release: 2.91.2 "Dragon's Breath"

Release produced with releng/create_release.xsh

Change-Id: Iaae75be51d51a553c29720e00f98565d314783bb

.clang-tidy

@@ -16,3 +16,20 @@ Checks:
   - -bugprone-unchecked-optional-access
     # many warnings, seems like a questionable lint
   - -bugprone-branch-clone
+    # all thrown exceptions must derive from std::exception
+  - hicpp-exception-baseclass
+    # capturing async lambdas are dangerous
+  - cppcoreguidelines-avoid-capturing-lambda-coroutines
+    # crimes must be appropriately declared as crimes
+  - cppcoreguidelines-pro-type-cstyle-cast
+  - lix-*
+    # This can not yet be applied to Lix itself since we need to do source
+    # reorganization so that lix/ include paths work.
+  - -lix-fixincludes
+    # This lint is included as an example, but the lib function it replaces is
+    # already gone.
+  - -lix-hasprefixsuffix
+
+
+CheckOptions:
+  bugprone-reserved-identifier.AllowedIdentifiers: '__asan_default_options'

README.md

@@ -26,4 +26,4 @@ See our [Hacking guide](https://git.lix.systems/lix-project/lix/src/branch/main/
 
 ## License
 
-Lix is released under the [LGPL v2.1](./COPYING).
+Lix is released under [LGPL-2.1-or-later](./COPYING).

boehmgc-coroutine-sp-fallback.diff

@@ -1,54 +0,0 @@
-diff --git a/pthread_stop_world.c b/pthread_stop_world.c
-index 2b45489..0e6d8ef 100644
---- a/pthread_stop_world.c
-+++ b/pthread_stop_world.c
-@@ -776,6 +776,8 @@ STATIC void GC_restart_handler(int sig)
- /* world is stopped.  Should not fail if it isn't.                      */
- GC_INNER void GC_push_all_stacks(void)
- {
-+    size_t stack_limit;
-+    pthread_attr_t pattr;
-     GC_bool found_me = FALSE;
-     size_t nthreads = 0;
-     int i;
-@@ -868,6 +870,40 @@ GC_INNER void GC_push_all_stacks(void)
-             hi = p->altstack + p->altstack_size;
- #         endif
-           /* FIXME: Need to scan the normal stack too, but how ? */
-+        } else {
-+          #ifdef HAVE_PTHREAD_ATTR_GET_NP
-+          if (pthread_attr_init(&pattr) != 0) {
-+            ABORT("GC_push_all_stacks: pthread_attr_init failed!");
-+          }
-+          if (pthread_attr_get_np(p->id, &pattr) != 0) {
-+            ABORT("GC_push_all_stacks: pthread_attr_get_np failed!");
-+          }
-+          #else
-+          if (pthread_getattr_np(p->id, &pattr)) {
-+            ABORT("GC_push_all_stacks: pthread_getattr_np failed!");
-+          }
-+          #endif
-+          if (pthread_attr_getstacksize(&pattr, &stack_limit)) {
-+            ABORT("GC_push_all_stacks: pthread_attr_getstacksize failed!");
-+          }
-+          if (pthread_attr_destroy(&pattr)) {
-+            ABORT("GC_push_all_stacks: pthread_attr_destroy failed!");
-+          }
-+          // When a thread goes into a coroutine, we lose its original sp until
-+          // control flow returns to the thread.
-+          // While in the coroutine, the sp points outside the thread stack,
-+          // so we can detect this and push the entire thread stack instead,
-+          // as an approximation.
-+          // We assume that the coroutine has similarly added its entire stack.
-+          // This could be made accurate by cooperating with the application
-+          // via new functions and/or callbacks.
-+          #ifndef STACK_GROWS_UP
-+            if (lo >= hi || lo < hi - stack_limit) { // sp outside stack
-+              lo = hi - stack_limit;
-+            }
-+          #else
-+          #error "STACK_GROWS_UP not supported in boost_coroutine2 (as of june 2021), so we don't support it in Nix."
-+          #endif
-         }
- #       ifdef STACKPTR_CORRECTOR_AVAILABLE
-           if (GC_sp_corrector != 0)

clang-tidy/meson.build

@@ -1,13 +0,0 @@
-project('lix-clang-tidy', ['cpp', 'c'],
-  version : '0.1',
-  default_options : ['warning_level=3', 'cpp_std=c++20'])
-
-llvm = dependency('Clang', version: '>= 14', modules: ['libclang'])
-sources = files(
-  'HasPrefixSuffix.cc',
-  'LixClangTidyChecks.cc',
-  'FixIncludes.cc',
-)
-
-shared_module('lix-clang-tidy', sources,
-  dependencies: llvm)

doc/internal-api/doxygen.cfg.in

@@ -20,7 +20,7 @@ OUTPUT_DIRECTORY       = @docdir@
 # for a project that appears at the top of each page and should give viewer a
 # quick idea about the purpose of the project. Keep the description short.
 
-PROJECT_BRIEF          = "Nix, the purely functional package manager; unstable internal interfaces"
+PROJECT_BRIEF          = "Lix: A modern, delicious implementation of the Nix package manager; unstable internal interfaces"
 
 # If the GENERATE_LATEX tag is set to YES, doxygen will generate LaTeX output.
 # The default value is: YES.

doc/manual/change-authors.yml

@@ -57,6 +57,11 @@ ericson:
   display_name: John Ericson
   github: ericson2314
 
+goldstein:
+  display_name: goldstein
+  forgejo: goldstein
+  github: GoldsteinE
+
 horrors:
   display_name: eldritch horrors
   forgejo: pennae
@@ -65,10 +70,17 @@ horrors:
 iFreilicht:
   github: iFreilicht
 
+isabelroses:
+  forgejo: isabelroses
+  github: isabelroses
+
 jade:
   forgejo: jade
   github: lf-
 
+kjeremy:
+  github: kjeremy
+
 kloenk:
   forgejo: kloenk
   github: kloenk
@@ -91,6 +103,11 @@ midnightveil:
 ncfavier:
   github: ncfavier
 
+piegames:
+  display_name: piegames
+  forgejo: piegames
+  github: piegamesde
+
 puck:
   display_name: puck
   forgejo: puck

doc/manual/redirects.js

@@ -345,7 +345,7 @@ const redirects = {
     "linux": "uninstall.html#linux",
     "macos": "uninstall.html#macos",
     "uninstalling": "uninstall.html",
-  }
+  },
   "contributing/hacking.html": {
     "nix-with-flakes": "#building-nix-with-flakes",
     "classic-nix": "#building-nix",

doc/manual/rl-next/better-attrpath-errors.md

@@ -1,90 +0,0 @@
----
-synopsis: "Trace which part of a `foo.bar.baz` expression errors"
-cls: 1505, 1506
-credits: Qyriad
-category: Improvements
----
-
-Previously, if an attribute path selection expression like `linux_4_9.meta.description` it wouldn't show you which one of those parts in the attribute path, or even that that line of code is what caused evaluation of the failing expression.
-The previous error looks like this:
-
-```
-pkgs.linuxKernel.kernels.linux_4_9.meta.description
-
-error:
-       … while evaluating the attribute 'linuxKernel.kernels.linux_4_9.meta.description'
-         at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:5:
-          277|   } // lib.optionalAttrs config.allowAliases {
-          278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
-             |     ^
-          279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
-
-       … while calling the 'throw' builtin
-         at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:17:
-          277|   } // lib.optionalAttrs config.allowAliases {
-          278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
-             |                 ^
-          279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
-
-       error: linux 4.9 was removed because it will reach its end of life within 22.11
-```
-
-Now, the error will look like this:
-
-```
-pkgs.linuxKernel.kernels.linux_4_9.meta.description
-
-error:
-       … while evaluating the attribute 'linuxKernel.kernels.linux_4_9.meta.description'
-         at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:5:
-          277|   } // lib.optionalAttrs config.allowAliases {
-          278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
-             |     ^
-          279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
-
-       … while evaluating 'pkgs.linuxKernel.kernels.linux_4_9' to select 'meta' on it
-         at «string»:1:1:
-            1| pkgs.linuxKernel.kernels.linux_4_9.meta.description
-             | ^
-
-       … caused by explicit throw
-         at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:17:
-          277|   } // lib.optionalAttrs config.allowAliases {
-          278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
-             |                 ^
-          279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
-
-       error: linux 4.9 was removed because it will reach its end of life within 22.11
-```
-
-Not only does the line of code that referenced the failing attribute show up in the trace, it also tells you that it was specifically the `linux_4_9` part that failed.
-
-This includes if the failing part is a top-level binding:
-
-```
-let
-  inherit (pkgs.linuxKernel.kernels) linux_4_9;
-in linux_4_9.meta.description
-error:
-       … while evaluating 'linux_4_9' to select 'meta.description' on it
-         at «string»:3:4:
-            2|   inherit (pkgs.linuxKernel.kernels) linux_4_9;
-            3| in linux_4_9.meta.description
-             |    ^
-
-       … while evaluating the attribute 'linux_4_9'
-         at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:5:
-          277|   } // lib.optionalAttrs config.allowAliases {
-          278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
-             |     ^
-          279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
-
-       … caused by explicit throw
-         at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:17:
-          277|   } // lib.optionalAttrs config.allowAliases {
-          278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
-             |                 ^
-          279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
-
-       error: linux 4.9 was removed because it will reach its end of life within 22.11
-```

doc/manual/rl-next/build-dir.md

@@ -1,12 +0,0 @@
----
-synopsis: "Add a `build-dir` setting to set the backing directory for builds"
-cls: 1514
-credits: [roberth, tomberek]
-category: Improvements
----
-
-`build-dir` can now be set in the Nix configuration to choose the backing directory for the build sandbox.
-This can be useful on systems with `/tmp` on tmpfs, or simply to relocate large builds to another disk.
-
-Also, `XDG_RUNTIME_DIR` is no longer considered when selecting the default temporary directory,
-as it's not intended to be used for large amounts of data.

doc/manual/rl-next/distinguish-throw-errors.md

@@ -1,70 +0,0 @@
----
-synopsis: "Distinguish between explicit throws and errors that happened while evaluating a throw"
-cls: 1511
-credits: Qyriad
-category: Improvements
----
-
-Previously, errors caused by an expression like `throw "invalid argument"` were treated like an error that happened simply while some builtin function was being called:
-
-```
-let
-  throwMsg = p: throw "${p} isn't the right package";
-in throwMsg "linuz"
-
-error:
-       … while calling the 'throw' builtin
-         at «string»:2:17:
-            1| let
-            2|   throwMsg = p: throw "${p} isn't the right package";
-             |                 ^
-            3| in throwMsg "linuz"
-
-       error: linuz isn't the right package
-```
-
-But the error didn't just happen "while" calling the `throw` builtin — it's a throw error!
-Now it looks like this:
-
-```
-let
-  throwMsg = p: throw "${p} isn't the right package";
-in throwMsg "linuz"
-
-error:
-       … caused by explicit throw
-         at «string»:2:17:
-            1| let
-            2|   throwMsg = p: throw "${p} isn't the right package";
-             |                 ^
-            3| in throwMsg "linuz"
-
-       error: linuz isn't the right package
-```
-
-This also means that incorrect usage of `throw` or errors evaluating its arguments are easily distinguishable from explicit throws:
-
-```
-let
-  throwMsg = p: throw "${p} isn't the right package";
-in throwMsg { attrs = "error when coerced in string interpolation"; }
-
-error:
-       … while calling the 'throw' builtin
-         at «string»:2:17:
-            1| let
-            2|   throwMsg = p: throw "${p} isn't the right package";
-             |                 ^
-            3| in throwMsg { attrs = "error when coerced in string interpolation"; }
-
-       … while evaluating a path segment
-         at «string»:2:24:
-            1| let
-            2|   throwMsg = p: throw "${p} isn't the right package";
-             |                        ^
-            3| in throwMsg { attrs = "error when coerced in string interpolation"; }
-
-       error: cannot coerce a set to a string: { attrs = "error when coerced in string interpolation"; }
-```
-
-Here, instead of an actual thrown error, a type error happens first (trying to coerce an attribute set to a string), but that type error happened *while* calling `throw`.

doc/manual/rl-next/fix-gc-dry-run.md

@@ -1,29 +0,0 @@
----
-synopsis: Fix nix-collect-garbage --dry-run
-issues: [fj#432]
-cls: [1566]
-category: Fixes
-credits: [quantumjump]
----
-
-`nix-collect-garbage --dry-run` did not previously give any output - it simply
-exited without even checking to see what paths would be deleted.
-
-```
-$ nix-collect-garbage --dry-run
-$
-```
-
-We updated the behaviour of the flag such that instead it prints out how many
-paths it *would* delete, but doesn't actually delete them.
-
-```
-$ nix-collect-garbage --dry-run
-finding garbage collector roots...
-determining live/dead paths...
-...
-<nix store paths>
-...
-2670 store paths deleted, 0.00MiB freed
-$
-```

doc/manual/rl-next/fod-failure-includes-url.md

@@ -1,16 +0,0 @@
----
-synopsis: "Hash mismatch diagnostics for fixed-output derivations include the URL"
-cls: [1536]
-credits: [jade]
-category: Improvements
----
-
-Now, when building fixed-output derivations, Lix will guess the URL that was used in the derivation using the `url` or `urls` properties in the derivation environment.
-This is a layering violation but making these diagnostics tractable when there are multiple instances of the `AAAA` hash is too significant of an improvement to pass it up.
-
-```
-error: hash mismatch in fixed-output derivation '/nix/store/sjfw324j4533lwnpmr5z4icpb85r63ai-x1.drv':
-        likely URL: https://meow.puppy.forge/puppy.tar.gz
-         specified: sha256-AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA=
-            got:    sha256-a1Qvp3FOOkWpL9kFHgugU1ok5UtRPSu+NwCZKbbaEro=
-```

doc/manual/rl-next/multiline-log-format.md

@@ -1,14 +0,0 @@
----
-synopsis: Add log formats `multiline` and `multiline-with-logs`
-cls: [1369]
-credits: [kloenk]
-category: Improvements
----
-
-Added two new log formats (`multiline` and `multiline-with-logs`) that display
-current activities below each other for better visibility.
-
-These formats attempt to use the maximum available lines
-(defaulting to 25 if unable to determine) and print up to that many lines.
-The status bar is displayed as the first line, with each subsequent
-activity on its own line.

doc/manual/rl-next/nix-copy-is-fast.md

@@ -1,12 +0,0 @@
----
-synopsis: "`nix copy` is now several times faster at `querying info about /nix/store/...`"
-cls: [1462]
-issues: [fj#366]
-credits: [jade]
-category: Fixes
----
-
-We fixed a locking bug that serialized `querying info about /nix/store/...`
-onto just one thread such that it was eating `O(paths to copy * latency)` time
-while setting up to copy paths to s3 and other stores. It is now `nproc` times
-faster.

doc/manual/rl-next/old-protocol-removal.md

@@ -1,21 +0,0 @@
----
-synopsis: "Lix no longer speaks the Nix remote-build worker protocol to clients or servers older than CppNix 2.3"
-cls: [1207, 1208, 1206, 1205, 1204, 1203, 1479]
-issues: [fj#325]
-credits: [jade]
-category: Breaking Changes
----
-
-CppNix 2.3 was released in 2019, and is the new oldest supported version. We
-will increase our support baseline in the future up to a final version of CppNix
-2.18 (which may happen soon given that it is the only still-packaged and thus
-still-tested >2.3 version), but this step already removes a significant amount
-of dead, untested, code paths.
-
-Lix speaks the same version of the protocol as CppNix 2.18 and that fact will
-never change in the future; the Lix plans to replace the protocol for evolution
-will entail a complete incompatible replacement that will be supported in
-parallel with the old protocol. Lix will thus retain remote build compatibility
-with CppNix as long as CppNix maintains protocol compatibility with 2.18, and
-as long as Lix retains legacy protocol support (which will likely be a long
-time given that we plan to convert it to a frozen-in-time shim).

doc/manual/rl-next/registry-add-shorthand-only.md

@@ -1,10 +0,0 @@
----
-synopsis: "`nix registry add` now requires a shorthand flakeref on the 'from' side"
-cls: 1494
-credits: delan
-category: Improvements
----
-
-The 'from' argument must now be a shorthand flakeref like `nixpkgs` or `nixpkgs/nixos-20.03`, making it harder to accidentally swap the 'from' and 'to' arguments.
-
-Registry entries that map from other flake URLs can still be specified in registry.json, the `nix.registry` option in NixOS, or the `--override-flake` option in the CLI, but they are not guaranteed to work correctly.

doc/manual/rl-next/reject-flake-config.md

@@ -1,9 +0,0 @@
----
-synopsis: Allow automatic rejection of configuration options from flakes
-cls: [1541]
-credits: [alois31]
-category: Improvements
----
-
-Setting `accept-flake-config` to `false` now respects user choice by automatically rejecting configuration options set by flakes.
-The old behaviour of asking each time is still available (and default) by setting it to the special value `ask`.

doc/manual/rl-next/repl-complete-colon.md

@@ -1,8 +0,0 @@
----
-synopsis: "`nix repl` now allows tab-completing the special repl :colon commands"
-cls: 1367
-credits: Qyriad
-category: Improvements
----
-
-The REPL (`nix repl`) supports pressing `<TAB>` to complete a partial expression, but now also supports completing the special :colon commands as well (`:b`, `:edit`, `:doc`, etc), if the line starts with a colon.

doc/manual/rl-next/sanitizers.md

@@ -1,10 +0,0 @@
----
-synopsis: "Lix now supports building with UndefinedBehaviorSanitizer"
-cls: [1483]
-credits: [jade]
-category: Development
----
-
-You can now build Lix with the configuration option `-Db_sanitize=undefined` and it will both work and pass tests. AddressSanitizer support is also coming soon.
-
-For a list of undefined behaviour fixed by sanitizer usage, see [the gerrit topic "undefined-behaviour"](https://gerrit.lix.systems/q/topic:%22undefined-behaviour%22).

doc/manual/src/SUMMARY.md

@@ -196,8 +196,9 @@
   - [C++ style guide](contributing/cxx.md)
 - [Release Notes](release-notes/release-notes.md)
   - [Upcoming release](release-notes/rl-next.md)
-  <!-- RELENG-AUTO-INSERTION-MARKER (see releng/release_notes.py) -->
-  - [Lix 2.90 (FIXME date)](release-notes/rl-2.90.md)
+    <!-- RELENG-AUTO-INSERTION-MARKER (see releng/release_notes.py) -->
+  - [Lix 2.91 (2024-08-12)](release-notes/rl-2.91.md)
+  - [Lix 2.90 (2024-07-10)](release-notes/rl-2.90.md)
   - [Nix 2.18 (2023-09-20)](release-notes/rl-2.18.md)
   - [Nix 2.17 (2023-07-24)](release-notes/rl-2.17.md)
   - [Nix 2.16 (2023-05-31)](release-notes/rl-2.16.md)

doc/manual/src/contributing/hacking.md

@@ -175,6 +175,7 @@ These are specified in `crossSystems` in `flake.nix`; feel free to submit change
 
 - `armv6l-linux`
 - `armv7l-linux`
+- `aarch64-linux`
 - `riscv64-linux`
 
 For example, to cross-compile Lix for `armv6l-linux` from another Linux, use the following:

doc/manual/src/contributing/testing.md

@@ -247,7 +247,6 @@ To ensure that characterization testing doesn't make it harder to intentionally
 
 The integration tests are defined in the Nix flake under the `hydraJobs.tests` attribute.
 These tests include everything that needs to interact with external services or run Lix in a non-trivial distributed setup.
-Because these tests are expensive and require more than what the standard github-actions setup provides, they only run on the master branch (on <https://hydra.nixos.org/jobset/nix/master>).
 
 You can run them manually with `nix build .#hydraJobs.tests.{testName}` or `nix-build -A hydraJobs.tests.{testName}`
 
@@ -428,6 +427,7 @@ I grepped `src/` for `get[eE]nv\("` to find the mentions in Lix code.
 - `NIX_SHOW_STATS_PATH` - Writes those statistics into a file at the given path instead of stdout. Undocumented.
 - `NIX_SHOW_SYMBOLS` - Dumps the symbol table into the show-stats json output.
 - `TERM` - If `dumb` or unset, disables ANSI colour output.
+- `FORCE_COLOR`, `CLICOLOR_FORCE` - Enables ANSI colour output if `NO_COLOR`/`NOCOLOR` not set.
 - `NO_COLOR`, `NOCOLOR` - Disables ANSI colour output.
 - `_NIX_DEVELOPER_SHOW_UNKNOWN_LOCATIONS` - Highlights unknown locations in errors.
 - `NIX_PROFILE` - Selects which profile `nix-env` will operate on. Documented elsewhere.

doc/manual/src/language/advanced-attributes.md

@@ -292,6 +292,12 @@ Derivations can declare some infrequently used optional attributes.
     (associative) arrays. For example, the attribute `hardening.format = true`
     ends up as the Bash associative array element `${hardening[format]}`.
 
+    > **Warning**
+    >
+    > If set to `true`, other advanced attributes such as [`allowedReferences`](#adv-attr-allowedReferences), [`allowedReferences`](#adv-attr-allowedReferences), [`allowedRequisites`](#adv-attr-allowedRequisites),
+    [`disallowedReferences`](#adv-attr-disallowedReferences) and [`disallowedRequisites`](#adv-attr-disallowedRequisites), maxSize, and maxClosureSize.
+    will have no effect.
+
   - [`outputChecks`]{#adv-attr-outputChecks}\
     When using [structured attributes](#adv-attr-structuredAttrs), the `outputChecks`
     attribute allows defining checks per-output.
@@ -326,7 +332,6 @@ Derivations can declare some infrequently used optional attributes.
     ```
 
   - [`unsafeDiscardReferences`]{#adv-attr-unsafeDiscardReferences}\
-
     When using [structured attributes](#adv-attr-structuredAttrs), the
     attribute `unsafeDiscardReferences` is an attribute set with a boolean value for each output name.
     If set to `true`, it disables scanning the output for runtime dependencies.

doc/manual/src/language/operators.md

@@ -26,6 +26,8 @@
 | Logical conjunction (`AND`)            | *bool* `&&` *bool*                         | left          | 12         |
 | Logical disjunction (`OR`)             | *bool* <code>\|\|</code> *bool*            | left          | 13         |
 | [Logical implication]                  | *bool* `->` *bool*                         | none          | 14         |
+| \[Experimental\] [Function piping]     | *expr* |> *func*                           | left          | 15         |
+| \[Experimental\] [Function piping]     | *expr* <| *func*                           | right         | 16         |
 
 [string]: ./values.md#type-string
 [path]: ./values.md#type-path
@@ -59,8 +61,10 @@ The result is a [Boolean] value.
 
 ## Arithmetic
 
-Numbers are type-compatible:
-Pure integer operations will always return integers, whereas any operation involving at least one floating point number return a floating point number.
+Numbers will retain their type unless mixed with other numeric types:
+Pure integer operations will always return integers, whereas any operation involving at least one floating point number returns a floating point number.
+
+Integer overflow (of 64-bit signed integers) and division by zero are defined to throw an error.
 
 See also [Comparison] and [Equality].
 
@@ -143,21 +147,103 @@ All comparison operators are implemented in terms of `<`, and the following equi
 | *a* `>`  *b* |       *b* `<` *a*     |
 | *a* `>=` *b* | `! (` *a* `<` *b* `)` |
 
+Note that the above behaviour violates IEEE 754 for floating point numbers with respect to NaN, for instance.
+This may be fixed in a future major language revision.
+
 [Comparison]: #comparison-operators
 
 ## Equality
 
-- [Attribute sets][attribute set] and [list]s are compared recursively, and therefore are fully evaluated.
-- Comparison of [function]s always returns `false`.
-- Numbers are type-compatible, see [arithmetic] operators.
-- Floating point numbers only differ up to a limited precision.
+The following equality comparison rules are followed in order:
+
+- Comparisons are first, sometimes, performed by identity (pointer value), and whether or not this occurs varies depending on the context in which the comparison is performed; for example, through `builtins.elem`, comparison of lists, or other cases.
+  The exact instances in which this occurs, aside from direct list and attribute set comparisons as discussed below, are too dependent on implementation details to meaningfully document.
+
+  See [note on identity comparison](#identity-comparison) below.
+- Comparisons between a combination of integers and floating point numbers are first converted to floating point then compared as floating point.
+- Comparisons between values of differing types, besides the ones mentioned in the above rule, are unequal.
+- Strings are compared as their string values, disregarding string contexts.
+- Paths are compared as their absolute form (since they are stored as such).
+- [Functions][function] are always considered unequal, including with themselves.
+- The following are compared in the typical manner:
+  - Integers
+  - Floating point numbers have equality comparison per IEEE 754.
+
+    Note that this means that just like in most languages, floating point arithmetic results are not typically equality comparable, and should instead be compared by checking that the absolute difference is less than some error margin.
+  - Booleans
+  - Null
+- [Attribute sets][attribute set] are compared following these rules in order:
+  - If both attribute sets have the same identity (via pointer equality), they are considered equal, regardless of whether the contents have reflexive equality (e.g. even if there are functions contained within).
+
+    See [note on identity comparison](#identity-comparison) below.
+  - If both attribute sets have `type = "derivation"` and have an attribute `outPath` that is equal, they are considered equal.
+
+    This means that two results of `builtins.derivation`, regardless of other things added to their attributes via `//` afterwards (or `passthru` in nixpkgs), will compare equal if they passed the same arguments to `builtins.derivation`.
+  - Otherwise, they are compared element-wise in an unspecified order.
+    Although this order *may* be deterministic in some cases, this is not guaranteed, and correct code must not rely on this ordering behaviour.
+
+    The order determines which elements are evaluated first and thus, if there are throwing values in the attribute set, which of those get evaluated, if any, before the comparison returns an unequal result.
+- Lists are compared following these rules in order:
+  - If both lists have the same identity (via pointer equality), they are considered equal, regardless of whether the contents have reflexive equality (e.g. even if there are functions contained within).
+
+    See [note on identity comparison](#identity-comparison) below.
+  - Otherwise, they are compared element-wise in list order.
 
 [function]: ./constructs.md#functions
 
 [Equality]: #equality
 
+### Identity comparison
+
+In the current revision of the Nix language, values are first compared by identity (pointer equality).
+This means that values that are not reflexively equal (that is, they do not satisfy `a == a`), such as functions, are nonetheless sometimes compared as equal with themselves if they are placed in attribute sets or lists, or are compared through other indirect means.
+
+Whether identity comparison applies to a given usage of the language aside from direct list and attribute set comparison is strongly dependent on implementation details to the point it is not feasible to document the exact instances.
+
+This is rather unfortunate behaviour which is regrettably load-bearing on nixpkgs (such as with the `type` attribute of NixOS options) and cannot be changed for the time being.
+It may be changed in a future major language revision.
+
+Correct code must not rely on this behaviour.
+
+For example:
+
+```
+nix-repl> let f = x: 1; s = { func = f; }; in [ (f == f) (s == s) ]
+[ false true ]
+```
+
 ## Logical implication
 
 Equivalent to `!`*b1* `||` *b2*.
 
 [Logical implication]: #logical-implication
+
+## \[Experimental\] Function piping
+
+*This language feature is still experimental and may change at any time. Enable `--extra-experimental-features pipe-operator` to use it.*
+
+Pipes are a dedicated operator for function application, but with reverse order and a lower binding strength.
+This allows you to chain function calls together in way that is more natural to read and requires less parentheses.
+
+`a |> f b |> g` is equivalent to `g (f b a)`.
+`g <| f b <| a` is equivalent to `g (f b a)`.
+
+Example code snippet:
+
+```nix
+defaultPrefsFile = defaultPrefs
+  |> lib.mapAttrsToList (
+    key: value: ''
+      // ${value.reason}
+      pref("${key}", ${builtins.toJSON value.value});
+    ''
+  )
+  |> lib.concatStringsSep "\n"
+  |> pkgs.writeText "nixos-default-prefs.js";
+```
+
+Note how `mapAttrsToList` is called with two arguments (the lambda and `defaultPrefs`),
+but moving the last argument in front of the rest improves the reading flow.
+This is common for functions with long first argument, including all `map`-like functions.
+
+[Function piping]: #experimental-function-piping

doc/manual/src/language/values.md

@@ -7,13 +7,16 @@
   *Strings* can be written in three ways.
 
   The most common way is to enclose the string between double quotes,
-  e.g., `"foo bar"`. Strings can span multiple lines. The special
-  characters `"` and `\` and the character sequence `${` must be
-  escaped by prefixing them with a backslash (`\`). Newlines, carriage
-  returns and tabs can be written as `\n`, `\r` and `\t`,
-  respectively.
+  e.g., `"foo bar"`. Strings can span multiple lines. The backslash
+  (`\`) can be used to escape characters: newlines, carriage returns
+  and tabs may be written as `\n`, `\r` and `\t` respectively; any
+  other characters can be preceded by a backslash to remove any
+  special meaning they may have, like the special characters `"` and
+  `\` and the character sequence `${`.
 
   You can include the results of other expressions into a string by enclosing them in `${ }`, a feature known as [string interpolation].
+  Due to a parser issue that has since come to be relied upon, the character sequence `$${` is interpreted literally and does not introduce an interpolation.
+  To express a `$` character immediately followed by an interpolation, the former must be escaped.
 
   [string interpolation]: ./string-interpolation.md
 
@@ -43,16 +46,16 @@
   Note that the whitespace and newline following the opening `''` is
   ignored if there is no non-whitespace text on the initial line.
 
-  Indented strings support [string interpolation].
-
   Since `${` and `''` have special meaning in indented strings, you
   need a way to quote them. `$` can be escaped by prefixing it with
   `''` (that is, two single quotes), i.e., `''$`. `''` can be escaped
-  by prefixing it with `'`, i.e., `'''`. `$` removes any special
-  meaning from the following `$`. Linefeed, carriage-return and tab
+  by prefixing it with `'`, i.e., `'''`. Linefeed, carriage-return and tab
   characters can be written as `''\n`, `''\r`, `''\t`, and `''\`
   escapes any other character.
 
+  Indented strings support [string interpolation] using `${ }` the same way regular strings do.
+  `$${` is interpreted literally in indented strings as well, so the `$` character must be escaped if it is to be followed by an interpolation.
+
   Indented strings are primarily useful in that they allow multi-line
   string literals to follow the indentation of the enclosing Nix
   expression, and that less escaping is typically necessary for
@@ -85,6 +88,9 @@
   Numbers, which can be *integers* (like `123`) or *floating point*
   (like `123.43` or `.27e13`).
 
+  Integers in the Nix language are 64-bit signed integers.
+  Integer overflow is defined to throw an error.
+
   See [arithmetic] and [comparison] operators for semantics.
 
   [arithmetic]: ./operators.md#arithmetic

doc/manual/src/release-notes/rl-2.90.md

@@ -1,7 +1,7 @@
-# Lix 2.90 "Vanilla Ice Cream" (FIXME date)
+# Lix 2.90 "Vanilla Ice Cream" (2024-07-10)
 
 
-# Lix 2.90.0 (FIXME date)
+# Lix 2.90.0 (2024-07-10)
 
 ## Breaking Changes
 - Deprecate the online flake registries and vendor the default registry [fj#183](https://git.lix.systems/lix-project/lix/issues/183) [fj#110](https://git.lix.systems/lix-project/lix/issues/110) [fj#116](https://git.lix.systems/lix-project/lix/issues/116) [#8953](https://github.com/NixOS/nix/issues/8953) [#9087](https://github.com/NixOS/nix/issues/9087) [cl/1127](https://gerrit.lix.systems/c/lix/+/1127)

doc/manual/src/release-notes/rl-2.91.md

@@ -0,0 +1,544 @@
+# Lix 2.91 "Dragon's Breath" (2024-08-12)
+# Lix 2.91.2 (2025-06-23)
+
+## Breaking Changes
+- Fixed output derivations can be run using `pasta` network isolation [fj#285](https://git.lix.systems/lix-project/lix/issues/285) [cl/3418](https://gerrit.lix.systems/c/lix/+/3418)
+
+  Fixed output derivations traditionally run in the host network namespace.
+  On Linux this allows such derivations to communicate with other sandboxes
+  or the host using the abstract Unix domains socket namespace; this hasn't
+  been unproblematic in the past and has been used in two distinct exploits
+  to break out of the sandbox. For this reason fixed output derivations can
+  now run in a network namespace (provided by [`pasta`]), restricted to TCP
+  and UDP communication with the rest of the world. When enabled this could
+  be a breaking change and we classify it as such, even though we don't yet
+  enable or require such isolation by default. We may enforce this in later
+  releases of Lix once we have sufficient confidence that breakage is rare.
+
+  [`pasta`]: https://passt.top/
+
+  Many thanks to [eldritch horrors](https://git.lix.systems/pennae) and [puck](https://git.lix.systems/puck) for this.
+
+## Fixes
+- Always clean up scratch paths after derivations failed to build [cl/3420](https://gerrit.lix.systems/c/lix/+/3420)
+
+  Previously, scratch paths created during builds were not always cleaned up if
+  the derivation failed, potentially leaving behind unnecessary temporary files
+  or directories in the Nix store.
+
+  This fix ensures that such paths are consistently removed after a failed build,
+  improving Nix store hygiene, hardening Lix against mis-reuse of failed builds
+  scratch paths.
+
+  Many thanks to [Raito Bezarius](https://git.lix.systems/raito) and [eldritch horrors](https://git.lix.systems/pennae) for this.
+- `build-dir` no longer defaults to `temp-dir` [cl/3419](https://gerrit.lix.systems/c/lix/+/3419)
+
+  The directory in which temporary build directories are created no longer defaults
+  to the value of the `temp-dir` setting to avoid builders making their directories
+  world-accessible. This behavior has been used to escape the build sandbox and can
+  cause build impurities even when not used maliciously. We now default to `builds`
+  in `NIX_STATE_DIR` (which is `/nix/var/nix/builds` in the default configuration).
+
+  Many thanks to [eldritch horrors](https://git.lix.systems/pennae) for this.
+
+
+# Lix 2.91.1 (2024-10-18)
+
+## Fixes
+- `<nix/fetchurl.nix>` now uses TLS verification [#11585](https://github.com/NixOS/nix/pull/11585)
+
+  Previously `<nix/fetchurl.nix>` did not do TLS verification. This was because the Nix sandbox in the past did not have access to TLS certificates, and Nix checks the hash of the fetched file anyway. However, this can expose authentication data from `netrc` and URLs to man-in-the-middle attackers. In addition, Nix now in some cases (such as when using impure derivations) does *not* check the hash. Therefore we have now enabled TLS verification. This means that downloads by `<nix/fetchurl.nix>` will now fail if you're fetching from a HTTPS server that does not have a valid certificate.
+
+  `<nix/fetchurl.nix>` is also known as the builtin derivation builder `builtin:fetchurl`. It's not to be confused with the evaluation-time function `builtins.fetchurl`, which was not affected by this issue.
+
+  Many thanks to [Eelco Dolstra](https://github.com/edolstra) for this.
+
+
+
+
+# Lix 2.91.0 (2024-08-12)
+
+## Breaking Changes
+- Block io_uring in the Linux sandbox [cl/1611](https://gerrit.lix.systems/c/lix/+/1611)
+
+  The io\_uring API has the unfortunate property that it is not possible to selectively decide which operations should be allowed.
+  This, together with the fact that new operations are routinely added, makes it a hazard to the proper function of the sandbox.
+
+  Therefore, any access to io\_uring has been made unavailable inside the sandbox.
+  As such, attempts to execute any system calls forming part of this API will fail with the error `ENOSYS`, as if io\_uring support had not been configured into the kernel.
+
+  Many thanks to [alois31](https://git.lix.systems/alois31) for this.
+- The `build-hook` setting is now deprecated
+
+  Build hooks communicate with the daemon using a custom, internal, undocumented protocol that is entirely unversioned and cannot be changed.
+  Since we intend to change it anyway we must unfortunately deprecate the current build hook infrastructure.
+  We do not expect this to impact most users—we have not found any uses of `build-hook` in the wild—but if this does affect you, we'd like to hear from you!
+- Lix no longer speaks the Nix remote-build worker protocol to clients or servers older than CppNix 2.3 [fj#325](https://git.lix.systems/lix-project/lix/issues/325) [cl/1207](https://gerrit.lix.systems/c/lix/+/1207) [cl/1208](https://gerrit.lix.systems/c/lix/+/1208) [cl/1206](https://gerrit.lix.systems/c/lix/+/1206) [cl/1205](https://gerrit.lix.systems/c/lix/+/1205) [cl/1204](https://gerrit.lix.systems/c/lix/+/1204) [cl/1203](https://gerrit.lix.systems/c/lix/+/1203) [cl/1479](https://gerrit.lix.systems/c/lix/+/1479)
+
+  CppNix 2.3 was released in 2019, and is the new oldest supported version. We
+  will increase our support baseline in the future up to a final version of CppNix
+  2.18 (which may happen soon given that it is the only still-packaged and thus
+  still-tested >2.3 version), but this step already removes a significant amount
+  of dead, untested, code paths.
+
+  Lix speaks the same version of the protocol as CppNix 2.18 and that fact will
+  never change in the future; the Lix plans to replace the protocol for evolution
+  will entail a complete incompatible replacement that will be supported in
+  parallel with the old protocol. Lix will thus retain remote build compatibility
+  with CppNix as long as CppNix maintains protocol compatibility with 2.18, and
+  as long as Lix retains legacy protocol support (which will likely be a long
+  time given that we plan to convert it to a frozen-in-time shim).
+
+  Many thanks to [jade](https://git.lix.systems/jade) for this.
+
+## Features
+- Pipe operator `|>` (experimental) [fj#438](https://git.lix.systems/lix-project/lix/issues/438) [cl/1654](https://gerrit.lix.systems/c/lix/+/1654)
+
+  Implementation of the pipe operator (`|>`) in the language as described in [RFC 148](https://github.com/NixOS/rfcs/pull/148).
+  The feature is still marked experimental, enable `--extra-experimental-features pipe-operator` to use it.
+
+  Many thanks to [piegames](https://git.lix.systems/piegames) and [eldritch horrors](https://git.lix.systems/pennae) for this.
+
+## Improvements
+- Trace which part of a `foo.bar.baz` expression errors [cl/1505](https://gerrit.lix.systems/c/lix/+/1505) [cl/1506](https://gerrit.lix.systems/c/lix/+/1506)
+
+  Previously, if an attribute path selection expression like `linux_4_9.meta.description` it wouldn't show you which one of those parts in the attribute path, or even that that line of code is what caused evaluation of the failing expression.
+  The previous error looks like this:
+
+  ```
+  pkgs.linuxKernel.kernels.linux_4_9.meta.description
+
+  error:
+         … while evaluating the attribute 'linuxKernel.kernels.linux_4_9.meta.description'
+           at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:5:
+            277|   } // lib.optionalAttrs config.allowAliases {
+            278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
+               |     ^
+            279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
+
+         … while calling the 'throw' builtin
+           at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:17:
+            277|   } // lib.optionalAttrs config.allowAliases {
+            278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
+               |                 ^
+            279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
+
+         error: linux 4.9 was removed because it will reach its end of life within 22.11
+  ```
+
+  Now, the error will look like this:
+
+  ```
+  pkgs.linuxKernel.kernels.linux_4_9.meta.description
+
+  error:
+         … while evaluating the attribute 'linuxKernel.kernels.linux_4_9.meta.description'
+           at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:5:
+            277|   } // lib.optionalAttrs config.allowAliases {
+            278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
+               |     ^
+            279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
+
+         … while evaluating 'pkgs.linuxKernel.kernels.linux_4_9' to select 'meta' on it
+           at «string»:1:1:
+              1| pkgs.linuxKernel.kernels.linux_4_9.meta.description
+               | ^
+
+         … caused by explicit throw
+           at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:17:
+            277|   } // lib.optionalAttrs config.allowAliases {
+            278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
+               |                 ^
+            279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
+
+         error: linux 4.9 was removed because it will reach its end of life within 22.11
+  ```
+
+  Not only does the line of code that referenced the failing attribute show up in the trace, it also tells you that it was specifically the `linux_4_9` part that failed.
+
+  This includes if the failing part is a top-level binding:
+
+  ```
+  let
+    inherit (pkgs.linuxKernel.kernels) linux_4_9;
+  in linux_4_9.meta.description
+  error:
+         … while evaluating 'linux_4_9' to select 'meta.description' on it
+           at «string»:3:4:
+              2|   inherit (pkgs.linuxKernel.kernels) linux_4_9;
+              3| in linux_4_9.meta.description
+               |    ^
+
+         … while evaluating the attribute 'linux_4_9'
+           at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:5:
+            277|   } // lib.optionalAttrs config.allowAliases {
+            278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
+               |     ^
+            279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
+
+         … caused by explicit throw
+           at /nix/store/dk2rpyb6ndvfbf19bkb2plcz5y3k8i5v-source/pkgs/top-level/linux-kernels.nix:278:17:
+            277|   } // lib.optionalAttrs config.allowAliases {
+            278|     linux_4_9 = throw "linux 4.9 was removed because it will reach its end of life within 22.11";
+               |                 ^
+            279|     linux_4_14 = throw "linux 4.14 was removed because it will reach its end of life within 23.11";
+
+         error: linux 4.9 was removed because it will reach its end of life within 22.11
+  ```
+
+  Many thanks to [Qyriad](https://git.lix.systems/Qyriad) for this.
+- Confusing 'invalid path' errors are now 'path does not exist' [cl/1161](https://gerrit.lix.systems/c/lix/+/1161) [cl/1160](https://gerrit.lix.systems/c/lix/+/1160) [cl/1159](https://gerrit.lix.systems/c/lix/+/1159)
+
+  Previously, if a path did not exist in a Nix store, it was referred to as the internal name "path is invalid".
+  This is, however, very confusing, and there were numerous such errors that were exactly the same, making it hard to debug.
+  These errors are now more specific and refer to the path not existing in the store.
+
+  Many thanks to [julia](https://git.lix.systems/midnightveil) for this.
+- Add a `build-dir` setting to set the backing directory for builds [gh#10303](https://github.com/NixOS/nix/pull/10303) [gh#10312](https://github.com/NixOS/nix/pull/10312) [gh#10883](https://github.com/NixOS/nix/pull/10883) [cl/1514](https://gerrit.lix.systems/c/lix/+/1514)
+
+  `build-dir` can now be set in the Nix configuration to choose the backing directory for the build sandbox.
+  This can be useful on systems with `/tmp` on tmpfs, or simply to relocate large builds to another disk.
+
+  Also, `XDG_RUNTIME_DIR` is no longer considered when selecting the default temporary directory,
+  as it's not intended to be used for large amounts of data.
+
+  Many thanks to [Robert Hensing](https://github.com/roberth) and [Tom Bereknyei](https://github.com/tomberek) for this.
+- Better usage of colour control environment variables [cl/1699](https://gerrit.lix.systems/c/lix/+/1699) [cl/1702](https://gerrit.lix.systems/c/lix/+/1702)
+
+  Lix now heeds `NO_COLOR`/`NOCOLOR` for more output types, such as that used in `nix search`, `nix flake metadata` and similar.
+
+  It also now supports `CLICOLOR_FORCE`/`FORCE_COLOR` to force colours regardless of whether there is a terminal on the other side.
+
+  It now follows rules compatible with those described on <https://bixense.com/clicolors/> with `CLICOLOR` defaulted to enabled.
+
+  That is to say, the following procedure is followed in order:
+  - NO_COLOR or NOCOLOR set
+
+    Always disable colour
+  - CLICOLOR_FORCE or FORCE_COLOR set
+
+    Enable colour
+  - The output is a tty; TERM != "dumb"
+
+    Enable colour
+  - Otherwise
+
+    Disable colour
+
+  Many thanks to [jade](https://git.lix.systems/jade) for this.
+- Distinguish between explicit throws and errors that happened while evaluating a throw [cl/1511](https://gerrit.lix.systems/c/lix/+/1511)
+
+  Previously, errors caused by an expression like `throw "invalid argument"` were treated like an error that happened simply while some builtin function was being called:
+
+  ```
+  let
+    throwMsg = p: throw "${p} isn't the right package";
+  in throwMsg "linuz"
+
+  error:
+         … while calling the 'throw' builtin
+           at «string»:2:17:
+              1| let
+              2|   throwMsg = p: throw "${p} isn't the right package";
+               |                 ^
+              3| in throwMsg "linuz"
+
+         error: linuz isn't the right package
+  ```
+
+  But the error didn't just happen "while" calling the `throw` builtin — it's a throw error!
+  Now it looks like this:
+
+  ```
+  let
+    throwMsg = p: throw "${p} isn't the right package";
+  in throwMsg "linuz"
+
+  error:
+         … caused by explicit throw
+           at «string»:2:17:
+              1| let
+              2|   throwMsg = p: throw "${p} isn't the right package";
+               |                 ^
+              3| in throwMsg "linuz"
+
+         error: linuz isn't the right package
+  ```
+
+  This also means that incorrect usage of `throw` or errors evaluating its arguments are easily distinguishable from explicit throws:
+
+  ```
+  let
+    throwMsg = p: throw "${p} isn't the right package";
+  in throwMsg { attrs = "error when coerced in string interpolation"; }
+
+  error:
+         … while calling the 'throw' builtin
+           at «string»:2:17:
+              1| let
+              2|   throwMsg = p: throw "${p} isn't the right package";
+               |                 ^
+              3| in throwMsg { attrs = "error when coerced in string interpolation"; }
+
+         … while evaluating a path segment
+           at «string»:2:24:
+              1| let
+              2|   throwMsg = p: throw "${p} isn't the right package";
+               |                        ^
+              3| in throwMsg { attrs = "error when coerced in string interpolation"; }
+
+         error: cannot coerce a set to a string: { attrs = "error when coerced in string interpolation"; }
+  ```
+
+  Here, instead of an actual thrown error, a type error happens first (trying to coerce an attribute set to a string), but that type error happened *while* calling `throw`.
+
+  Many thanks to [Qyriad](https://git.lix.systems/Qyriad) for this.
+- `nix flake metadata` prints modified date [cl/1700](https://gerrit.lix.systems/c/lix/+/1700)
+
+  Ever wonder "gee, when *did* I update nixpkgs"?
+  Wonder no more, because `nix flake metadata` now simply tells you the times every locked flake input was updated:
+
+  ```
+  <...>
+  Description:   The purely functional package manager
+  Path:          /nix/store/c91yi8sxakc2ry7y4ac1smzwka4l5p78-source
+  Revision:      c52cff582043838bbe29768e7da232483d52b61d-dirty
+  Last modified: 2024-07-31 22:15:54
+  Inputs:
+  ├───flake-compat: github:edolstra/flake-compat/0f9255e01c2351cc7d116c072cb317785dd33b33
+  │   Last modified: 2023-10-04 06:37:54
+  ├───nix2container: github:nlewo/nix2container/3853e5caf9ad24103b13aa6e0e8bcebb47649fe4
+  │   Last modified: 2024-07-10 13:15:56
+  ├───nixpkgs: github:NixOS/nixpkgs/e21630230c77140bc6478a21cd71e8bb73706fce
+  │   Last modified: 2024-07-25 11:26:27
+  ├───nixpkgs-regression: github:NixOS/nixpkgs/215d4d0fd80ca5163643b03a33fde804a29cc1e2
+  │   Last modified: 2022-01-24 11:20:45
+  └───pre-commit-hooks: github:cachix/git-hooks.nix/f451c19376071a90d8c58ab1a953c6e9840527fd
+      Last modified: 2024-07-15 04:21:09
+  ```
+
+  Many thanks to [jade](https://git.lix.systems/jade) for this.
+- Hash mismatch diagnostics for fixed-output derivations include the URL [cl/1536](https://gerrit.lix.systems/c/lix/+/1536)
+
+  Now, when building fixed-output derivations, Lix will guess the URL that was used in the derivation using the `url` or `urls` properties in the derivation environment.
+  This is a layering violation but making these diagnostics tractable when there are multiple instances of the `AAAA` hash is too significant of an improvement to pass it up.
+
+  ```
+  error: hash mismatch in fixed-output derivation '/nix/store/sjfw324j4533lwnpmr5z4icpb85r63ai-x1.drv':
+          likely URL: https://meow.puppy.forge/puppy.tar.gz
+           specified: sha256-AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA=
+              got:    sha256-a1Qvp3FOOkWpL9kFHgugU1ok5UtRPSu+NwCZKbbaEro=
+  ```
+
+  Many thanks to [jade](https://git.lix.systems/jade) for this.
+- Add log formats `multiline` and `multiline-with-logs` [cl/1369](https://gerrit.lix.systems/c/lix/+/1369)
+
+  Added two new log formats (`multiline` and `multiline-with-logs`) that display
+  current activities below each other for better visibility.
+
+  These formats attempt to use the maximum available lines
+  (defaulting to 25 if unable to determine) and print up to that many lines.
+  The status bar is displayed as the first line, with each subsequent
+  activity on its own line.
+
+  Many thanks to [kloenk](https://git.lix.systems/kloenk) for this.
+- Lix will now show the package descriptions in when running `nix flake show`. [cl/1540](https://gerrit.lix.systems/c/lix/+/1540)
+
+  When running `nix flake show`, Lix will now show the package descriptions, if they exist.
+
+  Before:
+
+  ```shell
+  $ nix flake show
+  path:/home/isabel/dev/lix-show?lastModified=1721736108&narHash=sha256-Zo8HP1ur7Q2b39hKUEG8EAh/opgq8xJ2jvwQ/htwO4Q%3D
+  └───packages
+      └───x86_64-linux
+          ├───aNoDescription: package 'simple'
+          ├───bOneLineDescription: package 'simple'
+          ├───cMultiLineDescription: package 'simple'
+          ├───dLongDescription: package 'simple'
+          └───eEmptyDescription: package 'simple'
+  ```
+
+  After:
+
+  ```shell
+  $ nix flake show
+  path:/home/isabel/dev/lix-show?lastModified=1721736108&narHash=sha256-Zo8HP1ur7Q2b39hKUEG8EAh/opgq8xJ2jvwQ/htwO4Q%3D
+  └───packages
+      └───x86_64-linux
+          ├───aNoDescription: package 'simple'
+          ├───bOneLineDescription: package 'simple' - 'one line'
+          ├───cMultiLineDescription: package 'simple' - 'line one'
+          ├───dLongDescription: package 'simple' - 'abcdefghijklmnopqrstuvwxyz'
+          └───eEmptyDescription: package 'simple'
+  ```
+
+  Many thanks to [kjeremy](https://github.com/kjeremy) and [isabelroses](https://git.lix.systems/isabelroses) for this.
+- Eliminate some pretty-printing surprises [#11100](https://github.com/NixOS/nix/pull/11100) [cl/1616](https://gerrit.lix.systems/c/lix/+/1616) [cl/1617](https://gerrit.lix.systems/c/lix/+/1617) [cl/1618](https://gerrit.lix.systems/c/lix/+/1618)
+
+  Some inconsistent and surprising behaviours have been eliminated from the pretty-printing used by the REPL and `nix eval`:
+  * Lists and attribute sets that contain only a single item without nested structures are no longer sometimes inappropriately indented in the REPL, depending on internal state of the evaluator.
+  * Empty attribute sets and derivations are no longer shown as `«repeated»`, since they are always cheap to print.
+    This matches the existing behaviour of `nix-instantiate` on empty attribute sets.
+    Empty lists were never printed as `«repeated»` already.
+  * The REPL by default does not print nested attribute sets and lists, and indicates elided items with an ellipsis.
+    Previously, the ellipsis was printed even when the structure was empty, so that such items do not in fact exist.
+    Since this behaviour was confusing, it does not happen any more.
+
+  Before:
+  ```
+  nix-repl> :p let x = 1 + 2; in [ [ x ] [ x ] ]
+  [
+    [
+      3
+    ]
+    [ 3 ]
+  ]
+
+  nix-repl> let inherit (import <nixpkgs> { }) hello; in [ hello hello ]
+  [
+    «derivation /nix/store/fqs92lzychkm6p37j7fnj4d65nq9fzla-hello-2.12.1.drv»
+    «repeated»
+  ]
+
+  nix-repl> let x = {}; in [ x ]
+  [
+    { ... }
+  ]
+  ```
+
+  After:
+  ```
+  nix-repl> :p let x = 1 + 2; in [ [ x ] [ x ] ]
+  [
+    [ 3 ]
+    [ 3 ]
+  ]
+
+  nix-repl> let inherit (import <nixpkgs> { }) hello; in [ hello hello ]
+  [
+    «derivation /nix/store/fqs92lzychkm6p37j7fnj4d65nq9fzla-hello-2.12.1.drv»
+    «derivation /nix/store/fqs92lzychkm6p37j7fnj4d65nq9fzla-hello-2.12.1.drv»
+  ]
+
+  nix-repl> let x = {}; in [ x ]
+  [
+    { }
+  ]
+  ```
+
+  Many thanks to [alois31](https://git.lix.systems/alois31) and [Robert Hensing](https://github.com/roberth) for this.
+- `nix registry add` now requires a shorthand flakeref on the 'from' side [cl/1494](https://gerrit.lix.systems/c/lix/+/1494)
+
+  The 'from' argument must now be a shorthand flakeref like `nixpkgs` or `nixpkgs/nixos-20.03`, making it harder to accidentally swap the 'from' and 'to' arguments.
+
+  Registry entries that map from other flake URLs can still be specified in registry.json, the `nix.registry` option in NixOS, or the `--override-flake` option in the CLI, but they are not guaranteed to work correctly.
+
+  Many thanks to [delan](https://git.lix.systems/delan) for this.
+- Allow automatic rejection of configuration options from flakes [cl/1541](https://gerrit.lix.systems/c/lix/+/1541)
+
+  Setting `accept-flake-config` to `false` now respects user choice by automatically rejecting configuration options set by flakes.
+  The old behaviour of asking each time is still available (and default) by setting it to the special value `ask`.
+
+  Many thanks to [alois31](https://git.lix.systems/alois31) for this.
+- `nix repl` now allows tab-completing the special repl :colon commands [cl/1367](https://gerrit.lix.systems/c/lix/+/1367)
+
+  The REPL (`nix repl`) supports pressing `<TAB>` to complete a partial expression, but now also supports completing the special :colon commands as well (`:b`, `:edit`, `:doc`, etc), if the line starts with a colon.
+
+  Many thanks to [Qyriad](https://git.lix.systems/Qyriad) for this.
+- `:edit`ing a file in Nix store no longer reloads the repl [fj#341](https://git.lix.systems/lix-project/lix/issues/341) [cl/1620](https://gerrit.lix.systems/c/lix/+/1620)
+
+  Calling `:edit` from the repl now only reloads if the file being edited was outside of Nix store.
+  That means that all the local variables are now preserved across `:edit`s of store paths.
+  This is always safe because the store is read-only.
+
+  Many thanks to [goldstein](https://git.lix.systems/goldstein) for this.
+- `:log` in repl now works on derivation paths [fj#51](https://git.lix.systems/lix-project/lix/issues/51) [cl/1716](https://gerrit.lix.systems/c/lix/+/1716)
+
+  `:log` can now accept store derivation paths in addition to derivation expressions.
+
+  Many thanks to [goldstein](https://git.lix.systems/goldstein) for this.
+
+## Fixes
+- Define integer overflow in the Nix language as an error [fj#423](https://git.lix.systems/lix-project/lix/issues/423) [cl/1594](https://gerrit.lix.systems/c/lix/+/1594) [cl/1595](https://gerrit.lix.systems/c/lix/+/1595) [cl/1597](https://gerrit.lix.systems/c/lix/+/1597) [cl/1609](https://gerrit.lix.systems/c/lix/+/1609)
+
+  Previously, integer overflow in the Nix language invoked C++ level signed overflow, which was undefined behaviour, but *probably* manifested as wrapping around on overflow.
+
+  Since prior to the public release of Lix, Lix had C++ signed overflow defined to crash the process and nobody noticed this having accidentally removed overflow from the Nix language for three months until it was caught by fiddling around.
+  Given the significant body of actual Nix code that has been evaluated by Lix in that time, it does not appear that nixpkgs or much of importance depends on integer overflow, so it is safe to turn into an error.
+
+  Some other overflows were fixed:
+  - `builtins.fromJSON` of values greater than the maximum representable value in a signed 64-bit integer will generate an error.
+  - `nixConfig` in flakes will no longer accept negative values for configuration options.
+
+  Integer overflow now looks like the following:
+
+  ```
+  » nix eval --expr '9223372036854775807 + 1'
+  error: integer overflow in adding 9223372036854775807 + 1
+  ```
+
+  Many thanks to [jade](https://git.lix.systems/jade) for this.
+- Fix nix-collect-garbage --dry-run [fj#432](https://git.lix.systems/lix-project/lix/issues/432) [cl/1566](https://gerrit.lix.systems/c/lix/+/1566)
+
+  `nix-collect-garbage --dry-run` did not previously give any output - it simply
+  exited without even checking to see what paths would be deleted.
+
+  ```
+  $ nix-collect-garbage --dry-run
+  $
+  ```
+
+  We updated the behaviour of the flag such that instead it prints out how many
+  paths it *would* delete, but doesn't actually delete them.
+
+  ```
+  $ nix-collect-garbage --dry-run
+  finding garbage collector roots...
+  determining live/dead paths...
+  ...
+  <nix store paths>
+  ...
+  2670 store paths deleted, 0.00MiB freed
+  $
+  ```
+
+  Many thanks to [Quantum Jump](https://github.com/QuantumBJump) for this.
+- Fix unexpectedly-successful GC failures on macOS [fj#446](https://git.lix.systems/lix-project/lix/issues/446) [cl/1723](https://gerrit.lix.systems/c/lix/+/1723)
+
+  Has the following happened to you on macOS? This failure has been successfully eliminated, thanks to our successful deployment of advanced successful-failure detection technology (it's just `if (failed && errno == 0)`. Patent pending<sup>not really</sup>):
+
+  ```
+  $ nix-store --gc --print-dead
+  finding garbage collector roots...
+  error: Listing pid 87261 file descriptors: Undefined error: 0
+  ```
+
+  Many thanks to [jade](https://git.lix.systems/jade) for this.
+- `nix copy` is now several times faster at `querying info about /nix/store/...` [fj#366](https://git.lix.systems/lix-project/lix/issues/366) [cl/1462](https://gerrit.lix.systems/c/lix/+/1462)
+
+  We fixed a locking bug that serialized `querying info about /nix/store/...`
+  onto just one thread such that it was eating `O(paths to copy * latency)` time
+  while setting up to copy paths to s3 and other stores. It is now `nproc` times
+  faster.
+
+  Many thanks to [jade](https://git.lix.systems/jade) for this.
+
+## Development
+- clang-tidy support [fj#147](https://git.lix.systems/lix-project/lix/issues/147) [cl/1697](https://gerrit.lix.systems/c/lix/+/1697)
+
+  `clang-tidy` can be used to lint Lix with a limited set of lints using `ninja -C build clang-tidy` and `ninja -C build clang-tidy-fix`.
+  In practice, this fixes the built-in meson rule that was used the same as above being broken ever since precompiled headers were introduced.
+
+  Many thanks to [jade](https://git.lix.systems/jade) for this.
+- Lix now supports building with UndefinedBehaviorSanitizer [cl/1483](https://gerrit.lix.systems/c/lix/+/1483) [cl/1481](https://gerrit.lix.systems/c/lix/+/1481) [cl/1669](https://gerrit.lix.systems/c/lix/+/1669)
+
+  You can now build Lix with the configuration option `-Db_sanitize=undefined,address` and it will both work and pass tests with both AddressSanitizer and UndefinedBehaviorSanitizer enabled.
+  To use ASan specifically, you have to set `-Dgc=disabled`, which an error message will tell you to do if necessary anyhow.
+
+  Furthermore, tests passing with Clang ASan+UBSan is checked on every change in CI.
+
+  For a list of undefined behaviour found by tooling usage, see [the gerrit topic "undefined-behaviour"](https://gerrit.lix.systems/q/topic:%22undefined-behaviour%22).
+
+  Many thanks to [jade](https://git.lix.systems/jade) for this.

flake.lock

@@ -19,11 +19,11 @@
     "nix2container": {
       "flake": false,
       "locked": {
-        "lastModified": 1712990762,
-        "narHash": "sha256-hO9W3w7NcnYeX8u8cleHiSpK2YJo7ecarFTUlbybl7k=",
+        "lastModified": 1749158376,
+        "narHash": "sha256-uirStFNxauh0lxzBowcp28X+Sq7JgsBIDnbwbAfZwf8=",
         "owner": "nlewo",
         "repo": "nix2container",
-        "rev": "20aad300c925639d5d6cbe30013c8357ce9f2a2e",
+        "rev": "0f8974c58755dba441df03598eefd1e1cd50e341",
         "type": "github"
       },
       "original": {
@@ -34,11 +34,11 @@
     },
     "nixpkgs": {
       "locked": {
-        "lastModified": 1718111384,
-        "narHash": "sha256-7tSst0S5FOmcgvNtfy6cjZX5w8CabCVAfAeCkhY4OVg=",
+        "lastModified": 1735651292,
+        "narHash": "sha256-YLbzcBtYo1/FEzFsB3AnM16qFc6fWPMIoOuSoDwvg9g=",
         "owner": "NixOS",
         "repo": "nixpkgs",
-        "rev": "a508a44af0c1b1b57785c34d8b54783536273eeb",
+        "rev": "0da3c44a9460a26d2025ec3ed2ec60a895eb1114",
         "type": "github"
       },
       "original": {
@@ -67,11 +67,11 @@
     "pre-commit-hooks": {
       "flake": false,
       "locked": {
-        "lastModified": 1712055707,
-        "narHash": "sha256-4XLvuSIDZJGS17xEwSrNuJLL7UjDYKGJSbK1WWX2AK8=",
+        "lastModified": 1721042469,
+        "narHash": "sha256-6FPUl7HVtvRHCCBQne7Ylp4p+dpP3P/OYuzjztZ4s70=",
         "owner": "cachix",
         "repo": "git-hooks.nix",
-        "rev": "e35aed5fda3cc79f88ed7f1795021e559582093a",
+        "rev": "f451c19376071a90d8c58ab1a953c6e9840527fd",
         "type": "github"
       },
       "original": {

flake.nix

@@ -1,5 +1,5 @@
 {
-  description = "The purely functional package manager";
+  description = "Lix: A modern, delicious implementation of the Nix package manager";
 
   inputs = {
     nixpkgs.url = "github:NixOS/nixpkgs/nixos-24.05-small";
@@ -33,7 +33,7 @@
 
       # This notice gets echoed as a dev shell hook, and can be turned off with
       # `touch .nocontribmsg`
-      sgr = ''['';
+      sgr = builtins.fromJSON ''"\u001b["'';
       freezePage = "https://wiki.lix.systems/books/lix-contributors/page/freezes-and-recommended-contributions";
       codebaseOverview = "https://wiki.lix.systems/books/lix-contributors/page/codebase-overview";
       contribNotice = builtins.toFile "lix-contrib-notice" ''
@@ -59,7 +59,8 @@
         (Run `touch .nocontribmsg` to hide this message.)
       '';
 
-      officialRelease = false;
+      versionJson = builtins.fromJSON (builtins.readFile ./version.json);
+      officialRelease = versionJson.official_release;
 
       # Set to true to build the release notes for the next release.
       buildUnreleasedNotes = true;
@@ -90,6 +91,7 @@
         "armv6l-linux"
         "armv7l-linux"
         "riscv64-linux"
+        "aarch64-linux"
         # FIXME: still broken in 24.05: fails to build rustc(??) due to missing -lstdc++ dep
         # "x86_64-freebsd"
         # FIXME: broken dev shell due to python
@@ -139,10 +141,7 @@
                     system = crossSystem;
                   }
                   // lib.optionalAttrs (crossSystem == "x86_64-freebsd") { useLLVM = true; };
-              overlays = [
-                (overlayFor (p: p.${stdenv}))
-                (final: prev: { nixfmt = final.callPackage ./nix-support/nixfmt.nix { }; })
-              ];
+              overlays = [ (overlayFor (p: p.${stdenv})) ];
             };
           stdenvs = forAllStdenvs (make-pkgs null);
           native = stdenvs.stdenvPackages;
@@ -166,6 +165,7 @@
           nixUnstable = prev.nixUnstable;
 
           check-headers = final.buildPackages.callPackage ./maintainers/check-headers.nix { };
+          check-syscalls = final.buildPackages.callPackage ./maintainers/check-syscalls.nix { };
 
           default-busybox-sandbox-shell = final.busybox.override {
             useMusl = true;
@@ -197,7 +197,7 @@
             busybox-sandbox-shell = final.busybox-sandbox-shell or final.default-busybox-sandbox-shell;
           };
 
-          pegtl = final.nix.passthru.pegtl;
+          lix-clang-tidy = final.callPackage ./subprojects/lix-clang-tidy { };
 
           # Export the patched version of boehmgc that Lix uses into the overlay
           # for consumers of this flake.
@@ -278,6 +278,52 @@
 
         # System tests.
         tests = import ./tests/nixos { inherit lib nixpkgs nixpkgsFor; } // {
+          # This is x86_64-linux only, just because we have significantly
+          # cheaper x86_64-linux compute in CI.
+          # It is clangStdenv because clang's sanitizers are nicer.
+          asanBuild = self.packages.x86_64-linux.nix-clangStdenv.override {
+            # Improve caching of non-code changes by not changing the
+            # derivation name every single time, since this will never be seen
+            # by users anyway.
+            versionSuffix = "";
+            sanitize = [
+              "address"
+              "undefined"
+            ];
+            # it is very hard to make *every* CI build use this option such
+            # that we don't wind up building Lix twice, so we do it here where
+            # we are already doing so.
+            werror = true;
+          };
+
+          # Although this might be nicer to do with pre-commit, that would
+          # require adding 12MB of nodejs to the dev shell, whereas building it
+          # in CI with Nix avoids that at a cost of slower feedback on rarely
+          # touched files.
+          jsSyntaxCheck =
+            let
+              nixpkgs = nixpkgsFor.x86_64-linux.native;
+              inherit (nixpkgs) pkgs;
+              docSources = lib.fileset.toSource {
+                root = ./doc;
+                fileset = lib.fileset.fileFilter (f: f.hasExt "js") ./doc;
+              };
+            in
+            pkgs.runCommand "js-syntax-check" { } ''
+              find ${docSources} -type f -print -exec ${pkgs.nodejs-slim}/bin/node --check '{}' ';'
+              touch $out
+            '';
+
+          # clang-tidy run against the Lix codebase using the Lix clang-tidy plugin
+          clang-tidy =
+            let
+              nixpkgs = nixpkgsFor.x86_64-linux.native;
+              inherit (nixpkgs) pkgs;
+            in
+            pkgs.callPackage ./package.nix {
+              versionSuffix = "";
+              lintInsteadOfBuild = true;
+            };
 
           # Make sure that nix-env still produces the exact same result
           # on a particular version of Nixpkgs.
@@ -328,6 +374,25 @@
         pkgs = nixpkgsFor.x86_64-linux.native;
       };
 
+      releaseTests = lib.foldl lib.recursiveUpdate { } [
+        (lib.genAttrs (linux64BitSystems ++ darwinSystems) (system: {
+          nativeBuild = self.packages.${system}.nix;
+        }))
+        (lib.genAttrs (linux64BitSystems) (system: {
+          staticBuild = self.packages.${system}.nix-static;
+        }))
+        {
+          x86_64-linux = {
+            # TODO add more cross/static release targets?
+            crossBuild.aarch64-linux = self.packages.x86_64-linux.nix-aarch64-linux;
+
+            # TODO wire up a nixos installer test with that lix and
+            # run it, once nixpkgs can actually do that (again). :/
+            # # nix build .#nixosTests.installer.{btrfsSimple,luksroot,lvm,simple,switchToFlake}
+          };
+        }
+      ];
+
       # NOTE *do not* add fresh derivations to checks, always add them to
       # hydraJobs first (so CI will pick them up) and only link them here
       checks = forAvailableSystems (
@@ -354,6 +419,8 @@
         rec {
           inherit (nixpkgsFor.${system}.native) nix;
           default = nix;
+
+          inherit (nixpkgsFor.${system}.native) lix-clang-tidy;
         }
         // (
           lib.optionalAttrs (builtins.elem system linux64BitSystems) {
@@ -390,7 +457,7 @@
             pkgs: stdenv:
             let
               nix = pkgs.callPackage ./package.nix {
-                inherit stdenv officialRelease versionSuffix;
+                inherit stdenv versionSuffix;
                 busybox-sandbox-shell = pkgs.busybox-sandbox-shell or pkgs.default-busybox-sandbox;
                 internalApiDocs = false;
               };
@@ -420,7 +487,7 @@
             makeShell pkgs pkgs.stdenv
           ))
           // {
-            default = self.devShells.${system}.native-stdenvPackages;
+            default = self.devShells.${system}.native-clangStdenvPackages;
           }
         );
     };

justfile

@@ -25,3 +25,13 @@ install *OPTIONS: (build OPTIONS)
 # Run tests
 test *OPTIONS:
     meson test -C build --print-errorlogs {{ OPTIONS }}
+
+alias clang-tidy := lint
+
+lint:
+    ninja -C build clang-tidy
+
+alias clang-tidy-fix := lint-fix
+
+lint-fix:
+    ninja -C build clang-tidy-fix

lix-doc/Cargo.lock

@@ -9,13 +9,10 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d468802bab17cbc0cc575e9b053f41e72aa36bfa6b7f55e3529ffa43161b97fa"
 
 [[package]]
-name = "cbitset"
-version = "0.2.0"
+name = "countme"
+version = "3.0.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "29b6ad25ae296159fb0da12b970b2fe179b234584d7cd294c891e2bbb284466b"
-dependencies = [
- "num-traits",
-]
+checksum = "7704b5fdd17b18ae31c4c1da5a2e0305a2bf17b5249300a9ee9ed7b72114c636"
 
 [[package]]
 name = "dissimilar"
@@ -33,19 +30,26 @@ dependencies = [
  "once_cell",
 ]
 
+[[package]]
+name = "hashbrown"
+version = "0.14.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e5274423e17b7c9fc20b6e7e208532f9b19825d82dfd615708b70edd83df41f1"
+
 [[package]]
 name = "lix-doc"
 version = "0.0.1"
 dependencies = [
  "expect-test",
  "rnix",
+ "rowan",
 ]
 
 [[package]]
-name = "num-traits"
-version = "0.2.18"
+name = "memoffset"
+version = "0.9.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "da0df0e5185db44f69b44f26786fe401b6c293d1907744beaa7fa62b2e5a517a"
+checksum = "488016bfae457b036d996092f6cb448677611ce4449e970ceaf42695203f218a"
 dependencies = [
  "autocfg",
 ]
@@ -56,44 +60,26 @@ version = "1.19.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "3fdb12b2476b595f9358c5161aa467c2438859caa136dec86c26fdd2efe17b92"
 
-[[package]]
-name = "proc-macro2"
-version = "1.0.79"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e835ff2298f5721608eb1a980ecaee1aef2c132bf95ecc026a11b7bf3c01c02e"
-dependencies = [
- "unicode-ident",
-]
-
-[[package]]
-name = "quote"
-version = "1.0.35"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "291ec9ab5efd934aaf503a6466c5d5251535d108ee747472c3977cc5acc868ef"
-dependencies = [
- "proc-macro2",
-]
-
 [[package]]
 name = "rnix"
-version = "0.8.1"
+version = "0.11.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0a9b645f0edba447dbfc6473dd22999f46a1d00ab39e777a2713a1cf34a1597b"
+checksum = "bb35cedbeb70e0ccabef2a31bcff0aebd114f19566086300b8f42c725fc2cb5f"
 dependencies = [
- "cbitset",
  "rowan",
 ]
 
 [[package]]
 name = "rowan"
-version = "0.9.1"
+version = "0.15.15"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1ea7cadf87a9d8432e85cb4eb86bd2e765ace60c24ef86e79084dcae5d1c5a19"
+checksum = "32a58fa8a7ccff2aec4f39cc45bf5f985cec7125ab271cf681c279fd00192b49"
 dependencies = [
+ "countme",
+ "hashbrown",
+ "memoffset",
  "rustc-hash",
- "smol_str",
- "text_unit",
- "thin-dst",
+ "text-size",
 ]
 
 [[package]]
@@ -103,59 +89,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "08d43f7aa6b08d49f382cde6a7982047c3426db949b1424bc4b7ec9ae12c6ce2"
 
 [[package]]
-name = "serde"
-version = "1.0.197"
+name = "text-size"
+version = "1.1.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3fb1c873e1b9b056a4dc4c0c198b24c3ffa059243875552b2bd0933b1aee4ce2"
-dependencies = [
- "serde_derive",
-]
-
-[[package]]
-name = "serde_derive"
-version = "1.0.197"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7eb0b34b42edc17f6b7cac84a52a1c5f0e1bb2227e997ca9011ea3dd34e8610b"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
-
-[[package]]
-name = "smol_str"
-version = "0.1.24"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fad6c857cbab2627dcf01ec85a623ca4e7dcb5691cbaa3d7fb7653671f0d09c9"
-dependencies = [
- "serde",
-]
-
-[[package]]
-name = "syn"
-version = "2.0.53"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7383cd0e49fff4b6b90ca5670bfd3e9d6a733b3f90c686605aa7eec8c4996032"
-dependencies = [
- "proc-macro2",
- "quote",
- "unicode-ident",
-]
-
-[[package]]
-name = "text_unit"
-version = "0.1.10"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "20431e104bfecc1a40872578dbc390e10290a0e9c35fffe3ce6f73c15a9dbfc2"
-
-[[package]]
-name = "thin-dst"
-version = "1.1.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "db3c46be180f1af9673ebb27bc1235396f61ef6965b3fe0dbb2e624deb604f0e"
-
-[[package]]
-name = "unicode-ident"
-version = "1.0.12"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3354b9ac3fae1ff6755cb6db53683adb661634f67557942dea4facebec0fee4b"
+checksum = "f18aa187839b2bdb1ad2fa35ead8c4c2976b64e4363c386d45ac0f7ee85c9233"

lix-doc/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 description = "Nix function documentation tool, stripped down into a library"
-edition = "2018"
+edition = "2021"
 name = "lix-doc"
 version = "0.0.1"
 license = "BSD-2-Clause OR MIT"
@@ -12,7 +12,9 @@ repository = "https://github.com/lf-/nix-doc"
 crate_type = ["staticlib"]
 
 [dependencies]
-rnix = "0.8.0"
+rnix = "0.11.0"
+# Necessary because rnix fails to export a critical trait (Rowan's AstNode).
+rowan = "0.15.0"
 
 [dev-dependencies]
 expect-test = "1.1.0"

lix-doc/src/lib.rs

@@ -1,5 +1,5 @@
 // SPDX-FileCopyrightText: 2024 Jade Lovelace
-//
+// SPDX-FileCopyrightText: 2024 Lunaphied
 // SPDX-License-Identifier: BSD-2-Clause OR MIT
 
 //! library components of nix-doc
@@ -7,13 +7,16 @@ pub mod pprint;
 
 use crate::pprint::pprint_args;
 
-use rnix::types::{Lambda, TypedNode};
-use rnix::SyntaxKind::*;
-use rnix::{NodeOrToken, SyntaxNode, TextUnit, WalkEvent};
+use rnix::ast::{self, Lambda};
+use rnix::{NodeOrToken, SyntaxKind};
+use rnix::SyntaxNode;
+
+
+// Needed because rnix fucked up and didn't reexport this, oops.
+use rowan::ast::AstNode;
 
 use std::ffi::{CStr, CString};
 use std::fs;
-use std::iter;
 use std::os::raw::c_char;
 use std::panic;
 
@@ -23,66 +26,104 @@ use std::{fmt::Display, str};
 
 pub type Result<T> = std::result::Result<T, Box<dyn std::error::Error>>;
 
-const DOC_INDENT: usize = 3;
-
 struct SearchResult {
     /// Name of the function
     identifier: String,
 
-    /// Dedented documentation comments
+    /// Dedented documentation comment
     doc: String,
 
     /// Parameter block for the function
     param_block: String,
 }
 
-fn find_pos(file: &str, line: usize, col: usize) -> usize {
-    let mut lines = 1;
-    let mut line_start = 0;
-    let mut it = file.chars().enumerate().peekable();
-    while let Some((count, ch)) = it.next() {
-        if ch == '\n' || ch == '\r' {
-            lines += 1;
-            let addend = if ch == '\r' && it.peek().map(|x| x.1) == Some('\n') {
-                it.next();
-                1
-            } else {
-                0
-            };
-            line_start = count + addend;
-        }
-
-        let col_diff = ((count as i32) - (line_start as i32)).abs() as usize;
-        if lines == line && col_diff == col {
-            return count;
-        }
-    }
-    unreachable!();
-}
-
 impl SearchResult {
     fn format<P: Display>(&self, filename: P, line: usize) -> String {
         format!(
             "**Synopsis:** `{}` = {}\n\n{}\n\n# {}",
             self.identifier.as_str(),
             self.param_block,
-            indented(&self.doc, DOC_INDENT),
+            self.doc,
             format!("{}:{}", filename, line).as_str(),
         )
     }
 }
 
-/// Emits a string `s` indented by `indent` spaces
-fn indented(s: &str, indent: usize) -> String {
-    let indent_s = iter::repeat(' ').take(indent).collect::<String>();
-    s.split('\n')
-        .map(|line| indent_s.clone() + line)
-        .collect::<Vec<_>>()
-        .join("\n")
+/// Converts Nix compatible line endings (Nix accepts `\r`, `\n`, *and* `\r\n` as endings), to
+/// standard `\n` endings for use within Rust land.
+fn convert_endings(s: &str) -> String {
+    let mut out = String::with_capacity(s.len());
+    let mut it = s.chars().peekable();
+
+    while let Some(ch) = it.next() {
+        if ch == '\n' || ch == '\r' {
+            out.push('\n');
+            if ch == '\r' && it.peek().map(|&c| c == '\n').unwrap_or(false) {
+                // Consume `\n` in `\r\n`.
+                it.next();
+            }
+        } else {
+            out.push(ch);
+        }
+    }
+
+    out
+}
+
+/// Converts the position information from Lix itself into an character index into the file itself.
+/// Expects an input string that's already had it's line endings normalized.
+///
+/// Note that this returns a *byte* offset, not a character offset.
+fn find_pos(s: &str, line: usize, col: usize) -> usize {
+    // Nix line positions are 1-indexed.
+    let mut lines = 1;
+    for (byte_pos, ch) in s.char_indices() {
+        // If we find a newline, increase the line count.
+        if ch == '\n' {
+            lines += 1;
+        }
+
+        // We've arrived at the correct line.
+        if lines == line {
+            // Column position is 1-indexed, and it's a *byte* offset, because Nix doesn't actually
+            // support UTF-8. Rust does though, so we need to convert to a proper byte index to
+            // match rnix. Lix also doesn't consider the line endings part of the column offset so
+            // we implicitly add one to advance to the character *after* that.
+            return byte_pos + col;
+        }
+    }
+
+    // If things never match that should be literally impossible.
+    unreachable!();
+}
+
+/// Represents a forwarded token from rnix's AST over to lix-doc.
+#[derive(Debug, Clone)]
+enum DocToken {
+    Comment(String),
+    Whitespace(String),
+}
+
+/// Determine if a given token string contains more than two newlines, this is used to determine when
+/// we hit blank lines between comments indicating a contextually unrelated comment.
+fn has_empty_line(tok: &DocToken) -> bool {
+    // It's either solely whitespace with two newlines inside somewhere, or it's
+    // contained inside a comment token and we don't want to count that as empty.
+    if let DocToken::Whitespace(s) = tok {
+        s.chars().filter(|&c| c == '\n').take(2).count() == 2
+    } else {
+        false
+    }
 }
 
 /// Cleans up a single line, erasing prefix single line comments but preserving indentation
-fn cleanup_single_line<'a>(s: &'a str) -> &'a str {
+// NOTE: We have a bit of a conflict of interest problem here due to the inconsistent format of
+// doc comments. Some doc comments will use a series of single line comments that may then contain `*`
+// characters to represent a list. Some will be multiline comments that don't prefix individual lines
+// with `*`, only using them for lists directly, and some will prefix lines with `*` as a leading
+// character to mark the block. There's no way to disambiguate all three, but we do our best to
+// make the common case pretty.
+fn cleanup_single_line(s: &str) -> &str {
     let mut cmt_new_start = 0;
     let mut iter = s.char_indices().peekable();
     while let Some((idx, ch)) = iter.next() {
@@ -90,7 +131,9 @@ fn cleanup_single_line<'a>(s: &'a str) -> &'a str {
         let (_, next_ch) = iter.peek().unwrap_or(&(0, '\n'));
 
         // if we find a character, save the byte position after it as our new string start
-        if ch == '#' || (ch == '*' && next_ch.is_whitespace()) {
+        // This has special handling for `>` because some Nixpkgs documentation has `*>` right
+        // after the start of their doc comments, and we want to strip the `*` still.
+        if ch == '#' || (ch == '*' && (*next_ch == '>' || next_ch.is_whitespace())) {
             cmt_new_start = idx + 1;
             break;
         }
@@ -103,15 +146,12 @@ fn cleanup_single_line<'a>(s: &'a str) -> &'a str {
     &s[cmt_new_start..]
 }
 
-/// Erases indents in comments. This is *almost* a normal dedent function, but it starts by looking
-/// at the second line if it can.
+/// Erases indents in comments based on the indentation of the first line.
 fn dedent_comment(s: &str) -> String {
     let mut whitespaces = 0;
-    let mut lines = s.lines();
-    let first = lines.next();
 
     // scan for whitespace
-    for line in lines.chain(first) {
+    for line in s.lines() {
         let line_whitespace = line.chars().take_while(|ch| ch.is_whitespace()).count();
 
         if line_whitespace != line.len() {
@@ -121,16 +161,6 @@ fn dedent_comment(s: &str) -> String {
         }
     }
 
-    // maybe the first considered line we found was indented further, so let's look for more lines
-    // that might have a shorter indent. In the case of one line, do nothing.
-    for line in s.lines().skip(1) {
-        let line_whitespace = line.chars().take_while(|ch| ch.is_whitespace()).count();
-
-        if line_whitespace != line.len() {
-            whitespaces = line_whitespace.min(whitespaces);
-        }
-    }
-
     // delete up to `whitespaces` whitespace characters from each line and reconstitute the string
     let mut out = String::new();
     for line in s.lines() {
@@ -143,69 +173,163 @@ fn dedent_comment(s: &str) -> String {
     out
 }
 
-/// Deletes whitespace and leading comment characters
+/// Takes a series of comment and whitespace strings and output a clean single block of text to use
+/// as the output documentation comment block.
 ///
-/// Oversight we are choosing to ignore: if you put # characters at the beginning of lines in a
-/// multiline comment, they will be deleted.
-fn cleanup_comments<S: AsRef<str>, I: DoubleEndedIterator<Item = S>>(comment: &mut I) -> String {
+/// This function expects to be given the tokens in reverse order (proceeding upwards from the
+/// first comment above the definitions), this allows us to properly enforce the below conditions.
+/// The output from this function will be reordered and ready for display.
+///
+/// The two types of documentation comments we expect are:
+///
+/// - A single multiline comment not whitespace separated from the start.
+/// - A series of back to back single line comments not separated by whitespace.
+///
+/// Any other combination will be filtered out.
+///
+/// Once an empty line is encountered, we know no more valid documentation comments remain and stop.
+fn cleanup_comments<I: Iterator<Item = DocToken>>(tokens: &mut I) -> String {
+    // Keep track of when we've found a single line and multiline comment, we use this to
+    // only process a single multiline or back to back single lines.
+    let mut found_single_line = false;
+
+    // Comments that have survived our filtering phase and should be cleaned up.
+    let mut valid = vec![];
+
+    // Filter out comments that don't meet the characteristics of documentation comments.
+    for tok in tokens {
+        if has_empty_line(&tok) {
+            // Take tokens until we hit whitespace containing an empty line.
+            break;
+        }
+
+        // Only care about comments from this point on.
+        if let DocToken::Comment(comment) = tok {
+            // Now determine if it's a single line comment.
+            let is_single_line = comment.starts_with('#');
+
+            // We've found a single line comment if we've found one before or we just found one.
+            found_single_line |= is_single_line;
+
+            // What we do next is only special when we hit a multiline comment.
+            if !is_single_line {
+                // If we've hit a multiline comment as our first comment, take that one alone.
+                if !found_single_line {
+                    // Otherwise we've hit a multiline comment immediately and this is our
+                    // one and only doc comment to worry about.
+                    valid.push(comment);
+                }
+                // Otherwise we've hit a multiline comment after single line comments, in either
+                // case this means we're done processing comments.
+                break;
+            }
+
+            // Otherwise this is a new single line comment to push to the stack.
+            valid.push(comment);
+        }
+    }
+
+    // Cleanup comments for user consumption.
     dedent_comment(
-        &comment
+        &valid
+            .into_iter()
             .rev()
             .map(|small_comment| {
                 small_comment
-                    .as_ref()
-                    // space before multiline start
-                    .trim_start()
-                    // multiline starts
+                    // Trim off start of multiline comments.
                     .trim_start_matches("/*")
-                    // trailing so we can grab multiline end
-                    .trim_end()
-                    // multiline ends
+                    // Trim off end of multiline comments.
                     .trim_end_matches("*/")
-                    // extra space that was in the multiline
+                    // Trim off any internal whitespace that's trapped inside comments themselves.
                     .trim()
+                    // Split comments by newlines to extract lines of multiline comments.
                     .split('\n')
-                    // erase single line comments and such
+                    // Cleanup single line comments and a few more tweaks for multiline comments.
                     .map(cleanup_single_line)
                     .collect::<Vec<_>>()
+                    // Reconstruct the multiline comment's whitespace.
                     .join("\n")
             })
             .collect::<Vec<_>>()
-            .join("\n"),
+            // We've found that when multiple back to back single line comments are used in Nixpkgs,
+            // they make more sense to represent as if someone inserted line breaks into the Markdown
+            // properly, so we join them with linebreaks that markdown will pass through.
+            .join("\n\n"),
     )
 }
 
-/// Get the docs for a specific function
+/// Get the docs for a specific function.
+// TODO: Improve error reporting?
 pub fn get_function_docs(filename: &str, line: usize, col: usize) -> Option<String> {
     let content = fs::read(filename).ok()?;
-    let decoded = str::from_utf8(&content).ok()?;
+    let decoded = convert_endings(str::from_utf8(&content).ok()?);
     let pos = find_pos(&decoded, line, col);
-    let rowan_pos = TextUnit::from_usize(pos);
-    let tree = rnix::parse(decoded);
+    let rowan_pos = rnix::TextSize::from(pos as u32);
 
+    // The minimum length of a lambda is 4 characters and thus the range we're looking for must be
+    // at least 4 characters long `_: 3` being an example of a minimal length lambda.
+    let rowan_range = rnix::TextRange::at(rowan_pos, 4.into());
+
+    // Parse the file  using rnix.
+    let root = rnix::Root::parse(&decoded).ok().ok()?;
+
+    // Extract the inner expression that represents the Root node and extract the top level expression.
+    let expr = root.expr()?;
+
+    // There are two cases we have to be able to handle
+    // 1. A straightforward definition with an attrset binding to a lambda that's defined inline.
+    // 2. A lambda defined in a standalone file where the attrset binding imports that file directly.
+    // The latter case will not be able to find the binding so we must be able to handle not finding it.
+
+    // Find the deepest node or token that covers the position given by Lix.
+    let covering = expr.syntax().covering_element(rowan_range);
+
+    // Climb up until we find the lambda node that contains that token.
     let mut lambda = None;
-    for node in tree.node().preorder() {
-        match node {
-            WalkEvent::Enter(n) => {
-                if n.text_range().start() >= rowan_pos && n.kind() == NODE_LAMBDA {
-                    lambda = Lambda::cast(n);
-                    break;
-                }
-            }
-            WalkEvent::Leave(_) => (),
+    for ancestor in covering.ancestors() {
+        if ancestor.kind() == SyntaxKind::NODE_LAMBDA {
+            lambda = Some(ancestor);
+            break;
         }
     }
-    let lambda = lambda?;
-    let res = visit_lambda("func".to_string(), &lambda);
-    Some(res.format(filename, line))
+
+    // There is literally always a lambda or something has gone very very wrong.
+    let lambda =
+        ast::Lambda::cast(
+            lambda.expect("no lambda found; what.")
+        ) .expect("not a rnix::ast::Lambda; what.");
+
+    // Search up, hopefully to find the binding so we can get the identifier name.
+    // TODO: Just provide this directly from the C++ code to make it possible to always have the correct identifier.
+    let mut binding = None;
+    for ancestor in lambda.syntax().ancestors() {
+        if ancestor.kind() == SyntaxKind::NODE_ATTRPATH_VALUE {
+            binding = Some(ancestor);
+        }
+    }
+
+    // Convert the binding to an identifier if it was found, otherwise use a placeholder.
+    let identifier;
+    identifier = match binding.clone() {
+        Some(binding) => ast::AttrpathValue::cast(binding)
+            .expect("not an rnix::ast::AttrpathValue; what")
+            .attrpath()
+            .expect("AttrpathValue has no attrpath; what.")
+            .to_string(),
+        _ => "<unknown binding>".to_string(),
+    };
+
+    // Find all the comments on the binding or the lambda if we have to fall back.
+    let comment_node = binding.as_ref().unwrap_or(lambda.syntax());
+    let comment = find_comment(comment_node).unwrap_or_else(String::new);
+
+    // And display them properly for the markdown function in Lix.
+    Some(visit_lambda(identifier, comment, &lambda).format(filename, line))
 }
 
-fn visit_lambda(name: String, lambda: &Lambda) -> SearchResult {
+fn visit_lambda(name: String, comment: String, lambda: &Lambda) -> SearchResult {
     // grab the arguments
-    let param_block = pprint_args(&lambda);
-
-    // find the doc comment
-    let comment = find_comment(lambda.node().clone()).unwrap_or_else(|| "".to_string());
+    let param_block = pprint_args(lambda);
 
     SearchResult {
         identifier: name,
@@ -214,39 +338,47 @@ fn visit_lambda(name: String, lambda: &Lambda) -> SearchResult {
     }
 }
 
-fn find_comment(node: SyntaxNode) -> Option<String> {
-    let mut node = NodeOrToken::Node(node);
-    let mut comments = Vec::new();
-    loop {
-        loop {
-            if let Some(new) = node.prev_sibling_or_token() {
-                node = new;
-                break;
-            } else {
-                node = NodeOrToken::Node(node.parent()?);
-            }
-        }
+fn find_comment(node: &SyntaxNode) -> Option<String> {
+    let mut it = node
+        .siblings_with_tokens(rowan::Direction::Prev)
+        // Skip ourselves as we're always the first token returned.
+        .skip(1)
+        .peekable();
 
-        match node.kind() {
-            TOKEN_COMMENT => match &node {
-                NodeOrToken::Token(token) => comments.push(token.text().clone()),
-                NodeOrToken::Node(_) => unreachable!(),
-            },
-            // This stuff is found as part of `the-fn = f: ...`
-            // here:                           ^^^^^^^^
-            NODE_KEY | TOKEN_ASSIGN => (),
-            t if t.is_trivia() => (),
-            _ => break,
-        }
+    // Consume up to one whitespace token before the first comment. There might not always be
+    // whitespace such as the (rather unusual) case of `/* meow */x = a: 3`.
+    if matches!(it.peek(), Some(NodeOrToken::Token(token)) if token.kind() == SyntaxKind::TOKEN_WHITESPACE) {
+        it.next();
     }
-    let doc = cleanup_comments(&mut comments.iter().map(|c| c.as_str()));
-    Some(doc).filter(|it| !it.is_empty())
+
+    let comments = it.map_while(|element| match element {
+            NodeOrToken::Token(token) => {
+                match token.kind() {
+                    // Map the tokens we're interested in to our internal token type.
+                    SyntaxKind::TOKEN_COMMENT => Some(DocToken::Comment(token.text().to_owned())),
+                    SyntaxKind::TOKEN_WHITESPACE => {
+                        Some(DocToken::Whitespace(token.text().to_owned()))
+                    }
+                    // If we hit a different token type, we know we've gone past relevant comments
+                    // and should stop.
+                    _ => None,
+                }
+            }
+            // If we hit a node entry we've definitely gone past comments that would be related to
+            // this node and we should retreat.
+            _ => None,
+        });
+
+    // For the curious, `into_iter()` here consumes the binding producing an owned value allowing us to avoid
+    // making the original binding mutable, we don't reuse it later so this is a cute way to handle it, though
+    // there's probably a better way we just can't remember.
+    Some(cleanup_comments(&mut comments.into_iter())).filter(|c| !c.is_empty())
 }
 
 /// Get the docs for a function in the given file path at the given file position and return it as
 /// a C string pointer
 #[no_mangle]
-pub extern "C" fn nd_get_function_docs(
+pub extern "C" fn lixdoc_get_function_docs(
     filename: *const c_char,
     line: usize,
     col: usize,
@@ -269,9 +401,9 @@ pub extern "C" fn nd_get_function_docs(
         .unwrap_or(ptr::null())
 }
 
-/// Call this to free a string from nd_get_function_docs
+/// Call this to free a string from `lixdoc_get_function_docs`.
 #[no_mangle]
-pub extern "C" fn nd_free_string(s: *const c_char) {
+pub extern "C" fn lixdoc_free_string(s: *const c_char) {
     unsafe {
         // cast note: this cast is turning something that was cast to const
         // back to mut
@@ -283,35 +415,57 @@ pub extern "C" fn nd_free_string(s: *const c_char) {
 mod tests {
     use super::*;
 
+    #[test]
+    fn test_line_conversion() {
+        let fakefile = "abc\rdef\r\nghi";
+        assert_eq!(convert_endings(fakefile), "abc\ndef\nghi");
+    }
+
     #[test]
     fn test_bytepos() {
         let fakefile = "abc\ndef\nghi";
         assert_eq!(find_pos(fakefile, 2, 2), 5);
     }
 
+    #[test]
+    fn test_bytepos_unusual() {
+        let fakefile = convert_endings("abc\rdef\r\nghi");
+        assert_eq!(find_pos(&fakefile, 2, 2), 5);
+        assert_eq!(find_pos(&fakefile, 3, 2), 9);
+    }
+
+    /// This test is to check that we correctly resolve byte positions even when inconsistent with
+    /// character positions.
     #[test]
     fn test_bytepos_cursed() {
-        let fakefile = "abc\rdef\r\nghi";
-        assert_eq!(find_pos(fakefile, 2, 2), 5);
-        assert_eq!(find_pos(fakefile, 3, 2), 10);
+        let fakefile = "hello\nwórld";
+        // Try to find the position of the `r` after world, which will be wrong if we don't handle
+        // UTF-8 properly.
+        let pos = find_pos(&fakefile, 2, 4);
+        dbg!(&fakefile[pos..]);
+        assert_eq!(pos, 9)
     }
 
     #[test]
     fn test_comment_stripping() {
-        let ex1 = ["/* blah blah blah\n      foooo baaar\n   blah */"];
+        let ex1 = [DocToken::Comment(
+            "/* blah blah blah\n      foooo baaar\n   blah */".to_string(),
+        )];
         assert_eq!(
-            cleanup_comments(&mut ex1.iter()),
-            "blah blah blah\n   foooo baaar\nblah"
+            cleanup_comments(&mut ex1.into_iter()),
+            "blah blah blah\n      foooo baaar\n   blah"
         );
 
-        let ex2 = ["# a1", "#    a2", "# aa"];
-        assert_eq!(cleanup_comments(&mut ex2.iter()), "aa\n   a2\na1");
+        let ex2 = ["# a1", "#    a2", "# aa"]
+            .into_iter()
+            .map(|s| DocToken::Comment(s.to_string()));
+        assert_eq!(cleanup_comments(&mut ex2.into_iter()), "aa\n\n   a2\n\na1");
     }
 
     #[test]
     fn test_dedent() {
         let ex1 = "a\n   b\n   c\n     d";
-        assert_eq!(dedent_comment(ex1), "a\nb\nc\n  d");
+        assert_eq!(dedent_comment(ex1), ex1);
         let ex2 = "a\nb\nc";
         assert_eq!(dedent_comment(ex2), ex2);
         let ex3 = "   a\n   b\n\n     c";
@@ -335,4 +489,31 @@ mod tests {
         let ex1 = "   **Foo**:";
         assert_eq!(cleanup_single_line(ex1), ex1);
     }
+
+    // TODO: Next CL
+    //#[test]
+    //fn comment_test_complex() {
+    //    let testcase = r#"
+    //    rec {
+    //        /*
+    //           Hello
+    //           23
+    //             This is a comment.
+    //             this is another comment.
+    //             and this is a third comment.
+    //                          Way
+    //              go
+    //        */
+    //        meow = { g }: {a, b ? 4, ...}: g: c: 5;
+    //        # And another comment.
+    //        cat = 34;
+    //        # inner layer.
+    //        "inner-layer" = outer: meow;
+    //    }
+    //    "#;
+    //    // Need to find the location of the lambda, we do a quick hack.
+    //    let location = dbg!(testcase.find("{ g }").unwrap() as u32);
+    //
+    //    //get_function_docs(filename, line, col)
+    //}
 }

lix-doc/src/pprint.rs

@@ -1,36 +1,59 @@
 // SPDX-FileCopyrightText: 2024 Jade Lovelace
-//
+// SPDX-FileCopyrightText: 2024 Lunaphied
 // SPDX-License-Identifier: BSD-2-Clause OR MIT
 
-use rnix::types::{Lambda, TypedNode};
-use rnix::SyntaxKind::*;
+use rnix::ast::{Expr, Lambda};
+use rowan::ast::AstNode;
 
 /// Pretty-prints the arguments to a function
 pub fn pprint_args(lambda: &Lambda) -> String {
     // TODO: handle docs directly on NODE_IDENT args (uncommon case)
     let mut lambda = lambda.clone();
+    let mut depth = 0;
     let mut out = String::new();
     loop {
-        let arg = lambda.arg().unwrap();
-        match arg.kind() {
-            NODE_IDENT => {
-                out += &format!("*{}*", &arg.to_string());
-                out.push_str(": ");
-                let body = lambda.body().unwrap();
-                if body.kind() == NODE_LAMBDA {
-                    lambda = Lambda::cast(body).unwrap();
-                } else {
-                    break;
+        let arg = lambda.param().unwrap();
+        for child in arg.syntax().children_with_tokens() {
+            //dbg!(child.kind());
+            match child {
+                rowan::NodeOrToken::Node(node) => {
+                    out.push_str(&node.text().to_string());
+                    if node.kind() == rnix::SyntaxKind::NODE_PAT_ENTRY {
+                        out.push_str(&",\n");
+                    }
+                }
+                rowan::NodeOrToken::Token(token) => {
+                    use rnix::SyntaxKind::{
+                        TOKEN_COMMENT, TOKEN_ELLIPSIS, TOKEN_L_BRACE, TOKEN_QUESTION, TOKEN_R_BRACE,
+                    };
+                    match token.kind() {
+                        TOKEN_COMMENT | TOKEN_ELLIPSIS | TOKEN_QUESTION | TOKEN_L_BRACE
+                        | TOKEN_R_BRACE => {
+                            //dbg!(&token);
+                            out.push_str(&token.text().to_string());
+                            if token.kind() == TOKEN_COMMENT {
+                                out.push('\n');
+                            }
+                        }
+                        _ => {}
+                    }
+                    //out.push_str(&token.text().to_string());
                 }
             }
-            NODE_PATTERN => {
-                out += &format!("*{}*", &arg.to_string());
-                out.push_str(": ");
-                break;
-            }
-            t => {
-                unreachable!("unhandled arg type {:?}", t);
+        }
+        out.push_str(": ");
+        let body = lambda.body().unwrap();
+        if let Expr::Lambda(inner) = body {
+            lambda = inner;
+            // If we recurse we want the next line of recursion to be indented and on a new line.
+            out.push('\n');
+            for _ in 0..=depth {
+                out.push('\t');
             }
+            depth += 1;
+        } else {
+            // If we don't find an inner lambda we're done with argument handling.
+            break;
         }
     }
     out.push_str("...");

maintainers/build-release-notes.py

@@ -106,7 +106,7 @@ def do_category(author_info: AuthorInfoDB, entries: list[Tuple[pathlib.Path, Any
             links = []
             links += [format_issue(str(s)) for s in listify(entry.metadata.get('issues', []))]
             links += [format_pr(str(s)) for s in listify(entry.metadata.get('prs', []))]
-            links += [format_cl(cl) for cl in listify(entry.metadata.get('cls', []))]
+            links += [format_cl(int(cl)) for cl in listify(entry.metadata.get('cls', []))]
             if links != []:
                 header += " " + " ".join(links)
             if header:
@@ -129,7 +129,7 @@ def run_on_dir(author_info: AuthorInfoDB, d):
     entries = defaultdict(list)
     for p in paths:
         try:
-            e = frontmatter.load(p)
+            e = frontmatter.load(p) # type: ignore
             if 'synopsis' not in e.metadata:
                 raise Exception('missing synopsis')
             unknownKeys = set(e.metadata.keys()) - set(KNOWN_KEYS)

maintainers/check-syscalls.nix

@@ -0,0 +1,16 @@
+{
+  runCommandNoCC,
+  lib,
+  libseccomp,
+  writeShellScriptBin,
+}:
+let
+  syscalls-csv = runCommandNoCC "syscalls.csv" { } ''
+    echo ${lib.escapeShellArg libseccomp.src}
+    tar -xf ${lib.escapeShellArg libseccomp.src} --strip-components=2 ${libseccomp.name}/src/syscalls.csv
+    mv syscalls.csv "$out"
+  '';
+in
+writeShellScriptBin "check-syscalls" ''
+  ${./check-syscalls.sh} ${syscalls-csv}
+''

maintainers/check-syscalls.sh

@@ -0,0 +1,7 @@
+#!/usr/bin/env bash
+
+set -e
+
+diff -u <(awk < src/libstore/platform/linux.cc '/BEGIN extract-syscalls/ { extracting = 1; next }
+match($0, /allowSyscall\(ctx, SCMP_SYS\(([^)]*)\)\);|\/\/ skip ([^ ]*)/, result) { print result[1] result[2] }
+/END extract-syscalls/ { extracting = 0; next }') <(tail -n+2 "$1" | cut -d, -f 1)

meson.build

@@ -199,23 +199,27 @@ configdata = { }
 # Dependencies
 #
 
-boehm = dependency('bdw-gc', required : get_option('gc'), version : '>=8.2.6')
+gc_opt = get_option('gc').disable_if(
+  'address' in get_option('b_sanitize'),
+  error_message: 'gc does far too many memory crimes for ASan'
+)
+boehm = dependency('bdw-gc', required : gc_opt, version : '>=8.2.6', include_type : 'system')
 configdata += {
   'HAVE_BOEHMGC': boehm.found().to_int(),
 }
 
-boost = dependency('boost', required : true, modules : ['context', 'coroutine', 'container'])
+boost = dependency('boost', required : true, modules : ['container'], include_type : 'system')
 
 # cpuid only makes sense on x86_64
 cpuid_required = is_x64 ? get_option('cpuid') : false
-cpuid = dependency('libcpuid', 'cpuid', required : cpuid_required)
+cpuid = dependency('libcpuid', 'cpuid', required : cpuid_required, include_type : 'system')
 configdata += {
   'HAVE_LIBCPUID': cpuid.found().to_int(),
 }
 
 # seccomp only makes sense on Linux
 seccomp_required = is_linux ? get_option('seccomp-sandboxing') : false
-seccomp = dependency('libseccomp', 'seccomp', required : seccomp_required, version : '>=2.5.5')
+seccomp = dependency('libseccomp', 'seccomp', required : seccomp_required, version : '>=2.5.5', include_type : 'system')
 if is_linux and not seccomp.found()
   warning('Sandbox security is reduced because libseccomp has not been found! Please provide libseccomp if it supports your CPU architecture.')
 endif
@@ -223,19 +227,29 @@ configdata += {
   'HAVE_SECCOMP': seccomp.found().to_int(),
 }
 
-libarchive = dependency('libarchive', required : true)
+# fcntl(F_GETPATH) returns the path of an fd on macOS and BSDs
+configdata += {
+  'HAVE_F_GETPATH': cxx.has_header_symbol('fcntl.h', 'F_GETPATH').to_int(),
+}
+
+libarchive = dependency('libarchive', required : true, include_type : 'system')
 
 brotli = [
-  dependency('libbrotlicommon', required : true),
-  dependency('libbrotlidec', required : true),
-  dependency('libbrotlienc', required : true),
+  dependency('libbrotlicommon', required : true, include_type : 'system'),
+  dependency('libbrotlidec', required : true, include_type : 'system'),
+  dependency('libbrotlienc', required : true, include_type : 'system'),
 ]
 
-openssl = dependency('libcrypto', 'openssl', required : true)
+openssl = dependency('libcrypto', 'openssl', required : true, include_type : 'system')
 
 # FIXME: confirm we actually support such old versions of aws-sdk-cpp
-aws_sdk = dependency('aws-cpp-sdk-core', required : false, version : '>=1.8')
-aws_sdk_transfer = dependency('aws-cpp-sdk-transfer', required : aws_sdk.found(), fallback : ['aws_sdk', 'aws_cpp_sdk_transfer_dep'])
+aws_sdk = dependency('aws-cpp-sdk-core', required : false, version : '>=1.8', include_type : 'system')
+aws_sdk_transfer = dependency(
+  'aws-cpp-sdk-transfer',
+  required : aws_sdk.found(),
+  fallback : ['aws_sdk', 'aws_cpp_sdk_transfer_dep'],
+  include_type : 'system',
+)
 if aws_sdk.found()
   # The AWS pkg-config adds -std=c++11.
   # https://github.com/aws/aws-sdk-cpp/issues/2673
@@ -255,7 +269,12 @@ if aws_sdk.found()
   )
 endif
 
-aws_s3 = dependency('aws-cpp-sdk-s3', required : aws_sdk.found(), fallback : ['aws_sdk', 'aws_cpp_sdk_s3_dep'])
+aws_s3 = dependency(
+  'aws-cpp-sdk-s3',
+  required : aws_sdk.found(),
+  fallback : ['aws_sdk', 'aws_cpp_sdk_s3_dep'],
+  include_type : 'system',
+)
 if aws_s3.found()
   # The AWS pkg-config adds -std=c++11.
   # https://github.com/aws/aws-sdk-cpp/issues/2673
@@ -272,30 +291,30 @@ configdata += {
   'ENABLE_S3': aws_s3.found().to_int(),
 }
 
-sqlite = dependency('sqlite3', 'sqlite', version : '>=3.6.19', required : true)
+sqlite = dependency('sqlite3', 'sqlite', version : '>=3.6.19', required : true, include_type : 'system')
 
-sodium = dependency('libsodium', 'sodium', required : true)
+sodium = dependency('libsodium', 'sodium', required : true, include_type : 'system')
 
-curl = dependency('libcurl', 'curl', required : true)
+curl = dependency('libcurl', 'curl', required : true, include_type : 'system')
 
-editline = dependency('libeditline', 'editline', version : '>=1.14', required : true)
+editline = dependency('libeditline', 'editline', version : '>=1.14', required : true, include_type : 'system')
 
-lowdown = dependency('lowdown', version : '>=0.9.0', required : true)
+lowdown = dependency('lowdown', version : '>=0.9.0', required : true, include_type : 'system')
 
 # HACK(Qyriad): rapidcheck's pkg-config doesn't include the libs lol
 # Note: technically we 'check' for rapidcheck twice, for the internal-api-docs handling above,
 # but Meson will cache the result of the first one, and the required : arguments are different.
-rapidcheck_meson = dependency('rapidcheck', required : enable_tests)
+rapidcheck_meson = dependency('rapidcheck', required : enable_tests, include_type : 'system')
 rapidcheck = declare_dependency(dependencies : rapidcheck_meson, link_args : ['-lrapidcheck'])
 
 gtest = [
-  dependency('gtest', required : enable_tests),
-  dependency('gtest_main', required : enable_tests),
-  dependency('gmock', required : enable_tests),
-  dependency('gmock_main', required : enable_tests),
+  dependency('gtest', required : enable_tests, include_type : 'system'),
+  dependency('gtest_main', required : enable_tests, include_type : 'system'),
+  dependency('gmock', required : enable_tests, include_type : 'system'),
+  dependency('gmock_main', required : enable_tests, include_type : 'system'),
 ]
 
-toml11 = dependency('toml11', version : '>=3.7.0', required : true, method : 'cmake')
+toml11 = dependency('toml11', version : '>=3.7.0', required : true, method : 'cmake', include_type : 'system')
 
 pegtl = dependency(
   'pegtl',
@@ -303,9 +322,10 @@ pegtl = dependency(
   required : true,
   method : 'cmake',
   modules : [ 'taocpp::pegtl' ],
+  include_type : 'system',
 )
 
-nlohmann_json = dependency('nlohmann_json', required : true)
+nlohmann_json = dependency('nlohmann_json', required : true, include_type : 'system')
 
 # lix-doc is a Rust project provided via buildInputs and unfortunately doesn't have any way to be detected.
 # Just declare it manually to resolve this.
@@ -314,6 +334,10 @@ nlohmann_json = dependency('nlohmann_json', required : true)
 # *absolutely* are not going to make it work)
 lix_doc = declare_dependency(link_args : [ '-llix_doc' ])
 
+if is_freebsd
+  libprocstat = declare_dependency(link_args : [ '-lprocstat' ])
+endif
+
 #
 # Build-time tools
 #
@@ -352,6 +376,13 @@ endif
 # FIXME(Qyriad): the autoconf system checks that busybox has the "standalone" feature, indicating
 # that busybox sh won't run busybox applets as builtins (which would break our sandbox).
 
+pasta_path = get_option('pasta-path')
+# we can't check the pasta version because passt misuses stdio (it calls _exit()
+# after printing the version, which will never print the version unless run from
+# a terminal). pasta isn't mandatory yet due to high fetcher breakage potential.
+# we *will* enable it in our own packaging, but distributions are not forced to.
+pasta = find_program(pasta_path, required : false, native : false)
+
 lsof = find_program('lsof', native : true)
 
 # This is how Nix does generated headers...
@@ -398,6 +429,11 @@ check_funcs = [
   'strsignal',
   'sysconf',
 ]
+if is_linux or is_freebsd
+  # musl does not have close_range as of 2024-08-10
+  # patch: https://www.openwall.com/lists/musl/2024/08/01/9
+  check_funcs += [ 'close_range' ]
+endif
 foreach funcspec : check_funcs
   define_name = 'HAVE_' + funcspec.underscorify().to_upper()
   define_value = cxx.has_function(funcspec).to_int()
@@ -445,6 +481,7 @@ add_project_arguments(
   '-Werror=unused-result',
   '-Wdeprecated-copy',
   '-Wignored-qualifiers',
+  '-Werror=suggest-override',
   # Enable assertions in libstdc++ by default. Harmless on libc++. Benchmarked
   # at ~1% overhead in `nix search`.
   #
@@ -477,7 +514,14 @@ if cxx.get_id() == 'clang' and get_option('b_sanitize') != ''
   add_project_link_arguments('-shared-libsan', language : 'cpp')
 endif
 
+# Clang gets grumpy about missing libasan symbols if -shared-libasan is not
+# passed when building shared libs, at least on Linux
+if cxx.get_id() == 'clang' and 'address' in get_option('b_sanitize')
+  add_project_link_arguments('-shared-libasan', language : 'cpp')
+endif
+
 add_project_link_arguments('-pthread', language : 'cpp')
+
 if cxx.get_linker_id() in ['ld.bfd', 'ld.gold']
   add_project_link_arguments('-Wl,--no-copy-dt-needed-entries', language : 'cpp')
 endif
@@ -492,7 +536,7 @@ endif
 # maintainers/buildtime_report.sh BUILD-DIR to simply work in clang builds.
 #
 # They can also be manually viewed at https://ui.perfetto.dev
-if get_option('profile-build').require(meson.get_compiler('cpp').get_id() == 'clang').enabled()
+if get_option('profile-build').require(cxx.get_id() == 'clang').enabled()
   add_project_arguments('-ftime-trace', language: 'cpp')
 endif
 
@@ -532,3 +576,5 @@ if enable_tests
   subdir('tests/unit')
   subdir('tests/functional')
 endif
+
+subdir('meson/clang-tidy')

meson.options

@@ -1,7 +1,7 @@
 # vim: filetype=meson
 
 option('enable-build', type : 'boolean', value : true,
-  description : 'Set to false to not actually build. Only really makes sense with -Dinternal-api-docs=true',
+  description : 'set to false to not actually build. Only really makes sense with -Dinternal-api-docs=true',
 )
 
 option('gc', type : 'feature',
@@ -24,6 +24,10 @@ option('sandbox-shell', type : 'string', value : 'busybox',
   description : 'path to a statically-linked shell to use as /bin/sh in sandboxes (usually busybox)',
 )
 
+option('pasta-path', type : 'string', value : 'pasta',
+  description : 'path to the location of pasta (provided by passt)',
+)
+
 option('enable-tests', type : 'boolean', value : true,
   description : 'whether to enable tests or not (requires rapidcheck and gtest)',
 )
@@ -37,7 +41,7 @@ option('tests-brief', type : 'boolean', value : false,
 )
 
 option('profile-build', type : 'feature', value: 'disabled',
-  description : 'whether to enable -ftime-trace in clang builds, allowing for speeding up the build.'
+  description : 'whether to enable -ftime-trace in clang builds, allowing for diagnosing the cause of build time.'
 )
 
 option('store-dir', type : 'string', value : '/nix/store',
@@ -68,3 +72,7 @@ option('profile-dir', type : 'string', value : 'etc/profile.d',
 option('enable-pch-std', type : 'boolean', value : true,
   description : 'whether to use precompiled headers for C++\'s standard library (breaks clangd if you\'re using GCC)',
 )
+
+option('lix-clang-tidy-checks-path', type : 'string', value : '',
+  description: 'path to lix-clang-tidy-checks library file, if providing it externally. Uses an internal one if this is not set',
+)

meson/clang-tidy/build_required_targets.py

@@ -0,0 +1,21 @@
+#!/usr/bin/env python3
+import subprocess
+
+def get_targets_of_rule(build_root: str, rule_name: str) -> list[str]:
+    return subprocess.check_output(['ninja', '-C', build_root, '-t', 'targets', 'rule', rule_name]).decode().strip().splitlines()
+
+def ninja_build(build_root: str, targets: list[str]):
+    subprocess.check_call(['ninja', '-C', build_root, '--', *targets])
+
+def main():
+    import argparse
+    ap = argparse.ArgumentParser(description='Builds required targets for clang-tidy')
+    ap.add_argument('build_root', help='Ninja build root', type=str)
+
+    args = ap.parse_args()
+
+    targets = [t for t in get_targets_of_rule(args.build_root, 'CUSTOM_COMMAND') if t.endswith('gen.hh')]
+    ninja_build(args.build_root, targets)
+
+if __name__ == '__main__':
+    main()

meson/clang-tidy/clean_compdb.py

@@ -0,0 +1,53 @@
+#!/usr/bin/env python3
+# Deletes the PCH arguments from a compilation database, to workaround nixpkgs
+# stdenv having a cc-wrapper that is impossible to use for anything except cc
+# itself, for example, clang-tidy.
+
+import json
+import shlex
+
+
+def process_compdb(compdb: list[dict]) -> list[dict]:
+
+    def munch_command(args: list[str]) -> list[str]:
+        out = []
+        eat_next = False
+        for i, arg in enumerate(args):
+            if arg == '-fpch-preprocess':
+                # as used with gcc
+                continue
+            elif arg == '-include-pch' or (arg == '-include' and args[i + 1] == 'precompiled-headers.hh'):
+                # -include-pch some-pch (clang), or -include some-pch (gcc)
+                eat_next = True
+                continue
+            if not eat_next:
+                out.append(arg)
+            eat_next = False
+        return out
+
+    def chomp(item: dict) -> dict:
+        item = item.copy()
+        item['command'] = shlex.join(munch_command(shlex.split(item['command'])))
+        return item
+
+    return [chomp(x) for x in compdb if not x['file'].endswith('precompiled-headers.hh')]
+
+
+def main():
+    import argparse
+    ap = argparse.ArgumentParser(
+        description='Delete pch arguments from compilation database')
+    ap.add_argument('input',
+                    type=argparse.FileType('r'),
+                    help='Input json file')
+    ap.add_argument('output',
+                    type=argparse.FileType('w'),
+                    help='Output json file')
+    args = ap.parse_args()
+
+    input_json = json.load(args.input)
+    json.dump(process_compdb(input_json), args.output, indent=2)
+
+
+if __name__ == '__main__':
+    main()

meson/clang-tidy/meson.build

@@ -0,0 +1,98 @@
+# The clang-tidy target for Lix
+
+run_clang_tidy = find_program('run-clang-tidy', required : false)
+# Although this looks like it wants to be pkg-config, pkg-config does not
+# really work for *plugins*, which are executable-like .so files that also
+# cannot be found via find_program. Fun!
+if get_option('lix-clang-tidy-checks-path') != ''
+  lix_clang_tidy_so = get_option('lix-clang-tidy-checks-path')
+  lix_clang_tidy_so_found = true
+else
+  lix_clang_tidy_subproj = subproject(
+    'lix-clang-tidy',
+    required : false,
+    default_options : {'build-by-default': false}
+  )
+  if lix_clang_tidy_subproj.found()
+    lix_clang_tidy_so = lix_clang_tidy_subproj.get_variable('lix_clang_tidy')
+    lix_clang_tidy_so_found = true
+  else
+    lix_clang_tidy_so_found = false
+  endif
+endif
+
+# Due to numerous problems, such as:
+# - Meson does not expose pch targets, but *fine*, I can just ask Ninja for
+#   them with `ninja -t targets rule cpp_PCH` and build them manually:
+#   https://github.com/mesonbuild/meson/issues/13499
+# - Nixpkgs stdenv buries the cc-wrapper under a giant pile of assumptions
+#   about the cc-wrapper actually being used on the cc of a stdenv, rather than
+#   independently for clang-tidy, and we need to use cc-wrapper to get the
+#   correct hardening flags so that clang-tidy can actually parse the PCH file
+#
+# I give up. I am going to delete the damn PCH args and then it will work.
+meson.add_postconf_script(
+  python,
+  meson.current_source_dir() / 'clean_compdb.py',
+  meson.global_build_root() / 'compile_commands.json',
+  meson.current_build_dir() / 'compile_commands.json',
+)
+
+# Horrible hack to get around not being able to depend on another target's
+# generated headers in any way in the meson DSL
+# https://github.com/mesonbuild/meson/issues/12817 which was incorrectly
+# closed, if you *actually* need to generate the files once.
+# Also related: https://github.com/mesonbuild/meson/issues/3667
+#
+# Or we could ban meson generators because their design is broken.
+build_all_generated_headers = custom_target(
+  command : [
+    python,
+    meson.current_source_dir() / 'build_required_targets.py',
+    meson.global_build_root(),
+  ],
+  output : 'generated_headers.stamp',
+  build_by_default : false,
+  build_always_stale : true,
+)
+
+if lix_clang_tidy_so_found
+  run_clang_tidy_args = [
+    '-load',
+    lix_clang_tidy_so,
+    '-p',
+    # We have to workaround a run-clang-tidy bug too, so we must give the
+    # directory name rather than the actual compdb file.
+    # https://github.com/llvm/llvm-project/issues/101440
+    meson.current_build_dir(),
+    '-quiet',
+  ]
+  run_target(
+    'clang-tidy',
+    command : [
+      # XXX: This explicitly invokes it with python because of a nixpkgs bug
+      # where clang-unwrapped does not patch interpreters in run-clang-tidy.
+      # However, making clang-unwrapped depend on python is also silly, so idk.
+      python,
+      run_clang_tidy,
+      run_clang_tidy_args,
+      '-warnings-as-errors',
+      '*',
+    ],
+    depends : [
+      build_all_generated_headers,
+    ],
+  )
+  run_target(
+    'clang-tidy-fix',
+    command : [
+      python,
+      run_clang_tidy,
+      run_clang_tidy_args,
+      '-fix',
+    ],
+    depends : [
+      build_all_generated_headers,
+    ],
+  )
+endif

misc/meson.build

@@ -4,3 +4,9 @@ subdir('zsh')
 
 subdir('systemd')
 subdir('flake-registry')
+
+runinpty = configure_file(
+  copy : true,
+  input : meson.current_source_dir() / 'runinpty.py',
+  output : 'runinpty.py',
+)

misc/passt.nix

@@ -0,0 +1,60 @@
+{
+  lib,
+  stdenv,
+  buildPackages,
+  fetchurl,
+  getconf,
+  gitUpdater,
+  testers,
+}:
+
+stdenv.mkDerivation (finalAttrs: {
+  pname = "passt";
+  version = "2025_02_17.a1e48a0";
+
+  src = fetchurl {
+    url = "https://passt.top/passt/snapshot/passt-${finalAttrs.version}.tar.gz";
+    hash = "sha256-/FUXxeYv3Lb0DiXmbS2PUzfLL5ZwHJ42tiuH7YnlljE=";
+  };
+
+  postPatch = ''
+    substituteInPlace Makefile --replace-fail \
+      'PAGE_SIZE=$(shell getconf PAGE_SIZE)' \
+      "PAGE_SIZE=$(${stdenv.hostPlatform.emulator buildPackages} ${lib.getExe getconf} PAGE_SIZE)"
+  '';
+
+  makeFlags = [
+    "prefix=${placeholder "out"}"
+    "VERSION=${finalAttrs.version}"
+  ];
+
+  passthru = {
+    tests.version = testers.testVersion { package = finalAttrs.finalPackage; };
+
+    updateScript = gitUpdater { url = "https://passt.top/passt"; };
+  };
+
+  meta = with lib; {
+    homepage = "https://passt.top/passt/about/";
+    description = "Plug A Simple Socket Transport";
+    longDescription = ''
+      passt implements a translation layer between a Layer-2 network interface
+      and native Layer-4 sockets (TCP, UDP, ICMP/ICMPv6 echo) on a host.
+      It doesn't require any capabilities or privileges, and it can be used as
+      a simple replacement for Slirp.
+
+      pasta (same binary as passt, different command) offers equivalent
+      functionality, for network namespaces: traffic is forwarded using a tap
+      interface inside the namespace, without the need to create further
+      interfaces on the host, hence not requiring any capabilities or
+      privileges.
+    '';
+    license = [
+      licenses.bsd3 # and
+      licenses.gpl2Plus
+    ];
+    platforms = platforms.linux;
+    maintainers = with maintainers; [ _8aed ];
+    mainProgram = "passt";
+  };
+})

misc/pegtl.nix

@@ -1,23 +0,0 @@
-{
-  stdenv,
-  cmake,
-  ninja,
-  fetchFromGitHub,
-}:
-
-stdenv.mkDerivation {
-  pname = "pegtl";
-  version = "3.2.7";
-
-  src = fetchFromGitHub {
-    repo = "PEGTL";
-    owner = "taocpp";
-    rev = "refs/tags/3.2.7";
-    hash = "sha256-IV5YNGE4EWVrmg2Sia/rcU8jCuiBynQGJM6n3DCWTQU=";
-  };
-
-  nativeBuildInputs = [
-    cmake
-    ninja
-  ];
-}

misc/pre-commit.nix

@@ -106,7 +106,7 @@ pre-commit-run {
     };
     treefmt = {
       enable = true;
-      settings.formatters = [ pkgs.nixfmt ];
+      settings.formatters = [ pkgs.nixfmt-rfc-style ];
     };
   };
 }

misc/runinpty.py

@@ -0,0 +1,77 @@
+#!/usr/bin/env python3
+# SPDX-FileCopyrightText: 2001, 2002, 2003, 2004, 2005, 2006 Python Software Foundation; All Rights Reserved
+# SPDX-FileCopyrightText: 2024 Jade Lovelace
+# SPDX-License-Identifier: LGPL-2.1-or-later
+"""
+This script exists to lose Lix a dependency on expect(1) for the ability to run
+something in a pty.
+
+Yes, it could be replaced by script(1) but macOS and Linux script(1) have
+diverged sufficiently badly that even specifying a subcommand to run is not the
+same.
+"""
+import pty
+import sys
+import os
+from termios import ONLCR, ONLRET, ONOCR, OPOST, TCSAFLUSH, tcgetattr, tcsetattr
+from tty import setraw
+import termios
+
+def setup_terminal():
+    # does not matter which fd we use because we are in a fresh pty
+    modi = tcgetattr(pty.STDOUT_FILENO)
+    [iflag, oflag, cflag, lflag, ispeed, ospeed, cc] = modi
+
+    # Turning \n into \r\n is not cool, Linux!
+    oflag &= ~ONLCR
+    # I don't know what "implementation dependent postprocessing means" but it
+    # sounds bad
+    oflag &= ~OPOST
+    # Assume that NL performs the role of CR; do not insert CRs at column 0
+    oflag |= ONLRET | ONOCR
+
+    modi = [iflag, oflag, cflag, lflag, ispeed, ospeed, cc]
+
+    tcsetattr(pty.STDOUT_FILENO, TCSAFLUSH, modi)
+
+
+def spawn(argv: list[str]):
+    """
+    As opposed to pty.spawn, this one more seriously controls the pty settings.
+    Necessary to turn off such fun functionality as onlcr (LF to CRLF).
+
+    This is essentially copy pasted from pty.spawn, since there is no way to
+    hook the child pre-execve
+    """
+    pid, master_fd = pty.fork()
+    if pid == pty.CHILD:
+        setup_terminal()
+        os.execlp(argv[0], *argv)
+
+    try:
+        mode = tcgetattr(pty.STDIN_FILENO)
+        setraw(pty.STDIN_FILENO)
+        restore = True
+    except termios.error:
+        restore = False
+
+    try:
+        pty._copy(master_fd, pty._read, pty._read) # type: ignore
+    finally:
+        if restore:
+            tcsetattr(pty.STDIN_FILENO, TCSAFLUSH, mode) # type: ignore
+
+    os.close(master_fd)
+    return os.waitpid(pid, 0)[1]
+
+
+def main():
+    if len(sys.argv) == 1:
+        print(f'Usage: {sys.argv[0]} [command args]', file=sys.stderr)
+        sys.exit(1)
+
+    sys.exit(os.waitstatus_to_exitcode(spawn(sys.argv[1:])))
+
+
+if __name__ == '__main__':
+    main()

misc/systemd/nix-daemon.conf.in

@@ -1 +1,2 @@
-d @localstatedir@/nix/daemon-socket 0755 root root - -
+d @localstatedir@/nix/daemon-socket 0755 root root -  -
+d @localstatedir@/nix/builds        0755 root root 7d -

misc/systemd/nix-daemon.service.in

@@ -1,6 +1,6 @@
 [Unit]
 Description=Nix Daemon
-Documentation=man:nix-daemon https://nixos.org/manual
+Documentation=man:nix-daemon https://docs.lix.systems/manual/lix/stable
 RequiresMountsFor=@storedir@
 RequiresMountsFor=@localstatedir@
 RequiresMountsFor=@localstatedir@/nix/db

nix-support/binary-tarball.nix

@@ -5,29 +5,33 @@
   system,
 }:
 let
-  installerClosureInfo = buildPackages.closureInfo {
-    rootPaths = [
-      nix
-      cacert
-    ];
-  };
+  rootPaths = [
+    nix
+    cacert
+  ];
+  installerClosureInfo = buildPackages.closureInfo { inherit rootPaths; };
 
   meta.description = "Distribution-independent Lix bootstrap binaries for ${system}";
 in
-buildPackages.runCommand "lix-binary-tarball-${nix.version}" { inherit meta; } ''
-  cp ${installerClosureInfo}/registration $TMPDIR/reginfo
+buildPackages.runCommand "lix-binary-tarball-${nix.version}"
+  {
+    inherit meta;
+    passthru.rootPaths = rootPaths;
+  }
+  ''
+    cp ${installerClosureInfo}/registration $TMPDIR/reginfo
 
-  dir=lix-${nix.version}-${system}
-  fn=$out/$dir.tar.xz
-  mkdir -p $out/nix-support
-  echo "file binary-dist $fn" >> $out/nix-support/hydra-build-products
-  tar cvfJ $fn \
-    --owner=0 --group=0 --mode=u+rw,uga+r \
-    --mtime='1970-01-01' \
-    --absolute-names \
-    --hard-dereference \
-    --transform "s,$TMPDIR/reginfo,$dir/.reginfo," \
-    --transform "s,$NIX_STORE,$dir/store,S" \
-    $TMPDIR/reginfo \
-    $(cat ${installerClosureInfo}/store-paths)
-''
+    dir=lix-${nix.version}-${system}
+    fn=$out/$dir.tar.xz
+    mkdir -p $out/nix-support
+    echo "file binary-dist $fn" >> $out/nix-support/hydra-build-products
+    tar cvfJ $fn \
+      --owner=0 --group=0 --mode=u+rw,uga+r \
+      --mtime='1970-01-01' \
+      --absolute-names \
+      --hard-dereference \
+      --transform "s,$TMPDIR/reginfo,$dir/.reginfo," \
+      --transform "s,$NIX_STORE,$dir/store,S" \
+      $TMPDIR/reginfo \
+      $(cat ${installerClosureInfo}/store-paths)
+  ''

nix-support/nixfmt.nix

@@ -1,65 +0,0 @@
-# Copy of `nixfmt-rfc-style` vendored from `nixpkgs` master:
-# https://github.com/NixOS/nixpkgs/blob/ab6071eb54cc9b66dda436111d4f569e4e56cbf4/pkgs/by-name/ni/nixfmt-rfc-style/package.nix
-{
-  haskell,
-  haskellPackages,
-  fetchFromGitHub,
-}:
-let
-  inherit (haskell.lib.compose) justStaticExecutables;
-  raw-pkg = haskellPackages.callPackage (
-    {
-      mkDerivation,
-      base,
-      cmdargs,
-      directory,
-      fetchzip,
-      filepath,
-      lib,
-      megaparsec,
-      mtl,
-      parser-combinators,
-      safe-exceptions,
-      scientific,
-      text,
-      transformers,
-      unix,
-    }:
-    mkDerivation {
-      pname = "nixfmt";
-      version = "0.6.0-unstable-2024-03-14";
-      src = fetchFromGitHub {
-        owner = "serokell";
-        repo = "nixfmt";
-        rev = "8d13b593fa8d8d6e5075f541f3231222a08e84df";
-        hash = "sha256-HtXvzmfN4wk45qiKZ7V+/5WBV7jnTHfd7iBwF4XGl64=";
-      };
-      isLibrary = true;
-      isExecutable = true;
-      libraryHaskellDepends = [
-        base
-        megaparsec
-        mtl
-        parser-combinators
-        scientific
-        text
-        transformers
-      ];
-      executableHaskellDepends = [
-        base
-        cmdargs
-        directory
-        filepath
-        safe-exceptions
-        text
-        unix
-      ];
-      jailbreak = true;
-      homepage = "https://github.com/serokell/nixfmt";
-      description = "An opinionated formatter for Nix";
-      license = lib.licenses.mpl20;
-      mainProgram = "nixfmt";
-    }
-  ) { };
-in
-justStaticExecutables raw-pkg

package.nix

@@ -27,7 +27,11 @@
   libcpuid,
   libseccomp,
   libsodium,
+  lix-clang-tidy ? null,
+  llvmPackages,
   lsof,
+  # FIXME: remove default after dropping NixOS 24.05
+  lowdown-unsandboxed ? lowdown,
   lowdown,
   mdbook,
   mdbook-linkcheck,
@@ -35,7 +39,9 @@
   meson,
   ninja,
   openssl,
-  pegtl ? __forDefaults.pegtl,
+  # FIXME: we need passt 2024_12_11.09478d5 or newer, i.e. nixos 25.05 or later
+  passt-lix ? __forDefaults.passt-lix,
+  pegtl,
   pkg-config,
   python3,
   rapidcheck,
@@ -52,25 +58,30 @@
 
   pname ? "lix",
   versionSuffix ? "",
-  officialRelease ? false,
+  officialRelease ? __forDefaults.versionJson.official_release,
   # Set to true to build the release notes for the next release.
   buildUnreleasedNotes ? true,
   internalApiDocs ? false,
 
+  # Support garbage collection in the evaluator.
+  enableGC ? sanitize == null || !builtins.elem "address" sanitize,
+  # List of Meson sanitize options. Accepts values of b_sanitize, e.g.
+  # "address", "undefined", "thread".
+  # Enabling the "address" sanitizer will disable garbage collection in the evaluator.
+  sanitize ? null,
+  # Turn compiler warnings into errors.
+  werror ? false,
+
+  lintInsteadOfBuild ? false,
+
   # Not a real argument, just the only way to approximate let-binding some
   # stuff for argument defaults.
   __forDefaults ? {
     canRunInstalled = stdenv.buildPlatform.canExecute stdenv.hostPlatform;
 
-    boehmgc-nix = (boehmgc.override { enableLargeConfig = true; }).overrideAttrs {
-      patches = [
-        # We do *not* include prev.patches (which doesn't exist in normal pkgs.boehmgc anyway)
-        # because if the caller of this package passed a patched boehm as `boehmgc` instead of
-        # `boehmgc-nix` then this will almost certainly have duplicate patches, which means
-        # the patches won't apply and we'll get a build failure.
-        ./boehmgc-coroutine-sp-fallback.diff
-      ];
-    };
+    versionJson = builtins.fromJSON (builtins.readFile ./version.json);
+
+    boehmgc-nix = boehmgc.override { enableLargeConfig = true; };
 
     editline-lix = editline.overrideAttrs (prev: {
       configureFlags = prev.configureFlags or [ ] ++ [ (lib.enableFeature true "sigstop") ];
@@ -79,7 +90,7 @@
     lix-doc = callPackage ./lix-doc/package.nix { };
     build-release-notes = callPackage ./maintainers/build-release-notes.nix { };
 
-    pegtl = callPackage ./misc/pegtl.nix { };
+    passt-lix = callPackage ./misc/passt.nix { };
   },
 }:
 let
@@ -87,51 +98,39 @@ let
   inherit (lib) fileset;
   inherit (stdenv) hostPlatform buildPlatform;
 
-  versionJson = builtins.fromJSON (builtins.readFile ./version.json);
-  version = versionJson.version + versionSuffix;
+  version = __forDefaults.versionJson.version + versionSuffix;
 
-  aws-sdk-cpp-nix = aws-sdk-cpp.override {
-    apis = [
-      "s3"
-      "transfer"
-    ];
-    customMemoryManagement = false;
-  };
+  aws-sdk-cpp-nix =
+    if aws-sdk-cpp == null then
+      null
+    else
+      aws-sdk-cpp.override {
+        apis = [
+          "s3"
+          "transfer"
+        ];
+        customMemoryManagement = false;
+      };
 
   # Reimplementation of Nixpkgs' Meson cross file, with some additions to make
   # it actually work.
-  mesonCrossFile =
-    let
-      cpuFamily =
-        platform:
-        with platform;
-        if isAarch32 then
-          "arm"
-        else if isx86_32 then
-          "x86"
-        else
-          platform.uname.processor;
-    in
-    builtins.toFile "lix-cross-file.conf" ''
-      [properties]
-      # Meson is convinced that if !buildPlatform.canExecute hostPlatform then we cannot
-      # build anything at all, which is not at all correct. If we can't execute the host
-      # platform, we'll just disable tests and doc gen.
-      needs_exe_wrapper = false
+  mesonCrossFile = builtins.toFile "lix-cross-file.conf" ''
+    [properties]
+    # Meson is convinced that if !buildPlatform.canExecute hostPlatform then we cannot
+    # build anything at all, which is not at all correct. If we can't execute the host
+    # platform, we'll just disable tests and doc gen.
+    needs_exe_wrapper = false
 
-      [binaries]
-      # Meson refuses to consider any CMake binary during cross compilation if it's
-      # not explicitly specified here, in the cross file.
-      # https://github.com/mesonbuild/meson/blob/0ed78cf6fa6d87c0738f67ae43525e661b50a8a2/mesonbuild/cmake/executor.py#L72
-      cmake = 'cmake'
-    '';
+    [binaries]
+    # Meson refuses to consider any CMake binary during cross compilation if it's
+    # not explicitly specified here, in the cross file.
+    # https://github.com/mesonbuild/meson/blob/0ed78cf6fa6d87c0738f67ae43525e661b50a8a2/mesonbuild/cmake/executor.py#L72
+    cmake = 'cmake'
+  '';
 
   # The internal API docs need these for the build, but if we're not building
   # Nix itself, then these don't need to be propagated.
-  maybePropagatedInputs = [
-    boehmgc-nix
-    nlohmann_json
-  ];
+  maybePropagatedInputs = lib.optional enableGC boehmgc-nix ++ [ nlohmann_json ];
 
   # .gitignore has already been processed, so any changes in it are irrelevant
   # at this point. It is not represented verbatim for test purposes because
@@ -154,6 +153,7 @@ let
     (fileset.fileFilter (f: lib.strings.hasPrefix "nix-profile" f.name) ./scripts)
   ];
 in
+assert (lintInsteadOfBuild -> lix-clang-tidy != null);
 stdenv.mkDerivation (finalAttrs: {
   inherit pname version;
 
@@ -166,13 +166,13 @@ stdenv.mkDerivation (finalAttrs: {
           topLevelBuildFiles
           functionalTestFiles
         ]
-        ++ lib.optionals (!finalAttrs.dontBuild || internalApiDocs) [
-          ./boehmgc-coroutine-sp-fallback.diff
+        ++ lib.optionals (!finalAttrs.dontBuild || internalApiDocs || lintInsteadOfBuild) [
           ./doc
           ./misc
           ./src
           ./COPYING
         ]
+        ++ lib.optionals lintInsteadOfBuild [ ./.clang-tidy ]
       )
     );
   };
@@ -186,26 +186,36 @@ stdenv.mkDerivation (finalAttrs: {
       "doc"
     ];
 
-  dontBuild = false;
+  dontBuild = lintInsteadOfBuild;
 
   mesonFlags =
+    let
+      sanitizeOpts = lib.optional (
+        sanitize != null
+      ) "-Db_sanitize=${builtins.concatStringsSep "," sanitize}";
+    in
     lib.optionals hostPlatform.isLinux [
       # You'd think meson could just find this in PATH, but busybox is in buildInputs,
       # which don't actually get added to PATH. And buildInputs is correct over
       # nativeBuildInputs since this should be a busybox executable on the host.
       "-Dsandbox-shell=${lib.getExe' busybox-sandbox-shell "busybox"}"
+      "-Dpasta-path=${lib.getExe' passt-lix "pasta"}"
     ]
     ++ lib.optional hostPlatform.isStatic "-Denable-embedded-sandbox-shell=true"
-    ++ lib.optional (finalAttrs.dontBuild) "-Denable-build=false"
+    ++ lib.optional (finalAttrs.dontBuild && !lintInsteadOfBuild) "-Denable-build=false"
+    ++ lib.optional lintInsteadOfBuild "-Dlix-clang-tidy-checks-path=${lix-clang-tidy}/lib/liblix-clang-tidy.so"
     ++ [
       # mesonConfigurePhase automatically passes -Dauto_features=enabled,
       # so we must explicitly enable or disable features that we are not passing
       # dependencies for.
+      (lib.mesonEnable "gc" enableGC)
       (lib.mesonEnable "internal-api-docs" internalApiDocs)
-      (lib.mesonBool "enable-tests" finalAttrs.finalPackage.doCheck)
+      (lib.mesonBool "enable-tests" (finalAttrs.finalPackage.doCheck || lintInsteadOfBuild))
       (lib.mesonBool "enable-docs" canRunInstalled)
+      (lib.mesonBool "werror" werror)
     ]
-    ++ lib.optional (hostPlatform != buildPlatform) "--cross-file=${mesonCrossFile}";
+    ++ lib.optional (hostPlatform != buildPlatform) "--cross-file=${mesonCrossFile}"
+    ++ sanitizeOpts;
 
   # We only include CMake so that Meson can locate toml11, which only ships CMake dependency metadata.
   dontUseCmakeConfigure = true;
@@ -218,7 +228,7 @@ stdenv.mkDerivation (finalAttrs: {
       cmake
     ]
     ++ [
-      (lib.getBin lowdown)
+      (lib.getBin lowdown-unsandboxed)
       mdbook
       mdbook-linkcheck
     ]
@@ -233,7 +243,13 @@ stdenv.mkDerivation (finalAttrs: {
     ]
     ++ lib.optional hostPlatform.isLinux util-linuxMinimal
     ++ lib.optional (!officialRelease && buildUnreleasedNotes) build-release-notes
-    ++ lib.optional internalApiDocs doxygen;
+    ++ lib.optional internalApiDocs doxygen
+    ++ lib.optionals lintInsteadOfBuild [
+      # required for a wrapped clang-tidy
+      llvmPackages.clang-tools
+      # required for run-clang-tidy
+      llvmPackages.clang-unwrapped
+    ];
 
   buildInputs =
     [
@@ -255,12 +271,16 @@ stdenv.mkDerivation (finalAttrs: {
     ++ lib.optionals hostPlatform.isLinux [
       libseccomp
       busybox-sandbox-shell
+      passt-lix
     ]
     ++ lib.optional internalApiDocs rapidcheck
     ++ lib.optional hostPlatform.isx86_64 libcpuid
     # There have been issues building these dependencies
     ++ lib.optional (hostPlatform.canExecute buildPlatform) aws-sdk-cpp-nix
-    ++ lib.optionals (finalAttrs.dontBuild) maybePropagatedInputs;
+    ++ lib.optionals (finalAttrs.dontBuild) maybePropagatedInputs
+    # I am so sorry. This is because checkInputs are required to pass
+    # configure, but we don't actually want to *run* the checks here.
+    ++ lib.optionals lintInsteadOfBuild finalAttrs.checkInputs;
 
   checkInputs = [
     gtest
@@ -299,13 +319,6 @@ stdenv.mkDerivation (finalAttrs: {
       install_name_tool -change ${boost}/lib/libboost_system.dylib $out/lib/libboost_system.dylib $out/lib/libboost_thread.dylib
     ''
     + ''
-      # Workaround https://github.com/NixOS/nixpkgs/issues/294890.
-      if [[ -n "''${doCheck:-}" ]]; then
-        appendToVar configureFlags "--enable-tests"
-      else
-        appendToVar configureFlags "--disable-tests"
-      fi
-
       # Fix up /usr/bin/env shebangs relied on by the build
       patchShebangs --build tests/ doc/manual/
     '';
@@ -316,7 +329,7 @@ stdenv.mkDerivation (finalAttrs: {
 
   enableParallelBuilding = true;
 
-  doCheck = canRunInstalled;
+  doCheck = canRunInstalled && !lintInsteadOfBuild;
 
   mesonCheckFlags = [
     "--suite=check"
@@ -328,8 +341,19 @@ stdenv.mkDerivation (finalAttrs: {
   # Make sure the internal API docs are already built, because mesonInstallPhase
   # won't let us build them there. They would normally be built in buildPhase,
   # but the internal API docs are conventionally built with doBuild = false.
-  preInstall = lib.optional internalApiDocs ''
-    meson ''${mesonBuildFlags:-} compile "$installTargets"
+  preInstall =
+    (lib.optionalString internalApiDocs ''
+      meson ''${mesonBuildFlags:-} compile "$installTargets"
+    '')
+    # evil, but like above, we do not want to run an actual build phase
+    + lib.optionalString lintInsteadOfBuild ''
+      ninja clang-tidy
+    '';
+
+  installPhase = lib.optionalString lintInsteadOfBuild ''
+    runHook preInstall
+    touch $out
+    runHook postInstall
   '';
 
   postInstall =
@@ -390,8 +414,6 @@ stdenv.mkDerivation (finalAttrs: {
       pegtl
       ;
 
-    inherit officialRelease;
-
     # The collection of dependency logic for this derivation is complicated enough that
     # it's easier to parameterize the devShell off an already called package.nix.
     mkDevShell =
@@ -399,19 +421,18 @@ stdenv.mkDerivation (finalAttrs: {
         mkShell,
 
         bashInteractive,
-        clang-tools,
         clangbuildanalyzer,
         doxygen,
         glibcLocales,
         just,
-        llvmPackages,
-        nixfmt,
+        nixfmt-rfc-style,
         skopeo,
         xonsh,
 
         # Lix specific packages
         pre-commit-checks,
         contribNotice,
+        check-syscalls,
       }:
       let
         glibcFix = lib.optionalAttrs (buildPlatform.isLinux && glibcLocales != null) {
@@ -425,6 +446,7 @@ stdenv.mkDerivation (finalAttrs: {
             p.python-frontmatter
             p.requests
             p.xdg-base-dirs
+            p.packaging
             (p.toPythonModule xonsh.passthru.unwrapped)
           ]
         );
@@ -456,17 +478,18 @@ stdenv.mkDerivation (finalAttrs: {
             ++ [ (lib.mesonBool "enable-pch-std" stdenv.cc.isClang) ];
 
           packages =
-            lib.optional (stdenv.cc.isClang && hostPlatform == buildPlatform) clang-tools
+            lib.optional (stdenv.cc.isClang && hostPlatform == buildPlatform) llvmPackages.clang-tools
             ++ [
               # Why are we providing a bashInteractive? Well, when you run
               # `bash` from inside `nix develop`, say, because you are using it
               # via direnv, you will by default get bash (unusable edition).
               bashInteractive
+              check-syscalls
               pythonEnv
               # docker image tool
               skopeo
               just
-              nixfmt
+              nixfmt-rfc-style
               # Included above when internalApiDocs is true, but we set that to
               # false intentionally to save dev build time.
               # To build them in a dev shell, you can set -Dinternal-api-docs=enabled when configuring.
@@ -474,6 +497,14 @@ stdenv.mkDerivation (finalAttrs: {
               # Load-bearing order. Must come before clang-unwrapped below, but after clang_tools above.
               stdenv.cc
             ]
+            ++ [
+              pkgs.rust-analyzer
+              pkgs.cargo
+              pkgs.rustc
+              pkgs.rustfmt
+              pkgs.rustPlatform.rustLibSrc
+              pkgs.rustPlatform.rustcSrc
+            ]
             ++ lib.optionals stdenv.cc.isClang [
               # Required for clang-tidy checks.
               llvmPackages.llvm
@@ -493,9 +524,9 @@ stdenv.mkDerivation (finalAttrs: {
                 return
               fi
 
-              PATH=$prefix/bin:$PATH
+              PATH=$prefix/bin''${PATH:+:''${PATH}}
               unset PYTHONPATH
-              export MANPATH=$out/share/man:$MANPATH
+              export MANPATH=$out/share/man:''${MANPATH:-}
 
               # Make bash completion work.
               XDG_DATA_DIRS+=:$out/share

perl/configure.ac

@@ -1,84 +0,0 @@
-AC_INIT(nix-perl, m4_esyscmd([bash -c "echo -n $(cat ../.version)$VERSION_SUFFIX"]))
-AC_CONFIG_SRCDIR(MANIFEST)
-AC_CONFIG_AUX_DIR(../config)
-
-CFLAGS=
-CXXFLAGS=
-AC_PROG_CC
-AC_PROG_CXX
-
-AC_CANONICAL_HOST
-
-# Use 64-bit file system calls so that we can support files > 2 GiB.
-AC_SYS_LARGEFILE
-
-AC_DEFUN([NEED_PROG],
-[
-AC_PATH_PROG($1, $2)
-if test -z "$$1"; then
-    AC_MSG_ERROR([$2 is required])
-fi
-])
-
-NEED_PROG(perl, perl)
-NEED_PROG(curl, curl)
-NEED_PROG(bzip2, bzip2)
-NEED_PROG(xz, xz)
-
-# Test that Perl has the open/fork feature (Perl 5.8.0 and beyond).
-AC_MSG_CHECKING([whether Perl is recent enough])
-if ! $perl -e 'open(FOO, "-|", "true"); while (<FOO>) { print; }; close FOO or die;'; then
-    AC_MSG_RESULT(no)
-    AC_MSG_ERROR([Your Perl version is too old.  Lix requires Perl 5.8.0 or newer.])
-fi
-AC_MSG_RESULT(yes)
-
-
-# Figure out where to install Perl modules.
-AC_MSG_CHECKING([for the Perl installation prefix])
-perlversion=$($perl -e 'use Config; print $Config{version};')
-perlarchname=$($perl -e 'use Config; print $Config{archname};')
-AC_SUBST(perllibdir, [${libdir}/perl5/site_perl/$perlversion/$perlarchname])
-AC_MSG_RESULT($perllibdir)
-
-# Look for libsodium.
-PKG_CHECK_MODULES([SODIUM], [libsodium], [CXXFLAGS="$SODIUM_CFLAGS $CXXFLAGS"])
-
-# Check for the required Perl dependencies (DBI and DBD::SQLite).
-perlFlags="-I$perllibdir"
-
-AC_ARG_WITH(dbi, AC_HELP_STRING([--with-dbi=PATH],
-  [prefix of the Perl DBI library]),
-  perlFlags="$perlFlags -I$withval")
-
-AC_ARG_WITH(dbd-sqlite, AC_HELP_STRING([--with-dbd-sqlite=PATH],
-  [prefix of the Perl DBD::SQLite library]),
-  perlFlags="$perlFlags -I$withval")
-
-AC_MSG_CHECKING([whether DBD::SQLite works])
-if ! $perl $perlFlags -e 'use DBI; use DBD::SQLite;' 2>&5; then
-    AC_MSG_RESULT(no)
-    AC_MSG_FAILURE([The Perl modules DBI and/or DBD::SQLite are missing.])
-fi
-AC_MSG_RESULT(yes)
-
-AC_SUBST(perlFlags)
-
-PKG_CHECK_MODULES([NIX], [nix-store])
-
-NEED_PROG([NIX], [nix])
-
-# Expand all variables in config.status.
-test "$prefix" = NONE && prefix=$ac_default_prefix
-test "$exec_prefix" = NONE && exec_prefix='${prefix}'
-for name in $ac_subst_vars; do
-    declare $name="$(eval echo "${!name}")"
-    declare $name="$(eval echo "${!name}")"
-    declare $name="$(eval echo "${!name}")"
-done
-
-rm -f Makefile.config
-ln -sfn ../mk mk
-
-AC_CONFIG_FILES([])
-AC_OUTPUT

perl/lib/Nix/Store.xs

@@ -77,7 +77,7 @@ SV * queryReferences(char * path)
 SV * queryPathHash(char * path)
     PPCODE:
         try {
-            auto s = store()->queryPathInfo(store()->parseStorePath(path))->narHash.to_string(Base32, true);
+            auto s = store()->queryPathInfo(store()->parseStorePath(path))->narHash.to_string(Base::Base32, true);
             XPUSHs(sv_2mortal(newSVpv(s.c_str(), 0)));
         } catch (Error & e) {
             croak("%s", e.what());
@@ -103,7 +103,7 @@ SV * queryPathInfo(char * path, int base32)
                 XPUSHs(&PL_sv_undef);
             else
                 XPUSHs(sv_2mortal(newSVpv(store()->printStorePath(*info->deriver).c_str(), 0)));
-            auto s = info->narHash.to_string(base32 ? Base32 : Base16, true);
+            auto s = info->narHash.to_string(base32 ? Base::Base32 : Base::Base16, true);
             XPUSHs(sv_2mortal(newSVpv(s.c_str(), 0)));
             mXPUSHi(info->registrationTime);
             mXPUSHi(info->narSize);
@@ -205,7 +205,7 @@ SV * hashPath(char * algo, int base32, char * path)
     PPCODE:
         try {
             Hash h = hashPath(parseHashType(algo), path).first;
-            auto s = h.to_string(base32 ? Base32 : Base16, false);
+            auto s = h.to_string(base32 ? Base::Base32 : Base::Base16, false);
             XPUSHs(sv_2mortal(newSVpv(s.c_str(), 0)));
         } catch (Error & e) {
             croak("%s", e.what());
@@ -216,7 +216,7 @@ SV * hashFile(char * algo, int base32, char * path)
     PPCODE:
         try {
             Hash h = hashFile(parseHashType(algo), path);
-            auto s = h.to_string(base32 ? Base32 : Base16, false);
+            auto s = h.to_string(base32 ? Base::Base32 : Base::Base16, false);
             XPUSHs(sv_2mortal(newSVpv(s.c_str(), 0)));
         } catch (Error & e) {
             croak("%s", e.what());
@@ -227,7 +227,7 @@ SV * hashString(char * algo, int base32, char * s)
     PPCODE:
         try {
             Hash h = hashString(parseHashType(algo), s);
-            auto s = h.to_string(base32 ? Base32 : Base16, false);
+            auto s = h.to_string(base32 ? Base::Base32 : Base::Base16, false);
             XPUSHs(sv_2mortal(newSVpv(s.c_str(), 0)));
         } catch (Error & e) {
             croak("%s", e.what());
@@ -238,7 +238,7 @@ SV * convertHash(char * algo, char * s, int toBase32)
     PPCODE:
         try {
             auto h = Hash::parseAny(s, parseHashType(algo));
-            auto s = h.to_string(toBase32 ? Base32 : Base16, false);
+            auto s = h.to_string(toBase32 ? Base::Base32 : Base::Base16, false);
             XPUSHs(sv_2mortal(newSVpv(s.c_str(), 0)));
         } catch (Error & e) {
             croak("%s", e.what());

releng/README.md

@@ -30,7 +30,7 @@ First, we prepare the release. `python -m releng prepare` is used for this.
 Then we tag the release with `python -m releng tag`:
 
 * Git HEAD is detached.
-* `officialRelease = true` is set in `flake.nix`, this is committed, and a
+* `"official_release": true` is set in `version.json`, this is committed, and a
   release is tagged.
 * The tag is merged back into the last branch (either `main` for new releases
   or `release-MAJOR` for maintenance releases) with `git merge -s ours VERSION`
@@ -57,12 +57,10 @@ Next, we do the publication with `python -m releng upload`:
     `nix upgrade-nix`.
   * s3://releases/lix/lix-VERSION/ gets the following contents
     * Binary tarballs
-    * Docs: `manual/` (FIXME: should we actually do this? what about putting it
-      on docs.lix.systems? I think doing both is correct, since the Web site
-      should not be an archive of random old manuals)
+    * Docs: `manual/`, primarily as an archive of old manuals
     * Docs as tarball in addition to web.
     * Source tarball
-    * Docker image (FIXME: upload to forgejo registry and github registry [in the future][upload-docker])
+    * Docker image
   * s3://docs/manual/lix/MAJOR
   * s3://docs/manual/lix/stable
 
@@ -80,6 +78,7 @@ Next, we do the publication with `python -m releng upload`:
   FIXME: automate branch-off to `release-*` branch.
 * **Manually** (FIXME?) switch back to the release branch, which now has the
   correct revision.
+* Deal with the external systems (see sections below).
 * Post!!
   * Merge release blog post to [lix-website].
   * Toot about it! https://chaos.social/@lix_project
@@ -87,22 +86,33 @@ Next, we do the publication with `python -m releng upload`:
 
 [lix-website]: https://git.lix.systems/lix-project/lix-website
 
-[upload-docker]: https://git.lix.systems/lix-project/lix/issues/252
-
 ### Installer
 
-The installer is cross-built to several systems from a Mac using
-`build-all.xsh` and `upload-to-lix.xsh` in the installer repo (FIXME: currently
-at least; maybe this should be moved here?) .
+The installer is cross-built to several systems from a Mac using `build-all.xsh` and `upload-to-lix.xsh` in the installer repo (FIXME: currently at least; maybe this should be moved here?).
 
-It installs a binary tarball (FIXME: [it should be taught to substitute from
-cache instead][installer-substitute])
-from some URL; this is the `hydraJobs.binaryTarball`. The default URLs differ
-by architecture and are [configured here][tarball-urls].
+It installs a binary tarball (FIXME: [it should be taught to substitute from cache instead][installer-substitute]) from some URL; this is the `hydraJobs.binaryTarball`.
+The default URLs differ by architecture and are [configured here][tarball-urls].
+
+To automatically do the file changes for a new version, run `python3 set_version.py NEW_VERSION`, and submit the result for review.
 
 [installer-substitute]: https://git.lix.systems/lix-project/lix-installer/issues/13
 [tarball-urls]: https://git.lix.systems/lix-project/lix-installer/src/commit/693592ed10d421a885bec0a9dd45e87ab87eb90a/src/settings.rs#L14-L28
 
+### Web site
+
+The website has various release-version dependent pieces.
+You can update them with `python3 update_version.py NEW_VERSION`, which will regenerate the affected page sources.
+
+These need the release to have been done first as they need hashes for tarballs and such.
+
+### NixOS module
+
+The NixOS module has underdeveloped releng in it.
+Currently you have to do the whole branch-off dance manually to a `release-VERSION` branch and update the tarball URLs to point to the release versions manually.
+
+FIXME: this should be unified with the `set_version.py` work in `lix-installer` and probably all the releng kept in here, or kept elsewhere.
+Related: https://git.lix.systems/lix-project/lix/issues/439
+
 ## Infrastructure summary
 
 * releases.lix.systems (`s3://releases`):

releng/cli.py

@@ -1,10 +1,13 @@
+import logging
+import argparse
+import sys
+
 from . import create_release
 from . import docker
 from .environment import RelengEnvironment
 from . import environment
-import argparse
-import sys
 
+log = logging.getLogger(__name__)
 
 def do_build(args):
     if args.target == 'all':
@@ -21,6 +24,9 @@ def do_tag(args):
     create_release.do_tag_merge(force_tag=args.force_tag,
                                 no_check_git=args.no_check_git)
 
+    log.info('Merged the release commit into your last branch, and switched to a detached HEAD of the artifact to be released.')
+    log.info('After you are done with releasing, switch to your previous branch and push that branch for review.')
+
 
 def do_upload(env: RelengEnvironment, args):
     create_release.setup_creds(env)

releng/create_release.xsh

@@ -2,6 +2,7 @@ import json
 import subprocess
 import itertools
 import textwrap
+import logging
 from pathlib import Path
 import tempfile
 import hashlib
@@ -11,10 +12,12 @@ from . import environment
 from .environment import RelengEnvironment
 from . import keys
 from . import docker
-from .version import VERSION, RELEASE_NAME, MAJOR
+from .version import VERSION, RELEASE_NAME, MAJOR, OFFICIAL_RELEASE
 from .gitutils import verify_are_on_tag, git_preconditions
 from . import release_notes
 
+log = logging.getLogger(__name__)
+
 $RAISE_SUBPROC_ERROR = True
 $XONSH_SHOW_TRACEBACK = True
 
@@ -39,16 +42,25 @@ def setup_creds(env: RelengEnvironment):
 
 
 def official_release_commit_tag(force_tag=False):
-    print('[+] Setting officialRelease in flake.nix and tagging')
+    print('[+] Setting officialRelease in version.json and tagging')
     prev_branch = $(git symbolic-ref --short HEAD).strip()
 
     git switch --detach
-    sed -i 's/officialRelease = false/officialRelease = true/' flake.nix
-    git add flake.nix
+
+    # Must be done in two parts due to buffering (opening the file immediately
+    # would truncate it).
+    new_version_json = $(jq --indent 4 '.official_release = true' version.json)
+    with open('version.json', 'w') as fh:
+        fh.write(new_version_json)
+    git add version.json
+
     message = f'release: {VERSION} "{RELEASE_NAME}"\n\nRelease produced with releng/create_release.xsh'
     git commit -m @(message)
     git tag @(['-f'] if force_tag else []) -a -m @(message) @(VERSION)
 
+    with open('releng/prev-git-branch.txt', 'w') as fh:
+        fh.write(prev_branch)
+
     return prev_branch
 
 
@@ -229,8 +241,24 @@ def upload_artifacts(env: RelengEnvironment, noconfirm=False, no_check_git=False
     print('[+] Upload manual')
     upload_manual(env)
 
-    print('[+] git push tag')
-    git push @(['-f'] if force_push_tag else []) @(env.git_repo) f'{VERSION}:refs/tags/{VERSION}'
+    prev_branch = None
+    try:
+        with open('releng/prev-git-branch.txt', 'r') as fh:
+            prev_branch = fh.read().strip()
+    except FileNotFoundError:
+        log.warn('Cannot find previous git branch file, skipping pushing git objects')
+
+    if prev_branch:
+        print('[+] git push to the repo')
+        # We have to push the ref to gerrit for review at least such that the
+        # commit is known, before we can push it as a tag.
+        if env.git_repo_is_gerrit:
+            git push @(env.git_repo) f'{prev_branch}:refs/for/{prev_branch}'
+        else:
+            git push @(env.git_repo) f'{prev_branch}:{prev_branch}'
+
+        print('[+] git push tag')
+        git push @(['-f'] if force_push_tag else []) @(env.git_repo) f'{VERSION}:refs/tags/{VERSION}'
 
 
 def do_tag_merge(force_tag=False, no_check_git=False):
@@ -250,15 +278,14 @@ def build_manual(eval_result):
 
 
 def upload_manual(env: RelengEnvironment):
-    stable = json.loads($(nix eval --json '.#nix.officialRelease'))
-    if stable:
+    if OFFICIAL_RELEASE:
         version = MAJOR
     else:
         version = 'nightly'
 
     print('[+] aws s3 sync manual')
     aws s3 sync @(MANUAL)/ @(env.docs_bucket)/manual/lix/@(version)/
-    if stable:
+    if OFFICIAL_RELEASE:
         aws s3 sync @(MANUAL)/ @(env.docs_bucket)/manual/lix/stable/
 
 

releng/docker.xsh

@@ -44,23 +44,8 @@ def upload_docker_images(target: DockerTarget, paths: list[Path]):
 
         for path in paths:
             digest_file = tmp / (path.name + '.digest')
-            tmp_image = tmp / 'tmp-image.tar.gz'
 
-            # insecure-policy: we don't have any signature policy, we are just uploading an image
-            #
-            # Absurd: we copy it into an OCI image first so we can get the hash
-            # we need to upload it untagged, because skopeo has no "don't tag
-            # this" option.
-            # The reason for this is that forgejo's container registry throws
-            # away old versions of tags immediately, so we cannot use a temp
-            # tag, and it *does* reduce confusion to not upload tags that
-            # should not be used.
-            #
-            # Workaround for: https://github.com/containers/skopeo/issues/2354
-            log.info('skopeo copy to temp oci-archive %s', tmp_image)
-            skopeo --insecure-policy copy --format oci --all --digestfile @(digest_file) docker-archive:@(path) oci-archive:@(tmp_image)
-
-            inspection = json.loads($(skopeo inspect oci-archive:@(tmp_image)))
+            inspection = json.loads($(skopeo inspect docker-archive:@(path)))
 
             docker_arch = inspection['Architecture']
             docker_os = inspection['Os']
@@ -68,8 +53,9 @@ def upload_docker_images(target: DockerTarget, paths: list[Path]):
 
             log.info('Pushing image %s for %s to %s', path, docker_arch, target.registry_path)
 
+            # insecure-policy: we don't have any signature policy, we are just uploading an image
+            skopeo --insecure-policy copy --digestfile @(digest_file) --all docker-archive:@(path) f'docker://{target.registry_path}@@unknown-digest@@'
             digest = digest_file.read_text().strip()
-            skopeo --insecure-policy copy --preserve-digests --all oci-archive:@(tmp_image) f'docker://{target.registry_path}@{digest}'
 
             # skopeo doesn't give us the manifest size directly, so we just ask the registry
             metadata = reg.image_info(target.registry_path, digest)

releng/environment.py

@@ -52,6 +52,7 @@ class RelengEnvironment:
     releases_bucket: str
     docs_bucket: str
     git_repo: str
+    git_repo_is_gerrit: bool
 
     docker_targets: list[DockerTarget]
 
@@ -79,6 +80,7 @@ STAGING = RelengEnvironment(
     cache_store_overlay={'secret-key': 'staging.key'},
     releases_bucket='s3://staging-releases',
     git_repo='ssh://git@git.lix.systems/lix-project/lix-releng-staging',
+    git_repo_is_gerrit=False,
     docker_targets=[
         # latest will be auto tagged if appropriate
         DockerTarget('git.lix.systems/lix-project/lix-releng-staging',
@@ -113,6 +115,7 @@ PROD = RelengEnvironment(
     cache_store_overlay={'secret-key': 'prod.key'},
     releases_bucket='s3://releases',
     git_repo=guess_gerrit_remote(),
+    git_repo_is_gerrit=True,
     docker_targets=[
         # latest will be auto tagged if appropriate
         DockerTarget('git.lix.systems/lix-project/lix',

releng/gitutils.xsh

@@ -1,11 +1,24 @@
 import subprocess
-import json
+from packaging.version import Version
 
 from .version import VERSION
 
 
+def remote_is_plausible(url: str) -> bool:
+    return ('git.lix.systems' in url and 'lix-project/lix' in url) or ('gerrit.lix.systems' in url and url.endswith('lix'))
+
+
 def version_compare(v1: str, v2: str):
-    return json.loads($(nix-instantiate --eval --json --argstr v1 @(v1) --argstr v2 @(v2) --expr '{v1, v2}: builtins.compareVersions v1 v2'))
+    v1 = Version(v1)
+    v2 = Version(v2)
+    if v1 < v2:
+        return -1
+    elif v1 > v2:
+        return 1
+    elif v1 == v2:
+        return 0
+    else:
+        raise ValueError('these versions are beyond each others celestial plane')
 
 
 def latest_tag_on_branch(branch: str) -> str:
@@ -13,16 +26,18 @@ def latest_tag_on_branch(branch: str) -> str:
 
 
 def is_maintenance_branch(branch: str) -> bool:
-    try:
-        main_tag = latest_tag_on_branch('main')
-        current_tag = latest_tag_on_branch(branch)
+    """
+    Returns whether the given branch is probably a maintenance branch.
 
-        return version_compare(current_tag, main_tag) < 0
-    except subprocess.CalledProcessError:
-        # This is the case before Lix releases 2.90, since main *has* no
-        # release tag on it.
-        # FIXME: delete this case after 2.91
-        return False
+    This uses a heuristic: `main` should have a newer tag than a given
+    maintenance branch if there has been a major release since that maintenance
+    branch.
+    """
+    assert remote_is_plausible($(git remote get-url origin).strip())
+    main_tag = latest_tag_on_branch('origin/main')
+    current_tag = latest_tag_on_branch(branch)
+
+    return version_compare(current_tag, main_tag) < 0
 
 
 def verify_are_on_tag():

releng/release-tests.sh

@@ -0,0 +1,36 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+shopt -s inherit_errexit failglob
+
+nixpkgss=(
+    "$(nix eval --impure --raw --expr '(import ./flake.nix).inputs.nixpkgs.url')"
+    "github:NixOS/nixpkgs/nixos-unstable-small"
+)
+jobs=(
+    $(nix eval \
+        --json --apply '
+          let f = n: t:
+            if builtins.isAttrs t
+            then (if t.type or "" == "derivation"
+                  then [ n ]
+                  else builtins.concatMap (m: f "${n}.${m}" t.${m}) (builtins.attrNames t))
+            else [];
+          in f ".#.releaseTests"
+        ' \
+        '.#.releaseTests' \
+        | jq -r '.[]'
+    )
+)
+
+for override in "${nixpkgss}"
+do
+    (
+        set -x
+        nix build \
+            --log-format multiline \
+            --no-link \
+            --override-input nixpkgs "$override" \
+            "${jobs[@]}"
+    )
+done

releng/release_notes.py

@@ -16,7 +16,7 @@ def add_to_summary(date: str):
     if VERSION_RL.exists():
         return
 
-    MARKER = '<!-- RELENG-AUTO-INSERTION-MARKER'
+    MARKER = '  <!-- RELENG-AUTO-INSERTION-MARKER'
 
     new_lines = []
     for line in SUMMARY.read_text().splitlines():

releng/version.py

@@ -4,3 +4,4 @@ version_json = json.load(open('version.json'))
 VERSION = version_json['version']
 MAJOR = '.'.join(VERSION.split('.')[:2])
 RELEASE_NAME = version_json['release_name']
+OFFICIAL_RELEASE = version_json['official_release']

src/asan-options/asan-options.cc

@@ -0,0 +1,17 @@
+/// @file This is very bothersome code that has to be included in every
+/// executable to get the correct default ASan options. I am so sorry.
+
+extern "C" [[gnu::retain]] const char *__asan_default_options()
+{
+    // We leak a bunch of memory knowingly on purpose. It's not worthwhile to
+    // diagnose that memory being leaked for now.
+    //
+    // Instruction bytes are useful for finding the actual code that
+    // corresponds to an ASan report.
+    //
+    // TODO: setting log_path=asan.log or not: neither works, since you can't
+    // write to the fs in certain places in the testsuite, but you also cannot
+    // write arbitrarily to stderr in other places so the reports get eaten.
+    // pain 🥖
+    return "halt_on_error=1:abort_on_error=1:detect_leaks=0:print_summary=1:dump_instruction_bytes=1";
+}

src/build-remote/build-remote.cc

@@ -42,7 +42,7 @@ static std::string makeLockFilename(const std::string & storeUri) {
     // This avoids issues with the escaped URI being very long and causing
     // path too long errors, while also avoiding any possibility of collision
     // caused by simple truncation.
-    auto hash = hashString(HashType::htSHA256, storeUri).to_string(Base::Base32, false);
+    auto hash = hashString(HashType::SHA256, storeUri).to_string(Base::Base32, false);
     return escapeUri(storeUri).substr(0, 48) + "-" + hash.substr(0, 16);
 }
 
@@ -236,9 +236,9 @@ static int main_build_remote(int argc, char * * argv)
                 }
 
 #if __APPLE__
-                futimes(bestSlotLock.get(), NULL);
+                futimes(bestSlotLock.get(), nullptr);
 #else
-                futimens(bestSlotLock.get(), NULL);
+                futimens(bestSlotLock.get(), nullptr);
 #endif
 
                 lock.reset();

src/libcmd/cmd-profiles.cc

@@ -246,7 +246,7 @@ StorePath ProfileManifest::build(ref<Store> store)
     StringSink sink;
     sink << dumpPath(tempDir);
 
-    auto narHash = hashString(htSHA256, sink.s);
+    auto narHash = hashString(HashType::SHA256, sink.s);
 
     ValidPathInfo info{
         *store,

src/libcmd/cmd-profiles.hh

@@ -37,7 +37,7 @@ struct ProfileElement
     StorePathSet storePaths;
     std::optional<ProfileElementSource> source;
     bool active = true;
-    int priority = DEFAULT_PRIORITY;
+    NixInt::Inner priority = DEFAULT_PRIORITY;
 
     std::string identifier() const;
 

src/libcmd/installable-attr-path.cc

@@ -1,23 +1,11 @@
-#include "globals.hh"
 #include "installable-attr-path.hh"
 #include "outputs-spec.hh"
 #include "command.hh"
 #include "attr-path.hh"
 #include "common-eval-args.hh"
-#include "derivations.hh"
-#include "eval-inline.hh"
 #include "eval.hh"
 #include "get-drvs.hh"
-#include "store-api.hh"
-#include "shared.hh"
 #include "flake/flake.hh"
-#include "eval-cache.hh"
-#include "url.hh"
-#include "registry.hh"
-#include "build-result.hh"
-
-#include <regex>
-#include <queue>
 
 #include <nlohmann/json.hpp>
 

src/libcmd/installable-attr-path.hh

@@ -1,25 +1,11 @@
 #pragma once
 ///@file
 
-#include "globals.hh"
 #include "installable-value.hh"
 #include "outputs-spec.hh"
 #include "command.hh"
-#include "attr-path.hh"
 #include "common-eval-args.hh"
-#include "derivations.hh"
-#include "eval-inline.hh"
 #include "eval.hh"
-#include "get-drvs.hh"
-#include "store-api.hh"
-#include "shared.hh"
-#include "eval-cache.hh"
-#include "url.hh"
-#include "registry.hh"
-#include "build-result.hh"
-
-#include <regex>
-#include <queue>
 
 #include <nlohmann/json.hpp>
 

src/libcmd/installable-flake.cc

@@ -116,12 +116,12 @@ DerivedPathsWithInfo InstallableFlake::toDerivedPaths()
 
     auto drvPath = attr->forceDerivation();
 
-    std::optional<NixInt> priority;
+    std::optional<NixInt::Inner> priority;
 
     if (attr->maybeGetAttr(state->sOutputSpecified)) {
     } else if (auto aMeta = attr->maybeGetAttr(state->sMeta)) {
         if (auto aPriority = aMeta->maybeGetAttr("priority"))
-            priority = aPriority->getInt();
+            priority = aPriority->getInt().value;
     }
 
     return {{

src/libcmd/installable-value.hh

@@ -40,7 +40,7 @@ struct ExtraPathInfoValue : ExtraPathInfo
         /**
          * An optional priority for use with "build envs". See Package
          */
-        std::optional<NixInt> priority;
+        std::optional<NixInt::Inner> priority;
 
         /**
          * The attribute path associated with this value. The idea is

src/libcmd/repl-interacter.cc

@@ -96,7 +96,7 @@ static int listPossibleCallback(char * s, char *** avp)
         return p;
     };
 
-    vp = check((char **) malloc(possible.size() * sizeof(char *)));
+    vp = check(static_cast<char **>(malloc(possible.size() * sizeof(char *))));
 
     for (auto & p : possible)
         vp[ac++] = check(strdup(p.c_str()));

src/libcmd/repl.cc

@@ -40,24 +40,24 @@
 #include <gc/gc_cpp.h>
 #endif
 
-// XXX: These are for nix-doc features and will be removed in a future rewrite where this functionality is integrated more natively.
+// XXX: These are for lix-doc features and will be removed in a future rewrite where this functionality is integrated more natively.
 extern "C" {
-    char const *nd_get_function_docs(char const *filename, size_t line, size_t col);
-    void nd_free_string(char const *str);
+    char const *lixdoc_get_function_docs(char const *filename, size_t line, size_t col);
+    void lixdoc_free_string(char const *str);
 }
 
 namespace nix {
 
 
 /** Wrapper around std::unique_ptr with a custom deleter for strings from nix-doc **/
-using NdString = std::unique_ptr<const char, decltype(&nd_free_string)>;
+using NdString = std::unique_ptr<const char, decltype(&lixdoc_free_string)>;
 
 /**
  * Fetch a string representing the doc comment using nix-doc and wrap it in an RAII wrapper.
  */
 NdString lambdaDocsForPos(SourcePath const path, nix::Pos const &pos) {
   std::string const file = path.to_string();
-  return NdString{nd_get_function_docs(file.c_str(), pos.line, pos.column), &nd_free_string};
+  return NdString{lixdoc_get_function_docs(file.c_str(), pos.line, pos.column), &lixdoc_free_string};
 }
 
 /**
@@ -251,7 +251,7 @@ void runNix(Path program, const Strings & args)
         .program = settings.nixBinDir+ "/" + program,
         .args = args,
         .environment = subprocessEnv,
-    }).wait();
+    }).waitAndCheck();
 
     return;
 }
@@ -536,7 +536,7 @@ ProcessLineResult NixRepl::processLine(std::string line)
              << "  :t <expr>                    Describe result of evaluation\n"
              << "  :u <expr>                    Build derivation, then start nix-shell\n"
              << "  :doc <expr>                  Show documentation for the provided function (experimental lambda support)\n"
-             << "  :log <expr>                  Show logs for a derivation\n"
+             << "  :log <expr | .drv path>      Show logs for a derivation\n"
              << "  :te, :trace-enable [bool]    Enable, disable or toggle showing traces for\n"
              << "                               errors\n"
              << "  :?, :help                    Brings up this help menu\n"
@@ -650,11 +650,14 @@ ProcessLineResult NixRepl::processLine(std::string line)
 
         // runProgram redirects stdout to a StringSink,
         // using runProgram2 to allow editors to display their UI
-        runProgram2(RunOptions { .program = editor, .searchPath = true, .args = args }).wait();
+        runProgram2(RunOptions { .program = editor, .searchPath = true, .args = args }).waitAndCheck();
 
-        // Reload right after exiting the editor
-        state->resetFileCache();
-        reloadFiles();
+        // Reload right after exiting the editor if path is not in store
+        // Store is immutable, so there could be no changes, so there's no need to reload
+        if (!state->store->isInStore(path.resolveSymlinks().path.abs())) {
+            state->resetFileCache();
+            reloadFiles();
+        }
     }
 
     else if (command == ":t") {
@@ -673,7 +676,49 @@ ProcessLineResult NixRepl::processLine(std::string line)
         runNix("nix-shell", {state->store->printStorePath(drvPath)});
     }
 
-    else if (command == ":b" || command == ":bl" || command == ":i" || command == ":sh" || command == ":log") {
+    else if (command == ":log") {
+        StorePath drvPath = ([&] {
+            auto maybeDrvPath = state->store->maybeParseStorePath(arg);
+            if (maybeDrvPath && maybeDrvPath->isDerivation()) {
+                return std::move(*maybeDrvPath);
+            } else {
+                Value v;
+                evalString(arg, v);
+                return getDerivationPath(v);
+            }
+        })();
+        Path drvPathRaw = state->store->printStorePath(drvPath);
+
+        settings.readOnlyMode = true;
+        Finally roModeReset([&]() {
+            settings.readOnlyMode = false;
+        });
+        auto subs = getDefaultSubstituters();
+
+        subs.push_front(state->store);
+
+        bool foundLog = false;
+        RunPager pager;
+        for (auto & sub : subs) {
+            auto * logSubP = dynamic_cast<LogStore *>(&*sub);
+            if (!logSubP) {
+                printInfo("Skipped '%s' which does not support retrieving build logs", sub->getUri());
+                continue;
+            }
+            auto & logSub = *logSubP;
+
+            auto log = logSub.getBuildLog(drvPath);
+            if (log) {
+                printInfo("got build log for '%s' from '%s'", drvPathRaw, logSub.getUri());
+                logger->writeToStdout(*log);
+                foundLog = true;
+                break;
+            }
+        }
+        if (!foundLog) throw Error("build log of '%s' is not available", drvPathRaw);
+    }
+
+    else if (command == ":b" || command == ":bl" || command == ":i" || command == ":sh") {
         Value v;
         evalString(arg, v);
         StorePath drvPath = getDerivationPath(v);
@@ -709,34 +754,6 @@ ProcessLineResult NixRepl::processLine(std::string line)
             }
         } else if (command == ":i") {
             runNix("nix-env", {"-i", drvPathRaw});
-        } else if (command == ":log") {
-            settings.readOnlyMode = true;
-            Finally roModeReset([&]() {
-                settings.readOnlyMode = false;
-            });
-            auto subs = getDefaultSubstituters();
-
-            subs.push_front(state->store);
-
-            bool foundLog = false;
-            RunPager pager;
-            for (auto & sub : subs) {
-                auto * logSubP = dynamic_cast<LogStore *>(&*sub);
-                if (!logSubP) {
-                    printInfo("Skipped '%s' which does not support retrieving build logs", sub->getUri());
-                    continue;
-                }
-                auto & logSub = *logSubP;
-
-                auto log = logSub.getBuildLog(drvPath);
-                if (log) {
-                    printInfo("got build log for '%s' from '%s'", drvPathRaw, logSub.getUri());
-                    logger->writeToStdout(*log);
-                    foundLog = true;
-                    break;
-                }
-            }
-            if (!foundLog) throw Error("build log of '%s' is not available", drvPathRaw);
         } else {
             runNix("nix-shell", {drvPathRaw});
         }

src/libexpr/attr-set.cc

@@ -1,5 +1,6 @@
 #include "attr-set.hh"
-#include "eval-inline.hh"
+#include "eval.hh"
+#include "gc-alloc.hh"
 
 #include <algorithm>
 
@@ -19,7 +20,7 @@ Bindings * EvalState::allocBindings(size_t capacity)
         throw Error("attribute set of size %d is too big", capacity);
     nrAttrsets++;
     nrAttrsInAttrsets += capacity;
-    return new (allocBytes(sizeof(Bindings) + sizeof(Attr) * capacity)) Bindings((Bindings::size_t) capacity);
+    return new (gcAllocBytes(sizeof(Bindings) + sizeof(Attr) * capacity)) Bindings((Bindings::size_t) capacity);
 }
 
 

src/libexpr/eval-cache.cc

@@ -50,7 +50,7 @@ struct AttrDb
         Path cacheDir = getCacheDir() + "/nix/eval-cache-v5";
         createDirs(cacheDir);
 
-        Path dbPath = cacheDir + "/" + fingerprint.to_string(Base16, false) + ".sqlite";
+        Path dbPath = cacheDir + "/" + fingerprint.to_string(Base::Base16, false) + ".sqlite";
 
         state->db = SQLite(dbPath);
         state->db.isCache();
@@ -306,7 +306,7 @@ struct AttrDb
             case AttrType::Bool:
                 return {{rowId, queryAttribute.getInt(2) != 0}};
             case AttrType::Int:
-                return {{rowId, int_t{queryAttribute.getInt(2)}}};
+                return {{rowId, int_t{NixInt{queryAttribute.getInt(2)}}}};
             case AttrType::ListOfStrings:
                 return {{rowId, tokenizeString<std::vector<std::string>>(queryAttribute.getStr(2), "\t")}};
             case AttrType::Missing:
@@ -449,7 +449,7 @@ Value & AttrCursor::forceValue()
         else if (v.type() == nBool)
             cachedValue = {root->db->setBool(getKey(), v.boolean), v.boolean};
         else if (v.type() == nInt)
-            cachedValue = {root->db->setInt(getKey(), v.integer), int_t{v.integer}};
+            cachedValue = {root->db->setInt(getKey(), v.integer.value), int_t{v.integer}};
         else if (v.type() == nAttrs)
             ; // FIXME: do something?
         else

src/libexpr/eval-inline.hh

@@ -4,26 +4,10 @@
 #include "print.hh"
 #include "eval.hh"
 #include "eval-error.hh"
+#include "gc-alloc.hh"
 
 namespace nix {
 
-/**
- * Note: Various places expect the allocated memory to be zeroed.
- */
-[[gnu::always_inline]]
-inline void * allocBytes(size_t n)
-{
-    void * p;
-#if HAVE_BOEHMGC
-    p = GC_MALLOC(n);
-#else
-    p = calloc(n, 1);
-#endif
-    if (!p) throw std::bad_alloc();
-    return p;
-}
-
-
 [[gnu::always_inline]]
 Value * EvalState::allocValue()
 {
@@ -43,7 +27,7 @@ Value * EvalState::allocValue()
     *valueAllocCache = GC_NEXT(p);
     GC_NEXT(p) = nullptr;
 #else
-    void * p = allocBytes(sizeof(Value));
+    void * p = gcAllocBytes(sizeof(Value));
 #endif
 
     nrValues++;
@@ -73,7 +57,7 @@ Env & EvalState::allocEnv(size_t size)
         env = (Env *) p;
     } else
 #endif
-        env = (Env *) allocBytes(sizeof(Env) + size * sizeof(Value *));
+        env = (Env *) gcAllocBytes(sizeof(Env) + size * sizeof(Value *));
 
     /* We assume that env->values has been cleared by the allocator; maybeThunk() and lookupVar fromWith expect this. */
 

src/libexpr/eval.cc

@@ -9,6 +9,7 @@
 #include "store-api.hh"
 #include "derivations.hh"
 #include "downstream-placeholder.hh"
+#include "gc-alloc.hh"
 #include "globals.hh"
 #include "eval-inline.hh"
 #include "filetransfer.hh"
@@ -20,7 +21,6 @@
 #include "flake/flakeref.hh"
 
 #include <algorithm>
-#include <chrono>
 #include <iostream>
 #include <sstream>
 #include <cstring>
@@ -42,58 +42,12 @@
 #include <gc/gc.h>
 #include <gc/gc_cpp.h>
 
-#include <boost/coroutine2/coroutine.hpp>
-#include <boost/coroutine2/protected_fixedsize_stack.hpp>
-#include <boost/context/stack_context.hpp>
-
 #endif
 
 using json = nlohmann::json;
 
 namespace nix {
 
-static char * allocString(size_t size)
-{
-    char * t;
-#if HAVE_BOEHMGC
-    t = (char *) GC_MALLOC_ATOMIC(size);
-#else
-    t = (char *) malloc(size);
-#endif
-    if (!t) throw std::bad_alloc();
-    return t;
-}
-
-
-static char * dupString(const char * s)
-{
-    char * t;
-#if HAVE_BOEHMGC
-    t = GC_STRDUP(s);
-#else
-    t = strdup(s);
-#endif
-    if (!t) throw std::bad_alloc();
-    return t;
-}
-
-
-// When there's no need to write to the string, we can optimize away empty
-// string allocations.
-// This function handles makeImmutableString(std::string_view()) by returning
-// the empty string.
-static const char * makeImmutableString(std::string_view s)
-{
-    const size_t size = s.size();
-    if (size == 0)
-        return "";
-    auto t = allocString(size + 1);
-    memcpy(t, s.data(), size);
-    t[size] = '\0';
-    return t;
-}
-
-
 RootValue allocRootValue(Value * v)
 {
 #if HAVE_BOEHMGC
@@ -116,11 +70,6 @@ std::string printValue(EvalState & state, Value & v)
     return out.str();
 }
 
-void Value::print(EvalState & state, std::ostream & str, PrintOptions options)
-{
-    printValue(state, str, *this, options);
-}
-
 const Value * getPrimOp(const Value &v) {
     const Value * primOp = &v;
     while (primOp->isPrimOpApp()) {
@@ -170,32 +119,6 @@ std::string showType(const Value & v)
     #pragma GCC diagnostic pop
 }
 
-PosIdx Value::determinePos(const PosIdx pos) const
-{
-    // Allow selecting a subset of enum values
-    #pragma GCC diagnostic push
-    #pragma GCC diagnostic ignored "-Wswitch-enum"
-    switch (internalType) {
-        case tAttrs: return attrs->pos;
-        case tLambda: return lambda.fun->pos;
-        case tApp: return app.left->determinePos(pos);
-        default: return pos;
-    }
-    #pragma GCC diagnostic pop
-}
-
-bool Value::isTrivial() const
-{
-    return
-        internalType != tApp
-        && internalType != tPrimOpApp
-        && (internalType != tThunk
-            || (dynamic_cast<ExprAttrs *>(thunk.expr)
-                && ((ExprAttrs *) thunk.expr)->dynamicAttrs.empty())
-            || dynamic_cast<ExprLambda *>(thunk.expr)
-            || dynamic_cast<ExprList *>(thunk.expr));
-}
-
 
 #if HAVE_BOEHMGC
 /* Called when the Boehm GC runs out of memory. */
@@ -205,42 +128,6 @@ static void * oomHandler(size_t requested)
     throw std::bad_alloc();
 }
 
-class BoehmGCStackAllocator : public StackAllocator {
-    boost::coroutines2::protected_fixedsize_stack stack {
-        // We allocate 8 MB, the default max stack size on NixOS.
-        // A smaller stack might be quicker to allocate but reduces the stack
-        // depth available for source filter expressions etc.
-        std::max(boost::context::stack_traits::default_size(), static_cast<std::size_t>(8 * 1024 * 1024))
-    };
-
-    // This is specific to boost::coroutines2::protected_fixedsize_stack.
-    // The stack protection page is included in sctx.size, so we have to
-    // subtract one page size from the stack size.
-    std::size_t pfss_usable_stack_size(boost::context::stack_context &sctx) {
-        return sctx.size - boost::context::stack_traits::page_size();
-    }
-
-  public:
-    boost::context::stack_context allocate() override {
-        auto sctx = stack.allocate();
-
-        // Stacks generally start at a high address and grow to lower addresses.
-        // Architectures that do the opposite are rare; in fact so rare that
-        // boost_routine does not implement it.
-        // So we subtract the stack size.
-        GC_add_roots(static_cast<char *>(sctx.sp) - pfss_usable_stack_size(sctx), sctx.sp);
-        return sctx;
-    }
-
-    void deallocate(boost::context::stack_context sctx) override {
-        GC_remove_roots(static_cast<char *>(sctx.sp) - pfss_usable_stack_size(sctx), sctx.sp);
-        stack.deallocate(sctx);
-    }
-
-};
-
-static BoehmGCStackAllocator boehmGCStackAllocator;
-
 #endif
 
 
@@ -256,23 +143,6 @@ static Symbol getName(const AttrName & name, EvalState & state, Env & env)
     }
 }
 
-#if HAVE_BOEHMGC
-/* Disable GC while this object lives. Used by CoroutineContext.
- *
- * Boehm keeps a count of GC_disable() and GC_enable() calls,
- * and only enables GC when the count matches.
- */
-class BoehmDisableGC {
-public:
-    BoehmDisableGC() {
-        GC_disable();
-    };
-    ~BoehmDisableGC() {
-        GC_enable();
-    };
-};
-#endif
-
 static bool gcInitialised = false;
 
 void initGC()
@@ -294,17 +164,6 @@ void initGC()
 
     GC_set_oom_fn(oomHandler);
 
-    StackAllocator::defaultAllocator = &boehmGCStackAllocator;
-
-
-#if NIX_BOEHM_PATCH_VERSION != 1
-    printTalkative("Unpatched BoehmGC, disabling GC inside coroutines");
-    /* Used to disable GC when entering coroutines on macOS */
-    create_coro_gc_hook = []() -> std::shared_ptr<void> {
-        return std::make_shared<BoehmDisableGC>();
-    };
-#endif
-
     /* Set the initial heap size to something fairly big (25% of
        physical RAM, up to a maximum of 384 MiB) so that in most cases
        we don't need to garbage collect at all.  (Collection has a
@@ -358,6 +217,12 @@ EvalState::EvalState(
     , sRight(symbols.create("right"))
     , sWrong(symbols.create("wrong"))
     , sStructuredAttrs(symbols.create("__structuredAttrs"))
+    , sAllowedReferences(symbols.create("allowedReferences"))
+    , sAllowedRequisites(symbols.create("allowedRequisites"))
+    , sDisallowedReferences(symbols.create("disallowedReferences"))
+    , sDisallowedRequisites(symbols.create("disallowedRequisites"))
+    , sMaxSize(symbols.create("maxSize"))
+    , sMaxClosureSize(symbols.create("maxClosureSize"))
     , sBuilder(symbols.create("builder"))
     , sArgs(symbols.create("args"))
     , sContentAddressed(symbols.create("__contentAddressed"))
@@ -602,29 +467,6 @@ std::ostream & operator<<(std::ostream & output, PrimOp & primOp)
     return output;
 }
 
-
-PrimOp * Value::primOpAppPrimOp() const
-{
-    Value * left = primOpApp.left;
-    while (left && !left->isPrimOp()) {
-        left = left->primOpApp.left;
-    }
-
-    if (!left)
-        return nullptr;
-    return left->primOp;
-}
-
-
-void Value::mkPrimOp(PrimOp * p)
-{
-    p->check();
-    clearValue();
-    internalType = tPrimOp;
-    primOp = p;
-}
-
-
 Value * EvalState::addPrimOp(PrimOp && primOp)
 {
     /* Hack to make constants lazy: turn them into a application of
@@ -876,43 +718,6 @@ DebugTraceStacker::DebugTraceStacker(EvalState & evalState, DebugTrace t)
         evalState.runDebugRepl(nullptr, trace.env, trace.expr);
 }
 
-void Value::mkString(std::string_view s)
-{
-    mkString(makeImmutableString(s));
-}
-
-
-static void copyContextToValue(Value & v, const NixStringContext & context)
-{
-    if (!context.empty()) {
-        size_t n = 0;
-        v.string.context = (const char * *)
-            allocBytes((context.size() + 1) * sizeof(char *));
-        for (auto & i : context)
-            v.string.context[n++] = dupString(i.to_string().c_str());
-        v.string.context[n] = 0;
-    }
-}
-
-void Value::mkString(std::string_view s, const NixStringContext & context)
-{
-    mkString(s);
-    copyContextToValue(*this, context);
-}
-
-void Value::mkStringMove(const char * s, const NixStringContext & context)
-{
-    mkString(s);
-    copyContextToValue(*this, context);
-}
-
-
-void Value::mkPath(const SourcePath & path)
-{
-    mkPath(makeImmutableString(path.path.abs()));
-}
-
-
 inline Value * EvalState::lookupVar(Env * env, const ExprVar & var, bool noEval)
 {
     for (auto l = var.level; l; --l, env = env->up) ;
@@ -943,7 +748,7 @@ void EvalState::mkList(Value & v, size_t size)
 {
     v.mkList(size);
     if (size > 2)
-        v.bigList.elems = (Value * *) allocBytes(size * sizeof(Value *));
+        v.bigList.elems = gcAllocType<Value *>(size);
     nrListElems += size;
 }
 
@@ -1982,7 +1787,7 @@ void EvalState::autoCallFunction(Bindings & args, Value & fun, Value & res)
 Nix attempted to evaluate a function as a top level expression; in
 this case it must have its arguments supplied either by default
 values, or passed explicitly with '--arg' or '--argstr'. See
-https://nixos.org/manual/nix/stable/language/constructs.html#functions.)", symbols[i.name])
+https://docs.lix.systems/manual/lix/stable/language/constructs.html#functions)", symbols[i.name])
                     .atPos(i.pos).withFrame(*fun.lambda.env, *fun.lambda.fun).debugThrow();
             }
         }
@@ -2141,7 +1946,7 @@ void ExprConcatStrings::eval(EvalState & state, Env & env, Value & v)
     NixStringContext context;
     std::vector<BackedStringView> s;
     size_t sSize = 0;
-    NixInt n = 0;
+    NixInt n{0};
     NixFloat nf = 0;
 
     bool first = !forceString;
@@ -2157,7 +1962,7 @@ void ExprConcatStrings::eval(EvalState & state, Env & env, Value & v)
        Value. allocating a GC'd string directly and moving it into a
        Value lets us avoid an allocation and copy. */
     const auto c_str = [&] {
-        char * result = allocString(sSize + 1);
+        char * result = gcAllocString(sSize + 1);
         char * tmp = result;
         for (const auto & part : s) {
             memcpy(tmp, part->data(), part->size());
@@ -2185,17 +1990,22 @@ void ExprConcatStrings::eval(EvalState & state, Env & env, Value & v)
 
         if (firstType == nInt) {
             if (vTmp.type() == nInt) {
-                n += vTmp.integer;
+                auto newN = n + vTmp.integer;
+                if (auto checked = newN.valueChecked(); checked.has_value()) {
+                    n = NixInt(*checked);
+                } else {
+                    state.error<EvalError>("integer overflow in adding %1% + %2%", n, vTmp.integer).atPos(i_pos).debugThrow();
+                }
             } else if (vTmp.type() == nFloat) {
                 // Upgrade the type from int to float;
                 firstType = nFloat;
-                nf = n;
+                nf = n.value;
                 nf += vTmp.fpoint;
             } else
                 state.error<EvalError>("cannot add %1% to an integer", showType(vTmp)).atPos(i_pos).withFrame(env, *this).debugThrow();
         } else if (firstType == nFloat) {
             if (vTmp.type() == nInt) {
-                nf += vTmp.integer;
+                nf += vTmp.integer.value;
             } else if (vTmp.type() == nFloat) {
                 nf += vTmp.fpoint;
             } else
@@ -2320,7 +2130,7 @@ NixFloat EvalState::forceFloat(Value & v, const PosIdx pos, std::string_view err
     try {
         forceValue(v, pos);
         if (v.type() == nInt)
-            return v.integer;
+            return v.integer.value;
         else if (v.type() != nFloat)
             error<TypeError>(
                 "expected a float but found %1%: %2%",
@@ -2507,7 +2317,7 @@ BackedStringView EvalState::coerceToString(
            shell scripting convenience, just like `null'. */
         if (v.type() == nBool && v.boolean) return "1";
         if (v.type() == nBool && !v.boolean) return "";
-        if (v.type() == nInt) return std::to_string(v.integer);
+        if (v.type() == nInt) return std::to_string(v.integer.value);
         if (v.type() == nFloat) return std::to_string(v.fpoint);
         if (v.type() == nNull) return "";
 
@@ -2651,9 +2461,9 @@ bool EvalState::eqValues(Value & v1, Value & v2, const PosIdx pos, std::string_v
 
     // Special case type-compatibility between float and int
     if (v1.type() == nInt && v2.type() == nFloat)
-        return v1.integer == v2.fpoint;
+        return v1.integer.value == v2.fpoint;
     if (v1.type() == nFloat && v2.type() == nInt)
-        return v1.fpoint == v2.integer;
+        return v1.fpoint == v2.integer.value;
 
     // All other types are not compatible with each other.
     if (v1.type() != v2.type()) return false;
@@ -2900,20 +2710,20 @@ Expr & EvalState::parseExprFromFile(const SourcePath & path, std::shared_ptr<Sta
 }
 
 
-Expr & EvalState::parseExprFromString(std::string s_, const SourcePath & basePath, std::shared_ptr<StaticEnv> & staticEnv)
+Expr & EvalState::parseExprFromString(std::string s_, const SourcePath & basePath, std::shared_ptr<StaticEnv> & staticEnv, const ExperimentalFeatureSettings & xpSettings)
 {
     // NOTE this method (and parseStdin) must take care to *fully copy* their input
     // into their respective Pos::Origin until the parser stops overwriting its input
     // data.
     auto s = make_ref<std::string>(s_);
     s_.append("\0\0", 2);
-    return *parse(s_.data(), s_.size(), Pos::String{.source = s}, basePath, staticEnv);
+    return *parse(s_.data(), s_.size(), Pos::String{.source = s}, basePath, staticEnv, xpSettings);
 }
 
 
-Expr & EvalState::parseExprFromString(std::string s, const SourcePath & basePath)
+Expr & EvalState::parseExprFromString(std::string s, const SourcePath & basePath, const ExperimentalFeatureSettings & xpSettings)
 {
-    return parseExprFromString(std::move(s), basePath, staticBaseEnv);
+    return parseExprFromString(std::move(s), basePath, staticBaseEnv, xpSettings);
 }
 
 

src/libexpr/eval.hh

@@ -3,6 +3,7 @@
 
 #include "attr-set.hh"
 #include "eval-error.hh"
+#include "gc-alloc.hh"
 #include "types.hh"
 #include "value.hh"
 #include "nixexpr.hh"
@@ -12,7 +13,6 @@
 #include "search-path.hh"
 #include "repl-exit-status.hh"
 
-#include <gc/gc_allocator.h>
 #include <map>
 #include <optional>
 #include <unordered_map>
@@ -37,22 +37,6 @@ namespace eval_cache {
     class EvalCache;
 }
 
-/** Alias for std::map which uses boehmgc's allocator conditional on us actually
- * using boehmgc in this build.
- */
-#if HAVE_BOEHMGC
-    template<typename KeyT, typename ValueT>
-    using GcMap = std::map<
-        KeyT,
-        ValueT,
-        std::less<KeyT>,
-        traceable_allocator<std::pair<KeyT const, ValueT>>
-    >;
-#else
-    using GcMap = std::map<KeyT, ValueT>
-#endif
-
-
 /**
  * Function that implements a primop.
  */
@@ -177,7 +161,10 @@ public:
     const Symbol sWith, sOutPath, sDrvPath, sType, sMeta, sName, sValue,
         sSystem, sOverrides, sOutputs, sOutputName, sIgnoreNulls,
         sFile, sLine, sColumn, sFunctor, sToString,
-        sRight, sWrong, sStructuredAttrs, sBuilder, sArgs,
+        sRight, sWrong, sStructuredAttrs,
+        sAllowedReferences, sAllowedRequisites, sDisallowedReferences, sDisallowedRequisites,
+        sMaxSize, sMaxClosureSize,
+        sBuilder, sArgs,
         sContentAddressed, sImpure,
         sOutputHash, sOutputHashAlgo, sOutputHashMode,
         sRecurseForDerivations,
@@ -357,8 +344,8 @@ public:
     /**
      * Parse a Nix expression from the specified string.
      */
-    Expr & parseExprFromString(std::string s, const SourcePath & basePath, std::shared_ptr<StaticEnv> & staticEnv);
-    Expr & parseExprFromString(std::string s, const SourcePath & basePath);
+    Expr & parseExprFromString(std::string s, const SourcePath & basePath, std::shared_ptr<StaticEnv> & staticEnv, const ExperimentalFeatureSettings & xpSettings = experimentalFeatureSettings);
+    Expr & parseExprFromString(std::string s, const SourcePath & basePath, const ExperimentalFeatureSettings & xpSettings = experimentalFeatureSettings);
 
     Expr & parseStdin();
 
@@ -581,7 +568,8 @@ private:
         size_t length,
         Pos::Origin origin,
         const SourcePath & basePath,
-        std::shared_ptr<StaticEnv> & staticEnv);
+        std::shared_ptr<StaticEnv> & staticEnv,
+        const ExperimentalFeatureSettings & xpSettings = experimentalFeatureSettings);
 
     /**
      * Current Nix call stack depth, used with `max-call-depth` setting to throw stack overflow hopefully before we run out of system stack.

src/libexpr/flake/config.cc

@@ -53,7 +53,7 @@ void ConfigFile::apply()
 
         bool trusted = whitelist.count(baseName);
         if (!trusted) {
-            switch (nix::fetchSettings.acceptFlakeConfig) {
+            switch (nix::fetchSettings.acceptFlakeConfig.get()) {
             case AcceptFlakeConfig::True: {
                 trusted = true;
                 break;

src/libexpr/flake/flake.cc

@@ -137,9 +137,16 @@ static FlakeInput parseFlakeInput(EvalState & state,
                     case nBool:
                         attrs.emplace(state.symbols[attr.name], Explicit<bool> { attr.value->boolean });
                         break;
-                    case nInt:
-                        attrs.emplace(state.symbols[attr.name], (long unsigned int)attr.value->integer);
+                    case nInt: {
+                        auto intValue = attr.value->integer.value;
+
+                        if (intValue < 0) {
+                            state.error<EvalError>("negative value given for flake input attribute %1%: %2%", state.symbols[attr.name], intValue).debugThrow();
+                        }
+                        uint64_t asUnsigned = intValue;
+                        attrs.emplace(state.symbols[attr.name], asUnsigned);
                         break;
+                    }
                     default:
                         state.error<TypeError>("flake input attribute '%s' is %s while a string, Boolean, or integer is expected",
                             state.symbols[attr.name], showType(*attr.value)).debugThrow();
@@ -284,7 +291,7 @@ static Flake getFlake(
             else if (setting.value->type() == nInt)
                 flake.config.settings.emplace(
                     state.symbols[setting.name],
-                    state.forceInt(*setting.value, setting.pos, ""));
+                    state.forceInt(*setting.value, setting.pos, "").value);
             else if (setting.value->type() == nBool)
                 flake.config.settings.emplace(
                     state.symbols[setting.name],
@@ -873,8 +880,14 @@ static void prim_flakeRefToString(
     for (const auto & attr : *args[0]->attrs) {
         auto t = attr.value->type();
         if (t == nInt) {
-            attrs.emplace(state.symbols[attr.name],
-                          (uint64_t) attr.value->integer);
+            auto intValue = attr.value->integer.value;
+
+            if (intValue < 0) {
+                state.error<EvalError>("negative value given for flake ref attr %1%: %2%", state.symbols[attr.name], intValue).atPos(pos).debugThrow();
+            }
+            uint64_t asUnsigned = intValue;
+
+            attrs.emplace(state.symbols[attr.name], asUnsigned);
         } else if (t == nBool) {
             attrs.emplace(state.symbols[attr.name],
                           Explicit<bool> { attr.value->boolean });
@@ -924,7 +937,7 @@ Fingerprint LockedFlake::getFingerprint() const
     // FIXME: as an optimization, if the flake contains a lock file
     // and we haven't changed it, then it's sufficient to use
     // flake.sourceInfo.storePath for the fingerprint.
-    return hashString(htSHA256,
+    return hashString(HashType::SHA256,
         fmt("%s;%s;%d;%d;%s",
             flake.sourceInfo->storePath.to_string(),
             flake.lockedRef.subdir,

src/libexpr/flake/flakeref.cc

@@ -169,14 +169,13 @@ std::pair<FlakeRef, std::string> parseFlakeRefWithFragment(
                         if (subdir != "") {
                             if (parsedURL.query.count("dir"))
                                 throw Error("flake URL '%s' has an inconsistent 'dir' parameter", url);
-                            parsedURL.query.insert_or_assign("dir", subdir);
                         }
 
                         if (pathExists(flakeRoot + "/.git/shallow"))
                             parsedURL.query.insert_or_assign("shallow", "1");
 
                         return std::make_pair(
-                            FlakeRef(Input::fromURL(parsedURL, isFlake), getOr(parsedURL.query, "dir", "")),
+                            FlakeRef(Input::fromURL(parsedURL, isFlake), subdir),
                             fragment);
                     }
 
@@ -204,7 +203,13 @@ std::pair<FlakeRef, std::string> parseFlakeRefWithFragment(
         std::string fragment;
         std::swap(fragment, parsedURL.fragment);
 
-        auto input = Input::fromURL(parsedURL, isFlake);
+        // This has a special meaning for flakes and must not be passed to libfetchers.
+        // Of course this means that libfetchers cannot have fetchers
+        // expecting an argument `dir` 🫠
+        ParsedURL urlForFetchers(parsedURL);
+        urlForFetchers.query.erase("dir");
+
+        auto input = Input::fromURL(urlForFetchers, isFlake);
         input.parent = baseDir;
 
         return std::make_pair(

src/libexpr/gc-alloc.cc

@@ -0,0 +1,23 @@
+#include "gc-alloc.hh"
+
+#include <cstring>
+#include <string_view>
+
+namespace nix
+{
+
+char const * gcCopyStringIfNeeded(std::string_view toCopyFrom)
+{
+    if (toCopyFrom.empty()) {
+        return "";
+    }
+
+    size_t const size = toCopyFrom.size();
+    char * cstr = gcAllocString(size + 1);
+    memcpy(cstr, toCopyFrom.data(), size);
+    cstr[size] = '\0';
+
+    return cstr;
+}
+
+}

src/libexpr/gc-alloc.hh

@@ -0,0 +1,150 @@
+#pragma once
+/// @file Aliases and wrapper functions that are transparently GC-enabled
+/// if Lix is compiled with BoehmGC enabled.
+
+#include <cstddef>
+#include <cstring>
+#include <list>
+#include <map>
+#include <new>
+#include <string_view>
+#include <vector>
+
+#include "checked-arithmetic.hh"
+
+#if HAVE_BOEHMGC
+#include <functional> // std::less
+#include <utility> // std::pair
+#define GC_INCLUDE_NEW
+#include <gc/gc.h>
+#include <gc/gc_allocator.h>
+#include <gc/gc_cpp.h>
+
+/// calloc, transparently GC-enabled.
+#define LIX_GC_CALLOC(size) GC_MALLOC(size)
+
+/// strdup, transaprently GC-enabled.
+#define LIX_GC_STRDUP(str) GC_STRDUP(str)
+
+/// Atomic GC malloc() with GC enabled, or regular malloc() otherwise.
+#define LIX_GC_MALLOC_ATOMIC(size) GC_MALLOC_ATOMIC(size)
+
+namespace nix
+{
+
+/// Alias for std::map which uses BoehmGC's allocator conditional on this Lix
+/// build having GC enabled.
+template<typename KeyT, typename ValueT>
+using GcMap = std::map<
+    KeyT,
+    ValueT,
+    std::less<KeyT>,
+    traceable_allocator<std::pair<KeyT const, ValueT>>
+>;
+
+/// Alias for std::vector which uses BoehmGC's allocator conditional on this Lix
+/// build having GC enabled.
+template<typename ItemT>
+using GcVector = std::vector<ItemT, traceable_allocator<ItemT>>;
+
+/// Alias for std::list which uses BoehmGC's allocator conditional on this Lix
+/// build having GC enabled.
+template<typename ItemT>
+using GcList = std::list<ItemT, traceable_allocator<ItemT>>;
+
+}
+
+#else
+
+#include <cstdlib>
+
+/// calloc, transparently GC-enabled.
+#define LIX_GC_CALLOC(size) calloc(size, 1)
+
+/// strdup, transparently GC-enabled.
+#define LIX_GC_STRDUP(str) strdup(str)
+
+/// Atomic GC malloc() with GC enabled, or regular malloc() otherwise.
+/// The returned memory must never contain pointers.
+#define LIX_GC_MALLOC_ATOMIC(size) malloc(size)
+
+namespace nix
+{
+
+/// Alias for std::map which uses BoehmGC's allocator conditional on this Lix
+/// build having GC enabled.
+template<typename KeyT, typename ValueT>
+using GcMap = std::map<KeyT, ValueT>;
+
+/// Alias for std::vector which uses BoehmGC's allocator conditional on this Lix
+/// build having GC enabled.
+template<typename ItemT>
+using GcVector = std::vector<ItemT>;
+
+/// Alias for std::list which uses BoehmGC's allocator conditional on this Lix
+/// build having GC enabled.
+template<typename ItemT>
+using GcList = std::list<ItemT>;
+
+}
+
+#endif
+
+namespace nix
+{
+
+[[gnu::always_inline]]
+inline void * gcAllocBytes(size_t n)
+{
+    // Note: various places expect the allocated memory to be zero.
+    // Hence: calloc().
+    void * ptr = LIX_GC_CALLOC(n);
+    if (ptr == nullptr) {
+        throw std::bad_alloc();
+    }
+
+    return ptr;
+}
+
+/// Typed, safe wrapper around calloc() (transparently GC-enabled). Allocates
+/// enough for the requested count of the specified type. Also checks for
+/// nullptr (and throws @ref std::bad_alloc), and casts the void pointer to
+/// a pointer of the specified type, for type-convenient goodness.
+template<typename T>
+[[gnu::always_inline]]
+inline T * gcAllocType(size_t howMany = 1)
+{
+    // NOTE: size_t * size_t, which can definitely overflow.
+    // Unsigned integer overflow is definitely a bug, but isn't undefined
+    // behavior, so we can just check if we overflowed after the fact.
+    // However, people can and do request zero sized allocations, so we need
+    // to check that neither of our multiplicands were zero before complaining
+    // about it.
+    auto checkedSz = checked::Checked<size_t>(howMany) * sizeof(T);
+    size_t sz = checkedSz.valueWrapping();
+    if (checkedSz.overflowed()) {
+        // Congrats, you done did an overflow.
+        throw std::bad_alloc();
+    }
+
+    return static_cast<T *>(gcAllocBytes(sz));
+}
+
+/// GC-transparently allocates a buffer for a C-string of @ref size *bytes*,
+/// meaning you should include the size needed by the NUL terminator in the
+/// passed size. Memory allocated with this function must never contain other
+/// pointers.
+inline char * gcAllocString(size_t size)
+{
+    char * cstr = static_cast<char *>(LIX_GC_MALLOC_ATOMIC(size));
+    if (cstr == nullptr) {
+        throw std::bad_alloc();
+    }
+    return cstr;
+}
+
+/// Returns a C-string copied from @ref toCopyFrom, or a single, static empty
+/// string if @ref toCopyFrom is also empty.
+char const * gcCopyStringIfNeeded(std::string_view toCopyFrom);
+
+}

src/libexpr/get-drvs.cc

@@ -1,6 +1,7 @@
 #include "get-drvs.hh"
 #include "eval-inline.hh"
 #include "derivations.hh"
+#include "eval.hh"
 #include "store-api.hh"
 #include "path-with-outputs.hh"
 
@@ -101,66 +102,134 @@ StorePath DrvInfo::queryOutPath()
     return *outPath;
 }
 
+void DrvInfo::fillOutputs(bool withPaths)
+{
+    auto fillDefault = [&]() {
+        std::optional<StorePath> outPath = std::nullopt;
+        if (withPaths) {
+            outPath.emplace(this->queryOutPath());
+        }
+        this->outputs.emplace("out", outPath);
+    };
+
+    // lol. lmao even.
+    if (this->attrs == nullptr) {
+        fillDefault();
+        return;
+    }
+
+    Attr * outputs = this->attrs->get(this->state->sOutputs);
+    if (outputs == nullptr) {
+        fillDefault();
+        return;
+    }
+
+    // NOTE(Qyriad): I don't think there is any codepath that can cause this to error.
+    this->state->forceList(
+        *outputs->value,
+        outputs->pos,
+        "while evaluating the 'outputs' attribute of a derivation"
+    );
+
+    for (auto [idx, elem] : enumerate(outputs->value->listItems())) {
+        // NOTE(Qyriad): This error should be *extremely* rare in practice.
+        // It is impossible to construct with `stdenv.mkDerivation`,
+        // `builtins.derivation`, or even `derivationStrict`. As far as we can tell,
+        // it is only possible by overriding a derivation attrset already created by
+        // one of those with `//` to introduce the failing `outputs` entry.
+        auto errMsg = fmt("while evaluating output %d of a derivation", idx);
+        std::string_view outputName = state->forceStringNoCtx(
+            *elem,
+            outputs->pos,
+            errMsg
+        );
+
+        if (withPaths) {
+            // Find the attr with this output's name...
+            Attr * out = this->attrs->get(this->state->symbols.create(outputName));
+            if (out == nullptr) {
+                // FIXME: throw error?
+                continue;
+            }
+
+            // Meanwhile we couldn't figure out any circumstances
+            // that cause this to error.
+            state->forceAttrs(*out->value, outputs->pos, errMsg);
+
+            // ...and evaluate its `outPath` attribute.
+            Attr * outPath = out->value->attrs->get(this->state->sOutPath);
+            if (outPath == nullptr) {
+                continue;
+                // FIXME: throw error?
+            }
+
+            NixStringContext context;
+            // And idk what could possibly cause this one to error
+            // that wouldn't error before here.
+            auto storePath = state->coerceToStorePath(
+                outPath->pos,
+                *outPath->value,
+                context,
+                errMsg
+            );
+            this->outputs.emplace(outputName, storePath);
+        } else {
+            this->outputs.emplace(outputName, std::nullopt);
+        }
+    }
+}
 
 DrvInfo::Outputs DrvInfo::queryOutputs(bool withPaths, bool onlyOutputsToInstall)
 {
+    // If we haven't already cached the outputs set, then do so now.
     if (outputs.empty()) {
-        /* Get the ‘outputs’ list. */
-        Bindings::iterator i;
-        if (attrs && (i = attrs->find(state->sOutputs)) != attrs->end()) {
-            state->forceList(*i->value, i->pos, "while evaluating the 'outputs' attribute of a derivation");
-
-            /* For each output... */
-            for (auto elem : i->value->listItems()) {
-                std::string output(state->forceStringNoCtx(*elem, i->pos, "while evaluating the name of an output of a derivation"));
-
-                if (withPaths) {
-                    /* Evaluate the corresponding set. */
-                    Bindings::iterator out = attrs->find(state->symbols.create(output));
-                    if (out == attrs->end()) continue; // FIXME: throw error?
-                    state->forceAttrs(*out->value, i->pos, "while evaluating an output of a derivation");
-
-                    /* And evaluate its ‘outPath’ attribute. */
-                    Bindings::iterator outPath = out->value->attrs->find(state->sOutPath);
-                    if (outPath == out->value->attrs->end()) continue; // FIXME: throw error?
-                    NixStringContext context;
-                    outputs.emplace(output, state->coerceToStorePath(outPath->pos, *outPath->value, context, "while evaluating an output path of a derivation"));
-                } else
-                    outputs.emplace(output, std::nullopt);
-            }
-        } else
-            outputs.emplace("out", withPaths ? std::optional{queryOutPath()} : std::nullopt);
+        // FIXME: this behavior seems kind of busted, since whether or not this
+        // DrvInfo will have paths is forever determined by the *first* call to
+        // this function??
+        fillOutputs(withPaths);
     }
 
+    // Things that operate on derivations like packages, like `nix-env` and `nix build`,
+    // allow derivations to specify which outputs should be used in those user-facing
+    // cases if the user didn't specify an output explicitly.
+    // If the caller just wanted all the outputs for this derivation, though,
+    // then we're done here.
     if (!onlyOutputsToInstall || !attrs)
         return outputs;
 
-    Bindings::iterator i;
-    if (attrs && (i = attrs->find(state->sOutputSpecified)) != attrs->end() && state->forceBool(*i->value, i->pos, "while evaluating the 'outputSpecified' attribute of a derivation")) {
-        Outputs result;
-        auto out = outputs.find(queryOutputName());
-        if (out == outputs.end())
-            throw Error("derivation does not have output '%s'", queryOutputName());
-        result.insert(*out);
-        return result;
+    // Regardless of `meta.outputsToInstall`, though, you can select into a derivation
+    // output by its attribute, e.g. `pkgs.lix.dev`, which (lol?) sets the magic
+    // attribute `outputSpecified = true`, and changes the `outputName` attr to the
+    // explicitly selected-into output.
+    if (Attr * outSpecAttr = attrs->get(state->sOutputSpecified)) {
+        bool outputSpecified = this->state->forceBool(
+            *outSpecAttr->value,
+            outSpecAttr->pos,
+            "while evaluating the 'outputSpecified' attribute of a derivation"
+        );
+        if (outputSpecified) {
+            auto maybeOut = outputs.find(queryOutputName());
+            if (maybeOut == outputs.end()) {
+                throw Error("derivation does not have output '%s'", queryOutputName());
+            }
+            return Outputs{*maybeOut};
+        }
     }
 
-    else {
-        /* Check for `meta.outputsToInstall` and return `outputs` reduced to that. */
-        const Value * outTI = queryMeta("outputsToInstall");
-        if (!outTI) return outputs;
-        auto errMsg = Error("this derivation has bad 'meta.outputsToInstall'");
-            /* ^ this shows during `nix-env -i` right under the bad derivation */
-        if (!outTI->isList()) throw errMsg;
-        Outputs result;
-        for (auto elem : outTI->listItems()) {
-            if (elem->type() != nString) throw errMsg;
-            auto out = outputs.find(elem->string.s);
-            if (out == outputs.end()) throw errMsg;
-            result.insert(*out);
-        }
-        return result;
+    /* Check for `meta.outputsToInstall` and return `outputs` reduced to that. */
+    const Value * outTI = queryMeta("outputsToInstall");
+    if (!outTI) return outputs;
+    auto errMsg = Error("this derivation has bad 'meta.outputsToInstall'");
+        /* ^ this shows during `nix-env -i` right under the bad derivation */
+    if (!outTI->isList()) throw errMsg;
+    Outputs result;
+    for (auto elem : outTI->listItems()) {
+        if (elem->type() != nString) throw errMsg;
+        auto out = outputs.find(elem->string.s);
+        if (out == outputs.end()) throw errMsg;
+        result.insert(*out);
     }
+    return result;
 }
 
 
@@ -241,8 +310,8 @@ NixInt DrvInfo::queryMetaInt(const std::string & name, NixInt def)
     if (v->type() == nString) {
         /* Backwards compatibility with before we had support for
            integer meta fields. */
-        if (auto n = string2Int<NixInt>(v->string.s))
-            return *n;
+        if (auto n = string2Int<NixInt::Inner>(v->string.s))
+            return NixInt{*n};
     }
     return def;
 }
@@ -350,56 +419,95 @@ static void getDerivations(EvalState & state, Value & vIn,
     Value v;
     state.autoCallFunction(autoArgs, vIn, v);
 
-    /* Process the expression. */
-    if (!getDerivation(state, v, pathPrefix, drvs, ignoreAssertionFailures)) ;
-
-    else if (v.type() == nAttrs) {
-
-        /* Dont consider sets we've already seen, e.g. y in
-           `rec { x.d = derivation {...}; y = x; }`. */
-        if (!done.insert(v.attrs).second) return;
-
-        /* !!! undocumented hackery to support combining channels in
-           nix-env.cc. */
-        bool combineChannels = v.attrs->find(state.symbols.create("_combineChannels")) != v.attrs->end();
-
-        /* Consider the attributes in sorted order to get more
-           deterministic behaviour in nix-env operations (e.g. when
-           there are names clashes between derivations, the derivation
-           bound to the attribute with the "lower" name should take
-           precedence). */
-        for (auto & i : v.attrs->lexicographicOrder(state.symbols)) {
-            debug("evaluating attribute '%1%'", state.symbols[i->name]);
-            if (!std::regex_match(std::string(state.symbols[i->name]), attrRegex))
-                continue;
-            std::string pathPrefix2 = addToPath(pathPrefix, state.symbols[i->name]);
-            if (combineChannels)
-                getDerivations(state, *i->value, pathPrefix2, autoArgs, drvs, done, ignoreAssertionFailures);
-            else if (getDerivation(state, *i->value, pathPrefix2, drvs, ignoreAssertionFailures)) {
-                /* If the value of this attribute is itself a set,
-                   should we recurse into it?  => Only if it has a
-                   `recurseForDerivations = true' attribute. */
-                if (i->value->type() == nAttrs) {
-                    Bindings::iterator j = i->value->attrs->find(state.sRecurseForDerivations);
-                    if (j != i->value->attrs->end() && state.forceBool(*j->value, j->pos, "while evaluating the attribute `recurseForDerivations`"))
-                        getDerivations(state, *i->value, pathPrefix2, autoArgs, drvs, done, ignoreAssertionFailures);
-                }
-            }
-        }
+    bool shouldRecurse = getDerivation(state, v, pathPrefix, drvs, ignoreAssertionFailures);
+    if (!shouldRecurse) {
+        // We're done here.
+        return;
     }
 
-    else if (v.type() == nList) {
+    if (v.type() == nList) {
         // NOTE we can't really deduplicate here because small lists don't have stable addresses
         // and can cause spurious duplicate detections due to v being on the stack.
         for (auto [n, elem] : enumerate(v.listItems())) {
-            std::string pathPrefix2 = addToPath(pathPrefix, fmt("%d", n));
-            if (getDerivation(state, *elem, pathPrefix2, drvs, ignoreAssertionFailures))
-                getDerivations(state, *elem, pathPrefix2, autoArgs, drvs, done, ignoreAssertionFailures);
+            std::string joinedAttrPath = addToPath(pathPrefix, fmt("%d", n));
+            bool shouldRecurse = getDerivation(state, *elem, joinedAttrPath, drvs, ignoreAssertionFailures);
+            if (shouldRecurse) {
+                getDerivations(
+                    state,
+                    *elem,
+                    joinedAttrPath,
+                    autoArgs,
+                    drvs,
+                    done,
+                    ignoreAssertionFailures
+                );
+            }
         }
+
+        return;
+    } else if (v.type() != nAttrs) {
+        state.error<TypeError>(
+            "expression was expected to be a derivation or collection of derivations, but instead was %s",
+            showType(v.type(), true)
+        ).debugThrow();
     }
 
-    else
-        state.error<TypeError>("expression does not evaluate to a derivation (or a set or list of those)").debugThrow();
+    /* Dont consider sets we've already seen, e.g. y in
+       `rec { x.d = derivation {...}; y = x; }`. */
+    auto const &[_, didInsert] = done.insert(v.attrs);
+    if (!didInsert) {
+        return;
+    }
+
+    // FIXME: what the fuck???
+    /* !!! undocumented hackery to support combining channels in
+       nix-env.cc. */
+    bool combineChannels = v.attrs->find(state.symbols.create("_combineChannels")) != v.attrs->end();
+
+    /* Consider the attributes in sorted order to get more
+       deterministic behaviour in nix-env operations (e.g. when
+       there are names clashes between derivations, the derivation
+       bound to the attribute with the "lower" name should take
+       precedence). */
+    for (auto & attr : v.attrs->lexicographicOrder(state.symbols)) {
+        debug("evaluating attribute '%1%'", state.symbols[attr->name]);
+        // FIXME: only consider attrs with identifier-like names?? Why???
+        if (!std::regex_match(std::string(state.symbols[attr->name]), attrRegex)) {
+            continue;
+        }
+        std::string joinedAttrPath = addToPath(pathPrefix, state.symbols[attr->name]);
+        if (combineChannels) {
+            getDerivations(state, *attr->value, joinedAttrPath, autoArgs, drvs, done, ignoreAssertionFailures);
+        } else if (getDerivation(state, *attr->value, joinedAttrPath, drvs, ignoreAssertionFailures)) {
+            /* If the value of this attribute is itself a set,
+               should we recurse into it?  => Only if it has a
+               `recurseForDerivations = true' attribute. */
+            if (attr->value->type() == nAttrs) {
+                Attr * recurseForDrvs = attr->value->attrs->get(state.sRecurseForDerivations);
+                if (recurseForDrvs == nullptr) {
+                    continue;
+                }
+                bool shouldRecurse = state.forceBool(
+                    *recurseForDrvs->value,
+                    attr->pos,
+                    fmt("while evaluating the '%s' attribute", Magenta("recurseForDerivations"))
+                );
+                if (!shouldRecurse) {
+                    continue;
+                }
+
+                getDerivations(
+                    state,
+                    *attr->value,
+                    joinedAttrPath,
+                    autoArgs,
+                    drvs,
+                    done,
+                    ignoreAssertionFailures
+                );
+            }
+        }
+    }
 }
 
 

src/libexpr/get-drvs.hh

@@ -37,13 +37,14 @@ private:
 
     bool checkMeta(Value & v);
 
+    void fillOutputs(bool withPaths = true);
+
 public:
     /**
      * path towards the derivation
      */
     std::string attrPath;
 
-    DrvInfo(EvalState & state) : state(&state) { };
     DrvInfo(EvalState & state, std::string attrPath, Bindings * attrs);
     DrvInfo(EvalState & state, ref<Store> store, const std::string & drvPathWithOutputs);
 
@@ -80,13 +81,7 @@ public:
     bool hasFailed() { return failed; };
 };
 
-
-#if HAVE_BOEHMGC
-typedef std::list<DrvInfo, traceable_allocator<DrvInfo>> DrvInfos;
-#else
-typedef std::list<DrvInfo> DrvInfos;
-#endif
-
+using DrvInfos = GcList<DrvInfo>;
 
 /**
  * If value `v` denotes a derivation, return a DrvInfo object

src/libexpr/json-to-value.cc

@@ -2,6 +2,7 @@
 #include "value.hh"
 #include "eval.hh"
 
+#include <limits>
 #include <variant>
 #include <nlohmann/json.hpp>
 
@@ -81,42 +82,46 @@ class JSONSax : nlohmann::json_sax<json> {
 public:
     JSONSax(EvalState & state, Value & v) : state(state), rs(new JSONState(&v)) {};
 
-    bool null()
+    bool null() override
     {
         rs->value(state).mkNull();
         rs->add();
         return true;
     }
 
-    bool boolean(bool val)
+    bool boolean(bool val) override
     {
         rs->value(state).mkBool(val);
         rs->add();
         return true;
     }
 
-    bool number_integer(number_integer_t val)
+    bool number_integer(number_integer_t val) override
     {
         rs->value(state).mkInt(val);
         rs->add();
         return true;
     }
 
-    bool number_unsigned(number_unsigned_t val)
+    bool number_unsigned(number_unsigned_t val_) override
     {
+        if (val_ > std::numeric_limits<NixInt::Inner>::max()) {
+            throw Error("unsigned json number %1% outside of Nix integer range", val_);
+        }
+        NixInt::Inner val = val_;
         rs->value(state).mkInt(val);
         rs->add();
         return true;
     }
 
-    bool number_float(number_float_t val, const string_t & s)
+    bool number_float(number_float_t val, const string_t & s) override
     {
         rs->value(state).mkFloat(val);
         rs->add();
         return true;
     }
 
-    bool string(string_t & val)
+    bool string(string_t & val) override
     {
         rs->value(state).mkString(val);
         rs->add();
@@ -124,7 +129,7 @@ public:
     }
 
 #if NLOHMANN_JSON_VERSION_MAJOR >= 3 && NLOHMANN_JSON_VERSION_MINOR >= 8
-    bool binary(binary_t&)
+    bool binary(binary_t&) override
     {
         // This function ought to be unreachable
         assert(false);
@@ -132,35 +137,37 @@ public:
     }
 #endif
 
-    bool start_object(std::size_t len)
+    bool start_object(std::size_t len) override
     {
         rs = std::make_unique<JSONObjectState>(std::move(rs));
         return true;
     }
 
-    bool key(string_t & name)
+    bool key(string_t & name) override
     {
         dynamic_cast<JSONObjectState*>(rs.get())->key(name, state);
         return true;
     }
 
-    bool end_object() {
+    bool end_object() override {
         rs = rs->resolve(state);
         rs->add();
         return true;
     }
 
-    bool end_array() {
+    bool end_array() override {
         return end_object();
     }
 
-    bool start_array(size_t len) {
+    bool start_array(size_t len) override {
         rs = std::make_unique<JSONListState>(std::move(rs),
             len != std::numeric_limits<size_t>::max() ? len : 128);
         return true;
     }
 
-    bool parse_error(std::size_t, const std::string&, const nlohmann::detail::exception& ex) {
+    bool
+    parse_error(std::size_t, const std::string &, const nlohmann::detail::exception & ex) override
+    {
         throw JSONParseError("%s", ex.what());
     }
 };

src/libexpr/meson.build

@@ -22,6 +22,7 @@ libexpr_sources = files(
   'eval.cc',
   'function-trace.cc',
   'get-drvs.cc',
+  'gc-alloc.cc',
   'json-to-value.cc',
   'nixexpr.cc',
   'parser/parser.cc',
@@ -30,6 +31,7 @@ libexpr_sources = files(
   'print-ambiguous.cc',
   'print.cc',
   'search-path.cc',
+  'value.cc',
   'value-to-json.cc',
   'value-to-xml.cc',
   'flake/config.cc',
@@ -58,6 +60,7 @@ libexpr_headers = files(
   'function-trace.hh',
   'gc-small-vector.hh',
   'get-drvs.hh',
+  'gc-alloc.hh',
   'json-to-value.hh',
   'nixexpr.hh',
   'parser/change_head.hh',

src/libexpr/nixexpr.hh

@@ -75,6 +75,7 @@ struct ExprInt : Expr
     NixInt n;
     Value v;
     ExprInt(NixInt n) : n(n) { v.mkInt(n); };
+    ExprInt(NixInt::Inner n) : n(n) { v.mkInt(n); };
     Value * maybeThunk(EvalState & state, Env & env) override;
     COMMON_METHODS
 };

src/libexpr/parser/grammar.hh

@@ -434,6 +434,8 @@ struct op {
     struct and_       : _op<TAO_PEGTL_STRING("&&"), 12> {};
     struct or_        : _op<TAO_PEGTL_STRING("||"), 13> {};
     struct implies    : _op<TAO_PEGTL_STRING("->"), 14, kind::rightAssoc> {};
+    struct pipe_right : _op<TAO_PEGTL_STRING("|>"), 15> {};
+    struct pipe_left  : _op<TAO_PEGTL_STRING("<|"), 16, kind::rightAssoc> {};
 };
 
 struct _expr {
@@ -521,6 +523,7 @@ struct _expr {
         app
     > {};
 
+    /* Order matters here. The order is the parsing order, not the precedence order: '<=' must be parsed before '<'. */
     struct _binary_operator : sor<
         operator_<op::implies>,
         operator_<op::update>,
@@ -529,6 +532,8 @@ struct _expr {
         operator_<op::minus>,
         operator_<op::mul>,
         operator_<op::div>,
+        operator_<op::pipe_right>,
+        operator_<op::pipe_left>,
         operator_<op::less_eq>,
         operator_<op::greater_eq>,
         operator_<op::less>,
@@ -649,6 +654,8 @@ struct operator_semantics {
             grammar::op::minus,
             grammar::op::mul,
             grammar::op::div,
+            grammar::op::pipe_right,
+            grammar::op::pipe_left,
             has_attr
         > op;
     };

src/libexpr/parser/parser.cc

@@ -12,7 +12,6 @@
 #include "state.hh"
 
 #include <charconv>
-#include <clocale>
 #include <memory>
 
 // flip this define when doing parser development to enable some g checks.
@@ -43,7 +42,6 @@ error_message_for(p::one<'}'>) = "expecting '}'";
 error_message_for(p::one<'"'>) = "expecting '\"'";
 error_message_for(p::one<';'>) = "expecting ';'";
 error_message_for(p::one<')'>) = "expecting ')'";
-error_message_for(p::one<'='>) = "expecting '='";
 error_message_for(p::one<']'>) = "expecting ']'";
 error_message_for(p::one<':'>) = "expecting ':'";
 error_message_for(p::string<'\'', '\''>) = "expecting \"''\"";
@@ -115,6 +113,29 @@ struct ExprState
         return std::make_unique<ExprCall>(pos, std::make_unique<ExprVar>(fn), std::move(args));
     }
 
+    std::unique_ptr<Expr> pipe(PosIdx pos, State & state, bool flip = false)
+    {
+        if (!state.xpSettings.isEnabled(Xp::PipeOperator))
+            throw ParseError({
+                .msg = HintFmt("Pipe operator is disabled"),
+                .pos = state.positions[pos]
+            });
+
+        // Reverse the order compared to normal function application: arg |> fn
+        std::unique_ptr<Expr> fn, arg;
+        if (flip) {
+            fn = popExprOnly();
+            arg = popExprOnly();
+        } else {
+            arg = popExprOnly();
+            fn = popExprOnly();
+        }
+        std::vector<std::unique_ptr<Expr>> args{1};
+        args[0] = std::move(arg);
+
+        return std::make_unique<ExprCall>(pos, std::move(fn), std::move(args));
+    }
+
     std::unique_ptr<Expr> order(PosIdx pos, bool less, State & state)
     {
         return call(pos, state.s.lessThan, !less);
@@ -164,6 +185,8 @@ struct ExprState
                 [&] (Op::concat)      { return applyBinary<ExprOpConcatLists>(pos); },
                 [&] (has_attr & a)    { return applyUnary<ExprOpHasAttr>(std::move(a.path)); },
                 [&] (Op::unary_minus) { return negate(pos, state); },
+                [&] (Op::pipe_right)  { return pipe(pos, state, true); },
+                [&] (Op::pipe_left)   { return pipe(pos, state); },
             })(op)
         };
     }
@@ -255,7 +278,8 @@ struct AttrState : SubexprState {
 
     std::vector<AttrName> attrs;
 
-    void pushAttr(auto && attr, PosIdx) { attrs.emplace_back(std::move(attr)); }
+    template <typename T>
+    void pushAttr(T && attr, PosIdx) { attrs.emplace_back(std::forward<T>(attr)); }
 };
 
 template<> struct BuildAST<grammar::attr::simple> {
@@ -291,7 +315,8 @@ struct InheritState : SubexprState {
     std::unique_ptr<Expr> from;
     PosIdx fromPos;
 
-    void pushAttr(auto && attr, PosIdx pos) { attrs.emplace_back(std::move(attr), pos); }
+    template <typename T>
+    void pushAttr(T && attr, PosIdx pos) { attrs.emplace_back(std::forward<T>(attr), pos); }
 };
 
 template<> struct BuildAST<grammar::inherit::from> {
@@ -631,7 +656,7 @@ template<> struct BuildAST<grammar::expr::path> : p::maybe_nothing {};
 
 template<> struct BuildAST<grammar::expr::uri> {
     static void apply(const auto & in, ExprState & s, State & ps) {
-       static bool noURLLiterals = experimentalFeatureSettings.isEnabled(Xp::NoUrlLiterals);
+       bool noURLLiterals = ps.xpSettings.isEnabled(Xp::NoUrlLiterals);
        if (noURLLiterals)
            throw ParseError({
                .msg = HintFmt("URL literals are disabled"),
@@ -832,7 +857,8 @@ Expr * EvalState::parse(
     size_t length,
     Pos::Origin origin,
     const SourcePath & basePath,
-    std::shared_ptr<StaticEnv> & staticEnv)
+    std::shared_ptr<StaticEnv> & staticEnv,
+    const ExperimentalFeatureSettings & xpSettings)
 {
     parser::State s = {
         symbols,
@@ -840,6 +866,7 @@ Expr * EvalState::parse(
         basePath,
         positions.addOrigin(origin, length),
         exprSymbols,
+        xpSettings
     };
     parser::ExprState x;
 

src/libexpr/parser/state.hh

@@ -19,6 +19,7 @@ struct State
     SourcePath basePath;
     PosTable::Origin origin;
     const Expr::AstSymbols & s;
+    const ExperimentalFeatureSettings & xpSettings;
 
     void dupAttr(const AttrPath & attrPath, const PosIdx pos, const PosIdx prevPos);
     void dupAttr(Symbol attr, const PosIdx pos, const PosIdx prevPos);

src/libexpr/primops.cc

@@ -346,7 +346,7 @@ void prim_importNative(EvalState & state, const PosIdx pos, Value * * args, Valu
         state.error<EvalError>("could not open '%1%': %2%", path, dlerror()).debugThrow();
 
     dlerror();
-    ValueInitializer func = (ValueInitializer) dlsym(handle, sym.c_str());
+    ValueInitializer func = reinterpret_cast<ValueInitializer>(dlsym(handle, sym.c_str()));
     if(!func) {
         char *message = dlerror();
         if (message)
@@ -582,9 +582,9 @@ struct CompareValues
     {
         try {
             if (v1->type() == nFloat && v2->type() == nInt)
-                return v1->fpoint < v2->integer;
+                return v1->fpoint < v2->integer.value;
             if (v1->type() == nInt && v2->type() == nFloat)
-                return v1->integer < v2->fpoint;
+                return v1->integer.value < v2->fpoint;
             if (v1->type() != v2->type())
                 state.error<EvalError>("cannot compare %s with %s", showType(*v1), showType(*v2)).debugThrow();
             // Allow selecting a subset of enum values
@@ -622,14 +622,13 @@ struct CompareValues
     }
 };
 
-
+/// NOTE: this type must NEVER be outside of GC-scanned memory.
 #if HAVE_BOEHMGC
-typedef std::list<Value *, gc_allocator<Value *>> ValueList;
+using UnsafeValueList = std::list<Value *, gc_allocator<Value *>>;
 #else
-typedef std::list<Value *> ValueList;
+using UnsafeValueList = std::list<Value *>;
 #endif
 
-
 static Bindings::iterator getAttr(
     EvalState & state,
     Symbol attrSym,
@@ -652,7 +651,7 @@ static void prim_genericClosure(EvalState & state, const PosIdx pos, Value * * a
 
     state.forceList(*startSet->value, noPos, "while evaluating the 'startSet' attribute passed as argument to builtins.genericClosure");
 
-    ValueList workSet;
+    UnsafeValueList workSet;
     for (auto elem : startSet->value->listItems())
         workSet.push_back(elem);
 
@@ -668,7 +667,7 @@ static void prim_genericClosure(EvalState & state, const PosIdx pos, Value * * a
     /* Construct the closure by applying the operator to elements of
        `workSet', adding the result to `workSet', continuing until
        no new elements are found. */
-    ValueList res;
+    UnsafeValueList res;
     // `doneKeys' doesn't need to be a GC root, because its values are
     // reachable from res.
     auto cmp = CompareValues(state, noPos, "while comparing the `key` attributes of two genericClosure elements");
@@ -1214,6 +1213,20 @@ drvName, Bindings * attrs, Value & v)
                         handleOutputs(ss);
                     }
 
+                    if (i->name == state.sAllowedReferences)
+                        warn("In a derivation named '%s', 'structuredAttrs' disables the effect of the derivation attribute 'allowedReferences'; use 'outputChecks.<output>.allowedReferences' instead", drvName);
+                    if (i->name == state.sAllowedRequisites)
+                        warn("In a derivation named '%s', 'structuredAttrs' disables the effect of the derivation attribute 'allowedRequisites'; use 'outputChecks.<output>.allowedRequisites' instead", drvName);
+                    if (i->name == state.sDisallowedReferences)
+                        warn("In a derivation named '%s', 'structuredAttrs' disables the effect of the derivation attribute 'disallowedReferences'; use 'outputChecks.<output>.disallowedReferences' instead", drvName);
+                    if (i->name == state.sDisallowedRequisites)
+                        warn("In a derivation named '%s', 'structuredAttrs' disables the effect of the derivation attribute 'disallowedRequisites'; use 'outputChecks.<output>.disallowedRequisites' instead", drvName);
+                    if (i->name == state.sMaxSize)
+                        warn("In a derivation named '%s', 'structuredAttrs' disables the effect of the derivation attribute 'maxSize'; use 'outputChecks.<output>.maxSize' instead", drvName);
+                    if (i->name == state.sMaxClosureSize)
+                        warn("In a derivation named '%s', 'structuredAttrs' disables the effect of the derivation attribute 'maxClosureSize'; use 'outputChecks.<output>.maxClosureSize' instead", drvName);
+
+
                 } else {
                     auto s = state.coerceToString(pos, *i->value, context, context_below, true).toOwned();
                     drv.env.emplace(key, s);
@@ -1323,7 +1336,7 @@ drvName, Bindings * attrs, Value & v)
             state.error<EvalError>("derivation cannot be both content-addressed and impure")
                 .atPos(v).debugThrow();
 
-        auto ht = parseHashTypeOpt(outputHashAlgo).value_or(htSHA256);
+        auto ht = parseHashTypeOpt(outputHashAlgo).value_or(HashType::SHA256);
         auto method = ingestionMethod.value_or(FileIngestionMethod::Recursive);
 
         for (auto & i : outputs) {
@@ -1751,7 +1764,7 @@ static void prim_hashFile(EvalState & state, const PosIdx pos, Value * * args, V
 
     auto path = realisePath(state, pos, *args[1]);
 
-    v.mkString(hashString(*ht, path.readFile()).to_string(Base16, false));
+    v.mkString(hashString(*ht, path.readFile()).to_string(Base::Base16, false));
 }
 
 static RegisterPrimOp primop_hashFile({
@@ -2333,7 +2346,7 @@ static void prim_path(EvalState & state, const PosIdx pos, Value * * args, Value
         else if (n == "recursive")
             method = FileIngestionMethod { state.forceBool(*attr.value, attr.pos, "while evaluating the `recursive` attribute passed to builtins.path") };
         else if (n == "sha256")
-            expectedHash = newHashAllowEmpty(state.forceStringNoCtx(*attr.value, attr.pos, "while evaluating the `sha256` attribute passed to builtins.path"), htSHA256);
+            expectedHash = newHashAllowEmpty(state.forceStringNoCtx(*attr.value, attr.pos, "while evaluating the `sha256` attribute passed to builtins.path"), HashType::SHA256);
         else
             state.error<EvalError>(
                 "unsupported argument '%1%' to 'addPath'",
@@ -2426,6 +2439,8 @@ static void prim_attrValues(EvalState & state, const PosIdx pos, Value * * args,
 
     state.mkList(v, args[0]->attrs->size());
 
+    // FIXME: this is incredibly evil, *why*
+    // NOLINTBEGIN(cppcoreguidelines-pro-type-cstyle-cast)
     unsigned int n = 0;
     for (auto & i : *args[0]->attrs)
         v.listElems()[n++] = (Value *) &i;
@@ -2439,6 +2454,7 @@ static void prim_attrValues(EvalState & state, const PosIdx pos, Value * * args,
 
     for (unsigned int i = 0; i < n; ++i)
         v.listElems()[i] = ((Attr *) v.listElems()[i])->value;
+    // NOLINTEND(cppcoreguidelines-pro-type-cstyle-cast)
 }
 
 static RegisterPrimOp primop_attrValues({
@@ -2514,13 +2530,13 @@ static struct LazyPosAcessors {
     PrimOp primop_lineOfPos{
         .arity = 1,
         .fun = [] (EvalState & state, PosIdx pos, Value * * args, Value & v) {
-            v.mkInt(state.positions[PosIdx(args[0]->integer)].line);
+            v.mkInt(state.positions[PosIdx(args[0]->integer.value)].line);
         }
     };
     PrimOp primop_columnOfPos{
         .arity = 1,
         .fun = [] (EvalState & state, PosIdx pos, Value * * args, Value & v) {
-            v.mkInt(state.positions[PosIdx(args[0]->integer)].column);
+            v.mkInt(state.positions[PosIdx(args[0]->integer.value)].column);
         }
     };
 
@@ -2990,7 +3006,8 @@ static void elemAt(EvalState & state, const PosIdx pos, Value & list, int n, Val
 /* Return the n-1'th element of a list. */
 static void prim_elemAt(EvalState & state, const PosIdx pos, Value * * args, Value & v)
 {
-    elemAt(state, pos, *args[0], state.forceInt(*args[1], pos, "while evaluating the second argument passed to builtins.elemAt"), v);
+    NixInt::Inner elem = state.forceInt(*args[1], pos, "while evaluating the second argument passed to builtins.elemAt").value;
+    elemAt(state, pos, *args[0], elem, v);
 }
 
 static RegisterPrimOp primop_elemAt({
@@ -3274,17 +3291,19 @@ static RegisterPrimOp primop_all({
 
 static void prim_genList(EvalState & state, const PosIdx pos, Value * * args, Value & v)
 {
-    auto len = state.forceInt(*args[1], pos, "while evaluating the second argument passed to builtins.genList");
+    auto len_ = state.forceInt(*args[1], pos, "while evaluating the second argument passed to builtins.genList").value;
 
-    if (len < 0)
-        state.error<EvalError>("cannot create list of size %1%", len).atPos(pos).debugThrow();
+    if (len_ < 0)
+        state.error<EvalError>("cannot create list of size %1%", len_).atPos(pos).debugThrow();
+
+    size_t len = len_;
 
     // More strict than striclty (!) necessary, but acceptable
     // as evaluating map without accessing any values makes little sense.
     state.forceFunction(*args[0], noPos, "while evaluating the first argument passed to builtins.genList");
 
     state.mkList(v, len);
-    for (unsigned int n = 0; n < (unsigned int) len; ++n) {
+    for (size_t n = 0; n < len; ++n) {
         auto arg = state.allocValue();
         arg->mkInt(n);
         (v.listElems()[n] = state.allocValue())->mkApp(args[0], arg);
@@ -3534,9 +3553,17 @@ static void prim_add(EvalState & state, const PosIdx pos, Value * * args, Value
     if (args[0]->type() == nFloat || args[1]->type() == nFloat)
         v.mkFloat(state.forceFloat(*args[0], pos, "while evaluating the first argument of the addition")
                 + state.forceFloat(*args[1], pos, "while evaluating the second argument of the addition"));
-    else
-        v.mkInt(  state.forceInt(*args[0], pos, "while evaluating the first argument of the addition")
-                + state.forceInt(*args[1], pos, "while evaluating the second argument of the addition"));
+    else {
+        auto i1 = state.forceInt(*args[0], pos, "while evaluating the first argument of the addition");
+        auto i2 = state.forceInt(*args[1], pos, "while evaluating the second argument of the addition");
+
+        auto result_ = i1 + i2;
+        if (auto result = result_.valueChecked(); result.has_value()) {
+            v.mkInt(*result);
+        } else {
+            state.error<EvalError>("integer overflow in adding %1% + %2%", i1, i2).atPos(pos).debugThrow();
+        }
+    }
 }
 
 static RegisterPrimOp primop_add({
@@ -3555,9 +3582,18 @@ static void prim_sub(EvalState & state, const PosIdx pos, Value * * args, Value
     if (args[0]->type() == nFloat || args[1]->type() == nFloat)
         v.mkFloat(state.forceFloat(*args[0], pos, "while evaluating the first argument of the subtraction")
                 - state.forceFloat(*args[1], pos, "while evaluating the second argument of the subtraction"));
-    else
-        v.mkInt(  state.forceInt(*args[0], pos, "while evaluating the first argument of the subtraction")
-                - state.forceInt(*args[1], pos, "while evaluating the second argument of the subtraction"));
+    else {
+        auto i1 = state.forceInt(*args[0], pos, "while evaluating the first argument of the subtraction");
+        auto i2 = state.forceInt(*args[1], pos, "while evaluating the second argument of the subtraction");
+
+        auto result_ = i1 - i2;
+
+        if (auto result = result_.valueChecked(); result.has_value()) {
+            v.mkInt(*result);
+        } else {
+            state.error<EvalError>("integer overflow in subtracting %1% - %2%", i1, i2).atPos(pos).debugThrow();
+        }
+    }
 }
 
 static RegisterPrimOp primop_sub({
@@ -3576,9 +3612,18 @@ static void prim_mul(EvalState & state, const PosIdx pos, Value * * args, Value
     if (args[0]->type() == nFloat || args[1]->type() == nFloat)
         v.mkFloat(state.forceFloat(*args[0], pos, "while evaluating the first of the multiplication")
                 * state.forceFloat(*args[1], pos, "while evaluating the second argument of the multiplication"));
-    else
-        v.mkInt(  state.forceInt(*args[0], pos, "while evaluating the first argument of the multiplication")
-                * state.forceInt(*args[1], pos, "while evaluating the second argument of the multiplication"));
+    else {
+        auto i1 = state.forceInt(*args[0], pos, "while evaluating the first argument of the multiplication");
+        auto i2 = state.forceInt(*args[1], pos, "while evaluating the second argument of the multiplication");
+
+        auto result_ = i1 * i2;
+
+        if (auto result = result_.valueChecked(); result.has_value()) {
+            v.mkInt(*result);
+        } else {
+            state.error<EvalError>("integer overflow in multiplying %1% * %2%", i1, i2).atPos(pos).debugThrow();
+        }
+    }
 }
 
 static RegisterPrimOp primop_mul({
@@ -3605,10 +3650,12 @@ static void prim_div(EvalState & state, const PosIdx pos, Value * * args, Value
         NixInt i1 = state.forceInt(*args[0], pos, "while evaluating the first operand of the division");
         NixInt i2 = state.forceInt(*args[1], pos, "while evaluating the second operand of the division");
         /* Avoid division overflow as it might raise SIGFPE. */
-        if (i1 == std::numeric_limits<NixInt>::min() && i2 == -1)
-            state.error<EvalError>("overflow in integer division").atPos(pos).debugThrow();
-
-        v.mkInt(i1 / i2);
+        auto result_ = i1 / i2;
+        if (auto result = result_.valueChecked(); result.has_value()) {
+            v.mkInt(*result);
+        } else {
+            state.error<EvalError>("integer overflow in dividing %1% / %2%", i1, i2).atPos(pos).debugThrow();
+        }
     }
 }
 
@@ -3623,8 +3670,9 @@ static RegisterPrimOp primop_div({
 
 static void prim_bitAnd(EvalState & state, const PosIdx pos, Value * * args, Value & v)
 {
-    v.mkInt(state.forceInt(*args[0], pos, "while evaluating the first argument passed to builtins.bitAnd")
-            & state.forceInt(*args[1], pos, "while evaluating the second argument passed to builtins.bitAnd"));
+    auto i1 = state.forceInt(*args[0], pos, "while evaluating the first argument passed to builtins.bitAnd");
+    auto i2 = state.forceInt(*args[1], pos, "while evaluating the second argument passed to builtins.bitAnd");
+    v.mkInt(i1.value & i2.value);
 }
 
 static RegisterPrimOp primop_bitAnd({
@@ -3638,8 +3686,10 @@ static RegisterPrimOp primop_bitAnd({
 
 static void prim_bitOr(EvalState & state, const PosIdx pos, Value * * args, Value & v)
 {
-    v.mkInt(state.forceInt(*args[0], pos, "while evaluating the first argument passed to builtins.bitOr")
-            | state.forceInt(*args[1], pos, "while evaluating the second argument passed to builtins.bitOr"));
+    auto i1 = state.forceInt(*args[0], pos, "while evaluating the first argument passed to builtins.bitOr");
+    auto i2 = state.forceInt(*args[1], pos, "while evaluating the second argument passed to builtins.bitOr");
+
+    v.mkInt(i1.value | i2.value);
 }
 
 static RegisterPrimOp primop_bitOr({
@@ -3653,8 +3703,10 @@ static RegisterPrimOp primop_bitOr({
 
 static void prim_bitXor(EvalState & state, const PosIdx pos, Value * * args, Value & v)
 {
-    v.mkInt(state.forceInt(*args[0], pos, "while evaluating the first argument passed to builtins.bitXor")
-            ^ state.forceInt(*args[1], pos, "while evaluating the second argument passed to builtins.bitXor"));
+    auto i1 = state.forceInt(*args[0], pos, "while evaluating the first argument passed to builtins.bitXor");
+    auto i2 = state.forceInt(*args[1], pos, "while evaluating the second argument passed to builtins.bitXor");
+
+    v.mkInt(i1.value ^ i2.value);
 }
 
 static RegisterPrimOp primop_bitXor({
@@ -3734,13 +3786,19 @@ static RegisterPrimOp primop_toString({
    non-negative. */
 static void prim_substring(EvalState & state, const PosIdx pos, Value * * args, Value & v)
 {
-    int start = state.forceInt(*args[0], pos, "while evaluating the first argument (the start offset) passed to builtins.substring");
+    NixInt::Inner start = state.forceInt(*args[0], pos, "while evaluating the first argument (the start offset) passed to builtins.substring").value;
 
     if (start < 0)
         state.error<EvalError>("negative start position in 'substring'").atPos(pos).debugThrow();
 
 
-    int len = state.forceInt(*args[1], pos, "while evaluating the second argument (the substring length) passed to builtins.substring");
+    NixInt::Inner len = state.forceInt(*args[1], pos, "while evaluating the second argument (the substring length) passed to builtins.substring").value;
+
+    // Negative length may be idiomatically passed to builtins.substring to get
+    // the tail of the string.
+    if (len < 0) {
+        len = std::numeric_limits<NixInt::Inner>::max();
+    }
 
     // Special-case on empty substring to avoid O(n) strlen
     // This allows for the use of empty substrings to efficently capture string context
@@ -3782,7 +3840,7 @@ static void prim_stringLength(EvalState & state, const PosIdx pos, Value * * arg
 {
     NixStringContext context;
     auto s = state.coerceToString(pos, *args[0], context, "while evaluating the argument passed to builtins.stringLength");
-    v.mkInt(s->size());
+    v.mkInt(NixInt::Inner(s->size()));
 }
 
 static RegisterPrimOp primop_stringLength({
@@ -3806,7 +3864,7 @@ static void prim_hashString(EvalState & state, const PosIdx pos, Value * * args,
     NixStringContext context; // discarded
     auto s = state.forceString(*args[1], context, pos, "while evaluating the second argument passed to builtins.hashString");
 
-    v.mkString(hashString(*ht, s).to_string(Base16, false));
+    v.mkString(hashString(*ht, s).to_string(Base::Base16, false));
 }
 
 static RegisterPrimOp primop_hashString({
@@ -4162,7 +4220,8 @@ static void prim_compareVersions(EvalState & state, const PosIdx pos, Value * *
 {
     auto version1 = state.forceStringNoCtx(*args[0], pos, "while evaluating the first argument passed to builtins.compareVersions");
     auto version2 = state.forceStringNoCtx(*args[1], pos, "while evaluating the second argument passed to builtins.compareVersions");
-    v.mkInt(compareVersions(version1, version2));
+    auto result = compareVersions(version1, version2);
+    v.mkInt(result < 0 ? -1 : result > 0 ? 1 : 0);
 }
 
 static RegisterPrimOp primop_compareVersions({

src/libexpr/primops/fetchMercurial.cc

@@ -31,7 +31,7 @@ static void prim_fetchMercurial(EvalState & state, const PosIdx pos, Value * * a
                 // be both a revision or a branch/tag name.
                 auto value = state.forceStringNoCtx(*attr.value, attr.pos, "while evaluating the `rev` attribute passed to builtins.fetchMercurial");
                 if (std::regex_match(value.begin(), value.end(), revRegex))
-                    rev = Hash::parseAny(value, htSHA1);
+                    rev = Hash::parseAny(value, HashType::SHA1);
                 else
                     ref = value;
             }
@@ -73,7 +73,7 @@ static void prim_fetchMercurial(EvalState & state, const PosIdx pos, Value * * a
         attrs2.alloc("branch").mkString(*input2.getRef());
     // Backward compatibility: set 'rev' to
     // 0000000000000000000000000000000000000000 for a dirty tree.
-    auto rev2 = input2.getRev().value_or(Hash(htSHA1));
+    auto rev2 = input2.getRev().value_or(Hash(HashType::SHA1));
     attrs2.alloc("rev").mkString(rev2.gitRev());
     attrs2.alloc("shortRev").mkString(rev2.gitRev().substr(0, 12));
     if (auto revCount = input2.getRevCount())

src/libexpr/primops/fetchTree.cc

@@ -32,7 +32,7 @@ void emitTreeAttrs(
 
     auto narHash = input.getNarHash();
     assert(narHash);
-    attrs.alloc("narHash").mkString(narHash->to_string(SRI, true));
+    attrs.alloc("narHash").mkString(narHash->to_string(Base::SRI, true));
 
     if (input.getType() == "git")
         attrs.alloc("submodules").mkBool(
@@ -45,7 +45,7 @@ void emitTreeAttrs(
             attrs.alloc("shortRev").mkString(rev->gitShortRev());
         } else if (emptyRevFallback) {
             // Backwards compat for `builtins.fetchGit`: dirty repos return an empty sha1 as rev
-            auto emptyHash = Hash(htSHA1);
+            auto emptyHash = Hash(HashType::SHA1);
             attrs.alloc("rev").mkString(emptyHash.gitRev());
             attrs.alloc("shortRev").mkString(emptyHash.gitShortRev());
         }
@@ -148,9 +148,16 @@ static void fetchTree(
             }
             else if (attr.value->type() == nBool)
                 attrs.emplace(state.symbols[attr.name], Explicit<bool>{attr.value->boolean});
-            else if (attr.value->type() == nInt)
-                attrs.emplace(state.symbols[attr.name], uint64_t(attr.value->integer));
-            else
+            else if (attr.value->type() == nInt) {
+                auto intValue = attr.value->integer.value;
+
+                if (intValue < 0) {
+                    state.error<EvalError>("negative value given for fetchTree attr %1%: %2%", state.symbols[attr.name], intValue).atPos(pos).debugThrow();
+                }
+                unsigned long asUnsigned = intValue;
+
+                attrs.emplace(state.symbols[attr.name], asUnsigned);
+            } else
                 state.error<TypeError>("fetchTree argument '%s' is %s while a string, Boolean or integer is expected",
                     state.symbols[attr.name], showType(*attr.value)).debugThrow();
         }
@@ -219,7 +226,7 @@ static void fetch(EvalState & state, const PosIdx pos, Value * * args, Value & v
             if (n == "url")
                 url = state.forceStringNoCtx(*attr.value, attr.pos, "while evaluating the url we should fetch");
             else if (n == "sha256")
-                expectedHash = newHashAllowEmpty(state.forceStringNoCtx(*attr.value, attr.pos, "while evaluating the sha256 of the content we should fetch"), htSHA256);
+                expectedHash = newHashAllowEmpty(state.forceStringNoCtx(*attr.value, attr.pos, "while evaluating the sha256 of the content we should fetch"), HashType::SHA256);
             else if (n == "name")
                 name = state.forceStringNoCtx(*attr.value, attr.pos, "while evaluating the name of the content we should fetch");
             else
@@ -245,7 +252,7 @@ static void fetch(EvalState & state, const PosIdx pos, Value * * args, Value & v
         state.error<EvalError>("in pure evaluation mode, '%s' requires a 'sha256' argument", who).atPos(pos).debugThrow();
 
     // early exit if pinned and already in the store
-    if (expectedHash && expectedHash->type == htSHA256) {
+    if (expectedHash && expectedHash->type == HashType::SHA256) {
         auto expectedPath = state.store->makeFixedOutputPath(
             name,
             FixedOutputInfo {
@@ -270,13 +277,13 @@ static void fetch(EvalState & state, const PosIdx pos, Value * * args, Value & v
     if (expectedHash) {
         auto hash = unpack
             ? state.store->queryPathInfo(storePath)->narHash
-            : hashFile(htSHA256, state.store->toRealPath(storePath));
+            : hashFile(HashType::SHA256, state.store->toRealPath(storePath));
         if (hash != *expectedHash) {
             state.error<EvalError>(
                 "hash mismatch in file downloaded from '%s':\n  specified: %s\n  got:       %s",
                 *url,
-                expectedHash->to_string(Base32, true),
-                hash.to_string(Base32, true)
+                expectedHash->to_string(Base::Base32, true),
+                hash.to_string(Base::Base32, true)
             ).withExitStatus(102)
             .debugThrow();
         }

src/libexpr/print.cc

@@ -264,22 +264,24 @@ private:
             return true;
         }
 
+        if (options.force) {
+            // The item is going to be forced during printing anyway, but we need its type now.
+            state.forceValue(*item, item->determinePos(noPos));
+        }
+
         // Pretty-print single-item attrsets only if they contain nested
         // structures.
         auto itemType = item->type();
-        return itemType == nList || itemType == nAttrs || itemType == nThunk;
+        return itemType == nList || itemType == nAttrs;
     }
 
     void printAttrs(Value & v, size_t depth)
     {
-        if (seen && !seen->insert(v.attrs).second) {
-            printRepeated();
-            return;
-        }
-
         if (options.force && options.derivationPaths && state.isDerivation(v)) {
             printDerivation(v);
-        } else if (depth < options.maxDepth) {
+        } else if (seen && !v.attrs->empty() && !seen->insert(v.attrs).second) {
+            printRepeated();
+        } else if (depth < options.maxDepth || v.attrs->empty()) {
             increaseIndent();
             output << "{";
 
@@ -335,10 +337,15 @@ private:
             return true;
         }
 
+        if (options.force) {
+            // The item is going to be forced during printing anyway, but we need its type now.
+            state.forceValue(*item, item->determinePos(noPos));
+        }
+
         // Pretty-print single-item lists only if they contain nested
         // structures.
         auto itemType = item->type();
-        return itemType == nList || itemType == nAttrs || itemType == nThunk;
+        return itemType == nList || itemType == nAttrs;
     }
 
     void printList(Value & v, size_t depth)
@@ -348,7 +355,7 @@ private:
             return;
         }
 
-        if (depth < options.maxDepth) {
+        if (depth < options.maxDepth || v.listSize() == 0) {
             increaseIndent();
             output << "[";
             auto listItems = v.listItems();

src/libexpr/value-to-json.cc

@@ -22,7 +22,7 @@ json printValueAsJSON(EvalState & state, bool strict,
     switch (v.type()) {
 
         case nInt:
-            out = v.integer;
+            out = v.integer.value;
             break;
 
         case nBool:

src/libexpr/value.cc

@@ -0,0 +1,106 @@
+#include "value.hh"
+
+#include <ostream>
+
+#include "eval.hh"
+#include "print.hh"
+
+
+namespace nix
+{
+
+static void copyContextToValue(Value & v, const NixStringContext & context)
+{
+    if (!context.empty()) {
+        size_t n = 0;
+        v.string.context = gcAllocType<char const *>(context.size() + 1);
+        for (auto & i : context)
+            v.string.context[n++] = gcCopyStringIfNeeded(i.to_string());
+        v.string.context[n] = 0;
+    }
+}
+
+Value::Value(primop_t, PrimOp & primop)
+    : internalType(tPrimOp)
+    , primOp(&primop)
+    , _primop_pad(0)
+{
+    primop.check();
+}
+
+
+void Value::print(EvalState & state, std::ostream & str, PrintOptions options)
+{
+    printValue(state, str, *this, options);
+}
+
+PosIdx Value::determinePos(const PosIdx pos) const
+{
+    // Allow selecting a subset of enum values
+    #pragma GCC diagnostic push
+    #pragma GCC diagnostic ignored "-Wswitch-enum"
+    switch (internalType) {
+        case tAttrs: return attrs->pos;
+        case tLambda: return lambda.fun->pos;
+        case tApp: return app.left->determinePos(pos);
+        default: return pos;
+    }
+    #pragma GCC diagnostic pop
+}
+
+bool Value::isTrivial() const
+{
+    return
+        internalType != tApp
+        && internalType != tPrimOpApp
+        && (internalType != tThunk
+            || (dynamic_cast<ExprAttrs *>(thunk.expr)
+                && static_cast<ExprAttrs *>(thunk.expr)->dynamicAttrs.empty())
+            || dynamic_cast<ExprLambda *>(thunk.expr)
+            || dynamic_cast<ExprList *>(thunk.expr));
+}
+
+PrimOp * Value::primOpAppPrimOp() const
+{
+    Value * left = primOpApp.left;
+    while (left && !left->isPrimOp()) {
+        left = left->primOpApp.left;
+    }
+
+    if (!left)
+        return nullptr;
+    return left->primOp;
+}
+
+void Value::mkPrimOp(PrimOp * p)
+{
+    p->check();
+    clearValue();
+    internalType = tPrimOp;
+    primOp = p;
+}
+
+void Value::mkString(std::string_view s)
+{
+    mkString(gcCopyStringIfNeeded(s));
+}
+
+void Value::mkString(std::string_view s, const NixStringContext & context)
+{
+    mkString(s);
+    copyContextToValue(*this, context);
+}
+
+void Value::mkStringMove(const char * s, const NixStringContext & context)
+{
+    mkString(s);
+    copyContextToValue(*this, context);
+}
+
+
+void Value::mkPath(const SourcePath & path)
+{
+    mkPath(gcCopyStringIfNeeded(path.path.abs()));
+}
+
+}

src/libexpr/value.hh

@@ -3,16 +3,19 @@
 
 #include <cassert>
 #include <climits>
+#include <functional>
+#include <ranges>
+#include <span>
 
+#include "gc-alloc.hh"
 #include "symbol-table.hh"
 #include "value/context.hh"
 #include "input-accessor.hh"
 #include "source-path.hh"
 #include "print-options.hh"
+#include "checked-arithmetic.hh"
+#include "concepts.hh"
 
-#if HAVE_BOEHMGC
-#include <gc/gc_allocator.h>
-#endif
 #include <nlohmann/json_fwd.hpp>
 
 namespace nix {
@@ -73,8 +76,8 @@ class EvalState;
 class XMLWriter;
 class Printer;
 
-typedef int64_t NixInt;
-typedef double NixFloat;
+using NixInt = checked::Checked<int64_t>;
+using NixFloat = double;
 
 /**
  * External values must descend from ExternalValueBase, so that
@@ -133,6 +136,55 @@ class ExternalValueBase
 
 std::ostream & operator << (std::ostream & str, const ExternalValueBase & v);
 
+extern ExprBlackHole eBlackHole;
+
+struct NewValueAs
+{
+    struct integer_t { };
+    constexpr static integer_t integer{};
+
+    struct floating_t { };
+    constexpr static floating_t floating{};
+
+    struct boolean_t { };
+    constexpr static boolean_t boolean{};
+
+    struct string_t { };
+    constexpr static string_t string{};
+
+    struct path_t { };
+    constexpr static path_t path{};
+
+    struct list_t { };
+    constexpr static list_t list{};
+
+    struct attrs_t { };
+    constexpr static attrs_t attrs{};
+
+    struct thunk_t { };
+    constexpr static thunk_t thunk{};
+
+    struct null_t { };
+    constexpr static null_t null{};
+
+    struct app_t { };
+    constexpr static app_t app{};
+
+    struct primop_t { };
+    constexpr static primop_t primop{};
+
+    struct primOpApp_t { };
+    constexpr static primOpApp_t primOpApp{};
+
+    struct lambda_t { };
+    constexpr static lambda_t lambda{};
+
+    struct external_t { };
+    constexpr static external_t external{};
+
+    struct blackhole_t { };
+    constexpr static blackhole_t blackhole{};
+};
 
 struct Value
 {
@@ -143,6 +195,315 @@ private:
 
 public:
 
+    // Discount `using NewValueAs::*;`
+#define USING_VALUETYPE(name) using name = NewValueAs::name
+    USING_VALUETYPE(integer_t);
+    USING_VALUETYPE(floating_t);
+    USING_VALUETYPE(boolean_t);
+    USING_VALUETYPE(string_t);
+    USING_VALUETYPE(path_t);
+    USING_VALUETYPE(list_t);
+    USING_VALUETYPE(attrs_t);
+    USING_VALUETYPE(thunk_t);
+    USING_VALUETYPE(primop_t);
+    USING_VALUETYPE(app_t);
+    USING_VALUETYPE(null_t);
+    USING_VALUETYPE(primOpApp_t);
+    USING_VALUETYPE(lambda_t);
+    USING_VALUETYPE(external_t);
+    USING_VALUETYPE(blackhole_t);
+#undef USING_VALUETYPE
+
+    /// Default constructor which is still used in the codebase but should not
+    /// be used in new code. Zero initializes its members.
+    [[deprecated]] Value()
+        : internalType(static_cast<InternalType>(0))
+        , _empty{ 0, 0 }
+    { }
+
+    /// Constructs a nix language value of type "int", with the integral value
+    /// of @ref i.
+    Value(integer_t, NixInt i)
+        : internalType(tInt)
+        , _empty{ 0, 0 }
+    {
+        // the NixInt ctor here is is special because NixInt has a ctor too, so
+        // we're not allowed to have it as an anonymous aggreagte member. we do
+        // however still have the option to clear the data members using _empty
+        // and leaving the second word of data cleared by setting only integer.
+        integer = i;
+    }
+
+    /// Constructs a nix language value of type "float", with the floating
+    /// point value of @ref f.
+    Value(floating_t, NixFloat f)
+        : internalType(tFloat)
+        , fpoint(f)
+        , _float_pad(0)
+    { }
+
+    /// Constructs a nix language value of type "bool", with the boolean
+    /// value of @ref b.
+    Value(boolean_t, bool b)
+        : internalType(tBool)
+        , boolean(b)
+        , _bool_pad(0)
+    { }
+
+    /// Constructs a nix language value of type "string", with the value of the
+    /// C-string pointed to by @ref strPtr, and optionally with an array of
+    /// string context pointed to by @ref contextPtr.
+    ///
+    /// Neither the C-string nor the context array are copied; this constructor
+    /// assumes suitable memory has already been allocated (with the GC if
+    /// enabled), and string and context data copied into that memory.
+    Value(string_t, char const * strPtr, char const ** contextPtr = nullptr)
+        : internalType(tString)
+        , string({ .s = strPtr, .context = contextPtr })
+    { }
+
+    /// Constructx a nix language value of type "string", with a copy of the
+    /// string data viewed by @ref copyFrom.
+    ///
+    /// The string data *is* copied from @ref copyFrom, and this constructor
+    /// performs a dynamic (GC) allocation to do so.
+    Value(string_t, std::string_view copyFrom, NixStringContext const & context = {})
+        : internalType(tString)
+        , string({ .s = gcCopyStringIfNeeded(copyFrom), .context = nullptr })
+    {
+        if (context.empty()) {
+            // It stays nullptr.
+            return;
+        }
+
+        // Copy the context.
+        this->string.context = gcAllocType<char const *>(context.size() + 1);
+
+        size_t n = 0;
+        for (NixStringContextElem const & contextElem : context) {
+            this->string.context[n] = gcCopyStringIfNeeded(contextElem.to_string());
+            n += 1;
+        }
+
+        // Terminator sentinel.
+        this->string.context[n] = nullptr;
+    }
+
+    /// Constructx a nix language value of type "string", with the value of the
+    /// C-string pointed to by @ref strPtr, and optionally with a set of string
+    /// context @ref context.
+    ///
+    /// The C-string is not copied; this constructor assumes suitable memory
+    /// has already been allocated (with the GC if enabled), and string data
+    /// has been copied into that memory. The context data *is* copied from
+    /// @ref context, and this constructor performs a dynamic (GC) allocation
+    /// to do so.
+    Value(string_t, char const * strPtr, NixStringContext const & context)
+        : internalType(tString)
+        , string({ .s = strPtr, .context = nullptr })
+    {
+        if (context.empty()) {
+            // It stays nullptr
+            return;
+        }
+
+        // Copy the context.
+        this->string.context = gcAllocType<char const *>(context.size() + 1);
+
+        size_t n = 0;
+        for (NixStringContextElem const & contextElem : context) {
+            this->string.context[n] = gcCopyStringIfNeeded(contextElem.to_string());
+            n += 1;
+        }
+
+        // Terminator sentinel.
+        this->string.context[n] = nullptr;
+    }
+
+    /// Constructs a nix language value of type "path", with the value of the
+    /// C-string pointed to by @ref strPtr.
+    ///
+    /// The C-string is not copied; this constructor assumes suitable memory
+    /// has already been allocated (with the GC if enabled), and string data
+    /// has been copied into that memory.
+    Value(path_t, char const * strPtr)
+        : internalType(tPath)
+        , _path(strPtr)
+        , _path_pad(0)
+    { }
+
+    /// Constructs a nix language value of type "path", with the path
+    /// @ref path.
+    ///
+    /// The data from @ref path *is* copied, and this constructor performs a
+    /// dynamic (GC) allocation to do so.
+    Value(path_t, SourcePath const & path)
+        : internalType(tPath)
+        , _path(gcCopyStringIfNeeded(path.path.abs()))
+        , _path_pad(0)
+    { }
+
+    /// Constructs a nix language value of type "list", with element array
+    /// @ref items.
+    ///
+    /// Generally, the data in @ref items is neither deep copied nor shallow
+    /// copied. This construct assumes the std::span @ref items is a region of
+    /// memory that has already been allocated (with the GC if enabled), and
+    /// an array of valid Value pointers has been copied into that memory.
+    ///
+    /// Howver, as an implementation detail, if @ref items is only 2 items or
+    /// smaller, the list is stored inline, and the Value pointers in
+    /// @ref items are shallow copied into this structure, without dynamically
+    /// allocating memory.
+    Value(list_t, std::span<Value *> items)
+    {
+        if (items.size() == 1) {
+            this->internalType = tList1;
+            this->smallList[0] = items[0];
+            this->smallList[1] = nullptr;
+        } else if (items.size() == 2) {
+            this->internalType = tList2;
+            this->smallList[0] = items[0];
+            this->smallList[1] = items[1];
+        } else {
+            this->internalType = tListN;
+            this->bigList.size = items.size();
+            this->bigList.elems = items.data();
+        }
+    }
+
+    /// Constructs a nix language value of type "list", with an element array
+    /// initialized by applying @ref transformer to each element in @ref items.
+    ///
+    /// This allows "in-place" construction of a nix list when some logic is
+    /// needed to get each Value pointer. This constructor dynamically (GC)
+    /// allocates memory for the size of @ref items, and the Value pointers
+    /// returned by @ref transformer are shallow copied into it.
+    template<
+        std::ranges::sized_range SizedIterableT,
+        InvocableR<Value *, typename SizedIterableT::value_type const &> TransformerT
+    >
+    Value(list_t, SizedIterableT & items, TransformerT const & transformer)
+    {
+        if (items.size() == 1) {
+            this->internalType = tList1;
+            this->smallList[0] = transformer(*items.begin());
+            this->smallList[1] = nullptr;
+        } else if (items.size() == 2) {
+            this->internalType = tList2;
+            auto it = items.begin();
+            this->smallList[0] = transformer(*it);
+            it++;
+            this->smallList[1] = transformer(*it);
+        } else {
+            this->internalType = tListN;
+            this->bigList.size = items.size();
+            this->bigList.elems = gcAllocType<Value *>(items.size());
+            auto it = items.begin();
+            for (size_t i = 0; i < items.size(); i++, it++) {
+                this->bigList.elems[i] = transformer(*it);
+            }
+        }
+    }
+
+    /// Constructs a nix language value of the singleton type "null".
+    Value(null_t)
+        : internalType(tNull)
+        , _empty{0, 0}
+    { }
+
+    /// Constructs a nix language value of type "set", with the attribute
+    /// bindings pointed to by @ref bindings.
+    ///
+    /// The bindings are not not copied; this constructor assumes @ref bindings
+    /// has already been suitably allocated by something like nix::buildBindings.
+    Value(attrs_t, Bindings * bindings)
+        : internalType(tAttrs)
+        , attrs(bindings)
+        , _attrs_pad(0)
+    { }
+
+    /// Constructs a nix language lazy delayed computation, or "thunk".
+    ///
+    /// The thunk stores the environment it will be computed in @ref env, and
+    /// the expression that will need to be evaluated @ref expr.
+    Value(thunk_t, Env & env, Expr & expr)
+        : internalType(tThunk)
+        , thunk({ .env = &env, .expr = &expr })
+    { }
+
+    /// Constructs a nix language value of type "lambda", which represents
+    /// a builtin, primitive operation ("primop"), from the primop
+    /// implemented by @ref primop.
+    Value(primop_t, PrimOp & primop);
+
+    /// Constructs a nix language value of type "lambda", which represents a
+    /// partially applied primop.
+    Value(primOpApp_t, Value & lhs, Value & rhs)
+        : internalType(tPrimOpApp)
+        , primOpApp({ .left = &lhs, .right = &rhs })
+    { }
+
+    /// Constructs a nix language value of type "lambda", which represents a
+    /// lazy partial application of another lambda.
+    Value(app_t, Value & lhs, Value & rhs)
+        : internalType(tApp)
+        , app({ .left = &lhs, .right = &rhs })
+    { }
+
+    /// Constructs a nix language value of type "external", which is only used
+    /// by plugins. Do any existing plugins even use this mechanism?
+    Value(external_t, ExternalValueBase & external)
+        : internalType(tExternal)
+        , external(&external)
+        , _external_pad(0)
+    { }
+
+    /// Constructs a nix language value of type "lambda", which represents a
+    /// run of the mill lambda defined in nix code.
+    ///
+    /// This takes the environment the lambda is closed over @ref env, and
+    /// the lambda expression itself @ref lambda, which will not be evaluated
+    /// until it is applied.
+    Value(lambda_t, Env & env, ExprLambda & lambda)
+        : internalType(tLambda)
+        , lambda({ .env = &env, .fun = &lambda })
+    { }
+
+    /// Constructs an evil thunk, whose evaluation represents infinite recursion.
+    explicit Value(blackhole_t)
+        : internalType(tThunk)
+        , thunk({ .env = nullptr, .expr = reinterpret_cast<Expr *>(&eBlackHole) })
+    { }
+
+    Value(Value const & rhs) = default;
+
+    /// Move constructor. Does the same thing as the copy constructor, but
+    /// also zeroes out the other Value.
+    Value(Value && rhs)
+        : internalType(rhs.internalType)
+        , _empty{ 0, 0 }
+    {
+        *this = std::move(rhs);
+    }
+
+    Value & operator=(Value const & rhs) = default;
+
+    /// Move assignment operator.
+    /// Does the same thing as the copy assignment operator, but also zeroes out
+    /// the rhs.
+    inline Value & operator=(Value && rhs)
+    {
+        *this = static_cast<const Value &>(rhs);
+        if (this != &rhs) {
+            // Kill `rhs`, because non-destructive move lol.
+            rhs.internalType = static_cast<InternalType>(0);
+            rhs._empty[0] = 0;
+            rhs._empty[1] = 0;
+        }
+        return *this;
+    }
+
     void print(EvalState &state, std::ostream &str, PrintOptions options = PrintOptions {});
 
     // Functions needed to distinguish the type
@@ -161,8 +522,15 @@ public:
 
     union
     {
+        /// Dummy field, which takes up as much space as the largest union variants
+        /// to set the union's memory to zeroed memory.
+        uintptr_t _empty[2];
+
         NixInt integer;
-        bool boolean;
+        struct {
+            bool boolean;
+            uintptr_t _bool_pad;
+        };
 
         /**
          * Strings in the evaluator carry a so-called `context` which
@@ -191,8 +559,14 @@ public:
             const char * * context; // must be in sorted order
         } string;
 
-        const char * _path;
-        Bindings * attrs;
+        struct {
+            const char * _path;
+            uintptr_t _path_pad;
+        };
+        struct {
+            Bindings * attrs;
+            uintptr_t _attrs_pad;
+        };
         struct {
             size_t size;
             Value * * elems;
@@ -209,12 +583,21 @@ public:
             Env * env;
             ExprLambda * fun;
         } lambda;
-        PrimOp * primOp;
+        struct {
+            PrimOp * primOp;
+            uintptr_t _primop_pad;
+        };
         struct {
             Value * left, * right;
         } primOpApp;
-        ExternalValueBase * external;
-        NixFloat fpoint;
+        struct {
+            ExternalValueBase * external;
+            uintptr_t _external_pad;
+        };
+        struct {
+            NixFloat fpoint;
+            uintptr_t _float_pad;
+        };
     };
 
     /**
@@ -254,6 +637,11 @@ public:
         app.left = app.right = 0;
     }
 
+    inline void mkInt(NixInt::Inner n)
+    {
+        mkInt(NixInt{n});
+    }
+
     inline void mkInt(NixInt n)
     {
         clearValue();
@@ -445,8 +833,6 @@ public:
 };
 
 
-extern ExprBlackHole eBlackHole;
-
 bool Value::isBlackhole() const
 {
     return internalType == tThunk && thunk.expr == (Expr*) &eBlackHole;
@@ -458,17 +844,9 @@ void Value::mkBlackhole()
     thunk.expr = (Expr*) &eBlackHole;
 }
 
-
-#if HAVE_BOEHMGC
-typedef std::vector<Value *, traceable_allocator<Value *>> ValueVector;
-typedef std::map<Symbol, Value *, std::less<Symbol>, traceable_allocator<std::pair<const Symbol, Value *>>> ValueMap;
-typedef std::map<Symbol, ValueVector, std::less<Symbol>, traceable_allocator<std::pair<const Symbol, ValueVector>>> ValueVectorMap;
-#else
-typedef std::vector<Value *> ValueVector;
-typedef std::map<Symbol, Value *> ValueMap;
-typedef std::map<Symbol, ValueVector> ValueVectorMap;
-#endif
-
+using ValueVector = GcVector<Value *>;
+using ValueMap = GcMap<Symbol, Value *>;
+using ValueVectorMap = std::map<Symbol, ValueVector>;
 
 /**
  * A value allocated in traceable memory.

src/libfetchers/fetch-to-store.cc

@@ -18,8 +18,8 @@ StorePath fetchToStore(
 
     return
         settings.readOnlyMode
-        ? store.computeStorePathForPath(name, path.path.abs(), method, htSHA256, filter2).first
-        : store.addToStore(name, path.path.abs(), method, htSHA256, filter2, repair);
+        ? store.computeStorePathForPath(name, path.path.abs(), method, HashType::SHA256, filter2).first
+        : store.addToStore(name, path.path.abs(), method, HashType::SHA256, filter2, repair);
 }
 
 

src/libfetchers/fetchers.cc

@@ -153,12 +153,12 @@ std::pair<Tree, Input> Input::fetch(ref<Store> store) const
     };
 
     auto narHash = store->queryPathInfo(tree.storePath)->narHash;
-    input.attrs.insert_or_assign("narHash", narHash.to_string(SRI, true));
+    input.attrs.insert_or_assign("narHash", narHash.to_string(Base::SRI, true));
 
     if (auto prevNarHash = getNarHash()) {
         if (narHash != *prevNarHash)
             throw Error((unsigned int) 102, "NAR hash mismatch in input '%s' (%s), expected '%s', got '%s'",
-                to_string(), tree.actualPath, prevNarHash->to_string(SRI, true), narHash.to_string(SRI, true));
+                to_string(), tree.actualPath, prevNarHash->to_string(Base::SRI, true), narHash.to_string(Base::SRI, true));
     }
 
     if (auto prevLastModified = getLastModified()) {
@@ -240,8 +240,8 @@ std::string Input::getType() const
 std::optional<Hash> Input::getNarHash() const
 {
     if (auto s = maybeGetStrAttr(attrs, "narHash")) {
-        auto hash = s->empty() ? Hash(htSHA256) : Hash::parseSRI(*s);
-        if (hash.type != htSHA256)
+        auto hash = s->empty() ? Hash(HashType::SHA256) : Hash::parseSRI(*s);
+        if (hash.type != HashType::SHA256)
             throw UsageError("narHash must use SHA-256");
         return hash;
     }
@@ -264,7 +264,7 @@ std::optional<Hash> Input::getRev() const
             hash = Hash::parseAnyPrefixed(*s);
         } catch (BadHash &e) {
             // Default to sha1 for backwards compatibility with existing flakes
-            hash = Hash::parseAny(*s, htSHA1);
+            hash = Hash::parseAny(*s, HashType::SHA1);
         }
     }