization test" into main 2024-04-05 20:57:42 +00:00
Rebecca Turner ee423f391d Rewrite REPL test parser
- Use a recursive descent parser so that it's easy to extend.
- Add `@args` to enable customizing command-line arguments
- Add `@should-start` to enable `nix repl` tests that error before
  entering the REPL
- Make sure to read all stdout output before comparing. This catches
  some extra output we were tossing out before!

Change-Id: I5522555df4c313024ab15cd10f9f04e7293bda3a
2024-04-05 13:14:21 -07:00
eldritch horrors 38dc6f5b69 Revert "libutil: drop Pool resources on exceptional free"
This reverts commit de2884b82b376d10de5c400d8e73bc7d98f195d2.

Change-Id: I1fa301149d7c2ed3d266a40c15b2d010e12e44e6
2024-04-05 20:13:02 +00:00
eldritch horrors 52f741c23a Revert "libutil: remove Pool::Handle::bad"
This reverts commit 792844fb861ea7367ac2316c78fec055363f2f9e.

Change-Id: I3ca208b62edfd5cd1199478f75cd2edf19a364f6
2024-04-05 20:13:02 +00:00
eldritch horrors 0b8a17cab6 Revert "libstore: remove one Resource::good flag"
This reverts commit 87249eb579bf57f4f09e9fca100588a4d6b90b4c.

Change-Id: Ide4c6e00c4155216a17e46671ff47151d7bb85b4
2024-04-05 20:13:02 +00:00
eldritch horrors ad30972867 Revert "libstore: using throwing finally in withFramedSink"
This reverts commit 491caad6f62c21ffbcdebe662e63ec0f72e6f3a2.

this is not actually legal for nix! throwing exceptions in destructors
is fine, but the way nix is set up we'll end up throwing the exception
we received from the remote *twice* in some cases, and such cases will
cause an immediate terminate without active exception.

Change-Id: I74c46b9f26fd791086e4193ec60eb1deb9a5bb2a
2024-04-05 20:13:02 +00:00
eldritch horrors c77b6e1fdd Revert "libutil: allow graceful dropping of Pool::Handle"
This reverts commit 8075541d82d05347321d35b9934ccee5f82142f4.

Change-Id: I05fa6a9de1308a4827a6557cf2807eb47ca64da6
2024-04-05 20:13:02 +00:00
eldritch horrors 821ad98beb Revert "libutil: drop Fs{Source,Sink}::good"
This reverts commit 1340807e30dba4b3972c31f02861bbaeaeb60e61.

Change-Id: I34d2a80eb3c3e9d79cb02b92cd1189da32d18cb6
2024-04-05 20:13:02 +00:00
eldritch horrors 6c777476c9 libutil: guard Finally against invalid exception throws
throwing exceptions is fine, but throwing exceptions during exception
handling is hard enough to do correctly that we should just forbid it
entirely out of an overabundance of caution. in cases where terminate
is the correct answer the users of Finally must call it manually now.

Change-Id: Ia51a2cb4a0638500550bfabc89cf01a6d8098983
2024-04-05 20:13:02 +00:00
Rebecca Turner 5081109592 Fix Boost with make build
This prevents the autotools build from discovering Boost in Homebrew
installation directories on macOS.

Change-Id: I624309165c9371c391fd657424ba4c4f3182b385
2024-04-05 11:09:59 -07:00