Compare commits

..

513 Commits

Author SHA1 Message Date
John Ericson
156061409a Do some parsing of git hashes based on length.
Note that for Nix-native information we should *not* do length tricks,
but instead always rely on an explicit algorithm. This hack should be
only for foreign hash literals.
2025-11-09 18:52:35 -05:00
John Ericson
6da56d7ed4 Make changes to support libgit2 experimental API
This experimental interface is likely to become stable with libgit2 2.0.
2025-11-09 18:52:35 -05:00
John Ericson
9846305fee Switch to libgit2 experimental
Just patch the package for now, but we should make Nixpkgs provide what
we need (whether that the experimental variant, or a 2.x release with
the new API made non-experimental) before releasing this change.
2025-11-09 18:52:35 -05:00
John Ericson
cbe8ec7bd7 Merge pull request #14470 from NixOS/ctx-type-alias
Encapsulate and slightly optimize string contexts
2025-11-09 21:21:15 +00:00
John Ericson
60667e9e5a Merge pull request #14525 from NixOS/reset-positions-repl
libexpr: Clear PosTable contents in EvalState::resetFileCache
2025-11-09 21:04:03 +00:00
John Ericson
318eea040f Encapsulate and slightly optimize string contexts
These steps are done (originally in order, but I squashed it as the end
result is still pretty small, and the churn in the code comments was a
bit annoying to keep straight).

1. Create proper struct type for string contexts on the heap

   This will make it easier to change this type in the future.

2. Make `Value::StringWithContext` iterable

   This make some for loops a lot more terse.

3. Encapsulate `Value::StringWithContext::Context::elems`

   It turns out the iterators we just exposed are sufficient.

4. Make `StringWithContext::Context` length-prefixed instead

   Rather than having a null pointer at the end, have a `size_t` at the
   beginning. This is the exact same size (note that null pointer is
   longer than null byte) and thus takes no more space!

Also, see the new TODO on naming. The thing we already so-named is a
builder type for string contexts, not the on-heap type. The
`fromBuilder` static method reflects what the names ought to be too.

Co-authored-by: Sergei Zimmerman <sergei@zimmerman.foo>
2025-11-09 23:35:38 +03:00
Sergei Zimmerman
a091a8100a libexpr: Clear PosTable contents in EvalState::resetFileCache
Otherwise PosTable grows indefinitely for each reload. Since
the total input size is limited to 4GB (uint32_t for byte offset PosIdx)
it can get exhausted pretty. This ensures that we don't waste memory
on reloads as well.
2025-11-09 22:09:18 +03:00
John Ericson
6ebaba50dd Merge pull request #14515 from NixOS/dirOf-dont-call-std-filesystem
libexpr: Don't use nix::dirOf in prim_dirOf (fix 2.23 regression)
2025-11-09 17:12:04 +00:00
John Ericson
18941cb8fa Merge pull request #14516 from NixOS/honest-characterization-message
tests/functional: Output an actually correct command to accept test c…
2025-11-09 17:10:49 +00:00
Sergei Zimmerman
a33fccf55a libexpr: Don't use nix::dirOf in prim_dirOf
This gets us back to pre-2.23 behavior of this primop.
Done by inlining the code of `nix::dirOf` from 2.2-maintenance.
2025-11-09 18:56:33 +03:00
Sergei Zimmerman
86f090837b tests/functional: Add tests for builtins.dirOf
These will change in the next commit to fix the silent regression from 2.23
in the handling of multiple subsequent path separators.
2025-11-09 18:55:11 +03:00
Jörg Thalheim
08a8bae8b3 Merge pull request #14518 from roberth/channel-subdomain
Change channel URLs to channels.nixos.org subdomain
2025-11-09 15:18:07 +00:00
Robert Hensing
f715992346 Change channel URLs to channels.nixos.org subdomain
Update all channel URLs from https://nixos.org/channels/ to
https://channels.nixos.org/ to use the more reliable subdomain.

The nixos.org domain apex lacks IPv6 support due to DNS hoster
limitations. Using the subdomain allows better CDN distribution
and improved reliability.

Updated files:
- Installation scripts (multi-user and tarball installers)
- Channel URL resolution in eval-settings.cc
- Documentation and examples
- Docker image default channel URL
- Release notes (added note about URL change)

Fixes #14517
2025-11-09 15:28:12 +01:00
Sergei Zimmerman
98e61c6da9 tests/functional: Output an actually correct command to accept test changes
I've run into this quite a few times when working with characterization test
infra. It would print an invalid command:

_NIX_TEST_ACCEPT=1 meson test main/lang

Which you'd then proceed to run and it would fail. This commit makes it
be honest about the command you need to run:

_NIX_TEST_ACCEPT=1 meson test --suite main lang
2025-11-09 16:52:51 +03:00
John Ericson
479b6b73a9 Merge pull request #14509 from Mic92/no-tbb
build: Disable libstdc++ TBB backend to avoid unnecessary dependency
2025-11-07 20:42:34 +00:00
John Ericson
3c2dcf42e9 Merge pull request #14477 from lovesegfault/http-upload-headers
refactor(libstore): pass headers into upload methods
2025-11-07 20:41:14 +00:00
John Ericson
5a97c00f29 Merge pull request #14499 from roberth/genericClosure-errors
`builtins.genericClosure`: improve errors
2025-11-07 20:10:34 +00:00
Jörg Thalheim
2f3ec16793 build: Disable libstdc++ TBB backend to avoid unnecessary dependency
boost::concurrent_flat_map (used in libutil and libstore) includes the
C++17 <execution> header. GCC's libstdc++ implements parallel algorithms
using Intel TBB as the backend, which creates a link-time dependency on
libtbb even though we don't actually use any parallel algorithms.

Disable the TBB backend for libstdc++ by setting
_GLIBCXX_USE_TBB_PAR_BACKEND=0. This makes parallel algorithms fall back
to serial execution, which is acceptable since we don't use them anyway.

This only affects libstdc++ (GCC's standard library); other standard
libraries like libc++ (LLVM) are unaffected.
2025-11-07 20:58:46 +01:00
Robert Hensing
3ee8e45f8e tests: Replace fragile genericClosure unit tests
We now have functional tests for these. The unit tests added negligible
value while imposing a much higher maintenance cost.

The maintenance cost is high:
  - No automatic accept option
  - They broke 5+ times during this session due to implementation changes (trace count, ordering)
  - They require understanding ANSI escape codes, Uncolored() wrappers, trace reversal
  - They test empty traces HintFmt("") from withTrace(pos, "") - pure implementation detail
  - They're fragile: adding any trace anywhere breaks the exact count assertions

The additional value over functional tests is minimal:
  - Functional tests already verify the error message
  - Functional tests already show trace order and content (as users see it, helps review)
  - Unit tests verify "exactly 3 traces, not 2 or 4" - but users don't count traces
  - Unit tests verify empty traces exist - but users never see them

The white-box testing catches the wrong things:
  - It catches "you added helpful context" as a failure
  - It doesn't catch "the context is confusing" (which functional tests would show)
  - It enforces implementation details that should be allowed to evolve
2025-11-07 00:27:39 +01:00
Robert Hensing
d262efc240 libexpr: improve error messages for builtins.genericClosure
Show which element(s) are involved at each error point:

- When an element is missing the "key" attribute, show the element
- When an element is not an attribute set, show the element
- When comparing keys fails, show both elements being compared
- When calling operator fails, show which element was being processed

This provides concrete context using ValuePrinter with errorPrintOptions.

Note: errorPrintOptions uses maxDepth=10 by default, which may print
quite deeply nested structures in error messages. This could potentially
be overwhelming, but follows the existing default for error contexts.
2025-11-06 22:28:49 +01:00
John Ericson
5b15544bdd Merge pull request #14493 from obsidiansystems/drv-and-path-info-new-fmts
Modifications to the JSON formats for `Derivation` and `ValidPathInfo`
2025-11-06 21:09:27 +00:00
Robert Hensing
ca787bc3e0 tests: add error tests for builtins.genericClosure
Covers error conditions for:
- Invalid argument types (not an attrset)
- Missing required attributes (startSet, operator)
- Type mismatches (startSet/operator not correct type)
- Element validation (elements not attrsets, missing key attribute)
- Key comparison errors (incompatible types, uncomparable types)
- Operator return value validation (not a list)
2025-11-06 21:33:41 +01:00
John Ericson
8cc3ede0fa Add change-log entry for derivation format changes 2025-11-06 15:19:44 -05:00
John Ericson
caa196e31d Make the store path info ca field structured in JSON
The old string format is a holdover from the pre JSON days. It is not
friendly to users who need to get the information out of it.

Also introduce the sort of versioning we have for derivation for this
format too.
2025-11-06 15:19:44 -05:00
John Ericson
0c37a62207 Change JSON derivation format in two ways
- Use canonical content address JSON format for floating content
  addressed derivation outputs

  This keeps it more consistent.

- Reorganize inputs into nested structure (`inputs.srcs` and
  `inputs.drvs`)

  This will allow for an easier to use, but less compact, alternative
  where `srcs` is just a list of derived paths.

  It also allows for other experiments for derivations with a different
  input structure, as I suspect will be needed for secure build traces.
2025-11-06 15:19:44 -05:00
John Ericson
147e183c68 Merge pull request #14426 from obsidiansystems/json-schema-build-result
JSON Impl and schema for BuildResult
2025-11-06 18:40:35 +00:00
Eelco Dolstra
52b2909fd2 Merge pull request #14491 from NixOS/fix-14311
Don't crash on flakerefs containing newlines
2025-11-06 18:29:44 +00:00
Jörg Thalheim
34c77ffe38 Merge pull request #14471 from obsidiansystems/derivation-options-json-test
FIx `DerivationOptions` JSON and clean up unit tests
2025-11-06 18:21:15 +00:00
John Ericson
af8e44821e Merge pull request #14490 from obsidiansystems/derivation-builder-no-inputs
`DerivationBuilderParams` have reference to `BasicDerivation`
2025-11-06 18:15:56 +00:00
Jörg Thalheim
70fbd1cdf4 Merge pull request #14465 from obsidiansystems/split-realisation-protocol-tests
Split realisation protocol unit tests
2025-11-06 18:14:25 +00:00
Jörg Thalheim
daace78239 Merge pull request #14425 from obsidiansystems/json-schema-build-trace
JSON Schema for build trace entry
2025-11-06 18:06:57 +00:00
Sergei Zimmerman
d596b9754e Merge pull request #14472 from Radvendii/exprs-alloc
libexpr: allocate the Exprs themselves in Exprs::alloc
2025-11-06 17:29:09 +00:00
Eelco Dolstra
c1317017e9 Don't crash on flakerefs containing newlines
Fixes #14311.
2025-11-06 13:06:38 +01:00
Jörg Thalheim
3f18cad5f1 Merge pull request #14459 from jfroche/fix/macos-ipcs
Fix macOS IPC cleanup in builder
2025-11-06 09:31:53 +00:00
Jörg Thalheim
41b62aa979 Merge pull request #14445 from CyberShadow/nix-flake-check-log-success
nix flake check: log success in verbose mode
2025-11-06 09:30:55 +00:00
Jörg Thalheim
af41eccb31 Merge pull request #14469 from roberth/doc-check-link-fragments
Manual: fix and check link fragments
2025-11-06 09:27:19 +00:00
John Ericson
e7b274f85a DerivationBuilderParams have reference to BasicDerivation
Have one to that instead of one to `Derivation`. `DerivationBuilder`
doesn't need `inputDrvs`, so `BasicDerivation` suffices.

(In fact, it doesn't need `inputSrcs` either, but we don't yet hve a
type to exclude that.)
2025-11-05 23:41:47 -05:00
John Ericson
6bd92d47e5 Merge pull request #14488 from Mic92/kaitai-struct
nix-kaitai-struct: make it not longer part of the devshell
2025-11-05 21:57:29 +00:00
Sergei Zimmerman
b5302fc111 Merge pull request #14487 from NixOS/git-show-progress
Git fetcher: Restore progress indication
2025-11-05 21:32:58 +00:00
Jörg Thalheim
724086005a nix-kaitai-struct: make it not longer part of the devshell
just now pulls in jdk in
2025-11-05 22:22:45 +01:00
Eelco Dolstra
038d74edf7 Git fetcher: Restore progress indication
We were calling git with `--quiet` in order not to mess up Nix's
progress bar. However, `runProgram()` already suspends the progress
bar (since git may be interactive) so that's no longer an issue. So we
can just run with `--progress` instead.
2025-11-05 21:59:07 +01:00
Eelco Dolstra
b177354c35 Merge pull request #14482 from NixOS/fix-nix-flake-check-crash-upstream
nix flake check: Remove incorrect assertion
2025-11-05 19:15:29 +00:00
John Ericson
2039235f6e Merge pull request #14484 from NixOS/fix-typo
manual: Fix MathJax typo
2025-11-05 18:24:51 +00:00
John Ericson
0fd3b6fee6 Merge pull request #14483 from NixOS/purge-toRealPath
Relegate `toRealPath` to `LocalFSStore`
2025-11-05 17:20:36 +00:00
Taeer Bar-Yam
b2f0472fe2 parser.y: allocate Exprs in the allocator 2025-11-05 17:10:35 +01:00
John Ericson
91af29f37a manual: Fix MathJax typo
Thanks to @cafkafk for catching my mistake.
2025-11-05 11:08:36 -05:00
John Ericson
099af7578f Relegate toRealPath to LocalFSStore
Fix #14480

This method is not well-defined for arbitrary stores, which do not have
a notion of a "real path" -- it is only well-defined for local file
systems stores, which do have exactly that notion, and so it is moved to
that sub-interface instead.

Some call-sites had to be fixed up for this, but in all cases the
changes are positive. Using `getFSSourceAccessor` allows for more other
stores to work properly. `nix-channel` was straight-up wrong in the case
of redirected local stores. And the building logic with remote building
and a non-local store is also fixed, properly gating some deletions on
store type.

Co-authored-by: Robert Hensing <robert@roberthensing.nl>
2025-11-05 10:44:25 -05:00
John Ericson
948c89b367 Merge pull request #14481 from Radvendii/exprs-alloc-pre
parser.y: abstract `new` into a function on Exprs
2025-11-05 14:49:00 +00:00
Robert Hensing
7e84ce3904 ci/gha: Disable linkcheck on darwin
Does not reproduce all settings on darwin. (Pre-existing issue)

Build with `nix build .#nix-manual.tests.linkcheck`
2025-11-05 15:38:23 +01:00
Eelco Dolstra
a828cf777a nix flake check: Remove incorrect assertion
The assumption that no unknown paths can be returned is incorrect. It
can happen if a derivation has outputs that are substitutable, but
that have references that cannot be substituted (i.e. an incomplete
closure in the binary cache). This can easily happen with
magic-nix-cache.
2025-11-05 15:24:36 +01:00
Taeer Bar-Yam
687dd38998 parser.y: abstract new into a function on Exprs
so it can easily be swapped out for other implementations
2025-11-04 23:59:45 +01:00
Taeer Bar-Yam
62729ff472 parser.y: pass all of Exprs in, not just alloc 2025-11-04 23:59:45 +01:00
Bernardo Meurer Costa
a0d4714073 refactor(libstore): pass headers into upload methods
This will make it easier to attach additional headers (e.g. storage
class) on the s3 side of things and makes `Content-Encoding` less
special.
2025-11-04 22:55:32 +00:00
Jean-François Roche
0507674a13 Document the new cleanup function using a Doxygen-style comment 2025-11-04 20:57:40 +00:00
Jean-François Roche
4f85cfe824 fix(darwin): extend IPC cleanup to message queues and semaphores
Previously, only shared memory segments were cleaned up.
This could lead to leaked message queues and semaphore sets when builds use System V IPC, exhausting kernel IPC limits over time.

This commit extends the cleanup to all three System V IPC types:
1. Shared memory segments
2. Message queues
3. Semaphores

Additionally, we stop removing IPC objects during iteration, as it could corrupt the kernel's iterator state and cause some objects to be skipped. The new implementation uses a two-pass approach where we list first and then remove them in a separate pass.

The IPC IDs are now extracted during iteration using actual system calls (shmget, msgget, semget) rather than being looked up later, ensuring the objects exist when we capture their IDs.
2025-11-04 20:57:40 +00:00
Jean-François Roche
7d5567a8d7 Fix macOS IPC cleanup using sysctl: shared memory segments
In Linux, IPC objects are automatically cleaned up when the IPC namespace is destroyed.
On Darwin, since there are no IPC namespaces, the IPC objects may sometimes persist after the build user's processes are killed.

This patch modifies the cleanup logic to use sysctl calls to identify and remove left over shm segments associated with the build user.

Fixes: #12548
2025-11-04 20:57:40 +00:00
Sergei Zimmerman
3ed42cd354 Merge pull request #14474 from xokdvium/shut-up-ai-slop
.coderabbit.yaml: Kill chats
2025-11-04 19:57:07 +00:00
Sergei Zimmerman
4a888b4138 .coderabbit.yaml: Kill chats
Stops stupid spam on issues.
2025-11-04 22:55:18 +03:00
John Ericson
f2436a47bb Merge pull request #14388 from NixOS/dependabot/github_actions/actions/upload-artifact-5
build(deps): bump actions/upload-artifact from 4 to 5
2025-11-04 18:14:06 +00:00
John Ericson
83ddfaebf4 Merge pull request #14389 from NixOS/dependabot/github_actions/actions/download-artifact-6
build(deps): bump actions/download-artifact from 5 to 6
2025-11-04 18:13:45 +00:00
John Ericson
2b382b171c Merge pull request #14453 from Radvendii/exprwith-alloc
libexpr: shrink ExprWith by 8 bytes
2025-11-04 18:04:25 +00:00
John Ericson
b7553378a4 Merge pull request #14467 from NixOS/dependabot/github_actions/cachix/install-nix-action-31.8.2
build(deps): bump cachix/install-nix-action from 31.5.1 to 31.8.2
2025-11-04 18:03:44 +00:00
John Ericson
d40f66109b Merge pull request #14462 from NixOS/parallel-revcount
GitRepo::getRevCount(): Compute revcount in parallel
2025-11-04 17:31:18 +00:00
Eelco Dolstra
9657feaf8c GitRepo::getRevCount(): Compute revcount in parallel
For repos with a lot of non-linearity in the commit graph (like
Nixpkgs), this speeds up getting the revcount a lot, e.g. `nix flake
metadata /path/to/nixpkgs?rev=9dc7035bbee85ffc740d893e02cb64460f11989f` went
from 9.1s to 3.7s.
2025-11-04 14:50:57 +01:00
John Ericson
d05e85e5be Fix DerivationOptions JSON implementation and test 2025-11-04 03:04:59 -05:00
John Ericson
9daef9cca2 Clean up DerivationOptions unit tests
We now test equality on whole strucks much more often, which avoids
forgetting to test for specific fields.
2025-11-04 02:48:16 -05:00
John Ericson
341c42f321 Merge pull request #14454 from Radvendii/exprconcatstrings-alloc
libexpr: move ExprConcatStrings data to Exprs::alloc
2025-11-04 04:21:16 +00:00
Taeer Bar-Yam
631fb6c9ad libexpr: move ExprConcatStrings data to Exprs::alloc
ExprConcatStrings no longer consumes the vector argument

Co-authored-by: John Ericson <git@JohnEricson.me>
2025-11-04 04:17:35 +01:00
Taeer Bar-Yam
11e19ee690 libexpr: shrink ExprWith by 8 bytes
Correct bound on prevWith

Co-authored-by: John Ericson <git@JohnEricson.me>
2025-11-04 04:16:20 +01:00
Robert Hensing
9f322398b4 Run linkcheck as regular passthru test
... and add nix-manual.site attribute for a nice and DRY aftertaste.
2025-11-04 01:17:39 +01:00
Robert Hensing
e07510e504 Make nix check .#linkcheck pass
It's not quite perfect yet, with two kinds of excludes, but at least
we won't regress!
2025-11-04 00:31:46 +01:00
Robert Hensing
ae15d4eaf3 Fix links in the manual 2025-11-04 00:31:46 +01:00
Robert Hensing
469123eda1 doc: Check link fragments with lychee 2025-11-04 00:31:46 +01:00
John Ericson
389bcba97a JSON Impl and schema for BuildResult 2025-11-03 18:25:16 -05:00
Sergei Zimmerman
3ef22a521d Merge pull request #14468 from fzakaria/fzakaria/nar-warning
Fix warning in kaitai spec
2025-11-03 22:54:55 +00:00
Farid Zakaria
c8e24491c0 Fix warning in kaitai spec
Warning:
```
[39/483] Generating src/kaitai-struct-checks/kaitai-generated-sources with a custom command
../src/kaitai-struct-checks/nar.ksy: /types/padded_str/seq/1/encoding:
        warning: use canonical encoding name `ASCII` instead of `ascii` (see https://doc.kaitai.io/ksy_style_guide.html#encoding-name)
```
2025-11-03 14:19:54 -08:00
dependabot[bot]
c3d4c5f69d build(deps): bump cachix/install-nix-action from 31.5.1 to 31.8.2
Bumps [cachix/install-nix-action](https://github.com/cachix/install-nix-action) from 31.5.1 to 31.8.2.
- [Release notes](https://github.com/cachix/install-nix-action/releases)
- [Changelog](https://github.com/cachix/install-nix-action/blob/master/RELEASE.md)
- [Commits](c134e4c9e3...456688f15b)

---
updated-dependencies:
- dependency-name: cachix/install-nix-action
  dependency-version: 31.8.2
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-11-03 22:00:54 +00:00
Jörg Thalheim
43ce9da6ad Merge pull request #14408 from obsidiansystems/hash-derivation-modulo
Document "hash derivation quotiented", resolution, and build trace
2025-11-03 21:41:49 +00:00
John Ericson
144c66215b JSON Schema for build trace entry
Note, starting to make progress on #11895 by calling it this in the
manual.
2025-11-03 15:59:50 -05:00
John Ericson
0d7b16da4d Split realisation protocol unit tests
This will allow us to more accurately test dropping support for
dependent realisations, by separating the tests that should not change
from the tests that should.

I do that change in PR #14247, but even if for some reasons we don't end
up doing this soon, I think it is still good to separate the test data
this way so we have the option of doing that at some point.
2025-11-03 15:43:38 -05:00
John Ericson
72d0f7b619 Document "hash derivation quotiented", resolution, and build trace
Progress on #13405, which asks for an explicit characterisation of the
equivalence relation like the one given here.

Also progress on #11895, because we're using the term "build trace
entry" instead of "realisation".

Mention #9259, a future work item.

Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
2025-11-03 15:18:18 -05:00
John Ericson
34ac1792f9 Merge pull request #14330 from lovesegfault/nix-s3-multipart
feat(libstore): add support for multipart s3 uploads
2025-11-03 18:36:48 +00:00
John Ericson
0586370e58 Merge pull request #14345 from fzakaria/fzakaria/nar-kaitai-spec
Add documentation for NAR spec in kaitai
2025-11-03 18:24:26 +00:00
John Ericson
f63bb5b338 Merge pull request #14427 from obsidiansystems/drv-output-better-schema
Better JSON schema for derivation outputs
2025-11-03 18:23:29 +00:00
Farid Zakaria
53b4ea6c85 Add documentation for NAR spec in kaitai
* Add a new flake check
* Add unit tests
* Add Kaitai spec
* Updated documentation
2025-11-03 12:59:16 -05:00
John Ericson
7c85ac23e2 Merge pull request #14444 from NixOS/less-c_str
Use less `c_str()` in the evaluator, and other cleanups
2025-11-03 17:56:22 +00:00
John Ericson
0539b58253 Merge pull request #14246 from obsidiansystems/dummy-store-derivations-separately
Make Dummy store store derivations separately
2025-11-03 17:29:28 +00:00
Eelco Dolstra
beace42e7a Merge pull request #14458 from NixOS/thread-pool-move
ThreadPool::enqueue(): Use move semantics
2025-11-03 17:27:14 +00:00
Eelco Dolstra
4a0ccc89d9 ThreadPool::enqueue(): Use move semantics
This avoids a superfluous copy of the work item.
2025-11-03 17:39:18 +01:00
Sergei Zimmerman
89fa8c09a9 Merge pull request #14450 from roberth/update
flake: Update, nixos-25.05-small -> nixos-25.05
2025-11-03 16:19:58 +00:00
Eelco Dolstra
5e025ce940 Merge pull request #14456 from NixOS/remove-infoAttrs
getAccessorFromCommit(): Remove superfluous infoAttrs variable
2025-11-03 13:33:58 +00:00
Eelco Dolstra
2f6c865e25 getAccessorFromCommit(): Remove superfluous infoAttrs variable 2025-11-03 13:23:09 +01:00
John Ericson
bd42092873 Use less c_str() in the evaluator, and other cleanups
It is better to avoid null termination for performance and memory
safety, wherever possible.

These are good cleanups extracted from the Pascal String work that we
can land by themselves first, shrinking the diff in that PR.

Co-Authored-By: Aspen Smith <root@gws.fyi>
Co-Authored-By: Sergei Zimmerman <sergei@zimmerman.foo>
2025-11-03 15:14:50 +03:00
Robert Hensing
81a2809a52 Apply updated nixfmt 2025-11-03 12:01:55 +01:00
Bernardo Meurer Costa
3448d4fa4c docs(rl-next/s3-curl-implementation): update with multipart uploads 2025-11-03 01:15:46 +00:00
Bernardo Meurer Costa
965d6be7c1 tests(nixos/s3-binary-cache-store): enable multipart 2025-11-03 01:15:46 +00:00
Bernardo Meurer Costa
040d1aae41 feat(libstore/s3-binary-cache-store): implement uploadMultipart()
Implement `uploadMultipart()`, the main method that orchestrates S3
multipart uploads
2025-11-03 01:15:46 +00:00
Bernardo Meurer Costa
bf947bfc26 feat(libstore/s3-binary-cache-store): add multipart upload config settings
Add three configuration settings to `S3BinaryCacheStoreConfig` to control
multipart upload behavior:

- `bool multipart-upload` (default `false`): Enable/disable multipart uploads
- `uint64_t multipart-chunk-size` (default 5 MiB): Size of each upload part
- `uint64_t multipart-threshold` (default 100 MiB): Minimum file size for multipart

The feature is disabled by default.
2025-11-02 18:41:48 +00:00
John Ericson
2d83bc6b83 Merge pull request #14446 from mjoerg/fix-docs
fix documentation issues
2025-11-02 16:59:33 +00:00
Sergei Zimmerman
e0debd61d5 Merge pull request #14449 from roberth/error-resolution-failed
Improve "resolution failed" error
2025-11-02 15:00:56 +00:00
Robert Hensing
233bd250d1 flake: Update, nixos-25.05-small -> nixos-25.05
Flake lock file updates:

• Updated input 'nixpkgs':
    'github:NixOS/nixpkgs/d98ce345cdab58477ca61855540999c86577d19d?narHash=sha256-O2CIn7HjZwEGqBrwu9EU76zlmA5dbmna7jL1XUmAId8%3D' (2025-08-26)
  → 'github:NixOS/nixpkgs/daf6dc47aa4b44791372d6139ab7b25269184d55?narHash=sha256-wxX7u6D2rpkJLWkZ2E932SIvDJW8%2BON/0Yy8%2Ba5vsDU%3D' (2025-10-27)
2025-11-02 14:10:36 +01:00
Robert Hensing
4ea32d0b03 Improve "resolution failed" error
Previously:

error: Cannot build '/nix/store/cqc798lwy2njwbdzgd0319z4r19j2d1w-nix-manual-2.33.0pre20251101_e4e4063.drv'.
       Reason: 1 dependency failed.
       Output paths:
         /nix/store/f1kln1c6z9r7rlhj0h9shcpch7j5g1fj-nix-manual-2.33.0pre20251101_e4e4063-man
         /nix/store/k65203rx5g1kcagpcz3c3a09bghcj92a-nix-manual-2.33.0pre20251101_e4e4063
error: Cannot build '/nix/store/ajk2fb6r7ijn2fc5c3h85n6zdi36xlfl-nixops-manual.drv'.
       Reason: 1 dependency failed.
       Output paths:
         /nix/store/0anr0998as8ry4hr5g3f3iarszx5aisx-nixops-manual
error: resolution failed

Now:

error: Cannot build '/nix/store/cqc798lwy2njwbdzgd0319z4r19j2d1w-nix-manual-2.33.0pre20251101_e4e4063.drv'.
       Reason: 1 dependency failed.
       Output paths:
         /nix/store/f1kln1c6z9r7rlhj0h9shcpch7j5g1fj-nix-manual-2.33.0pre20251101_e4e4063-man
         /nix/store/k65203rx5g1kcagpcz3c3a09bghcj92a-nix-manual-2.33.0pre20251101_e4e4063
error: Cannot build '/nix/store/ajk2fb6r7ijn2fc5c3h85n6zdi36xlfl-nixops-manual.drv'.
       Reason: 1 dependency failed.
       Output paths:
         /nix/store/0anr0998as8ry4hr5g3f3iarszx5aisx-nixops-manual
error: Build failed due to failed dependency
2025-11-02 14:03:27 +01:00
Martin Joerg
892eba4944 fix documentation issues 2025-11-02 09:17:40 +00:00
Sergei Zimmerman
e4e4063f16 Merge pull request #14443 from NixOS/inline-unreused-lambda
Inline only-used-once closures in `ExprConcatStrings::eval`
2025-11-01 21:55:03 +00:00
Vladimir Panteleev
d8cec03fce nix flake check: log success in verbose mode
The rule of silence can be a little surprising. As a compromise to
changing the default behavior, this adds printing a success message in
verbose mode, where we don't really have a reason to be silent about
our success.
2025-11-01 21:45:42 +00:00
Aspen Smith
b67c2f1572 Inline only-used-once closures in ExprConcatStrings::eval
Refactor `ExprConcatStrings::eval` by inlining two only-called-once
closures into the call-site, so that the code is easier to reason about
locally (especially since the variables that were closed over were
mutated all over the place within this function).

Also use curly braces with each branch for consistency in the the
resulting code.

This is a pure refactor, but also arguably causes us to depend less on
the optimizer; now, we don't have to make sure that this closure is
inlined.
2025-11-01 16:35:54 -04:00
John Ericson
ca9fde1b88 In EvalState::concatLists, local variable s -> strings
It deserves a better name.

Co-Authored-By: Aspen Smith <root@gws.fyi>
2025-11-01 13:41:50 -04:00
John Ericson
0ba1aa34dc Merge pull request #14440 from lovesegfault/cleanup-aws-sdk
chore(libstore/package): remove lingering aws-sdk-cpp
2025-11-01 15:08:21 +00:00
John Ericson
6fa7510055 Merge pull request #14439 from NixOS/no-buffer-overflows
libexpr: Do not overflow heap buffer when there are too many formal a…
2025-11-01 14:48:11 +00:00
Bernardo Meurer Costa
8151afb345 chore(libstore/package): remove lingering aws-sdk-cpp
This was left behind during the great s3 refactoring of 2025
2025-11-01 14:42:07 +00:00
Sergei Zimmerman
134613e885 libexpr: Do not overflow heap buffer when there are too many formal arguments
3a3c062982 introduced a buffer overflow for the
case when there are more than 65535 formal arguments. It is a perfectly reasonable
limitation, but we *must* not crash, corrupt memory or otherwise crash the process.

Add a test for the graceful behavior and switch to using an explicit uninitialized_copy_n
to further guard against buffer overflows.
2025-11-01 12:53:53 +03:00
John Ericson
9d1907fff7 Merge pull request #14434 from NixOS/improve-ipv6-zoneid-backcompat
libstore: Improve store-reference back-compat with IPv6 ZoneId literals
2025-10-31 23:10:40 +00:00
John Ericson
c29411ada9 Merge pull request #14431 from NixOS/git-url-fixes
libfetchers: Restore plain git inputs recognition
2025-10-31 22:28:14 +00:00
Sergei Zimmerman
8dbc2475f7 libstore: Improve store-reference back-compat with IPv6 ZoneId literals
This restores the pre-2.31 handling of ZoneID identifiers in store references.
It's the only place we reasonably care about this back-compat.
2025-11-01 00:36:15 +03:00
John Ericson
9e79e83cb5 Merge pull request #14384 from Radvendii/exprlambda-alloc
libexpr: store ExprLambda data in Expr::alloc
2025-10-31 21:12:30 +00:00
John Ericson
937a6df809 Merge pull request #14432 from NixOS/meson-darwin-soname
meson: Also split version string at '+' for Darwin
2025-10-31 21:03:47 +00:00
Sergei Zimmerman
1ca6e9ef54 meson: Also split version string at '+' for Darwin 2025-10-31 23:12:54 +03:00
Sergei Zimmerman
ade3d5d746 libfetchers: Restore plain git inputs recognition
Accidentally broken in dbc235cc62.
Adds a bit of tests for this, even though this protocol is mostly deprecated
everywhere.
2025-10-31 22:42:43 +03:00
John Ericson
d035d8ba8d Merge pull request #14428 from obsidiansystems/path-info-parse-json-cleanup
Clean up `PathInfo::fromJSON` using recent JSON utils changes
2025-10-31 19:27:09 +00:00
Taeer Bar-Yam
67be2df174 remove unnecessary constructor argument 2025-10-31 16:54:59 +01:00
Taeer Bar-Yam
34f780d747 safer interface for ExprLambda's formals 2025-10-31 16:54:59 +01:00
Taeer Bar-Yam
e43888890f restore proper handling of no formals vs. 0 formals
e.g. (foo@{}: 1) { a = 3; } should error, but wasn't with the previous
commit
2025-10-31 16:54:59 +01:00
Taeer Bar-Yam
4a80c92a4d add test for no formals case 2025-10-31 16:54:59 +01:00
Taeer Bar-Yam
3a3c062982 libexpr: store ExprLambda data in Expr::alloc 2025-10-31 16:54:59 +01:00
John Ericson
4a2fb18ba0 Merge pull request #14137 from lovesegfault/nix-debug-14130
fix(libstore/build/derivation-goal): don't assert on partially valid outputs
2025-10-31 02:45:50 +00:00
Bernardo Meurer Costa
9eecee3d4e fix(libstore/build/derivation-goal): don't assert on partially valid outputs
Fixes: #14130
2025-10-31 01:58:02 +00:00
John Ericson
089a222111 Clean up PathInfo::fromJSON using recent JSON utils changes
`optionalValueAt` and then `optionalValueAt` avoids looking up twice.
2025-10-30 18:38:27 -04:00
John Ericson
c2609df08c Better JSON schema for derivation outputs
It now uses a `oneOf` and properly models each type of output
(corresponding to each type of derivation) separately.
2025-10-30 17:05:00 -04:00
John Ericson
37c1ef52e6 Merge pull request #14412 from NixOS/recursive-lambdas
Cleanup: Use C++23 "this auto" for recursive lambdas
2025-10-30 20:41:52 +00:00
Sergei Zimmerman
e776a10db3 Merge pull request #14356 from lovesegfault/s3-upload
refactor(libstore/s3-binary-cache-store): implement `upload()`
2025-10-30 20:19:59 +00:00
Eelco Dolstra
1507843f6c Cleanup: Use C++23 "explicit this" for recursive lambdas
Try to pass by reference where possible.

Co-authored-by: Sergei Zimmerman <sergei@zimmerman.foo>
2025-10-30 15:56:06 -04:00
Bernardo Meurer Costa
e636888a09 refactor(libstore/s3-binary-cache-store): implement upload()
Stop delegating to `HttpBinaryCacheStore::upsertFile` and instead
handle compression in the S3 store's `upsertFile` override, then call
our own `upload()` method. This separation is necessary for future
multipart upload support.
2025-10-30 19:01:05 +00:00
Eelco Dolstra
3b2186e1c8 Merge pull request #14397 from fzakaria/fzakaria/issue-14315
Move docker documentation to docker.io
2025-10-30 17:38:22 +00:00
John Ericson
7e2d2db8ef Merge pull request #14399 from obsidiansystems/json-schema-path-info
Convert store path info JSON docs to formal JSON Schema, and test
2025-10-30 17:35:20 +00:00
Eelco Dolstra
2cc53201eb Merge pull request #14418 from lovesegfault/fix-curl-interrupt
fix(libstore/filetransfer): prevent double callback invocation on interrupt during retry
2025-10-30 17:12:15 +00:00
Eelco Dolstra
720f693627 Merge pull request #14416 from lovesegfault/fix-lexer-warn
fix(libexpr/lexer): fix flex warning about default rule
2025-10-30 17:11:23 +00:00
John Ericson
49084a7e9e Merge pull request #14421 from lovesegfault/http-upload
refactor(libstore): add `HttpBinaryCacheStore::upload` method
2025-10-30 15:11:58 +00:00
tomberek
6d87184a52 Merge pull request #14363 from cootshk/patch-1
fix(libstore): Rewrite hard linking message to be more clear
2025-10-30 09:23:00 +00:00
Henry
6985e9f2c3 fix(libstore): Rewrite hard linking message to be more clear 2025-10-30 03:05:06 -05:00
John Ericson
e6f0dd8df5 Merge pull request #14420 from lovesegfault/compressed-source
refactor(libutil): add `CompressedSource`
2025-10-30 05:21:41 +00:00
Bernardo Meurer Costa
d857a4be50 refactor(libstore): add HttpBinaryCacheStore::upload method
Introduce protected `upload` method overloads in `HttpBinaryCacheStore`
that handle the actual upload after compression has been applied. This
separates compression concerns (in `upsertFile`) from upload mechanics
(in `upload`).

Two overloads are provided:

1. `upload(path, RestartableSource &, sizeHint, mimeType, contentEncoding)`
2. `upload(path, CompressedSource &, mimeType)`
2025-10-30 04:35:43 +00:00
Bernardo Meurer Costa
93fe3354b5 refactor(libutil): add CompressedSource
Introduce a `CompressedSource` class in libutil's `serialise.hh` that
compresses a `RestartableSource` and owns the compressed data. This is a
general-purpose utility that can be used anywhere compressed data needs
to be treated as a source.
2025-10-30 04:35:27 +00:00
Bernardo Meurer Costa
8b3af40006 fix(libexpr/lexer): fix flex warning about default rule
We were getting this flex lexer warning during build:
```
../src/libexpr/lexer.l:333: warning, -s option given but default rule can be matched
```

The lexer uses `%option nodefault` but the `PATH_START` state only had
rules for specific patterns (`PATH_SEG` and `HPATH_START`) without a
catch-all rule to handle unexpected input.

Added a catch-all rule with `unreachable()`. This code path should never
be reached in normal operation since `PATH_START` is only entered after
matching `PATH_SEG` or `HPATH_START`, and we immediately rewind to
re-parse those same patterns. The catch-all exists solely to satisfy
flex's `%option nodefault` requirement.
2025-10-29 23:55:37 +00:00
John Ericson
bffbdcfddc Merge pull request #14390 from NixOS/constant-memory-uploads
libstore: Make HTTP binary cache uploads run in constant memory
2025-10-29 23:14:42 +00:00
John Ericson
495d1b8435 Merge pull request #14393 from lovesegfault/s3-multipart-tests
test(nixos): add S3 multipart upload integration tests
2025-10-29 22:56:21 +00:00
John Ericson
66d7b8fe1b Merge pull request #14396 from roberth/c-api-docs
doc: Improve libexpr-c docs
2025-10-29 22:38:04 +00:00
Sergei Zimmerman
cf75079bd8 libstore: Make uploads with filetransfer.cc consume a RestartableSource
Make uploads run in constant memory. Also change the callbacks to be
noexcept, since we really don't want to be unwinding the stack in the
curl thread. That will definitely corrupt that stack and make nix/curl
crash in very bad ways.
2025-10-29 18:34:56 -04:00
Sergei Zimmerman
b8d7f551e4 libutil: Add RestartableSource
This is necessary to make seeking work with libcurl.
2025-10-29 18:25:49 -04:00
Sergei Zimmerman
e947c895ec binary-cache-store: UpsertFile accept Source & instead of std::istream 2025-10-29 18:25:49 -04:00
Robert Hensing
f301669adc doc/dev/documentation: Use appendToVar
Co-authored-by: John Ericson <git@JohnEricson.me>
2025-10-29 22:53:43 +01:00
John Ericson
e3c41407f9 Merge pull request #14391 from lovesegfault/nix-s3-complete-multipart
feat(libstore/s3-binary-cache-store): implement `completeMultipartUpload()`
2025-10-29 20:49:12 +00:00
John Ericson
00f4a860e7 Merge pull request #14400 from obsidiansystems/json-schema-derivation-output
Enable JSON schema testing for derivation outputs
2025-10-29 19:40:29 +00:00
Bernardo Meurer Costa
560a596de7 fix(libstore/filetransfer): prevent double callback invocation on interrupt during retry
Fix a race condition where interrupting a download (via Ctrl-C) during a
retry attempt could cause a crash. When `enqueueItem()` throws because the
download thread is shutting down, the exception would propagate without
setting `done=true`, causing the `TransferItem` destructor to invoke the
callback a second time.

This triggered an assertion failure in `Callback::rethrow()` with:
`Assertion '!prev' failed` and the error message `cannot enqueue download
request because the download thread is shutting down`.

The fix catches the exception from `enqueueItem()` and calls `fail()` to
properly complete the transfer, ensuring the callback is invoked exactly
once.
2025-10-29 18:12:47 +00:00
Sergei Zimmerman
da637a05da Merge pull request #14410 from bryango/patch-1
zsh/completion: put compdef on first line
2025-10-29 15:24:02 +00:00
bryango
956fffdd6f zsh/completion: put compdef on first line
Some zsh setups (including mine) do not load the
completion if `#compdef` is not on the first line.

So we move the `# shellcheck` comment to the
second line to avoid this issue.
2025-10-29 18:09:42 +08:00
John Ericson
bac41d6989 Merge pull request #14289 from obsidiansystems/fix-14287
Fix issue #14287
2025-10-29 07:17:59 +00:00
John Ericson
de192794c9 Fix issue #14287
The test added in the previous commit now passes.

Co-authored-by: Eelco Dolstra <edolstra@gmail.com>
2025-10-29 02:15:46 -04:00
John Ericson
246dbe1c05 Regression test for issue #14287
This will currently fail, until the bug is fixed.

Co-Authored-By: Sergei Zimmerman <sergei@zimmerman.foo>
2025-10-29 02:15:41 -04:00
John Ericson
6280905638 Convert store path info JSON docs to formal JSON Schema, and test
This continues the work for formalizing our current JSON docs. Note that
in the process, a few bugs were caught:

 - `closureSize` was repeated twice, forgot `closureDownloadSize`

 - `file*` fields should be `download*`. They are in fact called that in
   the line-oriented `.narinfo` file, but were renamed in the JSON
   format.
2025-10-28 23:28:16 -04:00
Sergei Zimmerman
194c21fc82 Merge pull request #14407 from NixOS/fix-upload-put-http
libstore/filetransfer: Add HttpMethod::PUT
2025-10-29 03:24:10 +00:00
Sergei Zimmerman
e08853a67c Merge pull request #14406 from NixOS/better-error-message
libstore/http-binary-cache-store: Improve error messages in HttpBinar…
2025-10-29 03:23:49 +00:00
Sergei Zimmerman
ae49074548 libstore/filetransfer: Add HttpMethod::PUT
This got lost in f1968ea38e and
now we had incorrect logs that confused "downloading" when we were
in fact "uploading" things.
2025-10-29 02:48:26 +03:00
Sergei Zimmerman
f1d4fab1e5 Merge pull request #14405 from obsidiansystems/json-schema-store-path
Create JSON Schema for Store Paths
2025-10-28 23:24:05 +00:00
Sergei Zimmerman
c874e7071b libstore/http-binary-cache-store: Improve error messages in HttpBinaryCacheStore::upsertFile
Now the error message doesn't cram everything into a single line and we now instead get:

error:
       … while uploading to HTTP binary cache at 's3://my-cache?endpoint=http://localhost:9000?compression%3Dzstd&region=eu-west-1'

       error: unable to download 'http://localhost:9000/my-cache/nar/1125zqba8cx8wbfa632vy458a3j3xja0qpcqafsfdildyl9dqa7x.nar.xz': Operation was aborted by an application callback (42)
2025-10-29 02:05:14 +03:00
John Ericson
c67966418f Create JSON Schema for Store Paths
We immediately use this in the JSON schemas for Derivation and Deriving
Path, but we cannot yet use it in Store Object Info because those paths
*do* include the store dir currently.
2025-10-28 17:22:51 -04:00
John Ericson
be2572ed8d Make inputDrvs JSON schema more precise
It now captures the stable non-recursive format (just an output set) and
the unstable recursive form for dynamic derivations.
2025-10-28 17:22:30 -04:00
Sergei Zimmerman
be99a1c6bb Merge pull request #14404 from Mic92/test-settings
coderabbit: disable high_level_summary/poem/github status
2025-10-28 21:10:56 +00:00
Jörg Thalheim
fe8cdbc3e4 coderabbit: disable high_level_summary/poem/github status/sequence_diagrams 2025-10-28 22:09:05 +01:00
Sergei Zimmerman
70176ed317 Merge pull request #14402 from Mic92/test-settings
coderabbit: don't show review status
2025-10-28 20:46:31 +00:00
Jörg Thalheim
84a5bee424 coderabbit: disable reporting review status 2025-10-28 21:45:30 +01:00
John Ericson
e3246301a6 Enable JSON schema testing for derivation outputs
I figured out what the problem was: the fragment needs to start with a
`/`.
2025-10-28 16:07:44 -04:00
Sergei Zimmerman
d4c69c7b8f Merge pull request #14398 from roberth/quiet-coderabbit 2025-10-28 18:41:43 +00:00
Robert Hensing
f5aafbd6ed .coderabbit.yaml: Disable auto-review 2025-10-28 19:39:04 +01:00
Farid Zakaria
943788754f Add ghcr for pre-release 2025-10-28 11:16:37 -07:00
Farid Zakaria
883860c7ff Move docker documentation to docker.io 2025-10-28 11:14:31 -07:00
Robert Hensing
5fc0c4f102 doc: Improve libexpr-c docs
- Uses the more explicit `@ingroup` most of the time, to avoid problems
  with nested groups, and to make group membership more explicit.
  The division into headers is not great for documentation purposes,
  so this helps.
- More attention for memory management details
- Various other improvements to doc comments
2025-10-28 17:57:15 +01:00
Eelco Dolstra
1a4ad0706b Merge pull request #14394 from me-and/no-print-dead-space-usage
docs: remove incorrect claim re gc --print-dead
2025-10-28 15:10:54 +00:00
Adam Dinwoodie
972915cabd docs: remove incorrect claim re gc --print-dead
Per #7591, the `nix-store --gc --print-dead` command does not provide
any feedback about the amount of disk space that is used by dead store
paths.  It looks like this has been the case since 7ab68961e (* Garbage
collector: added an option `--use-atime' to delete paths in...,
2008-09-17).

Update the nix-store documentation to remove the claim that this is
function that `nix-store --gc --print-dead` performs.
2025-10-28 09:47:25 +00:00
Bernardo Meurer Costa
94965a3a3e test(nixos): add S3 multipart upload integration tests 2025-10-28 06:17:41 +00:00
Bernardo Meurer Costa
c77317b1a9 feat(libstore/s3-binary-cache-store): implement completeMultipartUpload()
`completeMultipartUpload()`: Build XML with part numbers and `ETags`,
POST to key with `?uploadId` to finalize the multipart upload
2025-10-28 01:13:28 +00:00
Jörg Thalheim
dd0d006517 Merge pull request #14375 from lovesegfault/nix-s3-upload-part
feat(libstore/s3-binary-cache-store): implement `uploadPart()`
2025-10-27 22:40:00 +00:00
dependabot[bot]
ccc06451df build(deps): bump actions/download-artifact from 5 to 6
Bumps [actions/download-artifact](https://github.com/actions/download-artifact) from 5 to 6.
- [Release notes](https://github.com/actions/download-artifact/releases)
- [Commits](https://github.com/actions/download-artifact/compare/v5...v6)

---
updated-dependencies:
- dependency-name: actions/download-artifact
  dependency-version: '6'
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-27 22:35:42 +00:00
dependabot[bot]
3775a2a226 build(deps): bump actions/upload-artifact from 4 to 5
Bumps [actions/upload-artifact](https://github.com/actions/upload-artifact) from 4 to 5.
- [Release notes](https://github.com/actions/upload-artifact/releases)
- [Commits](https://github.com/actions/upload-artifact/compare/v4...v5)

---
updated-dependencies:
- dependency-name: actions/upload-artifact
  dependency-version: '5'
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-27 22:22:54 +00:00
John Ericson
1d3f0ca22e Merge pull request #14383 from obsidiansystems/misc-cleanups
Two misc cleanups
2025-10-27 22:16:37 +00:00
John Ericson
1c41e07b46 Merge pull request #14385 from lovesegfault/ci-concurrency-group
ci: cancel previous workflow runs on PR updates
2025-10-27 17:35:50 -04:00
Bernardo Meurer Costa
c592090fff feat(libstore/s3-binary-cache-store): implement uploadPart()
Implement `uploadPart()` for uploading individual parts in S3 multipart
uploads:

- Constructs URL with `?partNumber=N&uploadId=ID` query parameters
- Uploads chunk data with `application/octet-stream` mime type
- Extracts and returns `ETag` from response
2025-10-27 21:09:39 +00:00
Bernardo Meurer Costa
4b6d07d642 feat(libstore/s3-binary-cache-store): implement createMultipartUpload()
POST to key with `?uploads` query parameter, optionally set
`Content-Encoding` header, parse `uploadId` from XML response using
regex
2025-10-27 21:07:29 +00:00
John Ericson
e177f42536 Merge pull request #14379 from Radvendii/exprlist-alloc
libexpr: store ExprList data in Exprs::alloc
2025-10-27 21:04:45 +00:00
John Ericson
ac8b1efcf9 Merge pull request #14379 from Radvendii/exprlist-alloc
libexpr: store ExprList data in Exprs::alloc
2025-10-27 21:04:45 +00:00
Bernardo Meurer Costa
ad664ce64e ci: cancel previous workflow runs on PR updates
Add concurrency group configuration to the CI workflow to automatically
cancel outdated runs when a PR receives new commits or is force-pushed.
This prevents wasting CI resources on superseded code.
2025-10-27 20:56:56 +00:00
John Ericson
18941a2421 Optimize DummyStore::isValidPathUncached
See the API docs for the rationale of why this is needed.
2025-10-27 16:49:18 -04:00
John Ericson
136825b4a2 Make Dummy store store derivations separately
This makes for more efficiency. Once we have JSON for the dummy store,
it will also make for better JSON, too.
2025-10-27 16:49:18 -04:00
John Ericson
28b73cabcc Make reading and writing derivations store methods
This allows for different representations.
2025-10-27 16:49:18 -04:00
John Ericson
aa4106fd68 Merge pull request #14360 from lovesegfault/scan-for-references-detailed
feat(libstore): add scanForReferencesDeep and use it for why-depends
2025-10-27 20:38:10 +00:00
John Ericson
7f1d92793e Merge pull request #14360 from lovesegfault/scan-for-references-detailed
feat(libstore): add scanForReferencesDeep and use it for why-depends
2025-10-27 20:38:10 +00:00
John Ericson
234f029940 Add consuming ref <-> std::share_ptr methods/ctrs
This can help churning ref counts when we don't need to.
2025-10-27 16:23:43 -04:00
John Ericson
dd716dc9be Create default Store::narFromPath implementation in terms of getFSAccessor
This is a good default (the methods that allow for an arbitrary choice
of source accessor are generally preferable both to implement and to
use). And it also pays its way by allowing us to delete *both* the
`DummyStore` and `LocalStore` implementations.
2025-10-27 15:57:26 -04:00
John Ericson
ea17cc1b57 Merge pull request #14376 from lovesegfault/nix-s3-abort-multipart
feat(libstore/s3-binary-cache-store): implement `abortMultipartUpload()`
2025-10-27 19:52:36 +00:00
John Ericson
0c1be3aabe Merge pull request #14309 from obsidiansystems/json-schema-content-address
` nlohmann::json` instance and JSON Schema for `ContentAddress`
2025-10-27 19:52:19 +00:00
John Ericson
6ca3434cac Merge pull request #14309 from obsidiansystems/json-schema-content-address
` nlohmann::json` instance and JSON Schema for `ContentAddress`
2025-10-27 19:52:19 +00:00
Bernardo Meurer Costa
6129aee988 refactor(nix/why-depends): use scanForReferencesDeep for --precise mode
Replaces manual tree-walking and reference scanning with the new
scanForReferencesDeep function.
2025-10-27 19:14:49 +00:00
Bernardo Meurer Costa
5e220271e2 feat(libstore): add scanForReferencesDeep for per-file reference tracking
Introduces `scanForReferencesDeep` to provide per-file granularity when
scanning for store path references, enabling better diagnostics for
cycle detection and `nix why-depends --precise`.
2025-10-27 19:14:49 +00:00
John Ericson
8e6b69de54 Merge pull request #14378 from Radvendii/parser-improvements
parser.y: remove some unnecessary copies
2025-10-27 19:11:11 +00:00
Bernardo Meurer Costa
3915b3a111 feat(libstore/s3-binary-cache-store): implement abortMultipartUpload()
Implement `abortMultipartUpload()` for cleaning up incomplete multipart
uploads on error:

- Constructs URL with `?uploadId=ID` query parameter
- Issues `DELETE` request to abort the multipart upload
2025-10-27 18:56:52 +00:00
Jörg Thalheim
c5515bb22e Merge pull request #14364 from MarcelCoding/human-sizes
diff-closures: print sizes with dynamic unit
2025-10-27 18:49:23 +00:00
John Ericson
91b69e9e70 nlohmann::json instance and JSON Schema for ContentAddress
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
2025-10-27 14:47:50 -04:00
Taeer Bar-Yam
9e9dfe36df libexpr: store ExprList data in Exprs::alloc 2025-10-27 19:38:03 +01:00
Taeer Bar-Yam
50e8d17f3c parser.y: use emplace_back() for vector<AttrName> 2025-10-27 19:30:32 +01:00
Taeer Bar-Yam
ef8dd58d9b parser.y: use std::move() to avoid unnecessary copies
With #14314, in some places in the parser we started using C++ objects
directly rather than pointers. In those places lines like `$$ = $1` now
imply a copy when we don't need one. This commit changes those to `$$ =
std::move($1)` to avoid those copies.
2025-10-27 19:30:32 +01:00
John Ericson
91ed3701fe Merge pull request #14377 from lovesegfault/makerequest-stringview
refactor(libstore): use string_view in HttpBinaryCacheStore::makeRequest
2025-10-27 17:58:53 +00:00
Jörg Thalheim
b8e5d1f290 Merge pull request #14369 from NixOS/copy-sigs-docs
nix store copy-sigs: Add docs
2025-10-27 17:55:56 +00:00
Jörg Thalheim
d44b33562f Merge pull request #14373 from NixOS/copy-sigs-parallel
nix store copy-sigs: Use http-connections setting to control parallelism
2025-10-27 17:53:23 +00:00
Jörg Thalheim
d46504a136 Merge pull request #14359 from obsidiansystems/structured-attrs-always-object
Use types to show that structured attrs are always JSON objects
2025-10-27 17:51:33 +00:00
Eelco Dolstra
126f30deb2 Merge pull request #14366 from NixOS/const-fields
EvalState: Make some more fields const
2025-10-27 17:47:21 +00:00
Bernardo Meurer Costa
5dcfa86910 refactor(libstore): use string_view in HttpBinaryCacheStore::makeRequest 2025-10-27 17:13:48 +00:00
Eelco Dolstra
6b6ceddf72 nix store copy-sigs: Use http-connections setting to control parallelism
Previously it used the `ThreadPool` default,
i.e. `std::thread::hardware_concurrency()`. But copying signatures is
not primarily CPU-bound so it makes more sense to use the
`http-connections` setting (since we're typically copying from/to a
binary cache).
2025-10-27 16:43:25 +01:00
Eelco Dolstra
60f9489b83 Merge pull request #14370 from NixOS/misc-cleanups
Miscellaneous cleanups
2025-10-27 15:04:53 +00:00
Marcel
584a8e8a00 treewide: replace manual MiB calculations with renderSize 2025-10-27 16:04:19 +01:00
Marcel
f234633e27 refactor(libutil): remove showBytes() in favor of renderSize()
The `showBytes()` function was redundant with `renderSize()` as the
latter automatically selects the appropriate unit (KiB, MiB, GiB, etc.)
based on the value, whereas `showBytes()` always formatted as MiB
regardless of size.

Co-authored-by: Bernardo Meurer Costa <beme@anthropic.com>
2025-10-27 16:04:08 +01:00
Sergei Zimmerman
6417863ce9 Merge pull request #14357 from lovesegfault/s3-setup-pub
refactor(libstore/filetransfer): make setupForS3 public
2025-10-27 14:36:41 +00:00
Eelco Dolstra
91cd42511e Introduce MINIMUM_PROTOCOL_VERSION constant 2025-10-27 15:11:20 +01:00
Eelco Dolstra
1af5a98955 Document removed WorkerProto ops 2025-10-27 15:09:03 +01:00
Eelco Dolstra
17777e3b70 Settings typos 2025-10-27 15:07:56 +01:00
Eelco Dolstra
9321669353 Make getDefaultCores() static 2025-10-27 15:07:01 +01:00
Eelco Dolstra
3742ae061e Typo 2025-10-27 15:04:56 +01:00
Eelco Dolstra
a91115bf22 Remove unnecessary virtual 2025-10-27 15:04:13 +01:00
Eelco Dolstra
8c8b706f6b Fix an update to a finished value 2025-10-27 15:01:46 +01:00
Eelco Dolstra
fb26285458 Fix #include 2025-10-27 14:53:46 +01:00
Eelco Dolstra
bbfaaf3a20 showHelp(): Use one callFunction 2025-10-27 14:52:18 +01:00
Sergei Zimmerman
f9b73185e4 Merge pull request #14362 from NixOS/k-way-merge-speedup
libexpr: Speed up BindingsBuilder::finishSizeIfNecessary
2025-10-27 13:45:29 +00:00
Eelco Dolstra
27e3d28ed8 Merge pull request #14340 from juhp/patch-1
nix-2.32 needs boost-1.87+ for `try_emplace_and_cvisit`
2025-10-27 13:44:37 +00:00
Eelco Dolstra
3994e5627f nix store copy-sigs: Add docs 2025-10-27 14:42:22 +01:00
Sergei Zimmerman
ec2fd2dc23 libexpr: Speed up BindingsBuilder::finishSizeIfNecessary
Instead of iterating over the newly built bindings we can
do a cheaper set_intersection to count duplicates or fall back
to a per-element binary search over the "base" bindings.

This speeds up `hello` evaluation by around 10ms (0.196s -> 0.187s) and
`nixos.closures.ec2.x86_64-linux` by 140ms (2.744s -> 2.609s).

This addresses a somewhat steep performance regression from 82315c3807
that reduced memory requirements of attribute set merges. With this patch
we get back around to 2.31 level of eval performance while keeping the memory
usage optimization.

Also document the optimization a bit more.
2025-10-27 16:14:19 +03:00
Eelco Dolstra
fdc5600fa7 makeRegexCache(): Return a ref 2025-10-27 14:11:59 +01:00
Eelco Dolstra
1f6ac88efc Mark some fields in EvalState as const 2025-10-27 14:10:34 +01:00
Marcel
9d4d10954a diff-closures: print sizes with dynamic unit 2025-10-27 02:05:03 +01:00
John Ericson
7e53afd8b9 Use types to show that structured attrs are always JSON objects
Before we just had partial code accessing it. Now, we use
`nlohmann::json::object_t`, which is a `std::map`, to enforce this by
construction.
2025-10-26 12:53:58 -04:00
John Ericson
bef3c37cb2 Merge pull request #14351 from obsidiansystems/json-project-reference
Clean up JSON utils in a few ways
2025-10-25 19:32:30 +00:00
John Ericson
0f0d9255c6 Clean up JSON utils in a few ways
In particular

- Remove `get`, it is redundant with `valueAt` and the `get` in
  `util.hh`.

- Remove `nullableValueAt`. It is morally just the function composition
  `getNullable . valueAt`, not an orthogonal combinator like the others.

- `optionalValueAt` return a pointer, not `std::optional`. This also
  expresses optionality, but without creating a needless copy. This
  brings it in line with the other combinators which also return
  references.

- Delete `valueAt` and `optionalValueAt` taking the map by value, as we
  did for `get` in 408c09a120, which
  prevents bugs / unnecessary copies.

`adl_serializer<DerivationOptions::OutputChecks>::from_json` was the one
use of `getNullable`. I give it a little static function for the
ultimate creation of a `std::optional` it does need to do (after
switching it to using `getNullable . valueAt`. That could go in
`json-utils.hh` eventually, but I didn't bother for now since only one
things needs it.

Co-authored-by: Sergei Zimmerman <sergei@zimmerman.foo>
2025-10-25 14:49:51 -04:00
Sergei Zimmerman
f0b95b6d5b Merge pull request #14274 from lovesegfault/nix-s3-versioned
feat(libstore): support S3 object versioning via versionId parameter
2025-10-25 08:39:12 +00:00
Bernardo Meurer Costa
e38128b90d feat(libstore): support S3 object versioning via versionId parameter
S3 buckets support object versioning to prevent unexpected changes,
but Nix previously lacked the ability to fetch specific versions of
S3 objects. This adds support for a `versionId` query parameter in S3
URLs, enabling users to pin to specific object versions:

```
s3://bucket/key?region=us-east-1&versionId=abc123
```
2025-10-25 07:57:58 +00:00
Bernardo Meurer Costa
78e98691d6 refactor(libstore/filetransfer): make setupForS3 public 2025-10-25 03:45:30 +00:00
John Ericson
e213fd64b6 Merge pull request #14352 from NixOS/source-paths-tests
tests/functional: Add source-paths tests
2025-10-24 23:50:07 +00:00
Eelco Dolstra
1cd8458c28 tests/functional: Add source-paths tests
This has already been implemented in 1e709554d5
as a side-effect of mounting the accessors in storeFS. Let's test this so it
doesn't regress.

(cherry-picked from https://github.com/NixOS/nix/pull/12915)
2025-10-25 02:13:30 +03:00
John Ericson
ecaf9470b9 Merge pull request #14344 from obsidiansystems/json-schema-deriving-path
JSON Schema for `DerivedPath`
2025-10-24 23:09:08 +00:00
Sergei Zimmerman
8b7e03f0f9 Merge pull request #14350 from lovesegfault/s3-binary-cache-store
refactor(libstore): expose HttpBinaryCacheStore and add S3BinaryCacheStore
2025-10-24 22:59:02 +00:00
Sergei Zimmerman
04606d50d1 Merge pull request #14343 from NixOS/epipe-graceful
Revert "libmain: Catch logger exceptions in `handleExceptions`"
2025-10-24 22:52:29 +00:00
Bernardo Meurer Costa
476c21d5ef refactor(libstore): expose HttpBinaryCacheStore and add S3BinaryCacheStore
Move HttpBinaryCacheStore class from .cc file to header to enable
inheritance by S3BinaryCacheStore. Create S3BinaryCacheStore class that
overrides upsertFile() to implement multipart upload logic.
2025-10-24 21:54:13 +00:00
John Ericson
1a9ba0d6fe Merge pull request #14333 from lovesegfault/upsert-size-hint
refactor(libstore): add sizeHint parameter to upsertFile()
2025-10-24 19:29:06 +00:00
John Ericson
648714cd44 Merge pull request #14336 from lovesegfault/filetransfer-delete
feat(libstore): add DELETE method support to FileTransfer
2025-10-24 18:50:53 +00:00
Bernardo Meurer Costa
6b7223b6b7 refactor(libstore): add sizeHint parameter to upsertFile()
Add a sizeHint parameter to BinaryCacheStore::upsertFile() to enable
size-based upload decisions in implementations. This lays the groundwork
for reintroducing S3 multipart upload support.
2025-10-24 18:49:28 +00:00
Bernardo Meurer Costa
afe5ed879f feat(libstore): add DELETE method support to FileTransfer
Add support for HTTP DELETE requests to FileTransfer infrastructure:

This enables S3 multipart upload abort functionality via DELETE requests
to S3 endpoints.
2025-10-24 18:03:14 +00:00
Bernardo Meurer Costa
d924374bf2 docs(libstore): document verb() method returns verb root for gerund form
Add documentation to FileTransferRequest::verb() explaining that it returns
a verb root intended to be concatenated with "ing" to form the gerund.
2025-10-24 18:03:13 +00:00
Bernardo Meurer Costa
f1968ea38e refactor(libstore): replace HTTP method boolean flags with enum
Replace the individual boolean flags (head, post) with a unified
HttpMethod enum struct in FileTransferRequest.
2025-10-24 18:03:12 +00:00
John Ericson
8d338c9234 JSON Schema for DerivedPath
Note that this is "deriving path" in the manual -- the great sed of the
code base to bring it in sync has yet to happen yet.
2025-10-24 12:08:00 -04:00
John Ericson
9a695f9067 Merge pull request #14348 from NixOS/fetchClosure-access
Allow access to the result of fetchClosure
2025-10-24 15:44:31 +00:00
Sergei Zimmerman
925c0fa4a2 Merge pull request #14346 from NixOS/remove-verify-tls
libstore/filetransfer: Remove verifyTLS from FileTransferRequest, sin…
2025-10-24 10:48:43 +00:00
Eelco Dolstra
7308fde0bc Allow access to the result of fetchClosure 2025-10-24 11:11:03 +02:00
Sergei Zimmerman
324bfd82dc Merge pull request #14337 from lovesegfault/fix-post-large
fix(libstore): use CURLOPT_POSTFIELDSIZE_LARGE for POST requests
2025-10-23 22:00:08 +00:00
Sergei Zimmerman
8e01e4ad5c Merge pull request #14347 from NixOS/mahic-nix-cache-hook-fix
ci: Bump magic-nix-cache with post-build-hook fix
2025-10-23 22:43:46 +00:00
Sergei Zimmerman
4c4eb5d07f ci: Bump magic-nix-cache with post-build-hook fix
No tagged release with the fix for [^].

[^]: 578f01e147
2025-10-24 01:34:09 +03:00
Sergei Zimmerman
b5ae3e10c2 libstore/filetransfer: Remove verifyTLS from FileTransferRequest, since it's always true
This variable is always true, so there's no use-case for it anymore.
2025-10-24 00:29:10 +03:00
Sergei Zimmerman
4f5af471fb Revert "libmain: Catch logger exceptions in handleExceptions"
This reverts commit 90d1ff4805.

The initial issue with EPIPE was solved in 9f680874c5.
Now this patch does move bad than good by eating up boost::io::format_error that are
bugs.
2025-10-23 23:49:41 +03:00
Sergei Zimmerman
b9af19cedf Merge pull request #14295 from NixOS/s3-store-human-readable-uri
libstore: Implement getHumanReadableURI for S3BinaryCacheStoreConfig
2025-10-23 19:33:49 +00:00
Eelco Dolstra
d6f1e2de21 Merge pull request #14323 from NixOS/skip-nar-parse
addToStore(): Don't parse the NAR

* StringSource: Implement skip()

This is slightly faster than doing a read() into a buffer just to
discard the data.

* LocalStore::addToStore(): Skip unnecessary NARs rather than parsing them

Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
2025-10-23 22:17:09 +03:00
John Ericson
5d365cd61f Merge pull request #14341 from obsidiansystems/fix-golden-tests
Fix some characterization tests
2025-10-23 19:08:43 +00:00
John Ericson
c87f29a0b6 Fix some characterization tests
A few changes had cropped up with `_NIX_TEST_ACCEPT=1`:

1. Blake hashing test JSON had a different indentation

2. Store URI had improper non-quoted spaces

(1) was is just fixed, as we trust nlohmann JSON to parse JSON
correctly, regardless of whitespace.

For (2), the existing URL was made a read-only test, since we very much
wish to continue parsing such invalid URLs directly. And then the
original read/write test was updated to properly percent-encode the
space, as the normal form should be.
2025-10-23 14:03:21 -04:00
Jens Petersen
f594a8e11e libexpr needs boost-1.87+ for try_emplace_and_cvisit
Since 2.32, nix now needs boost 1.87 or later to build,
due to using unordered::concurrent_flat_map try_emplace_and_cvisit

../src/libexpr/eval.cc: In member function ‘void nix::EvalState::evalFile(const nix::SourcePath&, nix::Value&, bool)’:
../src/libexpr/eval.cc:1096:20: error: ‘class boost::unordered::concurrent_flat_map<nix::SourcePath, nix::Value*, std::hash<nix::SourcePath>, std::equal_to<nix::SourcePath>, traceable_allocator<std::pair<const nix::SourcePath, nix::Value*> > >’ has no member named ‘try_emplace_and_cvisit’; did you mean ‘try_emplace_or_cvisit’?
 1096 |     fileEvalCache->try_emplace_and_cvisit(
      |                    ^~~~~~~~~~~~~~~~~~~~~~
      |                    try_emplace_or_cvisit

See 834580b539
2025-10-24 01:24:04 +08:00
Eelco Dolstra
0a74b4905c Merge pull request #14332 from NixOS/cleanup-ci
ci: Assorted collection of cleanups
2025-10-23 16:50:11 +00:00
Eelco Dolstra
d74177dccc Merge pull request #14328 from cachix/nar-substitutiongone
Fix misleading error messages for missing NARs due to stale cache
2025-10-23 16:48:31 +00:00
Sergei Zimmerman
36ee38efd1 Merge pull request #14338 from lovesegfault/s3-docs-listbucket
docs: add s3:ListBucket to S3 read permissions
2025-10-23 08:43:01 +00:00
Sergei Zimmerman
5d7912eb18 Merge pull request #14335 from lovesegfault/extract-getcompressionmethod
refactor(libstore): extract getCompressionMethod() in HttpBinaryCacheStore
2025-10-23 08:30:08 +00:00
Bernardo Meurer Costa
78888ec8a8 docs: add s3:ListBucket to S3 read permissions
The s3:ListBucket permission is required for read operations on S3
binary caches, not just for writes. Without this permission, users get
"Access Denied" errors when running nix-build.
2025-10-23 06:03:00 +00:00
Bernardo Meurer Costa
b047cecf5c refactor(libstore): extract getCompressionMethod() in HttpBinaryCacheStore
Extract the path-based compression method determination logic into a
protected method that returns std::optional<std::string>. This allows
subclasses to reuse the logic and makes the semantics clearer (nullopt
means no compression, not empty string).

This prepares for S3BinaryCacheStore to apply the same compression
rules when implementing multipart uploads.
2025-10-23 05:03:02 +00:00
John Ericson
d0217ec180 Merge pull request #14331 from NixOS/debug-build-fix
meson: Only enable b_lto for nixexpr-parser when b_lto is enabled glo…
2025-10-23 04:52:55 +00:00
Bernardo Meurer Costa
953929f899 fix(libstore): use CURLOPT_POSTFIELDSIZE_LARGE for POST requests
Fix POST requests with data to use the correct curl option for specifying
body size. Previously used CURLOPT_INFILESIZE_LARGE for both POST and PUT,
but POST requires CURLOPT_POSTFIELDSIZE_LARGE.

This caused POST request bodies to not be sent correctly, manifesting as
S3 multipart CompleteMultipartUpload requests failing with "You must
specify at least one part" even though the XML body contained valid parts.
2025-10-23 02:26:45 +00:00
Sergei Zimmerman
3c83856494 ci: Update pinned install_url 2.30.2 -> 2.32.1 2025-10-23 02:17:12 +03:00
Sergei Zimmerman
f3d8d1f719 ci: Reuse composite install-nix-action for docker_push_image job 2025-10-23 02:17:11 +03:00
Sergei Zimmerman
c8a15bf70d ci: Pin cachix action 2025-10-23 02:17:10 +03:00
Sergei Zimmerman
ad5c6a53b9 ci: Move magic-nix-cache-action into install-nix-action composite
This reduces duplication and pins the underlying version of magic-nix-cache,
as we already do with other actions.
2025-10-23 02:17:09 +03:00
Sergei Zimmerman
350d602832 meson: Only enable b_lto for nixexpr-parser when b_lto is enabled globally 2025-10-23 01:49:31 +03:00
John Ericson
115dea10b2 Merge pull request #14320 from roberth/open-manual-app
flake.nix: Add nix run .#open-manual
2025-10-22 21:37:21 +00:00
Eelco Dolstra
ddb8830c97 Merge pull request #14326 from adeci/githint
fetchers: add helpful hint for file+git URL scheme error
2025-10-22 20:39:16 +00:00
Domen Kožar
459f9e0185 Fix misleading error messages for missing NARs due to stale cache
When Nix's SQLite narinfo cache indicates a NAR exists, but the NAR
has been garbage collected from the binary cache, Nix displays error
messages even though the operation succeeds via fallback. This is
misleading because the cached narinfo is simply outdated.

This changes SubstituteGone exceptions to produce warnings instead of
errors, accurately reflecting that this is an expected cache coherency
issue, not an actual failure.

Fixes #11411

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-10-22 15:07:42 -05:00
Sergei Zimmerman
5390bba920 Merge pull request #14314 from Radvendii/parser-cpp-variant
libexpr: parser.y: use api.value.type variant
2025-10-22 18:49:14 +00:00
adeci
387eceff45 fetchers: Add helpful hint for file+git URL scheme error
At least one user has probably used `file+git://` when they mean `git+file://`, maybe thinking of it as "a file-based git repository". This adds a specific error message to hint at the correct URL scheme format and may save some users from resorting to `path:///` and copying an entire repo.
2025-10-22 13:57:51 -04:00
Sergei Zimmerman
96c8cc550f libexpr/meson: Rice the compiler inlining heuristics to improve perf of the bison generated parser
Turns out both GCC and Clang need a bit of hand-holding to optimize the bison generated
code well, otherwise parser performance tanks.

(Comparisons against baseline in 7e8db2eb59):

For GCC:

Benchmark 1 (15 runs): result/bin/nix-instantiate --parse ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
  measurement          mean ± σ            min … max           outliers         delta
  wall_time           335ms ± 2.89ms     332ms …  342ms          0 ( 0%)        0%

Benchmark 2 (16 runs): result-old/bin/nix-instantiate --parse ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
  measurement          mean ± σ            min … max           outliers         delta
  wall_time           330ms ± 2.87ms     326ms …  337ms          0 ( 0%)          -  1.4% ±  0.6%

For Clang:

Benchmark 1 (15 runs): result-clang/bin/nix-instantiate --parse ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
  measurement          mean ± σ            min … max           outliers         delta
  wall_time           340ms ± 1.43ms     338ms …  343ms          0 ( 0%)        0%

Benchmark 2 (15 runs): result-old-clang/bin/nix-instantiate --parse ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
  measurement          mean ± σ            min … max           outliers         delta
  wall_time           334ms ± 1.61ms     332ms …  338ms          0 ( 0%)        -  1.7% ±  0.3%
2025-10-22 02:25:11 +02:00
Taeer Bar-Yam
32b286e5d6 libexpr: parser.y: api.value.type variant 2025-10-22 02:25:11 +02:00
Robert Hensing
b558dac7a9 flake.nix: Add nix run .#open-manual
Great for reviewing the rendered manual
2025-10-22 00:42:18 +02:00
Sergei Zimmerman
7e8db2eb59 Merge pull request #14318 from cole-h/remove-useless-fmt
libstore: remove useless fmt
2025-10-21 17:50:56 +00:00
John Ericson
6ca2efc7d4 Merge pull request #14254 from roberth/upstream-RossComputerGuy/feat/expose-computefsclosure
libstore-c: add nix_store_get_fs_closure #14025 with tests and realise error fix
2025-10-21 17:41:29 +00:00
Cole Helbling
62247af363 libstore: remove useless fmt 2025-10-21 10:06:35 -07:00
John Ericson
a4a49a9dae Merge pull request #14316 from NixOS/fix-computeStorePath-arg
Fix computeStorePath() default argument
2025-10-21 14:48:56 +00:00
Eelco Dolstra
606c258c6f Fix computeStorePath() default argument 2025-10-21 15:58:44 +02:00
John Ericson
ef8218f2e3 Merge pull request #14307 from NixOS/json-schema-hash
`nlohmann::json` instance and JSON Schema for `Hash`
2025-10-21 06:03:20 +00:00
John Ericson
ada008a795 Merge pull request #14310 from obsidiansystems/inline-drv-output-subst-goal
Inline `realisationFetched`
2025-10-21 06:03:06 +00:00
John Ericson
2a2bb8330d Merge pull request #14312 from corngood/develop-structuredAttrs-fix
tests/functional/flakes/develop.sh: Add test for outputChecks stripping
2025-10-20 22:48:39 +00:00
David McFarland
645794b458 tests/functional/flakes/develop.sh: Add test for outputChecks stripping 2025-10-20 19:16:20 -03:00
John Ericson
1121f0d8ec Inline realisationFetched
Now that we are using coroutines, we don't need this to be a separate
method of `DrvOutputSubstitutionGoal`.
2025-10-20 16:45:41 -04:00
Sergei Zimmerman
6420879728 Merge pull request #14296 from lovesegfault/nix-s3-more-tests
fix(nix-prefetch-url): correctly extract filename from URLs with query parameters
2025-10-20 19:42:22 +00:00
Sergei Zimmerman
67f5cb97a3 Merge pull request #14306 from corngood/develop-structuredAttrs-fix
nix/develop: Strip outputChecks when structuredAttrs is enabled
2025-10-20 19:38:19 +00:00
John Ericson
5e7ee808de nlohmann::json instance and JSON Schema for Hash
Improving and codifying our experimental JSON interfacing.

Co-Authored-By: Robert Hensing <robert@roberthensing.nl>
2025-10-20 15:21:07 -04:00
John Ericson
270f20a505 Merge pull request #14305 from NixOS/alignment-utils
libutil: Add alignUp helper function, use in archive.cc
2025-10-20 19:08:20 +00:00
Bernardo Meurer Costa
1b1d7e3047 test(nixos): add nix-prefetch-url test for S3 URLs with query parameters
Adds a comprehensive test to verify that `nix-prefetch-url` correctly
handles S3 URLs with query parameters (e.g., custom endpoints and regions).

Previously, nix-prefetch-url would fail with "invalid store
path" errors when given S3 URLs with query parameters like
`?endpoint=http://server:9000&region=eu-west-1`, because it incorrectly
extracted the filename from the query parameters instead of the path.
2025-10-20 21:45:37 +03:00
David McFarland
0f28c76a44 nix/develop: Strip outputChecks when structuredAttrs is enabled 2025-10-20 15:40:05 -03:00
Bernardo Meurer Costa
e3b3f05e5d fix(nix-prefetch-url): correctly extract filename from URLs with query parameters
Previously, `prefetchFile()` used `baseNameOf()` directly on the URL string
to extract the filename. This caused issues with URLs containing query
parameters that include slashes, such as S3 URLs with custom endpoints:

```
s3://bucket/file.txt?endpoint=http://server:9000
```

The `baseNameOf()` function naively searches for the rightmost `/` in the
entire string, which would find the `/` in `http://server:9000` and extract
`server:9000&region=...` as the filename. This resulted in invalid store
path names containing illegal characters like `:`.

This commit fixes the issue by:

1. Adding a `VerbatimURL::lastPathSegment()` method that extracts the last
   non-empty path segment from a URL, using `pathSegments(true)` to filter
   empty segments
2. Changing `prefetchFile()` to accept `const VerbatimURL &` and use the new
   `lastPathSegment()` method instead of manual path parsing
3. Adding early validation with `checkName()` to fail quickly on invalid
   filenames
4. Maintains backward compatibility by falling back to `baseNameOf()` for
   unparsable `VerbatimURL`s
2025-10-20 21:40:03 +03:00
John Ericson
f05d240222 Merge pull request #14278 from obsidiansystems/adl-serializer-xp
Cleanup and JSON serializer and XP feature interations
2025-10-20 18:22:21 +00:00
Sergei Zimmerman
22c73868c3 libutil/archive: Use alignUp
With this change it's much more apparent what's going on.
2025-10-20 21:15:11 +03:00
Sergei Zimmerman
a91b787524 libutil: Add alignUp helper function 2025-10-20 21:11:00 +03:00
Eelco Dolstra
ddf7de0a76 Merge pull request #14291 from NixOS/skip-source
Add skip() method to Source interface to allow efficient seeks
2025-10-20 15:04:36 +00:00
Sergei Zimmerman
1fabed18b6 Merge pull request #14301 from NixOS/s3-terminate-unknown-profile
libstore: Fix reentrancy in AwsCredentialProviderImpl::getCredentialsRaw
2025-10-20 14:28:16 +00:00
Eelco Dolstra
6c9083db2c Use a smaller buffer 2025-10-20 13:40:19 +02:00
Sergei Zimmerman
c663f7ec79 libstore: Fix reentrancy in AwsCredentialProviderImpl::getCredentialsRaw
Old code would do very much incorrect reentrancy crimes (trying to do an
erase inside the emplace callback). This would fail miserably with an assertion
in Boost:

terminating due to unexpected unrecoverable internal error: Assertion '(!find(px))&&("reentrancy not allowed")' failed in boost::unordered::detail::foa::entry_trace::entry_trace(const void *) at include/boost/unordered/detail/foa/reentrancy_check.hpp:33

This is trivially reproduced by using any S3 URL with a non-empty profile:

nix-prefetch-url "s3://happy/crash?profile=default"
2025-10-19 21:03:13 +03:00
Sergei Zimmerman
d0fb03c35d Merge pull request #14282 from NixOS/s3-cleanup
Simplify meson for S3 support via aws-crt-cpp
2025-10-19 17:00:46 +00:00
Sergei Zimmerman
c847cd87f1 Merge pull request #14297 from lovesegfault/nix-s3-test-public
test(nixos/s3-binary-cache-store): misc improvements
2025-10-19 16:53:40 +00:00
tomberek
dbbdae926b Merge pull request #14299 from roberth/unlocked-msg
Clarify unlocked input warning message
2025-10-19 16:50:16 +00:00
Eelco Dolstra
3c03050cd6 Merge pull request #14290 from NixOS/dont-write-nar-to-tty
nix store dump-path: Refuse to write NARs to the terminal
2025-10-19 12:41:55 +00:00
Robert Hensing
e33cd5aa38 Clarify unlocked input warning message
The previous message was vague about what "deprecated" meant and why
unlocked inputs with NAR hashes "may not be reproducible". It also
used "verifiable" which was confusing.

The new message makes it clear that the NAR hash provides verification
(is checked by NAR hash) and explicitly states the failure modes:
garbage collection and sharing.
2025-10-19 14:08:34 +02:00
Bernardo Meurer Costa
d9c808f8a7 refactor(tests/nixos/s3-binary-cache-store): add verify_packages_in_store helper 2025-10-19 00:21:54 +00:00
Bernardo Meurer Costa
55ea3d3476 test(tests/nixos/s3-binary-cache-store): test public bucket operations
Add `test_public_bucket_operations` to validate that store operations
work correctly on public S3 buckets without requiring credentials.
Tests nix store info and nix copy operations.
2025-10-19 00:04:33 +00:00
Bernardo Meurer Costa
7d0c06f921 feat(tests/nixos/s3-binary-cache-store): add public parameter to setup_s3
Add optional 'public' parameter to setup_s3 decorator. When set to True,
the bucket will be made publicly accessible using mc anonymous set.
2025-10-18 23:57:51 +00:00
Bernardo Meurer Costa
5b4bd5bcb8 refactor(tests/nixos/s3-binary-cache-store): inline make_http_url fn
Remove make_http_url helper function and inline its single usage.
2025-10-18 23:51:44 +00:00
Bernardo Meurer Costa
4ae6c65bc5 test(tests/nixos/s3-binary-cache-store): verify credential caching in concurrent fetches
Add assertion to test_concurrent_fetches to verify that only one
credential provider is created even with 5 concurrent fetches.
2025-10-18 23:48:55 +00:00
Bernardo Meurer Costa
4f19e63a8f refactor(tests/nixos/s3-binary-cache-store): add --no-link to nix build commands
Prevent creation of result symlinks in all nix build commands by
adding the --no-link flag.
2025-10-18 23:44:13 +00:00
Bernardo Meurer Costa
f88c3055f8 refactor(tests/nixos/s3-binary-cache-store): clean client store in setup_s3
Add cleanup of client store in the finally block of setup_s3 decorator.
Uses `nix store delete --ignore-liveness` to properly handle GC roots
and only attempts deletion if the path exists.
2025-10-18 23:36:48 +00:00
Bernardo Meurer Costa
9058d90ab2 refactor(tests/nixos/s3-binary-cache-store): rename populate_with to populate_bucket 2025-10-18 23:27:03 +00:00
Bernardo Meurer Costa
c1a15d1a26 refactor(tests/nixos/s3-binary-cache-store): rename with_test_bucket to setup_s3 2025-10-18 23:24:30 +00:00
Bernardo Meurer Costa
22f4cccc71 refactor(tests/nixos/s3-binary-cache-store): use a PKGS dict
Replace individual PKG_A, PKG_B, and PKG_C variables with a PKGS
dictionary. This will enable `@with_clean_client_store` in the future.
2025-10-18 23:23:50 +00:00
John Ericson
b56e456b0d Merge pull request #14269 from roberth/json-schema
Add a JSON Schema for `Derivation`
2025-10-18 18:50:39 +00:00
Sergei Zimmerman
3d147c04a5 libstore: Implement getHumanReadableURI for S3BinaryCacheStoreConfig
This slightly improves the logs situation by including the region/profile/endpoint
in the logs when S3 store references get printed. Instead of:

copying path '/nix/store/lxnp9cs4cfh2g9r2bs4z7gwwz9kdj2r9-test-package-c' to 's3://bucketname'...

This now includes:

copying path '/nix/store/lxnp9cs4cfh2g9r2bs4z7gwwz9kdj2r9-test-package-c' to 's3://bucketname?endpoint=http://server:9000&region=eu-west-1'...
2025-10-18 19:11:39 +03:00
Sergei Zimmerman
61fbef42a6 libstore: Simplify check for S3-specific URI query parameters
Instead of hardcoding strings we should instead use the setting
objects to determine the query names that should be preserved.
2025-10-18 18:47:27 +03:00
Robert Hensing
c92ba4b9b7 Add titles in JSON schemas
This way, the description isn't rendered in the tables of contents,
leading to no more formatting errors.
2025-10-17 21:53:29 +02:00
Eelco Dolstra
67bffa19a5 NullFileSystemObjectSink: Skip over file contents 2025-10-17 20:44:02 +02:00
Eelco Dolstra
daa7e0d2e9 Source: Add skip() method
This allows FdSource to efficiently skip data we don't care about.
2025-10-17 20:41:33 +02:00
Eelco Dolstra
109f6449cc nix store dump-path: Refuse to write NARs to the terminal 2025-10-17 20:27:10 +02:00
John Ericson
ad2360c59f Merge pull request #14288 from lovesegfault/repl-skip-stack
fix(tests/functional/repl): skip test if stack size limit is insufficient
2025-10-17 17:35:52 +00:00
Bernardo Meurer Costa
20c7c551bf fix(tests/functional/repl): skip test if stack size limit is insufficient
Nix attempts to set the stack size to 64 MB during initialization, which is
required for the repl tests to run successfully. Skip the tests on systems
where the hard stack limit is less than this value rather than failing.
2025-10-17 17:05:12 +00:00
John Ericson
e78e6ca4f4 Merge pull request #14281 from NixOS/dead-code
libutil: Drop unused SubdirSourceAccessor
2025-10-17 03:01:17 +00:00
John Ericson
e34063cf21 Merge pull request #14283 from NixOS/nar-check
nix {cat,ls}: Add back missing checks for file descriptors
2025-10-17 02:58:23 +00:00
Sergei Zimmerman
e457ea7688 nix {cat,ls}: Add back missing checks for file descriptors
I didn't catch this during the review of https://github.com/NixOS/nix/pull/14273.
This fixes that mistake.
2025-10-17 02:26:24 +03:00
Farid Zakaria
64c55961eb Merge pull request #14273 from fzakaria/fzakaria/issue-13944
Make `nix nar [cat|ls]` lazy
2025-10-17 02:16:54 +03:00
Sergei Zimmerman
ffbc33fec6 libstore/meson: Rename curl-s3-store to s3-aws-auth
We now unconditionally compile support for s3:// URLs and stores
without authentication. The whole curl version check can be greatly
simplified by the previous commit, which bumps the minimum required curl
version.
2025-10-17 01:18:46 +03:00
Sergei Zimmerman
a80fc252e8 libstore/meson: Require curl >= 7.75.0
This version has been released a long time ago in 2021 and it's doubtful
that anybody actually uses it still, since it's full of vulnerabilities [^]

[^]: https://curl.se/docs/vuln-7.75.0.html
2025-10-17 01:18:14 +03:00
Sergei Zimmerman
bcd5a9d05c libutil: Drop unused SubdirSourceAccessor 2025-10-17 00:56:53 +03:00
Robert Hensing
01b001d5ba Add JSON Schema infrastructure, use for Derivation
For manual, and testing formats
2025-10-16 17:24:18 -04:00
John Ericson
27767a6094 Merge pull request #14276 from NixOS/fix-14193
libstore/registerOutputs: Don't try to optimize a non-existent actual…
2025-10-16 21:06:43 +00:00
John Ericson
1177d65094 Properly check xp features when deserializing deriving paths 2025-10-16 16:45:22 -04:00
John Ericson
a2c6f38e1f Remove now-redundant methods for JSON on Derivation 2025-10-16 16:45:22 -04:00
John Ericson
1c02dd5b9c Allow for standard nlohmann JSON serializers to take separate XP features
I realized that we can actually do this thing, even though it is not
what nlohmann expects at all, because the extra parameter has a default
argument so nlohmann doesn't need to care. Sneaky!
2025-10-16 16:45:22 -04:00
Sergei Zimmerman
4cbcaad435 libstore/registerOutputs: Don't try to optimize a non-existent actualPath
Since 3c610df550 this resulted in `getting status of`
errors on paths inside the chroot if a path was already valid. Careful inspection
of the logic shows that if buildMode != bmCheck actualPath gets reassigned to
store.toRealPath(finalDestPath). The only branch that cares about actualPath is
the buildMode == bmCheck case, which doesn't lead to optimisePath anyway.
2025-10-16 23:08:30 +03:00
John Ericson
d87a06af7a Merge pull request #14275 from NixOS/s3-cleanup
libstore: Miscellaneous S3 store cleanups
2025-10-16 19:36:59 +00:00
Eelco Dolstra
2dc9f2a2b7 Merge pull request #14272 from NixOS/use-store-path-serializer
Daemon protocol: Use the WorkerProto serializer for store paths
2025-10-16 19:35:25 +00:00
Eelco Dolstra
a7991d55cc Merge pull request #14270 from NixOS/use-optional-storepath-serializer
Use serializer for std::optional<StorePath>
2025-10-16 19:07:07 +00:00
Sergei Zimmerman
e7047fde25 libstore: Remove the unnecessary 'error: ' prefix in warning message 2025-10-16 21:49:38 +03:00
Sergei Zimmerman
33e94fe19f libstore: Make AwsAuthError more legible
Instead of the cryptic:

> error: Failed to resolve AWS credentials: error code 6153`

We now get more legible:

> error: AWS authentication error: 'Valid credentials could not be sourced by the IMDS provider' (6153)
2025-10-16 21:49:37 +03:00
Sergei Zimmerman
dc03c6a812 libstore: Put all the AWS credentials logic behind interface class AwsCredentialProvider
This makes it so we don't need to rely on global variables and hacky destructors to
clean up another global variable. Just putting it in the correct order in the class
is more than enough.
2025-10-16 21:49:36 +03:00
Sergei Zimmerman
b1d067c9bb tests/nixos: Rename back S3 store nixos test 2025-10-16 21:49:35 +03:00
Eelco Dolstra
d782c5e586 Daemon protocol: Use the WorkerProto serializer for store paths 2025-10-16 17:34:33 +02:00
Eelco Dolstra
f84b33644c Merge pull request #14271 from NixOS/no-check-sigs
Factor out `--no-check-sigs` into its own class
2025-10-16 15:07:29 +00:00
Eelco Dolstra
3bd2b76f6e nix store sign: Use required attribute 2025-10-16 16:35:13 +02:00
Eelco Dolstra
139df77440 Factor out --no-check-sigs 2025-10-16 16:35:09 +02:00
Eelco Dolstra
a48a737517 Use serializer for std::optional<StorePath> 2025-10-16 16:32:18 +02:00
John Ericson
0503a862ef Merge pull request #14268 from roberth/dev-doc-manual
doc/dev/doc: Update local build instructions for manual
2025-10-16 13:50:58 +00:00
Robert Hensing
61cb9c4832 doc/dev/doc: Update local build instructions for manual 2025-10-16 13:22:22 +02:00
John Ericson
721f5572a6 Merge pull request #14263 from NixOS/hydra-import-paths
Restore `ServeProto::Command::ImportPaths`
2025-10-15 22:57:50 +00:00
John Ericson
5a6864c027 Merge pull request #14264 from xokdvium/fix-splicing-test
tests: Fix splicing in functional tests for nix-cli
2025-10-15 22:32:45 +00:00
Sergei Zimmerman
0deb492b3d Restore ServeProto::Command::ImportPaths
This partially reverts commit 5e46df973f,
partially reversing changes made to
8c789db05b.

We do this because Hydra, while using the newer version of the protocol,
still uses this command, even though Nix (as a client) doesn't use it.
On that basis, we don't want to remove it (or consider it only part of
the older versions of the protocol) until Hydra no longer uses the
Legacy SSH Protocol.
2025-10-15 18:18:59 -04:00
Sergei Zimmerman
17b7fb383f tests: Fix splicing in functional tests for nix-cli
This is necessary to fix nix-everything-llvm.
The problem here is that nix-cli is taken from the previous
stage that is built with libstdc++, but this derivation builds
plugins with libc++ and the plugin load fails miserably.
2025-10-16 01:04:50 +03:00
John Ericson
94cfba7e84 Merge pull request #14226 from obsidiansystems/unkeyed-realisation
Reapply #14097
2025-10-15 21:27:13 +00:00
Sergei Zimmerman
0f1cfa4d60 Merge pull request #14262 from lovesegfault/ci-cleanup-s3
ci: cleanup s3 tests
2025-10-15 21:19:20 +00:00
Bernardo Meurer Costa
fa0d00e668 ci: cleanup s3 tests
This cleans up the work done in 8c2828387. Now that #13752 has landed,
there's no need to test configurations without AWS auth in CI.
2025-10-15 23:51:08 +03:00
Robert Hensing
6036aaf798 C API: Check output callback order 2025-10-15 22:04:21 +02:00
Sergei Zimmerman
d2b6499154 Merge pull request #13752 from lovesegfault/curl-based-s3-v2
feat(libstore): curl-based s3
2025-10-15 20:01:26 +00:00
Sergei Zimmerman
e3232af558 Merge pull request #14253 from NixOS/libgit2-refname-wa
libfetchers/git-utils: Be more correct about validating refnames
2025-10-15 19:30:53 +00:00
Bernardo Meurer Costa
e069c9892e docs(rl-next): add notes for curl-based s3 2025-10-15 19:26:53 +00:00
John Ericson
266fbebe66 Implement realisation operations on dummy store 2025-10-15 14:59:08 -04:00
John Ericson
6995d325ef Split out UnkeyedRealisation from Realisation
Realisations are conceptually key-value pairs, mapping `DrvOutputs` (the
key) to information about that derivation output.

This separate the value type, which will be useful in maps, etc., where
we don't want to denormalize by including the key twice.

This matches similar changes for existing types:

| keyed              | unkeyed                |
|--------------------|------------------------|
| `ValidPathInfo`    | `UnkeyedValidPathInfo` |
| `KeyedBuildResult` | `BuildResult`          |
| `Realisation`      | `UnkeyedRealisation`   |

Co-authored-by: Sergei Zimmerman <sergei@zimmerman.foo>
2025-10-15 14:59:04 -04:00
Sergei Zimmerman
5d1178b817 libfetchers/git-utils: Be more correct about validating refnames
Turns out there's a much better API for this that doesn't have the
footguns of the previous method.

isLegalRefName is somewhat of a misnomer, since it's mainly used to
validate user inputs that can be either references, branch names,
psedorefs or tags.
2025-10-15 21:54:09 +03:00
Bernardo Meurer Costa
3224636ab0 refactor(libstore): rename NIX_WITH_S3_SUPPORT to NIX_WITH_AWS_AUTH
The macro now accurately reflects its purpose: gating only AWS
authentication code, not all S3 functionality. S3 URL parsing, store
configuration, and public bucket access work regardless of this flag.

This rename clarifies that:
- S3 support is always available (URL parsing, store registration)
- Only AWS credential resolution requires the flag
- The flag controls AWS CRT SDK dependency, not S3 protocol support
2025-10-15 18:23:56 +00:00
Bernardo Meurer Costa
bb1f22a8df refactor(libstore): minimize NIX_WITH_S3_SUPPORT scope to auth only
Move S3 URL parsing, store configuration, and public bucket support
outside of NIX_WITH_S3_SUPPORT guards. Only AWS credential resolution
remains gated, allowing builds with withAWS = false to:

- Parse s3:// URLs
- Register S3 store types
- Access public S3 buckets (via HTTPS conversion)
- Use S3-compatible services without authentication

The setupForS3() function now always performs URL conversion, with
authentication code conditionally compiled based on NIX_WITH_S3_SUPPORT.
The aws-creds.cc file (only code using AWS CRT SDK) is now conditionally
compiled by meson.
2025-10-15 18:23:56 +00:00
Bernardo Meurer Costa
1f710300c9 refactor(libstore): withCurlS3 -> withAWS
Now that the legacy S3 implementation is gone, we can go back to calling
things `NIX_WITH_S3_SUPPORT`.
2025-10-15 18:23:56 +00:00
Bernardo Meurer Costa
9295c14a35 refactor(libstore): replace AWS SDK with curl-based S3 implementation
This commit replaces the AWS C++ SDK with a lighter curl-based approach
for S3 binary cache operations.

- Removed dependency on the heavy aws-cpp-sdk-s3 and aws-cpp-sdk-transfer
- Added lightweight aws-crt-cpp for credential resolution only
- Leverages curl's native AWS SigV4 authentication (requires curl >= 7.75.0)
- S3BinaryCacheStore now delegates to HttpBinaryCacheStore
- Function s3ToHttpsUrl converts ParsedS3URL to ParsedURL
- Multipart uploads are no longer supported (may be reimplemented later)
- Build now requires curl >= 7.75.0 for AWS SigV4 support

Fixes: #13084, #12671, #11748, #12403, #5947
2025-10-15 18:23:55 +00:00
John Ericson
a543519ca9 Merge pull request #14257 from obsidiansystems/misc-builder-tech-debt
Cleanup misc builder tech debt
2025-10-15 17:28:27 +00:00
John Ericson
632ccfb8c0 Remove dead outputPaths variable. 2025-10-15 12:16:53 -04:00
John Ericson
46357468a4 Remove unused parameters to DrvOutputSubstitutionGoal 2025-10-15 12:16:52 -04:00
John Ericson
b20cebf993 Remove unused typedef and field 2025-10-15 12:15:23 -04:00
Eelco Dolstra
5700112127 Merge pull request #14205 from GrahamDennis/gdennis/improve-dir-url-backcompat
Improved backwards compatibility hack for git URLs using dir=...
2025-10-15 15:20:02 +00:00
Robert Hensing
a9d9b50b72 Merge remote-tracking branch 'upstream/master' into upstream-RossComputerGuy/feat/expose-computefsclosure 2025-10-15 15:40:10 +02:00
Robert Hensing
6fa03765ed C API: Propagate nix_store_realise build errors 2025-10-15 15:20:24 +02:00
Robert Hensing
12293a8b11 C API: Document nix_store_copy_closure flags 2025-10-15 15:05:50 +02:00
Robert Hensing
3fb943d130 C API: Make store realise tests multi-platform
... and improve assertions.
2025-10-15 14:55:28 +02:00
Robert Hensing
aace1fb5d6 C API: test nix_store_get_fs_closure 2025-10-15 13:27:09 +02:00
John Ericson
606eb1dfb5 Merge pull request #14250 from fzakaria/patch-1
Remove duplicate shellcheck in dev-shell.nix
2025-10-15 05:03:19 +00:00
John Ericson
e07754d888 Merge pull request #14251 from fzakaria/fzakaria/iwyu-libflake
Clean-up libflake headers
2025-10-15 04:27:07 +00:00
Farid Zakaria
01a8499d2f Format cpp files 2025-10-14 23:51:40 -04:00
Farid Zakaria
e8b126fa90 Remove unecessary includes 2025-10-14 23:48:45 -04:00
Farid Zakaria
902faf4fe5 More fixes for iwyu 2025-10-14 23:20:35 -04:00
Farid Zakaria
7bc3d9b9a9 First attempt at uwyu for libflake 2025-10-14 22:53:13 -04:00
Farid Zakaria
092639709f Remove duplicate shellcheck in dev-shell.nix 2025-10-14 19:25:06 -07:00
John Ericson
620091bc8b Merge pull request #14223 from lovesegfault/curl-based-s3-tests
test(nixos): add comprehensive curl-based S3 VM tests
2025-10-14 23:08:55 +00:00
John Ericson
6dcc468253 Merge pull request #14249 from NixOS/more-to-real-path-cleanups
More toRealPath cleanups
2025-10-14 22:46:15 +00:00
Sergei Zimmerman
0347958dd2 nix/develop: Remove usage of toRealPath, replace with SourceAccessor 2025-10-15 00:52:13 +03:00
Sergei Zimmerman
918a3cebaa libexpr: Use Store::requireStoreObjectAccessor instead or toRealPath in fetch
This forces the code to go through proper abstractions instead of the raw filesystem
API.

This issue is evident from this reproducer:

nix eval --expr 'builtins.fetchurl { url = "https://example.com"; sha256 = ""; }' --json --eval-store "dummy://?read-only=false"

error:
       … while calling the 'fetchurl' builtin
         at «string»:1:1:
            1| builtins.fetchurl { url = "https://example.com"; sha256 = ""; }
             | ^

       error: opening file '/nix/store/r4f87yrl98f2m6v9z8ai2rbg4qwlcakq-example.com': No such file or directory
2025-10-15 00:27:41 +03:00
Sergei Zimmerman
69c005e805 libstore: Use getFSAccessor for store object in Worker::pathContentsGood
We only care about the accessor for a single store object anyway, but
the validity gets ignored. Also `pathExists(store.printStorePath(path))`
is definitely incorrect since it confuses the logical location vs physical
location in case of a chroot store.
2025-10-15 00:15:50 +03:00
Sergei Zimmerman
0c32fb3fa2 treewide: Add Store::requireStoreObjectAccessor, simplify uses of getFSAccessor
This is a simple wrapper around getFSAccessor that throws an InvalidPath
error. This simplifies usage in callsites that only care about getting
a non-null accessor.
2025-10-14 23:58:20 +03:00
Bernardo Meurer Costa
d18f959d4f test(nixos): add comprehensive curl-based S3 VM tests
Add `curl-s3-binary-cache-store.nix` with comprehensive test coverage
for the curl-based S3 implementation.

Depends-On: #14206, #14222
2025-10-14 20:55:14 +00:00
Sergei Zimmerman
4041bfdb40 Merge pull request #14206 from lovesegfault/curl-based-s3-pieces
feat(libstore): add builtin fetchurl S3 credential pre-resolution
2025-10-14 20:10:41 +00:00
John Ericson
1fb4ff8c0e Merge pull request #14232 from roberth/dyndrv-messages
Better dyndrv messages
2025-10-14 15:40:27 +00:00
Robert Hensing
1b96a704d3 Add lazy evaluation for experimental feature reasons
Wrap fmt() calls in lambdas to defer string formatting until the
feature check fails. This avoids unnecessary string formatting in
the common case where the feature is enabled.

Addresses performance concern raised by xokdvium in PR review.
2025-10-14 16:49:59 +02:00
John Ericson
959c244a12 Merge pull request #14243 from NixOS/canon-path-nul-bytes
libutil: Ensure that CanonPath does not contain NUL bytes
2025-10-14 14:30:24 +00:00
Eelco Dolstra
c44d2d5913 Merge pull request #14241 from NixOS/dependabot/github_actions/actions/create-github-app-token-2
build(deps): bump actions/create-github-app-token from 1 to 2
2025-10-14 11:55:43 +00:00
Eelco Dolstra
dd590eca74 Merge pull request #14242 from NixOS/dependabot/github_actions/actions/checkout-5
build(deps): bump actions/checkout from 4 to 5
2025-10-14 11:55:25 +00:00
Sergei Zimmerman
1633ceaff2 libutil: Ensure that CanonPath does not contain NUL bytes
This, alongside the other invariants of the CanonPath is important
to uphold. std::filesystem happily crashes on NUL bytes in the constructor,
as we've seen with `path:%00` prior to c436b7a32a.
Best to stay clear of NUL bytes when we're talking about syscalls, especially
on Unix where strings are null terminated.

Very nice to have if we decide to switch over to pascal-style strings.
2025-10-14 02:33:42 +03:00
John Ericson
16e946bfb1 Merge pull request #14225 from obsidiansystems/derivation-resolution-goal-2
Reapply the rest of #14022
2025-10-13 23:26:29 +00:00
Sergei Zimmerman
edf9163c22 libutil: Make CanonPath::root const
By all means CanonPath::root must be immutable. Let's enforce this with
in the code.
2025-10-14 02:24:40 +03:00
John Ericson
ad893acf46 Fix ca/eval-store.sh test
The refactor in the last commit fixed the bug it was supposed to fix,
but introduced a new bug in that sometimes we tried to write a resolved
derivation to a store before all its `inputSrcs` were in that store.

The solution is to defer writing the derivation until inside
`DerivationBuildingGoal`, just before we do an actual build. At this
point, we are sure that all inputs in are the store.

This does have the side effect of meaning we don't write down the
resolved derivation in the substituting case, only the building case,
but I think that is actually fine. The store that actually does the
building should make a record of what it built by storing the resolved
derivation. Other stores that just substitute from that store don't
necessary want that derivation however. They can trust the substituter
to keep the record around, or baring that, they can attempt to re
resolve everything, if they need to be audited.

(cherry picked from commit c97b050a6c)
2025-10-13 18:41:59 -04:00
John Ericson
06bb1c2f93 Remove some buildMode default parameters
Force the internals to be more explicit.
2025-10-13 18:40:10 -04:00
John Ericson
2ee41976c2 Fix #13247
Resolve the derivation before creating a building goal, in a context
where we know what output(s) we want. That way we have a chance just to
download the outputs we want.

Fix #13247

(cherry picked from commit 39f6fd9b46)
2025-10-13 18:37:14 -04:00
dependabot[bot]
b846f27682 build(deps): bump actions/checkout from 4 to 5
Bumps [actions/checkout](https://github.com/actions/checkout) from 4 to 5.
- [Release notes](https://github.com/actions/checkout/releases)
- [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md)
- [Commits](https://github.com/actions/checkout/compare/v4...v5)

---
updated-dependencies:
- dependency-name: actions/checkout
  dependency-version: '5'
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-13 22:00:59 +00:00
dependabot[bot]
962862e9e0 build(deps): bump actions/create-github-app-token from 1 to 2
Bumps [actions/create-github-app-token](https://github.com/actions/create-github-app-token) from 1 to 2.
- [Release notes](https://github.com/actions/create-github-app-token/releases)
- [Commits](https://github.com/actions/create-github-app-token/compare/v1...v2)

---
updated-dependencies:
- dependency-name: actions/create-github-app-token
  dependency-version: '2'
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
2025-10-13 22:00:55 +00:00
Robert Hensing
39c4665488 Store reason as a field in MissingExperimentalFeature
Store the reason string as a field in the exception class rather than
only embedding it in the error message. This supports better structured
error handling and future JSON error reporting.

Suggested by Ericson2314 in PR review.
2025-10-13 23:49:20 +02:00
Robert Hensing
71aa9a4798 Add reasons to dyndrv xp messages 2025-10-13 23:49:20 +02:00
Robert Hensing
0fd890a8d6 Add reason string support to MissingExperimentalFeature 2025-10-13 23:49:20 +02:00
John Ericson
6642ffb506 Merge pull request #14239 from NixOS/asan-stack-overflow
libstore/outputs-spec: Drop usage of std::regex
2025-10-13 21:44:49 +00:00
Sergei Zimmerman
3ba221025f libstore/outputs-spec: Drop usage of std::regex
std::regex is a really bad tool for parsing things, since
it tends to overflow the stack pretty badly. See the build failure
under ASan in [^].

[^]: https://hydra.nixos.org/build/310077167/nixlog/5
2025-10-13 23:58:20 +03:00
Eelco Dolstra
b56cc1808d Merge pull request #14237 from NixOS/url-parser-regression
Remove validation of URLs passed to FileTransferRequest verbatim
2025-10-13 20:01:01 +00:00
Sergei Zimmerman
47f427a172 Remove validation of URLs passed to FileTransferRequest verbatim
CURL is not very strict about validation of URLs passed to it. We
should reflect this in our handling of URLs that we get from the user
in <nix/fetchurl.nix> or builtins.fetchurl. ValidURL was an attempt to
rectify this, but it turned out to be too strict. The only good way to
resolve this is to pass (in some cases) the user-provided string verbatim
to CURL. Other usages in libfetchers still benefit from using structured
ParsedURL and validation though.

nix store prefetch-file --name foo 'https://cdn.skypack.dev/big.js@^5.2.2'
error: 'https://cdn.skypack.dev/big.js@^5.2.2' is not a valid URL: leftover
2025-10-13 22:23:26 +03:00
John Ericson
0f85ef3677 Merge pull request #14219 from lovesegfault/eval-copy-less
libstore: Avoid copying derivations to the store if they are already valid
2025-10-13 16:36:40 +00:00
Eelco Dolstra
be2c9ef44c Merge pull request #14229 from NixOS/reduce-hydra-load
packaging/hydra: buildNoGC is the same as buildWithSanitizers
2025-10-13 16:22:30 +00:00
John Ericson
d2c0c0607c Merge branch 'master' into eval-copy-less 2025-10-13 11:52:42 -04:00
John Ericson
480ce19011 Merge pull request #14233 from neuralsorcerer/fix-typo
Fix typos
2025-10-13 15:30:05 +00:00
John Ericson
3f876bcb61 Merge pull request #14231 from roberth/code-docs
Code docs
2025-10-13 15:05:30 +00:00
Soumyadip Sarkar
998f93f267 Fix typos 2025-10-13 18:15:52 +05:30
Robert Hensing
583f5e37fc Refactor: use optionalBracket in nix search 2025-10-13 14:02:59 +02:00
Robert Hensing
5dcfddf997 strings: Add optionalBracket helper 2025-10-13 13:59:39 +02:00
Robert Hensing
48a5e2dde2 EvalState: add doc comment 2025-10-13 13:14:05 +02:00
Robert Hensing
6db86389ce util/error: Document addTrace params
... and rename e -> pos. That was weird.
2025-10-13 12:57:22 +02:00
Bernardo Meurer Costa
000e6f6282 feat(libstore): add builtin fetchurl S3 credential pre-resolution
Add support for pre-resolving AWS credentials in the parent process
before forking for builtin:fetchurl. This avoids recreating credential
providers in the forked child process.
2025-10-12 23:01:13 +00:00
Bernardo Meurer Costa
18ec3d1094 libstore: Avoid copying derivations to the store if they are already valid
This avoids the quite costly copying of derivations to the daemon over the
wire in case it already exists in the eval store.

For a fresh instantiatation (after running nix-collect-garbage) this doesn't
significantly slow down eval:

taskset -c 2,3 hyperfine --reference "result-old/bin/nix eval -f ../nixpkgs hello --store unix:///tmp/nix_socket" --prepare "nix-collect-garbage --store /tmp/store1111 --no-keep-derivations" "result/bin/nix eval -f ../nixpkgs hello --store unix:///tmp/nix_socket"
Benchmark 1: result-old/bin/nix eval -f ../nixpkgs hello --store unix:///tmp/nix_socket
  Time (mean ± σ):     388.7 ms ±  10.5 ms    [User: 157.0 ms, System: 61.3 ms]
  Range (min … max):   379.4 ms … 415.9 ms    10 runs

Benchmark 2: result/bin/nix eval -f ../nixpkgs hello --store unix:///tmp/nix_socket
  Time (mean ± σ):     389.2 ms ±   4.8 ms    [User: 158.5 ms, System: 60.7 ms]
  Range (min … max):   381.2 ms … 397.6 ms    10 runs

But if the derivations are already instantiated this shows a pretty neat speedup:

taskset -c 2,3 hyperfine --reference "result-old/bin/nix eval -f ../nixpkgs hello --store unix:///tmp/nix_socket" "result/bin/nix eval -f ../nixpkgs hello --store unix:///tmp/nix_socket"
Benchmark 1: result-old/bin/nix eval -f ../nixpkgs hello --store unix:///tmp/nix_socket
  Time (mean ± σ):     240.4 ms ±   3.1 ms    [User: 148.1 ms, System: 57.0 ms]
  Range (min … max):   233.8 ms … 245.0 ms    12 runs

Benchmark 2: result/bin/nix eval -f ../nixpkgs hello --store unix:///tmp/nix_socket
  Time (mean ± σ):     226.5 ms ±   4.5 ms    [User: 147.8 ms, System: 55.2 ms]
  Range (min … max):   214.0 ms … 231.2 ms    13 runs

Co-authored-by: Sergei Zimmerman <sergei@zimmerman.foo>
2025-10-13 01:59:38 +03:00
John Ericson
f77094715f Merge pull request #14222 from lovesegfault/curl-based-s3-fix-race
fix(libstore): fix race condition in AWS credential provider caching
2025-10-12 22:07:10 +00:00
Bernardo Meurer Costa
f0e1f65260 fix(libstore): fix race condition in AWS credential provider caching
The previous implementation had a check-then-create race condition where
multiple threads could simultaneously:
1. Check the cache and find no provider (line 122)
2. Create their own providers (lines 126-145)
3. Insert into cache (line 161)

This resulted in multiple credential providers being created when
downloading multiple packages in parallel, as each .narinfo download
would trigger provider creation on its own thread.

Fix by using boost::concurrent_flat_map's try_emplace_and_cvisit, which
provides atomic get-or-create semantics:
- f1 callback: Called atomically during insertion, creates the provider
- f2 callback: Called if key exists, returns cached provider
- Other threads are blocked during f1, so no nullptr is ever visible
2025-10-12 20:16:02 +00:00
Sergei Zimmerman
89b35ec0dc packaging/hydra: buildNoGC is the same as buildWithSanitizers
This will reduce the load on hydra. It doesn't make sense to
build 2 slightly different variations where the difference
is only in the nix-perl-bindings and additional sanitizers.
2025-10-12 22:10:35 +03:00
Sergei Zimmerman
4f585dedbe Merge pull request #14228 from obsidiansystems/fix-windows-build
Fix windows build
2025-10-12 18:07:50 +00:00
John Ericson
a01df8de21 Merge pull request #14227 from NixOS/asan-hydra
packaging: Add buildWithSanitizers to hydraJobs
2025-10-12 17:43:54 +00:00
John Ericson
10223fae86 Fix windows build
I forget to add some CPP in b57caaa1a2.

Hopefully, as we relyon RAII more, these explicit resets become
unneeded.
2025-10-12 13:22:35 -04:00
John Ericson
9150ccb89e Fix Windows dev shell (mostly)
gbenchmark still has too-narrow supported systems, however. That needs
to be fixed in Nixpkgs.
2025-10-12 13:16:50 -04:00
Sergei Zimmerman
de75a180cb packaging: Add buildWithSanitizers to hydraJobs 2025-10-12 19:38:01 +03:00
Sergei Zimmerman
a491173369 packaging: Add withASan,withUBSan options to the scope 2025-10-12 19:16:08 +03:00
Sergei Zimmerman
199b6ff3fb Disable detect_odr_violation for ASan
There's some unfortunate ODR violations that get dianosed with GCC but not Clang
for static inline constexpr variables defined inside the class body:

template<typename T>
struct static_const
{
    static JSON_INLINE_VARIABLE constexpr T value{};
};

This can be ignored pretty much. There is the same problem for std::piecewise_construct:

http://lists.boost.org/Archives/boost/2007/06/123353.php

==2455704==ERROR: AddressSanitizer: odr-violation (0x7efddc460e20):
  [1] size=1 'value' /nix/store/235hvgzcbl06fxy53515q8sr6lljvf68-nlohmann_json-3.11.3/include/nlohmann/detail/meta/cpp_future.hpp:156:45 in /nix/store/pkmljfq97a83dbanr0n64zbm8cyhna33-nix-store-2.33.0pre/lib/libnixstore.so.2.33.0
  [2] size=1 'value' /nix/store/235hvgzcbl06fxy53515q8sr6lljvf68-nlohmann_json-3.11.3/include/nlohmann/detail/meta/cpp_future.hpp:156:45 in /nix/store/gbjpkjj0g8vk20fzlyrwj491gwp6g1qw-nix-util-2.33.0pre/lib/libnixutil.so.2.33.0
2025-10-12 19:16:07 +03:00
Sergei Zimmerman
711e738bf9 meson: Simplify asan-options handling even more
Instead of specifying env variables all the time
we can instead embed the __asan_default_options symbol
in all executables / shared objects. This reduces code
duplication.
2025-10-12 19:16:06 +03:00
Sergei Zimmerman
d9cabddd17 Merge pull request #14214 from obsidiansystems/derivation-resolution-goal
Split out `DerivationResolutionGoal`
2025-10-12 11:31:54 +00:00
John Ericson
0da430be35 Split out DerivationResolutionGoal
This prepares the way for fixing a few issues.

Take 2: was landed before in 8f4a739d0f.
2025-10-11 19:52:37 -04:00
John Ericson
07df87652c Make keys of Derivation*Goal more legible
The property that substitution goals come first is still preserved
2025-10-11 19:50:45 -04:00
John Ericson
a629ce3dec Use member initializer list for Derivation*Goal::drv 2025-10-11 18:37:04 -04:00
John Ericson
ba7bbcd1da Cleanup Derivation*Goal names 2025-10-11 18:30:47 -04:00
Sergei Zimmerman
97e770ad01 Merge pull request #14209 from lovesegfault/curl-based-s3-pieces-fix
fix(libstore): improve http-binary-cache-store S3 compatibility
2025-10-11 21:19:18 +00:00
John Ericson
682cf0b266 Merge pull request #14216 from NixOS/simplify-asan-meson
meson: Move asan-options to common
2025-10-11 18:21:24 +00:00
John Ericson
f1cb837888 Merge pull request #14217 from NixOS/remove-libgit2-patches
packaging: Remove no longer necessary libgit2 patches
2025-10-11 18:20:36 +00:00
Sergei Zimmerman
47705139c9 packaging: Remove no longer necessary libgit2 patches
25.05 already has 1.9.0 and we don't support older nixpkgs versions.
2025-10-11 16:30:55 +03:00
Sergei Zimmerman
d26a337c09 meson: Move asan-options to common
This way we don't have to duplicate the subdir everywhere.
Less copy-pasta is good.
2025-10-11 16:08:35 +03:00
John Ericson
8064c75694 Merge pull request #14194 from xokdvium/stacktrace-assertion
libutil: Print stack trace on assertion failure
2025-10-10 23:25:16 +00:00
John Ericson
d75614a315 Merge pull request #14208 from obsidiansystems/consolidate-builder-dispatch
Consolidate logic choosing where we can/should build a bit
2025-10-10 23:05:40 +00:00
Sergei Zimmerman
46382ade74 libutil: Print stack trace on assertion failure
This change overrides __assert_fail on glibc/musl
to instead call std::terminate that we have a custom
handler for. This ensures that we have more context
to diagnose issues encountered by users in the wild.
2025-10-11 01:35:58 +03:00
John Ericson
9cb686f816 Merge pull request #14210 from NixOS/double-quotes
libstore: Fix double-quoting of paths in logs
2025-10-10 22:10:36 +00:00
Bernardo Meurer Costa
f02218873e fix(libstore): improve http-binary-cache-store S3 compatibility
This commit adds two key fixes to http-binary-cache-store.cc to
properly support the new curl-based S3 implementation:

1. **Consistent cache key handling**: Use `getReference().render(withParams=false)`
   for disk cache keys instead of `cacheUri.to_string()`. This ensures cache
   keys are consistent with the S3 implementation and don't include query
   parameters, which matches the behavior expected by Store::queryPathInfo()
   lookups.

2. **S3 query parameter preservation**: When generating file transfer requests
   for S3 URLs, preserve query parameters from the base URL (region, endpoint,
   etc.) when the relative path doesn't have its own query parameters. This
   ensures S3-specific configuration is propagated to all requests.
2025-10-10 22:05:57 +00:00
John Ericson
b57caaa1a2 Consolidate logic choosing where we can/should build a bit
I want to separate "policy" from "mechanism".

Now the logic to decide how to build (a policy choice, though with some
hard constraints) is all in derivation building goal, and all in the
same spot. build hook, external builder, or local builder --- the choice
between all three is made in the same spot --- pure policy.

Now, if you want to use the external deriation builder, you simply
provide the `ExternalBuilder` you wish to use, and there is no
additional checking --- pure mechanism. It is the responsibility of the
caller to choose an external builder that works for the derivation in
question.

Also, `checkSystem()` was the only thing throwing `BuildError` from
`startBuilder`. Now that that is gone, we can now remove the
`try...catch` around that.
2025-10-10 17:28:57 -04:00
John Ericson
2ff59ec3e0 Use std::ranges::find_if for finding external builders
Co-authored-by: Sergei Zimmerman <sergei@zimmerman.foo>
2025-10-10 17:27:41 -04:00
Sergei Zimmerman
f30cb8667b libstore: Fix double-quoting of paths in logs
std::filesystem::path is already quoted by boost::format with double quotes (").
2025-10-11 00:25:51 +03:00
John Ericson
b56dd21c31 Settings::ExternalBuilder::systems make set
Nothing cares about the order, actually.
2025-10-10 17:23:45 -04:00
Sergei Zimmerman
2308aaf192 Merge pull request #14198 from lovesegfault/curl-based-s3-pieces
feat(libstore): add curl-based S3 store implementation
2025-10-10 07:55:44 +00:00
Graham Dennis
8d9e9bc400 Improve comment 2025-10-10 15:00:10 +11:00
Graham Dennis
43b01b6790 Improved backwards compatibility hack for git URLs using dir=... attribute 2025-10-10 14:54:47 +11:00
Bernardo Meurer Costa
0855b715a9 feat(libstore): add curl-based S3 store implementation
Add a new S3BinaryCacheStore implementation that inherits from
HttpBinaryCacheStore.

The implementation is activated with NIX_WITH_CURL_S3, keeping the
existing NIX_WITH_S3_SUPPORT (AWS SDK) implementation unchanged.
2025-10-10 00:13:29 +00:00
John Ericson
2fc88ec114 Merge pull request #14204 from NixOS/kill-to-real-path-fetchers
libfetchers: Remove toRealPath in SourceHutInputScheme::getRevFromRef
2025-10-09 22:01:03 +00:00
Sergei Zimmerman
9c8480becb Merge pull request #14203 from roberth/move-eval-cache-open-to-libflake
Move eval cache open to libflake
2025-10-09 21:42:06 +00:00
Sergei Zimmerman
c58acff42a libfetchers: Remove toRealPath in SourceHutInputScheme::getRevFromRef
This code had several issues:

1. Not going through the SourceAccessor means that we can only work
   with physical paths.

2. It did not actually check that the file exists. (std::ifstream does not check
   it by default).
2025-10-10 00:26:57 +03:00
Robert Hensing
abcceafbce Use const for lock in openEvalCache 2025-10-09 21:25:40 +02:00
Robert Hensing
42c9cbf9ca Use ref<LockedFlake> where non-null 2025-10-09 21:25:40 +02:00
Robert Hensing
0387b7d6db Move openEvalCache to libflake
Most of the eval cache logic is flake-independent and libexpr,
but the loading part is not.
`nix-flake` is the right component for this, as the eval cache
isn't exactly specific to the command line.
2025-10-09 20:40:40 +02:00
Sergei Zimmerman
dfafd8bc38 Merge pull request #14199 from getchoo-contrib/getchoo/toml11-fix
packaging: only override `toml11` when necessary
2025-10-09 17:20:35 +00:00
Sergei Zimmerman
44486871e9 Merge pull request #14200 from Mic92/less-ci-builds
only build on push to master
2025-10-09 15:07:06 +00:00
Jörg Thalheim
118acc84ba only build on push to master
we have now merge queues for maintainance branches. We still build it
for master to have our installer beeing updated. In future this part
could go in new workflow instead.
2025-10-09 14:16:14 +01:00
Seth Flynn
0f016f9bf5 packaging: only override toml11 when necessary
v4.4.0 hit Nixpkgs in https://github.com/NixOS/nixpkgs/pull/442682.
Ideally we'd just use that, but this keeps the fallback behavior until
it's more widespread
2025-10-09 03:24:57 -04:00
John Ericson
ce38b46e06 Merge pull request #14170 from lovesegfault/curl-based-s3-pieces
feat(libstore/filetransfer): add S3 signing support
2025-10-08 22:35:32 +00:00
Sergei Zimmerman
33d9270109 Merge pull request #14191 from NixOS/fix-14188
libutil: Fix renderAuthorityAndPath unreachable for path:/ URLs
2025-10-08 21:34:28 +00:00
Bernardo Meurer Costa
00c2a57666 feat(libstore/filetransfer): add S3 signing support 2025-10-08 21:31:34 +00:00
Eelco Dolstra
d591f17ecb Merge pull request #14189 from NixOS/fix-exportReferencesGraph
exportReferencesGraph: Handle heterogeneous arrays
2025-10-08 21:19:30 +00:00
John Ericson
bb1945a090 Merge pull request #14182 from NixOS/simplify-archive-tests
tests: Move invalid nar tests from tests/functional to libutil-tests
2025-10-08 20:46:37 +00:00
Eelco Dolstra
94f410b628 exportReferencesGraph: Handle heterogeneous arrays
This barfed with

   error: [json.exception.type_error.302] type must be string, but is array

on `nix build github:malt3/bazel-env#bazel-env` because it has a `exportReferencesGraph` with a value like `["string",...["string"]]`.
2025-10-08 22:15:33 +02:00
Bernardo Meurer Costa
3c1e2e56ea feat(libstore/filetransfer): add username/password authentication support
Add a `UsernameAuth` struct and optional `usernameAuth` field to
`FileTransferRequest` to support programmatic username/password
authentication.

This uses curl's `CURLOPT_USERNAME`/`CURLOPT_PASSWORD` options, which
works with multiple protocols (HTTP, FTP, etc.) and is not specific to
any particular authentication scheme.

The primary motivation is to enable S3 authentication refactoring where
AWS credentials (access key ID and secret access key) can be passed
through this general-purpose mechanism, reducing the amount of
S3-specific code behind `#if NIX_WITH_CURL_S3` guards.
2025-10-08 20:10:53 +00:00
Eelco Dolstra
925f10d5ea Merge pull request #14087 from NixOS/required-args
Args::Flag: Add `required` attribute
2025-10-08 19:33:22 +00:00
Sergei Zimmerman
1d8dd77e1d libutil: Fix renderAuthorityAndPath unreachable for path:/ URLs
This was mistakenly triggered by path:/ URL, since the `//` would
correspond to 3 empty segments.
2025-10-08 22:14:49 +03:00
Jörg Thalheim
090f7fb05e Merge pull request #14002 from getchoo-contrib/getchoo/dogfood-experimental-installer
Add experimental installer to installer tests
2025-10-08 04:05:46 +00:00
Jörg Thalheim
2ce343716d Merge pull request #14169 from NixOS/dev-shell-no-separate-debug-info
dev-shell: Disable separateDebugInfo
2025-10-08 04:03:43 +00:00
Jörg Thalheim
e69e621578 Merge pull request #14180 from NixOS/release-process-drop-mergify
maintainers: Remove mergify note from release-process.md
2025-10-08 03:34:05 +00:00
Jörg Thalheim
d2a0a11a8e Merge pull request #14184 from lovesegfault/nix-better-ci
ci: integrate vm_tests into main tests job
2025-10-08 05:23:53 +02:00
Bernardo Meurer Costa
a400ea4257 ci: integrate vm_tests into main tests job
This consolidates the separate vm_tests job into the main tests job,
simplifying the CI workflow. VM tests now run as part of the regular
test matrix.
2025-10-08 02:46:56 +00:00
Sergei Zimmerman
fc8b784924 Merge pull request #14147 from lovesegfault/nix-multi-ci
ci: test without s3 and with curl-based-s3
2025-10-08 02:38:03 +03:00
Sergei Zimmerman
0619351326 tests: Move invalid nar tests from tests/functional to libutil-tests
Since 242f362567 we have better infrastructure
for this kind of tests.
2025-10-08 02:02:33 +03:00
Sergei Zimmerman
c5b88c22fa dev-shell: Disable separateDebugInfo
This breaks gdb pretty-printers inserted into .debug_gdb_scripts section,
because it implies --compress-debug-sections=zlib, -Wa,--compress-debug-sections.
This is very unfortunate, because then gdb can't use pretty printers for
Boost.Unordered (which are very useful, since boost::unoredred_flat_map is
impossible to debug). This seems perfectly fine to disable in the dev-shell for
the time being.

See [1-3] for further references.

With this change I'm able to use boost's pretty-printers out-of-the box:

```
p *importResolutionCache
$2 = boost::concurrent_flat_map with 1 elements = {[{accessor = {p = std::shared_ptr<nix::SourceAccessor> (use count 5, weak count 1) = {
        get() = 0x555555d830a8}}, path = {static root = {static root = <same as static member of an already seen type>, path = "/"},
      path = "/derivation-internal.nix"}}] = {accessor = {p = std::shared_ptr<nix::SourceAccessor> (use count 5, weak count 1) = {
        get() = 0x555555d830a8}}, path = {static root = {static root = <same as static member of an already seen type>, path = "/"},
      path = "/derivation-internal.nix"}}}
```

When combined with a simple `add-auto-load-safe-path ~/code` in .gdbinit

[1]: https://gerrit.lix.systems/c/lix/+/3880
[2]: https://git.lix.systems/lix-project/lix/issues/1003
[3]: https://sourceware.org/pipermail/gdb-patches/2025-October/221398.html
2025-10-08 00:57:34 +03:00
Sergei Zimmerman
75b18a6e47 maintainers: Remove mergify note from release-process.md 2025-10-08 00:51:50 +03:00
Sergei Zimmerman
fb117e0cac Merge pull request #14155 from Mic92/backport-action
ci: Switch away from mergify to backport action
2025-10-08 00:32:00 +03:00
Jörg Thalheim
63e8b5f94a ci: Switch away from mergify to backport action
We want to use github native queues.
2025-10-07 23:43:03 +03:00
Eelco Dolstra
559061bf5a Merge pull request #14173 from NixOS/bump-2.33.0
Bump version
2025-10-07 20:42:25 +02:00
Eelco Dolstra
b0f567e18b Update mergify.yml 2025-10-07 17:16:57 +02:00
Eelco Dolstra
875f7ec25f Bump version 2025-10-07 17:15:28 +02:00
Bernardo Meurer Costa
8c28283876 ci: test without s3 and with curl-based-s3 2025-10-06 16:24:21 +00:00
Bernardo Meurer Costa
7f22a40e3b build(libstore): assert withAWS xor withCurlS3 2025-10-06 16:22:01 +00:00
Tristan Ross
9abcc68ad1 libstore-c: add nix_store_get_fs_closure 2025-10-06 09:12:02 -07:00
Seth Flynn
92d7381826 ci: allow for using the latest build of the experimental installer
Until these repos are potentially merged, this is good for dogfooding
alongside the experimental installer. It also uses the more official
`artifacts.nixos.org` endpoint to install stable releases now

More immediately though, we need a patch for the experimental installer
to really work in CI at all, and that hasn't landed in a tag yet. So,
this lets us use it right from `main`!
2025-10-03 03:37:59 -04:00
Seth Flynn
d2293fb458 ci: enable experimental installer tests 2025-10-03 01:26:55 -04:00
Seth Flynn
2cbbb63628 ci: enable use of the experimental installer 2025-10-03 01:26:52 -04:00
Eelco Dolstra
eec4dece33 Args::Flag: Add required attribute 2025-09-26 16:01:59 +02:00
495 changed files with 12177 additions and 6015 deletions

18
.coderabbit.yaml Normal file
View File

@@ -0,0 +1,18 @@
# yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json
# Disable CodeRabbit auto-review to prevent verbose comments on PRs.
# When enabled: false, CodeRabbit won't attempt reviews and won't post
# "Review skipped" or other automated comments.
reviews:
auto_review:
enabled: false
review_status: false
high_level_summary: false
poem: false
sequence_diagrams: false
changed_files_summary: false
tools:
github-checks:
enabled: false
chat:
art: false
auto_reply: false

View File

@@ -4,15 +4,29 @@ inputs:
dogfood:
description: "Whether to use Nix installed from the latest artifact from master branch"
required: true # Be explicit about the fact that we are using unreleased artifacts
experimental-installer:
description: "Whether to use the experimental installer to install Nix"
default: false
experimental-installer-version:
description: "Version of the experimental installer to use. If `latest`, the newest artifact from the default branch is used."
# TODO: This should probably be pinned to a release after https://github.com/NixOS/experimental-nix-installer/pull/49 lands in one
default: "latest"
extra_nix_config:
description: "Gets appended to `/etc/nix/nix.conf` if passed."
install_url:
description: "URL of the Nix installer"
required: false
default: "https://releases.nixos.org/nix/nix-2.30.2/install"
default: "https://releases.nixos.org/nix/nix-2.32.1/install"
tarball_url:
description: "URL of the Nix tarball to use with the experimental installer"
required: false
github_token:
description: "Github token"
required: true
use_cache:
description: "Whether to setup magic-nix-cache"
default: true
required: false
runs:
using: "composite"
steps:
@@ -37,14 +51,81 @@ runs:
gh run download "$RUN_ID" --repo "$DOGFOOD_REPO" -n "$INSTALLER_ARTIFACT" -D "$INSTALLER_DOWNLOAD_DIR"
echo "installer-path=file://$INSTALLER_DOWNLOAD_DIR" >> "$GITHUB_OUTPUT"
TARBALL_PATH="$(find "$INSTALLER_DOWNLOAD_DIR" -name 'nix*.tar.xz' -print | head -n 1)"
echo "tarball-path=file://$TARBALL_PATH" >> "$GITHUB_OUTPUT"
echo "::notice ::Dogfooding Nix installer from master (https://github.com/$DOGFOOD_REPO/actions/runs/$RUN_ID)"
env:
GH_TOKEN: ${{ inputs.github_token }}
DOGFOOD_REPO: "NixOS/nix"
- name: "Gather system info for experimental installer"
shell: bash
if: ${{ inputs.experimental-installer == 'true' }}
run: |
echo "::notice Using experimental installer from $EXPERIMENTAL_INSTALLER_REPO (https://github.com/$EXPERIMENTAL_INSTALLER_REPO)"
if [ "$RUNNER_OS" == "Linux" ]; then
EXPERIMENTAL_INSTALLER_SYSTEM="linux"
echo "EXPERIMENTAL_INSTALLER_SYSTEM=$EXPERIMENTAL_INSTALLER_SYSTEM" >> "$GITHUB_ENV"
elif [ "$RUNNER_OS" == "macOS" ]; then
EXPERIMENTAL_INSTALLER_SYSTEM="darwin"
echo "EXPERIMENTAL_INSTALLER_SYSTEM=$EXPERIMENTAL_INSTALLER_SYSTEM" >> "$GITHUB_ENV"
else
echo "::error ::Unsupported RUNNER_OS: $RUNNER_OS"
exit 1
fi
if [ "$RUNNER_ARCH" == "X64" ]; then
EXPERIMENTAL_INSTALLER_ARCH=x86_64
echo "EXPERIMENTAL_INSTALLER_ARCH=$EXPERIMENTAL_INSTALLER_ARCH" >> "$GITHUB_ENV"
elif [ "$RUNNER_ARCH" == "ARM64" ]; then
EXPERIMENTAL_INSTALLER_ARCH=aarch64
echo "EXPERIMENTAL_INSTALLER_ARCH=$EXPERIMENTAL_INSTALLER_ARCH" >> "$GITHUB_ENV"
else
echo "::error ::Unsupported RUNNER_ARCH: $RUNNER_ARCH"
exit 1
fi
echo "EXPERIMENTAL_INSTALLER_ARTIFACT=nix-installer-$EXPERIMENTAL_INSTALLER_ARCH-$EXPERIMENTAL_INSTALLER_SYSTEM" >> "$GITHUB_ENV"
env:
EXPERIMENTAL_INSTALLER_REPO: "NixOS/experimental-nix-installer"
- name: "Download latest experimental installer"
shell: bash
id: download-latest-experimental-installer
if: ${{ inputs.experimental-installer == 'true' && inputs.experimental-installer-version == 'latest' }}
run: |
RUN_ID=$(gh run list --repo "$EXPERIMENTAL_INSTALLER_REPO" --workflow ci.yml --branch main --status success --json databaseId --jq ".[0].databaseId")
EXPERIMENTAL_INSTALLER_DOWNLOAD_DIR="$GITHUB_WORKSPACE/$EXPERIMENTAL_INSTALLER_ARTIFACT"
mkdir -p "$EXPERIMENTAL_INSTALLER_DOWNLOAD_DIR"
gh run download "$RUN_ID" --repo "$EXPERIMENTAL_INSTALLER_REPO" -n "$EXPERIMENTAL_INSTALLER_ARTIFACT" -D "$EXPERIMENTAL_INSTALLER_DOWNLOAD_DIR"
# Executable permissions are lost in artifacts
find $EXPERIMENTAL_INSTALLER_DOWNLOAD_DIR -type f -exec chmod +x {} +
echo "installer-path=$EXPERIMENTAL_INSTALLER_DOWNLOAD_DIR" >> "$GITHUB_OUTPUT"
env:
GH_TOKEN: ${{ inputs.github_token }}
EXPERIMENTAL_INSTALLER_REPO: "NixOS/experimental-nix-installer"
- uses: cachix/install-nix-action@c134e4c9e34bac6cab09cf239815f9339aaaf84e # v31.5.1
if: ${{ inputs.experimental-installer != 'true' }}
with:
# Ternary operator in GHA: https://www.github.com/actions/runner/issues/409#issuecomment-752775072
install_url: ${{ inputs.dogfood == 'true' && format('{0}/install', steps.download-nix-installer.outputs.installer-path) || inputs.install_url }}
install_options: ${{ inputs.dogfood == 'true' && format('--tarball-url-prefix {0}', steps.download-nix-installer.outputs.installer-path) || '' }}
extra_nix_config: ${{ inputs.extra_nix_config }}
- uses: DeterminateSystems/nix-installer-action@786fff0690178f1234e4e1fe9b536e94f5433196 # v20
if: ${{ inputs.experimental-installer == 'true' }}
with:
diagnostic-endpoint: ""
# TODO: It'd be nice to use `artifacts.nixos.org` for both of these, maybe through an `/experimental-installer/latest` endpoint? or `/commit/<hash>`?
local-root: ${{ inputs.experimental-installer-version == 'latest' && steps.download-latest-experimental-installer.outputs.installer-path || '' }}
source-url: ${{ inputs.experimental-installer-version != 'latest' && 'https://artifacts.nixos.org/experimental-installer/tag/${{ inputs.experimental-installer-version }}/${{ env.EXPERIMENTAL_INSTALLER_ARTIFACT }}' || '' }}
nix-package-url: ${{ inputs.dogfood == 'true' && steps.download-nix-installer.outputs.tarball-path || (inputs.tarball_url || '') }}
extra-conf: ${{ inputs.extra_nix_config }}
- uses: DeterminateSystems/magic-nix-cache-action@565684385bcd71bad329742eefe8d12f2e765b39 # v13
if: ${{ inputs.use_cache == 'true' }}
with:
diagnostic-endpoint: ''
use-flakehub: false
use-gha-cache: true
source-revision: 92d9581367be2233c2d5714a2640e1339f4087d8 # main

37
.github/workflows/backport.yml vendored Normal file
View File

@@ -0,0 +1,37 @@
name: Backport
on:
pull_request_target:
types: [closed, labeled]
permissions:
contents: read
jobs:
backport:
name: Backport Pull Request
permissions:
# for korthout/backport-action
contents: write
pull-requests: write
if: github.repository_owner == 'NixOS' && github.event.pull_request.merged == true && (github.event_name != 'labeled' || startsWith('backport', github.event.label.name))
runs-on: ubuntu-24.04-arm
steps:
- name: Generate GitHub App token
id: generate-token
uses: actions/create-github-app-token@v2
with:
app-id: ${{ vars.CI_APP_ID }}
private-key: ${{ secrets.CI_APP_PRIVATE_KEY }}
- uses: actions/checkout@v5
with:
ref: ${{ github.event.pull_request.head.sha }}
# required to find all branches
fetch-depth: 0
- name: Create backport PRs
uses: korthout/backport-action@d07416681cab29bf2661702f925f020aaa962997 # v3.4.1
id: backport
with:
# Config README: https://github.com/korthout/backport-action#backport-action
github_token: ${{ steps.generate-token.outputs.token }}
github_workspace: ${{ github.workspace }}
auto_merge_enabled: true
pull_description: |-
Automatic backport to `${target_branch}`, triggered by a label in #${pull_number}.

View File

@@ -4,6 +4,8 @@ on:
pull_request:
merge_group:
push:
branches:
- master
workflow_dispatch:
inputs:
dogfood:
@@ -12,6 +14,10 @@ on:
default: true
type: boolean
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
cancel-in-progress: true
permissions: read-all
jobs:
@@ -27,6 +33,7 @@ jobs:
extra_nix_config:
experimental-features = nix-command flakes
github_token: ${{ secrets.GITHUB_TOKEN }}
use_cache: false
- run: nix flake show --all-systems --json
pre-commit-checks:
@@ -39,7 +46,6 @@ jobs:
dogfood: ${{ github.event_name == 'workflow_dispatch' && inputs.dogfood || github.event_name != 'workflow_dispatch' }}
extra_nix_config: experimental-features = nix-command flakes
github_token: ${{ secrets.GITHUB_TOKEN }}
- uses: DeterminateSystems/magic-nix-cache-action@main
- run: ./ci/gha/tests/pre-commit-checks
basic-checks:
@@ -90,7 +96,6 @@ jobs:
dogfood: ${{ github.event_name == 'workflow_dispatch' && inputs.dogfood || github.event_name != 'workflow_dispatch' }}
# The sandbox would otherwise be disabled by default on Darwin
extra_nix_config: "sandbox = true"
- uses: DeterminateSystems/magic-nix-cache-action@main
# Since ubuntu 22.30, unprivileged usernamespaces are no longer allowed to map to the root user:
# https://ubuntu.com/blog/ubuntu-23-10-restricted-unprivileged-user-namespaces
- run: sudo sysctl -w kernel.apparmor_restrict_unprivileged_userns=0
@@ -100,6 +105,12 @@ jobs:
nix build --file ci/gha/tests/wrapper.nix componentTests -L \
--arg withInstrumentation ${{ matrix.instrumented }} \
--argstr stdenv "${{ matrix.stdenv }}"
- name: Run VM tests
run: |
nix build --file ci/gha/tests/wrapper.nix vmTests -L \
--arg withInstrumentation ${{ matrix.instrumented }} \
--argstr stdenv "${{ matrix.stdenv }}"
if: ${{ matrix.os == 'linux' }}
- name: Run flake checks and prepare the installer tarball
run: |
ci/gha/tests/build-checks
@@ -114,13 +125,13 @@ jobs:
cat coverage-reports/index.txt >> $GITHUB_STEP_SUMMARY
if: ${{ matrix.instrumented }}
- name: Upload coverage reports
uses: actions/upload-artifact@v4
uses: actions/upload-artifact@v5
with:
name: coverage-reports
path: coverage-reports/
if: ${{ matrix.instrumented }}
- name: Upload installer tarball
uses: actions/upload-artifact@v4
uses: actions/upload-artifact@v5
with:
name: installer-${{matrix.os}}
path: out/*
@@ -135,25 +146,46 @@ jobs:
- scenario: on ubuntu
runs-on: ubuntu-24.04
os: linux
experimental-installer: false
- scenario: on macos
runs-on: macos-14
os: darwin
experimental-installer: false
- scenario: on ubuntu (experimental)
runs-on: ubuntu-24.04
os: linux
experimental-installer: true
- scenario: on macos (experimental)
runs-on: macos-14
os: darwin
experimental-installer: true
name: installer test ${{ matrix.scenario }}
runs-on: ${{ matrix.runs-on }}
steps:
- uses: actions/checkout@v5
- name: Download installer tarball
uses: actions/download-artifact@v5
uses: actions/download-artifact@v6
with:
name: installer-${{matrix.os}}
path: out
- name: Looking up the installer tarball URL
id: installer-tarball-url
run: echo "installer-url=file://$GITHUB_WORKSPACE/out" >> "$GITHUB_OUTPUT"
- uses: cachix/install-nix-action@v31
run: |
echo "installer-url=file://$GITHUB_WORKSPACE/out" >> "$GITHUB_OUTPUT"
TARBALL_PATH="$(find "$GITHUB_WORKSPACE/out" -name 'nix*.tar.xz' -print | head -n 1)"
echo "tarball-path=file://$TARBALL_PATH" >> "$GITHUB_OUTPUT"
- uses: cachix/install-nix-action@456688f15bc354bef6d396e4a35f4f89d40bf2b7 # v31.8.2
if: ${{ !matrix.experimental-installer }}
with:
install_url: ${{ format('{0}/install', steps.installer-tarball-url.outputs.installer-url) }}
install_options: ${{ format('--tarball-url-prefix {0}', steps.installer-tarball-url.outputs.installer-url) }}
- uses: ./.github/actions/install-nix-action
if: ${{ matrix.experimental-installer }}
with:
dogfood: false
experimental-installer: true
tarball_url: ${{ steps.installer-tarball-url.outputs.tarball-path }}
github_token: ${{ secrets.GITHUB_TOKEN }}
- run: sudo apt install fish zsh
if: matrix.os == 'linux'
- run: brew install fish
@@ -185,7 +217,7 @@ jobs:
echo "docker=${{ env._DOCKER_SECRETS != '' }}" >> $GITHUB_OUTPUT
docker_push_image:
needs: [tests, vm_tests, check_secrets]
needs: [tests, check_secrets]
permissions:
contents: read
packages: write
@@ -198,12 +230,13 @@ jobs:
- uses: actions/checkout@v5
with:
fetch-depth: 0
- uses: cachix/install-nix-action@v31
- uses: ./.github/actions/install-nix-action
with:
install_url: https://releases.nixos.org/nix/nix-2.20.3/install
- uses: DeterminateSystems/magic-nix-cache-action@main
- run: echo NIX_VERSION="$(nix --experimental-features 'nix-command flakes' eval .\#nix.version | tr -d \")" >> $GITHUB_ENV
- run: nix --experimental-features 'nix-command flakes' build .#dockerImage -L
dogfood: false
extra_nix_config: |
experimental-features = flakes nix-command
- run: echo NIX_VERSION="$(nix eval .\#nix.version | tr -d \")" >> $GITHUB_ENV
- run: nix build .#dockerImage -L
- run: docker load -i ./result/image.tar.gz
- run: docker tag nix:$NIX_VERSION ${{ secrets.DOCKERHUB_USERNAME }}/nix:$NIX_VERSION
- run: docker tag nix:$NIX_VERSION ${{ secrets.DOCKERHUB_USERNAME }}/nix:master
@@ -238,28 +271,8 @@ jobs:
docker tag nix:$NIX_VERSION $IMAGE_ID:master
docker push $IMAGE_ID:master
vm_tests:
needs: basic-checks
runs-on: ubuntu-24.04
steps:
- uses: actions/checkout@v5
- uses: ./.github/actions/install-nix-action
with:
dogfood: ${{ github.event_name == 'workflow_dispatch' && inputs.dogfood || github.event_name != 'workflow_dispatch' }}
extra_nix_config:
experimental-features = nix-command flakes
github_token: ${{ secrets.GITHUB_TOKEN }}
- uses: DeterminateSystems/magic-nix-cache-action@main
- run: |
nix build -L \
.#hydraJobs.tests.functional_user \
.#hydraJobs.tests.githubFlakes \
.#hydraJobs.tests.nix-docker \
.#hydraJobs.tests.tarballFlakes \
;
flake_regressions:
needs: vm_tests
needs: tests
runs-on: ubuntu-24.04
steps:
- name: Checkout nix
@@ -280,7 +293,6 @@ jobs:
extra_nix_config:
experimental-features = nix-command flakes
github_token: ${{ secrets.GITHUB_TOKEN }}
- uses: DeterminateSystems/magic-nix-cache-action@main
- run: nix build -L --out-link ./new-nix && PATH=$(pwd)/new-nix/bin:$PATH MAX_FLAKES=25 flake-regressions/eval-all.sh
profile_build:
@@ -301,7 +313,6 @@ jobs:
extra_nix_config: |
experimental-features = flakes nix-command ca-derivations impure-derivations
max-jobs = 1
- uses: DeterminateSystems/magic-nix-cache-action@main
- run: |
nix build -L --file ./ci/gha/profile-build buildTimeReport --out-link build-time-report.md
cat build-time-report.md >> $GITHUB_STEP_SUMMARY

View File

@@ -1,174 +0,0 @@
queue_rules:
- name: default
# all required tests need to go here
merge_conditions:
- check-success=tests on macos
- check-success=tests on ubuntu
- check-success=installer test on macos
- check-success=installer test on ubuntu
- check-success=vm_tests
batch_size: 5
pull_request_rules:
- name: merge using the merge queue
conditions:
- base~=master|.+-maintenance
- label~=merge-queue|dependencies
actions:
queue: {}
# The rules below will first create backport pull requests and put those in a merge queue.
- name: backport patches to 2.18
conditions:
- label=backport 2.18-maintenance
actions:
backport:
branches:
- 2.18-maintenance
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.19
conditions:
- label=backport 2.19-maintenance
actions:
backport:
branches:
- 2.19-maintenance
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.20
conditions:
- label=backport 2.20-maintenance
actions:
backport:
branches:
- 2.20-maintenance
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.21
conditions:
- label=backport 2.21-maintenance
actions:
backport:
branches:
- 2.21-maintenance
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.22
conditions:
- label=backport 2.22-maintenance
actions:
backport:
branches:
- 2.22-maintenance
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.23
conditions:
- label=backport 2.23-maintenance
actions:
backport:
branches:
- 2.23-maintenance
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.24
conditions:
- label=backport 2.24-maintenance
actions:
backport:
branches:
- "2.24-maintenance"
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.25
conditions:
- label=backport 2.25-maintenance
actions:
backport:
branches:
- "2.25-maintenance"
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.26
conditions:
- label=backport 2.26-maintenance
actions:
backport:
branches:
- "2.26-maintenance"
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.27
conditions:
- label=backport 2.27-maintenance
actions:
backport:
branches:
- "2.27-maintenance"
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.28
conditions:
- label=backport 2.28-maintenance
actions:
backport:
branches:
- "2.28-maintenance"
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.29
conditions:
- label=backport 2.29-maintenance
actions:
backport:
branches:
- "2.29-maintenance"
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.30
conditions:
- label=backport 2.30-maintenance
actions:
backport:
branches:
- "2.30-maintenance"
labels:
- automatic backport
- merge-queue
- name: backport patches to 2.31
conditions:
- label=backport 2.31-maintenance
actions:
backport:
branches:
- "2.31-maintenance"
labels:
- automatic backport
- merge-queue

View File

@@ -1 +1 @@
2.32.4
2.33.0

View File

@@ -21,16 +21,6 @@ let
packages' = nixFlake.packages.${system};
stdenv = (getStdenv pkgs);
enableSanitizersLayer = finalAttrs: prevAttrs: {
mesonFlags =
(prevAttrs.mesonFlags or [ ])
++ [ (lib.mesonOption "b_sanitize" "address,undefined") ]
++ (lib.optionals stdenv.cc.isClang [
# https://www.github.com/mesonbuild/meson/issues/764
(lib.mesonBool "b_lundef" false)
]);
};
collectCoverageLayer = finalAttrs: prevAttrs: {
env =
let
@@ -53,14 +43,15 @@ let
'';
};
componentOverrides =
(lib.optional withSanitizers enableSanitizersLayer)
++ (lib.optional withCoverage collectCoverageLayer);
componentOverrides = (lib.optional withCoverage collectCoverageLayer);
in
rec {
nixComponentsInstrumented = nixComponents.overrideScope (
final: prev: {
withASan = withSanitizers;
withUBSan = withSanitizers;
nix-store-tests = prev.nix-store-tests.override { withBenchmarks = true; };
# Boehm is incompatible with ASAN.
nix-expr = prev.nix-expr.override { enableGC = !withSanitizers; };
@@ -71,6 +62,14 @@ rec {
}
);
# Import NixOS tests using the instrumented components
nixosTests = import ../../../tests/nixos {
inherit lib pkgs;
nixComponents = nixComponentsInstrumented;
nixpkgs = nixFlake.inputs.nixpkgs;
inherit (nixFlake.inputs) nixpkgs-23-11;
};
/**
Top-level tests for the flake outputs, as they would be built by hydra.
These tests generally can't be overridden to run with sanitizers.
@@ -108,15 +107,33 @@ rec {
};
};
disable =
let
inherit (pkgs.stdenv) hostPlatform;
in
args@{
pkgName,
testName,
test,
}:
lib.any (b: b) [
# FIXME: Nix manual is impure and does not produce all settings on darwin
(hostPlatform.isDarwin && pkgName == "nix-manual" && testName == "linkcheck")
];
componentTests =
(lib.concatMapAttrs (
pkgName: pkg:
lib.concatMapAttrs (testName: test: {
"${componentTestsPrefix}${pkgName}-${testName}" = test;
}) (pkg.tests or { })
lib.concatMapAttrs (
testName: test:
lib.optionalAttrs (!disable { inherit pkgName testName test; }) {
"${componentTestsPrefix}${pkgName}-${testName}" = test;
}
) (pkg.tests or { })
) nixComponentsInstrumented)
// lib.optionalAttrs (pkgs.stdenv.hostPlatform == pkgs.stdenv.buildPlatform) {
"${componentTestsPrefix}nix-functional-tests" = nixComponentsInstrumented.nix-functional-tests;
"${componentTestsPrefix}nix-json-schema-checks" = nixComponentsInstrumented.nix-json-schema-checks;
};
codeCoverage =
@@ -221,4 +238,20 @@ rec {
{
inherit coverageProfileDrvs mergedProfdata coverageReports;
};
vmTests = {
inherit (nixosTests) s3-binary-cache-store;
}
// lib.optionalAttrs (!withSanitizers && !withCoverage) {
# evalNixpkgs uses non-instrumented components from hydraJobs, so only run it
# when not testing with sanitizers to avoid rebuilding nix
inherit (hydraJobs.tests) evalNixpkgs;
# FIXME: CI times out when building vm tests instrumented
inherit (nixosTests)
functional_user
githubFlakes
nix-docker
tarballFlakes
;
};
}

View File

@@ -3,7 +3,7 @@
def transform_anchors_html:
. | gsub($empty_anchor_regex; "<a name=\"" + .anchor + "\"></a>")
. | gsub($empty_anchor_regex; "<a id=\"" + .anchor + "\"></a>")
| gsub($anchor_regex; "<a href=\"#" + .anchor + "\" id=\"" + .anchor + "\">" + .text + "</a>");

View File

@@ -7,6 +7,7 @@ additional-css = ["custom.css"]
additional-js = ["redirects.js"]
edit-url-template = "https://github.com/NixOS/nix/tree/master/doc/manual/{path}"
git-repository-url = "https://github.com/NixOS/nix"
mathjax-support = true
# Handles replacing @docroot@ with a path to ./source relative to that markdown file,
# {{#include handlebars}}, and the @generated@ syntax used within these. it mostly

View File

@@ -15,7 +15,6 @@ pymod = import('python')
python = pymod.find_installation('python3')
nix_env_for_docs = {
'ASAN_OPTIONS' : 'abort_on_error=1:print_summary=1:detect_leaks=0',
'HOME' : '/dummy',
'NIX_CONF_DIR' : '/dummy',
'NIX_SSL_CERT_FILE' : '/dummy/no-ca-bundle.crt',
@@ -89,10 +88,12 @@ manual = custom_target(
@0@ @INPUT0@ @CURRENT_SOURCE_DIR@ > @DEPFILE@
@0@ @INPUT1@ summary @2@ < @CURRENT_SOURCE_DIR@/source/SUMMARY.md.in > @2@/source/SUMMARY.md
sed -e 's|@version@|@3@|g' < @INPUT2@ > @2@/book.toml
@4@ -r --include='*.md' @CURRENT_SOURCE_DIR@/ @2@/
@4@ -r -L --include='*.md' @CURRENT_SOURCE_DIR@/ @2@/
(cd @2@; RUST_LOG=warn @1@ build -d @2@ 3>&2 2>&1 1>&3) | { grep -Fv "because fragment resolution isn't implemented" || :; } 3>&2 2>&1 1>&3
rm -rf @2@/manual
mv @2@/html @2@/manual
# Remove Mathjax 2.7, because we will actually use MathJax 3.x
find @2@/manual | grep .html | xargs sed -i -e '/2.7.1.MathJax.js/d'
find @2@/manual -iname meson.build -delete
'''.format(
python.full_path(),
@@ -116,6 +117,7 @@ manual = custom_target(
builtins_md,
rl_next_generated,
summary_rl_next,
json_schema_generated_files,
nix_input,
],
output : [

View File

@@ -12,11 +12,15 @@
rsync,
nix-cli,
changelog-d,
json-schema-for-humans,
officialRelease,
# Configuration Options
version,
# `tests` attribute
testers,
}:
let
@@ -32,6 +36,15 @@ mkMesonDerivation (finalAttrs: {
fileset.difference
(fileset.unions [
../../.version
# For example JSON
../../src/libutil-tests/data/hash
../../src/libstore-tests/data/content-address
../../src/libstore-tests/data/store-path
../../src/libstore-tests/data/realisation
../../src/libstore-tests/data/derived-path
../../src/libstore-tests/data/path-info
../../src/libstore-tests/data/nar-info
../../src/libstore-tests/data/build-result
# Too many different types of files to filter for now
../../doc/manual
./.
@@ -55,6 +68,7 @@ mkMesonDerivation (finalAttrs: {
jq
python3
rsync
json-schema-for-humans
changelog-d
]
++ lib.optionals (!officialRelease) [
@@ -78,6 +92,29 @@ mkMesonDerivation (finalAttrs: {
echo "doc manual ''$out/share/doc/nix/manual" >> ''$out/nix-support/hydra-build-products
'';
/**
The root of the HTML manual.
E.g. "${nix-manual.site}/index.html" exists.
*/
passthru.site = finalAttrs.finalPackage + "/share/doc/nix/manual";
passthru.tests = {
# https://nixos.org/manual/nixpkgs/stable/index.html#tester-lycheeLinkCheck
linkcheck = testers.lycheeLinkCheck {
inherit (finalAttrs.finalPackage) site;
extraConfig = {
exclude = [
# Exclude auto-generated JSON schema documentation which has
# auto-generated fragment IDs that don't match the link references
".*/protocols/json/.*\\.html"
# Exclude undocumented builtins
".*/language/builtins\\.html#builtins-addErrorContext"
".*/language/builtins\\.html#builtins-appendContext"
];
};
};
};
meta = {
platforms = lib.platforms.all;
};

View File

@@ -0,0 +1,9 @@
---
synopsis: Channel URLs migrated to channels.nixos.org subdomain
prs: [14518]
issues: [14517]
---
Channel URLs have been updated from `https://nixos.org/channels/` to `https://channels.nixos.org/` throughout Nix.
The subdomain provides better reliability with IPv6 support and improved CDN distribution. The old domain apex (`nixos.org/channels/`) currently redirects to the new location but may be deprecated in the future.

View File

@@ -0,0 +1,47 @@
---
synopsis: "JSON format changes for store path info and derivations"
prs: []
issues: []
---
JSON formats for store path info and derivations have been updated with new versions and structured fields.
## Store Path Info JSON (Version 2)
The store path info JSON format has been updated from version 1 to version 2:
- **Added `version` field**:
All store path info JSON now includes `"version": 2`.
- **Structured `ca` field**:
Content address is now a structured JSON object instead of a string:
- Old: `"ca": "fixed:r:sha256:1abc..."`
- New: `"ca": {"method": "nar", "hash": {"algorithm": "sha256", "format": "base64", "hash": "EMIJ+giQ..."}}`
- Still `null` values for input-addressed store objects
Version 1 format is still accepted when reading for backward compatibility.
**Affected command**: `nix path-info --json`
## Derivation JSON (Version 4)
The derivation JSON format has been updated from version 3 to version 4:
- **Restructured inputs**:
Inputs are now nested under an `inputs` object:
- Old: `"inputSrcs": [...], "inputDrvs": {...}`
- New: `"inputs": {"srcs": [...], "drvs": {...}}`
- **Consistent content addresses**:
Floating content-addressed outputs now use structured JSON format.
This is the same format as `ca` in in store path info (after the new version).
Version 3 and earlier formats are *not* accepted when reading.
**Affected command**: `nix derivation`, namely it's `show` and `add` sub-commands.

View File

@@ -0,0 +1,40 @@
---
synopsis: "Improved S3 binary cache support via HTTP"
prs: [13752, 13823, 14026, 14120, 14131, 14135, 14144, 14170, 14190, 14198, 14206, 14209, 14222, 14223, 14330, 14333, 14335, 14336, 14337, 14350, 14356, 14357, 14374, 14375, 14376, 14377, 14391, 14393, 14420, 14421]
issues: [13084, 12671, 11748, 12403]
---
S3 binary cache operations now happen via HTTP, leveraging `libcurl`'s native
AWS SigV4 authentication instead of the AWS C++ SDK, providing significant
improvements:
- **Reduced memory usage**: Eliminates memory buffering issues that caused
segfaults with large files
- **Fixed upload reliability**: Resolves AWS SDK chunking errors
(`InvalidChunkSizeError`)
- **Lighter dependencies**: Uses lightweight `aws-crt-cpp` instead of full
`aws-cpp-sdk`, reducing build complexity
The new implementation requires curl >= 7.75.0 and `aws-crt-cpp` for credential
management.
All existing S3 URL formats and parameters remain supported, however the store
settings for configuring multipart uploads have changed:
- **`multipart-upload`** (default: `false`): Enable multipart uploads for large
files. When enabled, files exceeding the multipart threshold will be uploaded
in multiple parts.
- **`multipart-threshold`** (default: `100 MiB`): Minimum file size for using
multipart uploads. Files smaller than this will use regular PUT requests.
Only takes effect when `multipart-upload` is enabled.
- **`multipart-chunk-size`** (default: `5 MiB`): Size of each part in multipart
uploads. Must be at least 5 MiB (AWS S3 requirement). Larger chunk sizes
reduce the number of requests but use more memory.
- **`buffer-size`**: Has been replaced by `multipart-chunk-size` and is now an alias to it.
Note that this change also means Nix now supports S3 binary cache stores even
if built without `aws-crt-cpp`, but only for public buckets which do not
require authentication.

View File

@@ -0,0 +1,14 @@
---
synopsis: "S3 URLs now support object versioning via versionId parameter"
prs: [14274]
issues: [13955]
---
S3 URLs now support a `versionId` query parameter to fetch specific versions
of objects from S3 buckets with versioning enabled. This allows pinning to
exact object versions for reproducibility and protection against unexpected
changes:
```
s3://bucket/key?region=us-east-1&versionId=abc123def456
```

View File

@@ -26,9 +26,12 @@
- [Derivation Outputs and Types of Derivations](store/derivation/outputs/index.md)
- [Content-addressing derivation outputs](store/derivation/outputs/content-address.md)
- [Input-addressing derivation outputs](store/derivation/outputs/input-address.md)
- [Build Trace](store/build-trace.md)
- [Derivation Resolution](store/resolution.md)
- [Building](store/building.md)
- [Store Types](store/types/index.md)
{{#include ./store/types/SUMMARY.md}}
- [Appendix: Math notation](store/math-notation.md)
- [Nix Language](language/index.md)
- [Data Types](language/types.md)
- [String context](language/string-context.md)
@@ -117,11 +120,17 @@
- [Architecture and Design](architecture/architecture.md)
- [Formats and Protocols](protocols/index.md)
- [JSON Formats](protocols/json/index.md)
- [Hash](protocols/json/hash.md)
- [Content Address](protocols/json/content-address.md)
- [Store Path](protocols/json/store-path.md)
- [Store Object Info](protocols/json/store-object-info.md)
- [Derivation](protocols/json/derivation.md)
- [Deriving Path](protocols/json/deriving-path.md)
- [Build Trace Entry](protocols/json/build-trace-entry.md)
- [Build Result](protocols/json/build-result.md)
- [Serving Tarball Flakes](protocols/tarball-fetcher.md)
- [Store Path Specification](protocols/store-path.md)
- [Nix Archive (NAR) Format](protocols/nix-archive.md)
- [Nix Archive (NAR) Format](protocols/nix-archive/index.md)
- [Derivation "ATerm" file format](protocols/derivation-aterm.md)
- [C API](c-api.md)
- [Glossary](glossary.md)

View File

@@ -11,10 +11,10 @@
Channels are a mechanism for referencing remote Nix expressions and conveniently retrieving their latest version.
The moving parts of channels are:
- The official channels listed at <https://nixos.org/channels>
- The official channels listed at <https://channels.nixos.org>
- The user-specific list of [subscribed channels](#subscribed-channels)
- The [downloaded channel contents](#channels)
- The [Nix expression search path](@docroot@/command-ref/conf-file.md#conf-nix-path), set with the [`-I` option](#opt-i) or the [`NIX_PATH` environment variable](#env-NIX_PATH)
- The [Nix expression search path](@docroot@/command-ref/conf-file.md#conf-nix-path), set with the [`-I` option](#opt-I) or the [`NIX_PATH` environment variable](#env-NIX_PATH)
> **Note**
>
@@ -88,9 +88,9 @@ This command has the following operations:
Subscribe to the Nixpkgs channel and run `hello` from the GNU Hello package:
```console
$ nix-channel --add https://nixos.org/channels/nixpkgs-unstable
$ nix-channel --add https://channels.nixos.org/nixpkgs-unstable
$ nix-channel --list
nixpkgs https://nixos.org/channels/nixpkgs
nixpkgs https://channels.nixos.org/nixpkgs
$ nix-channel --update
$ nix-shell -p hello --run hello
hello

View File

@@ -22,7 +22,7 @@ left untouched; this is not an error. It is also not an error if an
element of *args* matches no installed derivations.
For a description of how *args* is mapped to a set of store paths, see
[`--install`](#operation---install). If *args* describes multiple
[`--install`](./install.md). If *args* describes multiple
store paths with the same symbolic name, only the one with the highest
version is installed.

View File

@@ -19,7 +19,7 @@
This man page describes the command `nix-shell`, which is distinct from `nix
shell`. For documentation on the latter, run `nix shell --help` or see `man
nix3-shell`.
nix3-env-shell`.
# Description

View File

@@ -48,8 +48,7 @@ The behaviour of the collector is also influenced by the
configuration file.
By default, the collector prints the total number of freed bytes when it
finishes (or when it is interrupted). With `--print-dead`, it prints the
number of bytes that would be freed.
finishes (or when it is interrupted).
{{#include ./opt-common.md}}

View File

@@ -66,7 +66,7 @@ You can also build Nix for one of the [supported platforms](#platforms).
This section assumes you are using Nix with the [`flakes`] and [`nix-command`] experimental features enabled.
[`flakes`]: @docroot@/development/experimental-features.md#xp-feature-flakes
[`nix-command`]: @docroot@/development/experimental-features.md#xp-nix-command
[`nix-command`]: @docroot@/development/experimental-features.md#xp-feature-nix-command
To build all dependencies and start a shell in which all environment variables are set up so that those dependencies can be found:
@@ -256,7 +256,7 @@ You can use any of the other supported environments in place of `nix-cli-ccacheS
## Editor integration
The `clangd` LSP server is installed by default on the `clang`-based `devShell`s.
See [supported compilation environments](#compilation-environments) and instructions how to set up a shell [with flakes](#nix-with-flakes) or in [classic Nix](#classic-nix).
See [supported compilation environments](#compilation-environments) and instructions how to set up a shell [with flakes](#building-nix-with-flakes) or in [classic Nix](#building-nix).
To use the LSP with your editor, you will want a `compile_commands.json` file telling `clangd` how we are compiling the code.
Meson's configure always produces this inside the build directory.

View File

@@ -25,20 +25,31 @@ nix build .#nix-manual
and open `./result/share/doc/nix/manual/index.html`.
To build the manual incrementally, [enter the development shell](./building.md) and run:
To build the manual incrementally, [enter the development shell](./building.md) and configure with `doc-gen` enabled:
**If using interactive `nix develop`:**
```console
make manual-html-open -j $NIX_BUILD_CORES
$ nix develop
$ mesonFlags="$mesonFlags -Ddoc-gen=true" mesonConfigurePhase
```
In order to reflect changes to the [Makefile for the manual], clear all generated files before re-building:
[Makefile for the manual]: https://github.com/NixOS/nix/blob/master/doc/manual/local.mk
**If using direnv:**
```console
rm $(git ls-files doc/manual/ -o | grep -F '.md') && rmdir doc/manual/source/command-ref/new-cli && make manual-html -j $NIX_BUILD_CORES
$ direnv allow
$ bash -c 'source $stdenv/setup && mesonFlags="$mesonFlags -Ddoc-gen=true" mesonConfigurePhase'
```
Then build the manual:
```console
$ cd build
$ meson compile manual
```
The HTML manual will be generated at `build/src/nix-manual/manual/index.html`.
## Style guide
The goal of this style guide is to make it such that
@@ -229,3 +240,9 @@ $ configurePhase
$ ninja src/external-api-docs/html
$ xdg-open src/external-api-docs/html/index.html
```
If you use direnv, or otherwise want to run `configurePhase` in a transient shell, use:
```bash
nix-shell -A devShells.x86_64-linux.native-clangStdenv --command 'appendToVar mesonFlags "-Ddoc-gen=true"; mesonConfigurePhase'
```

View File

@@ -119,7 +119,7 @@ This will:
3. Stop the program when the test fails, allowing the user to then issue arbitrary commands to GDB.
### Characterisation testing { #characaterisation-testing-unit }
### Characterisation testing { #characterisation-testing-unit }
See [functional characterisation testing](#characterisation-testing-functional) for a broader discussion of characterisation testing.

View File

@@ -208,7 +208,7 @@
- [impure derivation]{#gloss-impure-derivation}
[An experimental feature](#@docroot@/development/experimental-features.md#xp-feature-impure-derivations) that allows derivations to be explicitly marked as impure,
[An experimental feature](@docroot@/development/experimental-features.md#xp-feature-impure-derivations) that allows derivations to be explicitly marked as impure,
so that they are always rebuilt, and their outputs not reused by subsequent calls to realise them.
- [Nix database]{#gloss-nix-database}
@@ -279,7 +279,7 @@
See [References](@docroot@/store/store-object.md#references) for details.
- [referrer]{#gloss-reference}
- [referrer]{#gloss-referrer}
A reversed edge from one [store object] to another.
@@ -367,8 +367,8 @@
Nix represents files as [file system objects][file system object], and how they belong together is encoded as [references][reference] between [store objects][store object] that contain these file system objects.
The [Nix language] allows denoting packages in terms of [attribute sets](@docroot@/language/types.md#attribute-set) containing:
- attributes that refer to the files of a package, typically in the form of [derivation outputs](#output),
The [Nix language] allows denoting packages in terms of [attribute sets](@docroot@/language/types.md#type-attrs) containing:
- attributes that refer to the files of a package, typically in the form of [derivation outputs](#gloss-output),
- attributes with metadata, such as information about how the package is supposed to be used.
The exact shape of these attribute sets is up to convention.
@@ -383,7 +383,7 @@
[string]: ./language/types.md#type-string
[path]: ./language/types.md#type-path
[attribute name]: ./language/types.md#attribute-set
[attribute name]: ./language/types.md#type-attrs
- [base directory]{#gloss-base-directory}

View File

@@ -3,19 +3,21 @@
To run the latest stable release of Nix with Docker run the following command:
```console
$ docker run -ti ghcr.io/nixos/nix
Unable to find image 'ghcr.io/nixos/nix:latest' locally
latest: Pulling from ghcr.io/nixos/nix
$ docker run -ti docker.io/nixos/nix
Unable to find image 'docker.io/nixos/nix:latest' locally
latest: Pulling from docker.io/nixos/nix
5843afab3874: Pull complete
b52bf13f109c: Pull complete
1e2415612aa3: Pull complete
Digest: sha256:27f6e7f60227e959ee7ece361f75d4844a40e1cc6878b6868fe30140420031ff
Status: Downloaded newer image for ghcr.io/nixos/nix:latest
Status: Downloaded newer image for docker.io/nixos/nix:latest
35ca4ada6e96:/# nix --version
nix (Nix) 2.3.12
35ca4ada6e96:/# exit
```
> If you want the latest pre-release you can use ghcr.io/nixos/nix and view them at https://github.com/nixos/nix/pkgs/container/nix
# What is included in Nix's Docker image?
The official Docker image is created using `pkgs.dockerTools.buildLayeredImage`

View File

@@ -333,7 +333,7 @@ Here is more information on the `output*` attributes, and what values they may b
`outputHashAlgo` can only be `null` when `outputHash` follows the SRI format, because in that case the choice of hash algorithm is determined by `outputHash`.
- [`outputHash`]{#adv-attr-outputHashAlgo}; [`outputHash`]{#adv-attr-outputHashMode}
- [`outputHash`]{#adv-attr-outputHash}
This will specify the output hash of the single output of a [fixed-output derivation].

View File

@@ -16,7 +16,7 @@ It outputs an attribute set, and produces a [store derivation] as a side effect
- [`name`]{#attr-name} ([String](@docroot@/language/types.md#type-string))
A symbolic name for the derivation.
See [derivation outputs](@docroot@/store/derivation/index.md#outputs) for what this is affects.
See [derivation outputs](@docroot@/store/derivation/outputs/index.md#outputs) for what this is affects.
[store path]: @docroot@/store/store-path.md

View File

@@ -16,7 +16,7 @@ An *identifier* is an [ASCII](https://en.wikipedia.org/wiki/ASCII) character seq
# Names
A *name* can be written as an [identifier](#identifier) or a [string literal](./string-literals.md).
A *name* can be written as an [identifier](#identifiers) or a [string literal](./string-literals.md).
> **Syntax**
>

View File

@@ -137,7 +137,7 @@ This is an incomplete overview of language features, by example.
</td>
<td>
[Booleans](@docroot@/language/types.md#type-boolean)
[Booleans](@docroot@/language/types.md#type-bool)
</td>
</tr>
@@ -245,7 +245,7 @@ This is an incomplete overview of language features, by example.
</td>
<td>
An [attribute set](@docroot@/language/types.md#attribute-set) with attributes named `x` and `y`
An [attribute set](@docroot@/language/types.md#type-attrs) with attributes named `x` and `y`
</td>
</tr>
@@ -285,7 +285,7 @@ This is an incomplete overview of language features, by example.
</td>
<td>
[Lists](@docroot@/language/types.md#list) with three elements.
[Lists](@docroot@/language/types.md#type-list) with three elements.
</td>
</tr>
@@ -369,7 +369,7 @@ This is an incomplete overview of language features, by example.
</td>
<td>
[Attribute selection](@docroot@/language/types.md#attribute-set) (evaluates to `1`)
[Attribute selection](@docroot@/language/types.md#type-attrs) (evaluates to `1`)
</td>
</tr>
@@ -381,7 +381,7 @@ This is an incomplete overview of language features, by example.
</td>
<td>
[Attribute selection](@docroot@/language/types.md#attribute-set) with default (evaluates to `3`)
[Attribute selection](@docroot@/language/types.md#type-attrs) with default (evaluates to `3`)
</td>
</tr>

View File

@@ -111,7 +111,7 @@ It creates an [attribute set] representing the string context, which can be insp
[`builtins.hasContext`]: ./builtins.md#builtins-hasContext
[`builtins.getContext`]: ./builtins.md#builtins-getContext
[attribute set]: ./types.md#attribute-set
[attribute set]: ./types.md#type-attrs
## Clearing string contexts

View File

@@ -6,7 +6,7 @@ Such a construct is called *interpolated string*, and the expression inside is a
[string]: ./types.md#type-string
[path]: ./types.md#type-path
[attribute set]: ./types.md#attribute-set
[attribute set]: ./types.md#type-attrs
> **Syntax**
>

View File

@@ -51,7 +51,7 @@ See [String literals](string-literals.md).
Path literals can also include [string interpolation], besides being [interpolated into other expressions].
[interpolated into other expressions]: ./string-interpolation.md#interpolated-expressions
[interpolated into other expressions]: ./string-interpolation.md#interpolated-expression
At least one slash (`/`) must appear *before* any interpolated expression for the result to be recognized as a path.
@@ -235,7 +235,7 @@ of object-oriented programming, for example.
## Recursive sets
Recursive sets are like normal [attribute sets](./types.md#attribute-set), but the attributes can refer to each other.
Recursive sets are like normal [attribute sets](./types.md#type-attrs), but the attributes can refer to each other.
> *rec-attrset* = `rec {` [ *name* `=` *expr* `;` `]`... `}`
@@ -287,7 +287,7 @@ This evaluates to `"foobar"`.
## Inheriting attributes
When defining an [attribute set](./types.md#attribute-set) or in a [let-expression](#let-expressions) it is often convenient to copy variables from the surrounding lexical scope (e.g., when you want to propagate attributes).
When defining an [attribute set](./types.md#type-attrs) or in a [let-expression](#let-expressions) it is often convenient to copy variables from the surrounding lexical scope (e.g., when you want to propagate attributes).
This can be shortened using the `inherit` keyword.
Example:

View File

@@ -1,3 +1,6 @@
# Process JSON schema documentation
subdir('protocols')
summary_rl_next = custom_target(
command : [
bash,

View File

@@ -1,6 +1,8 @@
# Derivation "ATerm" file format
For historical reasons, [store derivations][store derivation] are stored on-disk in [ATerm](https://homepages.cwi.nl/~daybuild/daily-books/technology/aterm-guide/aterm-guide.html) format.
For historical reasons, [store derivations][store derivation] are stored on-disk in "Annotated Term" (ATerm) format
([guide](https://homepages.cwi.nl/~daybuild/daily-books/technology/aterm-guide/aterm-guide.html),
[paper](https://doi.org/10.1002/(SICI)1097-024X(200003)30:3%3C259::AID-SPE298%3E3.0.CO;2-Y)).
## The ATerm format used

View File

@@ -0,0 +1,21 @@
{{#include build-result-v1-fixed.md}}
## Examples
### Successful build
```json
{{#include schema/build-result-v1/success.json}}
```
### Failed build (output rejected)
```json
{{#include schema/build-result-v1/output-rejected.json}}
```
### Failed build (non-deterministic)
```json
{{#include schema/build-result-v1/not-deterministic.json}}
```

View File

@@ -0,0 +1,27 @@
{{#include build-trace-entry-v1-fixed.md}}
## Examples
### Simple build trace entry
```json
{{#include schema/build-trace-entry-v1/simple.json}}
```
### Build trace entry with dependencies
```json
{{#include schema/build-trace-entry-v1/with-dependent-realisations.json}}
```
### Build trace entry with signature
```json
{{#include schema/build-trace-entry-v1/with-signature.json}}
```
<!--
## Raw Schema
[JSON Schema for Build Trace Entry v1](schema/build-trace-entry-v1.json)
-->

View File

@@ -0,0 +1,21 @@
{{#include content-address-v1-fixed.md}}
## Examples
### [Text](@docroot@/store/store-object/content-address.html#method-text) method
```json
{{#include schema/content-address-v1/text.json}}
```
### [Nix Archive](@docroot@/store/store-object/content-address.html#method-nix-archive) method
```json
{{#include schema/content-address-v1/nar.json}}
```
<!-- need to convert YAML to JSON first
## Raw Schema
[JSON Schema for Hash v1](schema/content-address-v1.json)
-->

View File

@@ -1,120 +1,7 @@
# Derivation JSON Format
{{#include derivation-v4-fixed.md}}
> **Warning**
>
> This JSON format is currently
> [**experimental**](@docroot@/development/experimental-features.md#xp-feature-nix-command)
> and subject to change.
<!-- need to convert YAML to JSON first
## Raw Schema
The JSON serialization of a
[derivations](@docroot@/glossary.md#gloss-store-derivation)
is a JSON object with the following fields:
* `name`:
The name of the derivation.
This is used when calculating the store paths of the derivation's outputs.
* `version`:
Must be `3`.
This is a guard that allows us to continue evolving this format.
The choice of `3` is fairly arbitrary, but corresponds to this informal version:
- Version 0: A-Term format
- Version 1: Original JSON format, with ugly `"r:sha256"` inherited from A-Term format.
- Version 2: Separate `method` and `hashAlgo` fields in output specs
- Verison 3: Drop store dir from store paths, just include base name.
Note that while this format is experimental, the maintenance of versions is best-effort, and not promised to identify every change.
* `outputs`:
Information about the output paths of the derivation.
This is a JSON object with one member per output, where the key is the output name and the value is a JSON object with these fields:
* `path`:
The output path, if it is known in advanced.
Otherwise, `null`.
* `method`:
For an output which will be [content addressed], a string representing the [method](@docroot@/store/store-object/content-address.md) of content addressing that is chosen.
Valid method strings are:
- [`flat`](@docroot@/store/store-object/content-address.md#method-flat)
- [`nar`](@docroot@/store/store-object/content-address.md#method-nix-archive)
- [`text`](@docroot@/store/store-object/content-address.md#method-text)
- [`git`](@docroot@/store/store-object/content-address.md#method-git)
Otherwise, `null`.
* `hashAlgo`:
For an output which will be [content addressed], the name of the hash algorithm used.
Valid algorithm strings are:
- `blake3`
- `md5`
- `sha1`
- `sha256`
- `sha512`
* `hash`:
For fixed-output derivations, the expected content hash in base-16.
> **Example**
>
> ```json
> "outputs": {
> "out": {
> "method": "nar",
> "hashAlgo": "sha256",
> "hash": "6fc80dcc62179dbc12fc0b5881275898f93444833d21b89dfe5f7fbcbb1d0d62"
> }
> }
> ```
* `inputSrcs`:
A list of store paths on which this derivation depends.
> **Example**
>
> ```json
> "inputSrcs": [
> "47y241wqdhac3jm5l7nv0x4975mb1975-separate-debug-info.sh",
> "56d0w71pjj9bdr363ym3wj1zkwyqq97j-fix-pop-var-context-error.patch"
> ]
> ```
* `inputDrvs`:
A JSON object specifying the derivations on which this derivation depends, and what outputs of those derivations.
> **Example**
>
> ```json
> "inputDrvs": {
> "6lkh5yi7nlb7l6dr8fljlli5zfd9hq58-curl-7.73.0.drv": ["dev"],
> "fn3kgnfzl5dzym26j8g907gq3kbm8bfh-unzip-6.0.drv": ["out"]
> }
> ```
specifies that this derivation depends on the `dev` output of `curl`, and the `out` output of `unzip`.
* `system`:
The system type on which this derivation is to be built
(e.g. `x86_64-linux`).
* `builder`:
The absolute path of the program to be executed to run the build.
Typically this is the `bash` shell
(e.g. `/nix/store/r3j288vpmczbl500w6zz89gyfa4nr0b1-bash-4.4-p23/bin/bash`).
* `args`:
The command-line arguments passed to the `builder`.
* `env`:
The environment passed to the `builder`.
* `structuredAttrs`:
[Strucutured Attributes](@docroot@/store/derivation/index.md#structured-attrs), only defined if the derivation contains them.
Structured attributes are JSON, and thus embedded as-is.
[JSON Schema for Derivation v3](schema/derivation-v4.json)
-->

View File

@@ -0,0 +1,21 @@
{{#include deriving-path-v1-fixed.md}}
## Examples
### Constant
```json
{{#include schema/deriving-path-v1/single_opaque.json}}
```
### Output of static derivation
```json
{{#include schema/deriving-path-v1/single_built.json}}
```
### Output of dynamic derivation
```json
{{#include schema/deriving-path-v1/single_built_built.json}}
```

View File

@@ -0,0 +1,17 @@
# For some reason, backticks in the JSON schema are being escaped rather
# than being kept as intentional code spans. This removes all backtick
# escaping, which is an ugly solution, but one that is fine, because we
# are not using backticks for any other purpose.
s/\\`/`/g
# The way that semi-external references are rendered (i.e. ones to
# sibling schema files, as opposed to separate website ones, is not nice
# for humans. Replace it with a nice relative link within the manual
# instead.
#
# As we have more such relative links, more replacements of this nature
# should appear below.
s^\(./hash-v1.yaml\)\?#/$defs/algorithm^[JSON format for `Hash`](./hash.html#algorithm)^g
s^\(./hash-v1.yaml\)^[JSON format for `Hash`](./hash.html)^g
s^\(./content-address-v1.yaml\)\?#/$defs/method^[JSON format for `ContentAddress`](./content-address.html#method)^g
s^\(./content-address-v1.yaml\)^[JSON format for `ContentAddress`](./content-address.html)^g

View File

@@ -0,0 +1,33 @@
{{#include hash-v1-fixed.md}}
## Examples
### SHA-256 with Base64 encoding
```json
{{#include schema/hash-v1/sha256-base64.json}}
```
### SHA-256 with Base16 (hexadecimal) encoding
```json
{{#include schema/hash-v1/sha256-base16.json}}
```
### SHA-256 with Nix32 encoding
```json
{{#include schema/hash-v1/sha256-nix32.json}}
```
### BLAKE3 with Base64 encoding
```json
{{#include schema/hash-v1/blake3-base64.json}}
```
<!-- need to convert YAML to JSON first
## Raw Schema
[JSON Schema for Hash v1](schema/hash-v1.json)
-->

View File

@@ -0,0 +1,17 @@
# Configuration file for json-schema-for-humans
#
# https://github.com/coveooss/json-schema-for-humans/blob/main/docs/examples/examples_md_default/Configuration.md
template_name: md
show_toc: true
# impure timestamp and distracting
with_footer: false
recursive_detection_depth: 3
show_breadcrumbs: false
description_is_markdown: true
template_md_options:
properties_table_columns:
- Property
- Type
- Pattern
- Title/Description

View File

@@ -0,0 +1,80 @@
# Tests in: ../../../../src/json-schema-checks
fs = import('fs')
# Find json-schema-for-humans if available
json_schema_for_humans = find_program('generate-schema-doc', required : false)
# Configuration for json-schema-for-humans
json_schema_config = files('json-schema-for-humans-config.yaml')
schemas = [
'hash-v1',
'content-address-v1',
'store-path-v1',
'store-object-info-v2',
'derivation-v4',
'deriving-path-v1',
'build-trace-entry-v1',
'build-result-v1',
]
schema_files = files()
foreach schema_name : schemas
schema_files += files('schema' / schema_name + '.yaml')
endforeach
schema_outputs = []
foreach schema_name : schemas
schema_outputs += schema_name + '.md'
endforeach
json_schema_generated_files = []
# Generate markdown documentation from JSON schema
# Note: output must be just a filename, not a path
gen_file = custom_target(
schema_name + '-schema-docs.tmp',
command : [
json_schema_for_humans,
'--config-file',
json_schema_config,
meson.current_source_dir() / 'schema',
meson.current_build_dir(),
],
input : schema_files + [
json_schema_config,
],
output : schema_outputs,
capture : false,
build_by_default : true,
)
idx = 0
if json_schema_for_humans.found()
foreach schema_name : schemas
#schema_file = 'schema' / schema_name + '.yaml'
# There is one so-so hack, and one horrible hack being done here.
sedded_file = custom_target(
schema_name + '-schema-docs',
command : [
'sed',
'-f',
# Out of line to avoid https://github.com/mesonbuild/meson/issues/1564
files('fixup-json-schema-generated-doc.sed'),
'@INPUT@',
],
capture : true,
input : gen_file[idx],
output : schema_name + '-fixed.md',
)
idx += 1
json_schema_generated_files += [ sedded_file ]
endforeach
else
warning(
'json-schema-for-humans not found, skipping JSON schema documentation generation',
)
endif

View File

@@ -0,0 +1 @@
../../../../../../src/libstore-tests/data/build-result

View File

@@ -0,0 +1,136 @@
"$schema": "http://json-schema.org/draft-04/schema"
"$id": "https://nix.dev/manual/nix/latest/protocols/json/schema/build-result-v1.json"
title: Build Result
description: |
This schema describes the JSON representation of Nix's `BuildResult` type, which represents the result of building a derivation or substituting store paths.
Build results can represent either successful builds (with built outputs) or various types of failures.
oneOf:
- "$ref": "#/$defs/success"
- "$ref": "#/$defs/failure"
type: object
required:
- success
- status
properties:
timesBuilt:
type: integer
minimum: 0
title: Times built
description: |
How many times this build was performed.
startTime:
type: integer
minimum: 0
title: Start time
description: |
The start time of the build (or one of the rounds, if it was repeated), as a Unix timestamp.
stopTime:
type: integer
minimum: 0
title: Stop time
description: |
The stop time of the build (or one of the rounds, if it was repeated), as a Unix timestamp.
cpuUser:
type: integer
minimum: 0
title: User CPU time
description: |
User CPU time the build took, in microseconds.
cpuSystem:
type: integer
minimum: 0
title: System CPU time
description: |
System CPU time the build took, in microseconds.
"$defs":
success:
type: object
title: Successful Build Result
description: |
Represents a successful build with built outputs.
required:
- success
- status
- builtOutputs
properties:
success:
const: true
title: Success indicator
description: |
Always true for successful build results.
status:
type: string
title: Success status
description: |
Status string for successful builds.
enum:
- "Built"
- "Substituted"
- "AlreadyValid"
- "ResolvesToAlreadyValid"
builtOutputs:
type: object
title: Built outputs
description: |
A mapping from output names to their build trace entries.
additionalProperties:
"$ref": "build-trace-entry-v1.yaml"
failure:
type: object
title: Failed Build Result
description: |
Represents a failed build with error information.
required:
- success
- status
- errorMsg
properties:
success:
const: false
title: Success indicator
description: |
Always false for failed build results.
status:
type: string
title: Failure status
description: |
Status string for failed builds.
enum:
- "PermanentFailure"
- "InputRejected"
- "OutputRejected"
- "TransientFailure"
- "CachedFailure"
- "TimedOut"
- "MiscFailure"
- "DependencyFailed"
- "LogLimitExceeded"
- "NotDeterministic"
- "NoSubstituters"
- "HashMismatch"
errorMsg:
type: string
title: Error message
description: |
Information about the error if the build failed.
isNonDeterministic:
type: boolean
title: Non-deterministic flag
description: |
If timesBuilt > 1, whether some builds did not produce the same result.
Note that 'isNonDeterministic = false' does not mean the build is deterministic,
just that we don't have evidence of non-determinism.

View File

@@ -0,0 +1 @@
../../../../../../src/libstore-tests/data/realisation

View File

@@ -0,0 +1,74 @@
"$schema": "http://json-schema.org/draft-04/schema"
"$id": "https://nix.dev/manual/nix/latest/protocols/json/schema/build-trace-entry-v1.json"
title: Build Trace Entry
description: |
A record of a successful build outcome for a specific derivation output.
This schema describes the JSON representation of a [build trace entry](@docroot@/store/build-trace.md) entry.
> **Warning**
>
> This JSON format is currently
> [**experimental**](@docroot@/development/experimental-features.md#xp-feature-ca-derivations)
> and subject to change.
type: object
required:
- id
- outPath
- dependentRealisations
- signatures
properties:
id:
type: string
title: Derivation Output ID
pattern: "^sha256:[0-9a-f]{64}![a-zA-Z_][a-zA-Z0-9_-]*$"
description: |
Unique identifier for the derivation output that was built.
Format: `{hash-quotient-drv}!{output-name}`
- **hash-quotient-drv**: SHA-256 [hash of the quotient derivation](@docroot@/store/derivation/outputs/input-address.md#hash-quotient-drv).
Begins with `sha256:`.
- **output-name**: Name of the specific output (e.g., "out", "dev", "doc")
Example: `"sha256:ba7816bf8f01cfea414140de5dae2223b00361a396177a9cb410ff61f20015ad!foo"`
outPath:
"$ref": "store-path-v1.yaml"
title: Output Store Path
description: |
The path to the store object that resulted from building this derivation for the given output name.
dependentRealisations:
type: object
title: Underlying Base Build Trace
description: |
This is for [*derived*](@docroot@/store/build-trace.md#derived) build trace entries to ensure coherence.
Keys are derivation output IDs (same format as the main `id` field).
Values are the store paths that those dependencies resolved to.
As described in the linked section on derived build trace traces, derived build trace entries must be kept in addition and not instead of the underlying base build entries.
This is the set of base build trace entries that this derived build trace is derived from.
(The set is also a map since this miniature base build trace must be coherent, mapping each key to a single value.)
patternProperties:
"^sha256:[0-9a-f]{64}![a-zA-Z_][a-zA-Z0-9_-]*$":
$ref: "store-path-v1.yaml"
title: Dependent Store Path
description: Store path that this dependency resolved to during the build
additionalProperties: false
signatures:
type: array
title: Build Signatures
description: |
A set of cryptographic signatures attesting to the authenticity of this build trace entry.
items:
type: string
title: Signature
description: A single cryptographic signature
additionalProperties: false

View File

@@ -0,0 +1 @@
../../../../../../src/libstore-tests/data/content-address

View File

@@ -0,0 +1,55 @@
"$schema": "http://json-schema.org/draft-04/schema"
"$id": "https://nix.dev/manual/nix/latest/protocols/json/schema/content-address-v1.json"
title: Content Address
description: |
This schema describes the JSON representation of Nix's `ContentAddress` type, which conveys information about [content-addressing store objects](@docroot@/store/store-object/content-address.md).
> **Note**
>
> For current methods of content addressing, this data type is a bit suspicious, because it is neither simply a content address of a file system object (the `method` is richer), nor simply a content address of a store object (the `hash` doesn't account for the references).
> It should thus only be used in contexts where the references are also known / otherwise made tamper-resistant.
<!--
TODO currently `ContentAddress` is used in both of these, and so same rationale applies, but actually in both cases the JSON is currently ad-hoc.
That will be fixed, and as each is fixed, the example (along with a more precise link to the field in question) should be become part of the above note, so what is is saying is more clear.
> For example:
> - Fixed outputs of derivations are not allowed to have any references, so an empty reference set is statically known by assumption.
> - [Store object info](./store-object-info.md) includes the set of references along side the (optional) content address.
> This data type is thus safely used in both of these contexts.
-->
type: object
properties:
method:
"$ref": "#/$defs/method"
hash:
title: Content Address
description: |
This would be the content-address itself.
For all current methods, this is just a content address of the file system object of the store object, [as described in the store chapter](@docroot@/store/file-system-object/content-address.md), and not of the store object as a whole.
In particular, the references of the store object are *not* taken into account with this hash (and currently-supported methods).
"$ref": "./hash-v1.yaml"
required:
- method
- hash
additionalProperties: false
"$defs":
method:
type: string
enum: [flat, nar, text, git]
title: Content-Addressing Method
description: |
A string representing the [method](@docroot@/store/store-object/content-address.md) of content addressing that is chosen.
Valid method strings are:
- [`flat`](@docroot@/store/store-object/content-address.md#method-flat) (provided the contents are a single file)
- [`nar`](@docroot@/store/store-object/content-address.md#method-nix-archive)
- [`text`](@docroot@/store/store-object/content-address.md#method-text)
- [`git`](@docroot@/store/store-object/content-address.md#method-git)

View File

@@ -0,0 +1,299 @@
"$schema": "http://json-schema.org/draft-04/schema"
"$id": "https://nix.dev/manual/nix/latest/protocols/json/schema/derivation-v4.json"
title: Derivation
description: |
Experimental JSON representation of a Nix derivation (version 4).
This schema describes the JSON representation of Nix's `Derivation` type.
> **Warning**
>
> This JSON format is currently
> [**experimental**](@docroot@/development/experimental-features.md#xp-feature-nix-command)
> and subject to change.
type: object
required:
- name
- version
- outputs
- inputs
- system
- builder
- args
- env
properties:
name:
type: string
title: Derivation name
description: |
The name of the derivation.
Used when calculating store paths for the derivations outputs.
version:
const: 4
title: Format version (must be 4)
description: |
Must be `4`.
This is a guard that allows us to continue evolving this format.
The choice of `3` is fairly arbitrary, but corresponds to this informal version:
- Version 0: ATerm format
- Version 1: Original JSON format, with ugly `"r:sha256"` inherited from ATerm format.
- Version 2: Separate `method` and `hashAlgo` fields in output specs
- Version 3: Drop store dir from store paths, just include base name.
- Version 4: Two cleanups, batched together to lesson churn:
- Reorganize inputs into nested structure (`inputs.srcs` and `inputs.drvs`)
- Use canonical content address JSON format for floating content addressed derivation outputs.
Note that while this format is experimental, the maintenance of versions is best-effort, and not promised to identify every change.
outputs:
type: object
title: Output specifications
description: |
Information about the output paths of the derivation.
This is a JSON object with one member per output, where the key is the output name and the value is a JSON object as described.
> **Example**
>
> ```json
> "outputs": {
> "out": {
> "method": "nar",
> "hashAlgo": "sha256",
> "hash": "6fc80dcc62179dbc12fc0b5881275898f93444833d21b89dfe5f7fbcbb1d0d62"
> }
> }
> ```
additionalProperties:
"$ref": "#/$defs/output/overall"
inputs:
type: object
title: Derivation inputs
description: |
Input dependencies for the derivation, organized into source paths and derivation dependencies.
required:
- srcs
- drvs
properties:
srcs:
type: array
title: Input source paths
description: |
List of store paths on which this derivation depends.
> **Example**
>
> ```json
> "srcs": [
> "47y241wqdhac3jm5l7nv0x4975mb1975-separate-debug-info.sh",
> "56d0w71pjj9bdr363ym3wj1zkwyqq97j-fix-pop-var-context-error.patch"
> ]
> ```
items:
$ref: "store-path-v1.yaml"
drvs:
type: object
title: Input derivations
description: |
Mapping of derivation paths to lists of output names they provide.
> **Example**
>
> ```json
> "drvs": {
> "6lkh5yi7nlb7l6dr8fljlli5zfd9hq58-curl-7.73.0.drv": ["dev"],
> "fn3kgnfzl5dzym26j8g907gq3kbm8bfh-unzip-6.0.drv": ["out"]
> }
> ```
>
> specifies that this derivation depends on the `dev` output of `curl`, and the `out` output of `unzip`.
patternProperties:
"^[0123456789abcdfghijklmnpqrsvwxyz]{32}-.+\\.drv$":
title: Store Path
description: |
A store path to a derivation, mapped to the outputs of that derivation.
oneOf:
- "$ref": "#/$defs/outputNames"
- "$ref": "#/$defs/dynamicOutputs"
additionalProperties: false
additionalProperties: false
system:
type: string
title: Build system type
description: |
The system type on which this derivation is to be built
(e.g. `x86_64-linux`).
builder:
type: string
title: Build program path
description: |
Absolute path of the program used to perform the build.
Typically this is the `bash` shell
(e.g. `/nix/store/r3j288vpmczbl500w6zz89gyfa4nr0b1-bash-4.4-p23/bin/bash`).
args:
type: array
title: Builder arguments
description: |
Command-line arguments passed to the `builder`.
items:
type: string
env:
type: object
title: Environment variables
description: |
Environment variables passed to the `builder`.
additionalProperties:
type: string
structuredAttrs:
title: Structured attributes
description: |
[Structured Attributes](@docroot@/store/derivation/index.md#structured-attrs), only defined if the derivation contains them.
Structured attributes are JSON, and thus embedded as-is.
type: object
additionalProperties: true
"$defs":
output:
overall:
title: Derivation Output
description: |
A single output of a derivation, with different variants for different output types.
oneOf:
- "$ref": "#/$defs/output/inputAddressed"
- "$ref": "#/$defs/output/caFixed"
- "$ref": "#/$defs/output/caFloating"
- "$ref": "#/$defs/output/deferred"
- "$ref": "#/$defs/output/impure"
inputAddressed:
title: Input-Addressed Output
description: |
The traditional non-fixed-output derivation type.
The output path is determined from the derivation itself.
See [Input-addressing derivation outputs](@docroot@/store/derivation/outputs/input-address.md) for more details.
type: object
required:
- path
properties:
path:
$ref: "store-path-v1.yaml"
title: Output path
description: |
The output path determined from the derivation itself.
additionalProperties: false
caFixed:
title: Fixed Content-Addressed Output
description: |
The output is content-addressed, and the content-address is fixed in advance.
See [Fixed-output content-addressing](@docroot@/store/derivation/outputs/content-address.md#fixed) for more details.
"$ref": "./content-address-v1.yaml"
required:
- method
- hash
properties:
method:
description: |
Method of content addressing used for this output.
hash:
title: Expected hash value
description: |
The expected content hash.
additionalProperties: false
caFloating:
title: Floating Content-Addressed Output
description: |
Floating-output derivations, whose outputs are content
addressed, but not fixed, and so the output paths are dynamically calculated from
whatever the output ends up being.
See [Floating Content-Addressing](@docroot@/store/derivation/outputs/content-address.md#floating) for more details.
type: object
required:
- method
- hashAlgo
properties:
method:
"$ref": "./content-address-v1.yaml#/$defs/method"
description: |
Method of content addressing used for this output.
hashAlgo:
title: Hash algorithm
"$ref": "./hash-v1.yaml#/$defs/algorithm"
description: |
What hash algorithm to use for the given method of content-addressing.
additionalProperties: false
deferred:
title: Deferred Output
description: |
Input-addressed output which depends on a (CA) derivation whose outputs (and thus their content-address
are not yet known.
type: object
properties: {}
additionalProperties: false
impure:
title: Impure Output
description: |
Impure output which is just like a floating content-addressed output, but this derivation runs without sandboxing.
As such, we don't record it in the build trace, under the assumption that if we need it again, we should rebuild it, as it might produce something different.
required:
- impure
- method
- hashAlgo
properties:
impure:
const: true
method:
"$ref": "./content-address-v1.yaml#/$defs/method"
description: |
How the file system objects will be serialized for hashing.
hashAlgo:
title: Hash algorithm
"$ref": "./hash-v1.yaml#/$defs/algorithm"
description: |
How the serialization will be hashed.
additionalProperties: false
outputName:
type: string
title: Output name
description: Name of the derivation output to depend on
outputNames:
type: array
title: Output Names
description: Set of names of derivation outputs to depend on
items:
"$ref": "#/$defs/outputName"
dynamicOutputs:
type: object
title: Dynamic Outputs
description: |
**Experimental feature**: [`dynamic-derivations`](@docroot@/development/experimental-features.md#xp-feature-dynamic-derivations)
This recursive data type allows for depending on outputs of outputs.
properties:
outputs:
"$ref": "#/$defs/outputNames"
dynamicOutputs:
"$ref": "#/$defs/dynamicOutputs"

View File

@@ -0,0 +1 @@
../../../../../../src/libstore-tests/data/derived-path

View File

@@ -0,0 +1,27 @@
"$schema": "http://json-schema.org/draft-04/schema"
"$id": "https://nix.dev/manual/nix/latest/protocols/json/schema/deriving-path-v1.json"
title: Deriving Path
description: |
This schema describes the JSON representation of Nix's [Deriving Path](@docroot@/store/derivation/index.md#deriving-path).
oneOf:
- title: Constant
description: |
See [Constant](@docroot@/store/derivation/index.md#deriving-path-constant) deriving path.
$ref: "store-path-v1.yaml"
- title: Output
description: |
See [Output](@docroot@/store/derivation/index.md#deriving-path-output) deriving path.
type: object
properties:
drvPath:
"$ref": "#"
description: |
A deriving path to a [Derivation](@docroot@/store/derivation/index.md#store-derivation), whose output is being referred to.
output:
type: string
description: |
The name of an output produced by that derivation (e.g. "out", "doc", etc.).
required:
- drvPath
- output
additionalProperties: false

View File

@@ -0,0 +1 @@
../../../../../../src/libutil-tests/data/hash/

View File

@@ -0,0 +1,54 @@
"$schema": "http://json-schema.org/draft-04/schema"
"$id": "https://nix.dev/manual/nix/latest/protocols/json/schema/hash-v1.json"
title: Hash
description: |
A cryptographic hash value used throughout Nix for content addressing and integrity verification.
This schema describes the JSON representation of Nix's `Hash` type.
type: object
properties:
algorithm:
"$ref": "#/$defs/algorithm"
format:
type: string
enum:
- base64
- nix32
- base16
- sri
title: Hash format
description: |
The encoding format of the hash value.
- `base64` uses standard Base64 encoding [RFC 4648, section 4](https://datatracker.ietf.org/doc/html/rfc4648#section-4)
- `nix32` is Nix-specific base-32 encoding
- `base16` is lowercase hexadecimal
- `sri` is the [Subresource Integrity format](https://developer.mozilla.org/en-US/docs/Web/Security/Subresource_Integrity).
hash:
type: string
title: Hash
description: |
The encoded hash value, itself.
It is specified in the format specified by the `format` field.
It must be the right length for the hash algorithm specified in the `algorithm` field, also.
The hash value does not include any algorithm prefix.
required:
- algorithm
- format
- hash
additionalProperties: false
"$defs":
algorithm:
type: string
enum:
- blake3
- md5
- sha1
- sha256
- sha512
title: Hash algorithm
description: |
The hash algorithm used to compute the hash value.
`blake3` is currently experimental and requires the [`blake-hashing`](@docroot@/development/experimental-features.md#xp-feature-blake3-hashes) experimental feature.

View File

@@ -0,0 +1 @@
../../../../../../src/libstore-tests/data/nar-info

View File

@@ -0,0 +1 @@
../../../../../../src/libstore-tests/data/path-info

View File

@@ -0,0 +1,258 @@
"$schema": "http://json-schema.org/draft-04/schema"
"$id": "https://nix.dev/manual/nix/latest/protocols/json/schema/store-object-info-v2.json"
title: Store Object Info v2
description: |
Information about a [store object](@docroot@/store/store-object.md).
This schema describes the JSON representation of store object metadata as returned by commands like [`nix path-info --json`](@docroot@/command-ref/new-cli/nix3-path-info.md).
> **Warning**
>
> This JSON format is currently
> [**experimental**](@docroot@/development/experimental-features.md#xp-feature-nix-command)
> and subject to change.
### Field Categories
Store object information can come in a few different variations.
Firstly, "impure" fields, which contain non-intrinsic information about the store object, may or may not be included.
Second, binary cache stores have extra non-intrinsic infomation about the store objects they contain.
Thirdly, [`nix path-info --json --closure-size`](@docroot@/command-ref/new-cli/nix3-path-info.html#opt-closure-size) can compute some extra information about not just the single store object in question, but the store object and its [closure](@docroot@/glossary.md#gloss-closure).
The impure and NAR fields are grouped into separate variants below.
See their descriptions for additional information.
The closure fields however as just included as optional fields, to avoid a combinatorial explosion of variants.
oneOf:
- $ref: "#/$defs/base"
- $ref: "#/$defs/impure"
- $ref: "#/$defs/narInfo"
$defs:
base:
title: Store Object Info
description: |
Basic store object metadata containing only intrinsic properties.
This is the minimal set of fields that describe what a store object contains.
type: object
required:
- version
- narHash
- narSize
- references
- ca
properties:
version:
type: integer
const: 2
title: Format version (must be 2)
description: |
Must be `2`.
This is a guard that allows us to continue evolving this format.
Here is the rough version history:
- Version 0: `.narinfo` line-oriented format
- Version 1: Original JSON format, with ugly `"r:sha256"` inherited from `.narinfo` format.
- Version 2: Use structured JSON type for `ca`
path:
type: string
title: Store Path
description: |
[Store path](@docroot@/store/store-path.md) to the given store object.
Note: This field may not be present in all contexts, such as when the path is used as the key and the the store object info the value in map.
narHash:
type: string
title: NAR Hash
description: |
Hash of the [file system object](@docroot@/store/file-system-object.md) part of the store object when serialized as a [Nix Archive](@docroot@/store/file-system-object/content-address.md#serial-nix-archive).
narSize:
type: integer
minimum: 0
title: NAR Size
description: |
Size of the [file system object](@docroot@/store/file-system-object.md) part of the store object when serialized as a [Nix Archive](@docroot@/store/file-system-object/content-address.md#serial-nix-archive).
references:
type: array
title: References
description: |
An array of [store paths](@docroot@/store/store-path.md), possibly including this one.
items:
type: string
ca:
oneOf:
- type: "null"
const: null
- "$ref": "./content-address-v1.yaml"
title: Content Address
description: |
If the store object is [content-addressed](@docroot@/store/store-object/content-address.md),
this is the content address of this store object's file system object, used to compute its store path.
Otherwise (i.e. if it is [input-addressed](@docroot@/glossary.md#gloss-input-addressed-store-object)), this is `null`.
additionalProperties: false
impure:
title: Store Object Info with Impure Fields
description: |
Store object metadata including impure fields that are not *intrinsic* properties.
In other words, the same store object in different stores could have different values for these impure fields.
type: object
required:
- version
- narHash
- narSize
- references
- ca
# impure
- deriver
- registrationTime
- ultimate
- signatures
properties:
version: { $ref: "#/$defs/base/properties/version" }
path: { $ref: "#/$defs/base/properties/path" }
narHash: { $ref: "#/$defs/base/properties/narHash" }
narSize: { $ref: "#/$defs/base/properties/narSize" }
references: { $ref: "#/$defs/base/properties/references" }
ca: { $ref: "#/$defs/base/properties/ca" }
deriver:
type: ["string", "null"]
title: Deriver
description: |
If known, the path to the [store derivation](@docroot@/glossary.md#gloss-store-derivation) from which this store object was produced.
Otherwise `null`.
> This is an "impure" field that may not be included in certain contexts.
registrationTime:
type: ["integer", "null"]
title: Registration Time
description: |
If known, when this derivation was added to the store (Unix timestamp).
Otherwise `null`.
> This is an "impure" field that may not be included in certain contexts.
ultimate:
type: boolean
title: Ultimate
description: |
Whether this store object is trusted because we built it ourselves, rather than substituted a build product from elsewhere.
> This is an "impure" field that may not be included in certain contexts.
signatures:
type: array
title: Signatures
description: |
Signatures claiming that this store object is what it claims to be.
Not relevant for [content-addressed](@docroot@/store/store-object/content-address.md) store objects,
but useful for [input-addressed](@docroot@/glossary.md#gloss-input-addressed-store-object) store objects.
> This is an "impure" field that may not be included in certain contexts.
items:
type: string
# Computed closure fields
closureSize:
type: integer
minimum: 0
title: Closure Size
description: |
The total size of this store object and every other object in its [closure](@docroot@/glossary.md#gloss-closure).
> This field is not stored at all, but computed by traversing the other fields across all the store objects in a closure.
additionalProperties: false
narInfo:
title: Store Object Info with Impure fields and NAR Info
description: |
The store object info in the "binary cache" family of Nix store type contain extra information pertaining to *downloads* of the store object in question.
(This store info is called "NAR info", since the downloads take the form of [Nix Archives](@docroot@/store/file-system-object/content-address.md#serial-nix-archive, and the metadata is served in a file with a `.narinfo` extension.)
This download information, being specific to how the store object happens to be stored and transferred, is also considered to be non-intrinsic / impure.
type: object
required:
- version
- narHash
- narSize
- references
- ca
# impure
- deriver
- registrationTime
- ultimate
- signatures
# nar
- url
- compression
- downloadHash
- downloadSize
properties:
version: { $ref: "#/$defs/base/properties/version" }
path: { $ref: "#/$defs/base/properties/path" }
narHash: { $ref: "#/$defs/base/properties/narHash" }
narSize: { $ref: "#/$defs/base/properties/narSize" }
references: { $ref: "#/$defs/base/properties/references" }
ca: { $ref: "#/$defs/base/properties/ca" }
deriver: { $ref: "#/$defs/impure/properties/deriver" }
registrationTime: { $ref: "#/$defs/impure/properties/registrationTime" }
ultimate: { $ref: "#/$defs/impure/properties/ultimate" }
signatures: { $ref: "#/$defs/impure/properties/signatures" }
closureSize: { $ref: "#/$defs/impure/properties/closureSize" }
url:
type: string
title: URL
description: |
Where to download a compressed archive of the file system objects of this store object.
> This is an impure "`.narinfo`" field that may not be included in certain contexts.
compression:
type: string
title: Compression
description: |
The compression format that the archive is in.
> This is an impure "`.narinfo`" field that may not be included in certain contexts.
downloadHash:
type: string
title: Download Hash
description: |
A digest for the compressed archive itself, as opposed to the data contained within.
> This is an impure "`.narinfo`" field that may not be included in certain contexts.
downloadSize:
type: integer
minimum: 0
title: Download Size
description: |
The size of the compressed archive itself.
> This is an impure "`.narinfo`" field that may not be included in certain contexts.
closureDownloadSize:
type: integer
minimum: 0
title: Closure Download Size
description: |
The total size of the compressed archive itself for this object, and the compressed archive of every object in this object's [closure](@docroot@/glossary.md#gloss-closure).
> This is an impure "`.narinfo`" field that may not be included in certain contexts.
> This field is not stored at all, but computed by traversing the other fields across all the store objects in a closure.
additionalProperties: false

View File

@@ -0,0 +1 @@
../../../../../../src/libstore-tests/data/store-path

View File

@@ -0,0 +1,32 @@
"$schema": "http://json-schema.org/draft-07/schema"
"$id": "https://nix.dev/manual/nix/latest/protocols/json/schema/store-path-v1.json"
title: Store Path
description: |
A [store path](@docroot@/store/store-path.md) identifying a store object.
This schema describes the JSON representation of store paths as used in various Nix JSON APIs.
> **Warning**
>
> This JSON format is currently
> [**experimental**](@docroot@/development/experimental-features.md#xp-feature-nix-command)
> and subject to change.
## Format
Store paths in JSON are represented as strings containing just the hash and name portion, without the store directory prefix.
For example: `"g1w7hy3qg1w7hy3qg1w7hy3qg1w7hy3q-foo.drv"`
(If the store dir is `/nix/store`, then this corresponds to the path `/nix/store/g1w7hy3qg1w7hy3qg1w7hy3qg1w7hy3q-foo.drv`.)
## Structure
The format follows this pattern: `${digest}-${name}`
- **hash**: Digest rendered in a custom variant of [Base32](https://en.wikipedia.org/wiki/Base32) (20 arbitrary bytes become 32 ASCII characters)
- **name**: The package name and optional version/suffix information
type: string
pattern: "^[0123456789abcdfghijklmnpqrsvwxyz]{32}-.+$"
minLength: 34

View File

@@ -1,102 +1,45 @@
# Store object info JSON format
{{#include store-object-info-v2-fixed.md}}
> **Warning**
>
> This JSON format is currently
> [**experimental**](@docroot@/development/experimental-features.md#xp-feature-nix-command)
> and subject to change.
## Examples
Info about a [store object].
### Minimal store object (content-addressed)
* `path`:
```json
{{#include schema/store-object-info-v2/pure.json}}
```
[Store path][store path] to the given store object.
### Store object with impure fields
* `narHash`:
```json
{{#include schema/store-object-info-v2/impure.json}}
```
Hash of the [file system object] part of the store object when serialized as a [Nix Archive].
### Minimal store object (empty)
* `narSize`:
```json
{{#include schema/store-object-info-v2/empty_pure.json}}
```
Size of the [file system object] part of the store object when serialized as a [Nix Archive].
### Store object with all impure fields
* `references`:
```json
{{#include schema/store-object-info-v2/empty_impure.json}}
```
An array of [store paths][store path], possibly including this one.
### NAR info (minimal)
* `ca`:
```json
{{#include schema/nar-info-v1/pure.json}}
```
If the store object is [content-addressed],
this is the content address of this store object's file system object, used to compute its store path.
Otherwise (i.e. if it is [input-addressed]), this is `null`.
### NAR info (with binary cache fields)
[store path]: @docroot@/store/store-path.md
[file system object]: @docroot@/store/file-system-object.md
[Nix Archive]: @docroot@/store/file-system-object/content-address.md#serial-nix-archive
```json
{{#include schema/nar-info-v1/impure.json}}
```
## Impure fields
<!-- need to convert YAML to JSON first
## Raw Schema
These are not intrinsic properties of the store object.
In other words, the same store object residing in different store could have different values for these properties.
* `deriver`:
If known, the path to the [store derivation] from which this store object was produced.
Otherwise `null`.
[store derivation]: @docroot@/glossary.md#gloss-store-derivation
* `registrationTime` (optional):
If known, when this derivation was added to the store.
Otherwise `null`.
* `ultimate`:
Whether this store object is trusted because we built it ourselves, rather than substituted a build product from elsewhere.
* `signatures`:
Signatures claiming that this store object is what it claims to be.
Not relevant for [content-addressed] store objects,
but useful for [input-addressed] store objects.
[content-addressed]: @docroot@/store/store-object/content-address.md
[input-addressed]: @docroot@/glossary.md#gloss-input-addressed-store-object
### `.narinfo` extra fields
This meta data is specific to the "binary cache" family of Nix store types.
This information is not intrinsic to the store object, but about how it is stored.
* `url`:
Where to download a compressed archive of the file system objects of this store object.
* `compression`:
The compression format that the archive is in.
* `fileHash`:
A digest for the compressed archive itself, as opposed to the data contained within.
* `fileSize`:
The size of the compressed archive itself.
## Computed closure fields
These fields are not stored at all, but computed by traversing the other fields across all the store objects in a [closure].
* `closureSize`:
The total size of the compressed archive itself for this object, and the compressed archive of every object in this object's [closure].
### `.narinfo` extra fields
* `closureSize`:
The total size of this store object and every other object in its [closure].
[closure]: @docroot@/glossary.md#gloss-closure
[JSON Schema for Store Object Info v1](schema/store-object-info-v2.json)
-->

View File

@@ -0,0 +1,15 @@
{{#include store-path-v1-fixed.md}}
## Examples
### Simple store path
```json
{{#include schema/store-path-v1/simple.json}}
```
<!-- need to convert YAML to JSON first
## Raw Schema
[JSON Schema for Store Path v1](schema/store-path-v1.json)
-->

View File

@@ -0,0 +1,2 @@
# Process JSON schema documentation
subdir('json')

View File

@@ -4,7 +4,7 @@ This is the complete specification of the [Nix Archive] format.
The Nix Archive format closely follows the abstract specification of a [file system object] tree,
because it is designed to serialize exactly that data structure.
[Nix Archive]: @docroot@/store/file-system-object/content-address.md#nix-archive
[Nix Archive]: @docroot@/store/file-system-object/content-address.md#serial-nix-archive
[file system object]: @docroot@/store/file-system-object.md
The format of this specification is close to [Extended BackusNaur form](https://en.wikipedia.org/wiki/Extended_Backus%E2%80%93Naur_form), with the exception of the `str(..)` function / parameterized rule, which length-prefixes and pads strings.
@@ -41,3 +41,15 @@ The `str` function / parameterized rule is defined as follows:
- `int(n)` = the 64-bit little endian representation of the number `n`
- `pad(s)` = the byte sequence `s`, padded with 0s to a multiple of 8 byte
## Kaitai Struct Specification
The Nix Archive (NAR) format is also formally described using [Kaitai Struct](https://kaitai.io/), an Interface Description Language (IDL) for defining binary data structures.
> Kaitai Struct provides a language-agnostic, machine-readable specification that can be compiled into parsers for various programming languages (e.g., C++, Python, Java, Rust).
```yaml
{{#include nar.ksy}}
```
The source of the spec can be found [here](https://github.com/nixos/nix/blob/master/src/nix-manual/source/protocols/nix-archive/nar.ksy). Contributions and improvements to the spec are welcomed.

View File

@@ -0,0 +1,169 @@
meta:
id: nix_nar
title: Nix Archive (NAR)
file-extension: nar
endian: le
doc: |
Nix Archive (NAR) format. A simple, reproducible binary archive
format used by the Nix package manager to serialize file system objects.
doc-ref: 'https://nixos.org/manual/nix/stable/command-ref/nix-store.html#nar-format'
seq:
- id: magic
type: padded_str
doc: "Magic string, must be 'nix-archive-1'."
valid:
expr: _.body == 'nix-archive-1'
- id: root_node
type: node
doc: "The root of the archive, which is always a single node."
types:
padded_str:
doc: |
A string, prefixed with its length (u8le) and
padded with null bytes to the next 8-byte boundary.
seq:
- id: len_str
type: u8
- id: body
type: str
size: len_str
encoding: 'ASCII'
- id: padding
size: (8 - (len_str % 8)) % 8
node:
doc: "A single filesystem node (file, directory, or symlink)."
seq:
- id: open_paren
type: padded_str
doc: "Must be '(', a token starting the node definition."
valid:
expr: _.body == '('
- id: type_key
type: padded_str
doc: "Must be 'type'."
valid:
expr: _.body == 'type'
- id: type_val
type: padded_str
doc: "The type of the node: 'regular', 'directory', or 'symlink'."
- id: body
type:
switch-on: type_val.body
cases:
"'directory'": type_directory
"'regular'": type_regular
"'symlink'": type_symlink
- id: close_paren
type: padded_str
valid:
expr: _.body == ')'
if: "type_val.body != 'directory'"
doc: "Must be ')', a token ending the node definition."
type_directory:
doc: "A directory node, containing a list of entries. Entries must be ordered by their names."
seq:
- id: entries
type: dir_entry
repeat: until
repeat-until: _.kind.body == ')'
types:
dir_entry:
doc: "A single entry within a directory, or a terminator."
seq:
- id: kind
type: padded_str
valid:
expr: _.body == 'entry' or _.body == ')'
doc: "Must be 'entry' (for a child node) or '' (for terminator)."
- id: open_paren
type: padded_str
valid:
expr: _.body == '('
if: 'kind.body == "entry"'
- id: name_key
type: padded_str
valid:
expr: _.body == 'name'
if: 'kind.body == "entry"'
- id: name
type: padded_str
if: 'kind.body == "entry"'
- id: node_key
type: padded_str
valid:
expr: _.body == 'node'
if: 'kind.body == "entry"'
- id: node
type: node
if: 'kind.body == "entry"'
doc: "The child node, present only if kind is 'entry'."
- id: close_paren
type: padded_str
valid:
expr: _.body == ')'
if: 'kind.body == "entry"'
instances:
is_terminator:
value: kind.body == ')'
type_regular:
doc: "A regular file node."
seq:
# Read attributes (like 'executable') until we hit 'contents'
- id: attributes
type: reg_attribute
repeat: until
repeat-until: _.key.body == "contents"
# After the 'contents' token, read the file data
- id: file_data
type: file_content
instances:
is_executable:
value: 'attributes[0].key.body == "executable"'
doc: "True if the file has the 'executable' attribute."
types:
reg_attribute:
doc: "An attribute of the file, e.g., 'executable' or 'contents'."
seq:
- id: key
type: padded_str
doc: "Attribute key, e.g., 'executable' or 'contents'."
valid:
expr: _.body == 'executable' or _.body == 'contents'
- id: value
type: padded_str
if: 'key.body == "executable"'
valid:
expr: _.body == ''
doc: "Must be '' if key is 'executable'."
file_content:
doc: "The raw data of the file, prefixed by length."
seq:
- id: len_contents
type: u8
# # This relies on the property of instances that they are lazily evaluated and cached.
- size: 0
if: nar_offset < 0
- id: contents
size: len_contents
- id: padding
size: (8 - (len_contents % 8)) % 8
instances:
nar_offset:
value: _io.pos
type_symlink:
doc: "A symbolic link node."
seq:
- id: target_key
type: padded_str
doc: "Must be 'target'."
valid:
expr: _.body == 'target'
- id: target_val
type: padded_str
doc: "The destination path of the symlink."

View File

@@ -358,7 +358,7 @@ This release has the following new features:
they are needed for evaluation.
- You can now use `channel:` as a short-hand for
<https://nixos.org/channels//nixexprs.tar.xz>. For example,
<https://nixos.org/channels//nixexprs.tar.xz> [now <https://channels.nixos.org//nixexprs.tar.xz>]. For example,
`nix-build channel:nixos-15.09 -A hello` will build the GNU Hello
package from the `nixos-15.09` channel. In the future, this may
use Git to fetch updates more efficiently.

View File

@@ -13,7 +13,7 @@
- The `discard-references` feature has been stabilized.
This means that the
[unsafeDiscardReferences](@docroot@/development/experimental-features.md#xp-feature-discard-references)
[unsafeDiscardReferences](@docroot@/language/advanced-attributes.md#adv-attr-unsafeDiscardReferences)
attribute is no longer guarded by an experimental flag and can be used
freely.

View File

@@ -17,8 +17,8 @@
- `nix-shell` shebang lines now support single-quoted arguments.
- `builtins.fetchTree` is now its own experimental feature, [`fetch-tree`](@docroot@/development/experimental-features.md#xp-fetch-tree).
This allows stabilising it independently of the rest of what is encompassed by [`flakes`](@docroot@/development/experimental-features.md#xp-fetch-tree).
- `builtins.fetchTree` is now its own experimental feature, [`fetch-tree`](@docroot@/development/experimental-features.md#xp-feature-fetch-tree).
This allows stabilising it independently of the rest of what is encompassed by [`flakes`](@docroot@/development/experimental-features.md#xp-feature-flakes).
- The interface for creating and updating lock files has been overhauled:

View File

@@ -14,7 +14,7 @@
- Modify `nix derivation {add,show}` JSON format [#9866](https://github.com/NixOS/nix/issues/9866) [#10722](https://github.com/NixOS/nix/pull/10722)
The JSON format for derivations has been slightly revised to better conform to our [JSON guidelines](@docroot@/development/cli-guideline.md#returning-future-proof-json).
The JSON format for derivations has been slightly revised to better conform to our [JSON guidelines](@docroot@/development/json-guideline.md).
In particular, the hash algorithm and content addressing method of content-addressed derivation outputs are now separated into two fields `hashAlgo` and `method`,
rather than one field with an arcane `:`-separated format.

View File

@@ -93,7 +93,7 @@
- Support unit prefixes in configuration settings [#10668](https://github.com/NixOS/nix/pull/10668)
Configuration settings in Nix now support unit prefixes, allowing for more intuitive and readable configurations. For example, you can now specify [`--min-free 1G`](@docroot@/command-ref/opt-common.md#opt-min-free) to set the minimum free space to 1 gigabyte.
Configuration settings in Nix now support unit prefixes, allowing for more intuitive and readable configurations. For example, you can now specify [`--min-free 1G`](@docroot@/command-ref/conf-file.md#conf-min-free) to set the minimum free space to 1 gigabyte.
This enhancement was extracted from [#7851](https://github.com/NixOS/nix/pull/7851) and is also useful for PR [#10661](https://github.com/NixOS/nix/pull/10661).

View File

@@ -0,0 +1,53 @@
# Build Trace
> **Warning**
>
> This entire concept is currently
> [**experimental**](@docroot@/development/experimental-features.md#xp-feature-ca-derivations)
> and subject to change.
The *build trace* is a [memoization table](https://en.wikipedia.org/wiki/Memoization) for builds.
It maps the inputs of builds to the outputs of builds.
Concretely, that means it maps [derivations][derivation] to maps of [output] names to [store objects][store object].
In general the derivations used as a key should be [*resolved*](./resolution.md).
A build trace with all-resolved-derivation keys is also called a *base build trace* for extra clarity.
If all the resolved inputs of a derivation are content-addressed, that means the inputs will be fully determined, leaving no ambiguity for what build was performed.
(Input-addressed inputs however are still ambiguous. They too should be locked down, but this is left as future work.)
Accordingly, to look up an unresolved derivation, one must first resolve it to get a resolved derivation.
Resolving itself involves looking up entries in the build trace, so this is a mutually recursive process that will end up inspecting possibly many entries.
Except for the issue with input-addressed paths called out above, base build traces are trivially *coherent* -- incoherence is not possible.
That means that the claims that each key-value base build try entry makes are independent, and no mapping invalidates another mapping.
Whether the mappings are *true*, i.e. the faithful recording of actual builds performed, is another matter.
Coherence is about the multiple claims of the build trace being mutually consistent, not about whether the claims are individually true or false.
In general, there is no way to audit a build trace entry except for by performing the build again from scratch.
And even in that case, a different result doesn't mean the original entry was a "lie", because the derivation being built may be non-deterministic.
As such, the decision of whether to trust a counterparty's build trace is a fundamentally subject policy choice.
Build trace entries are typically *signed* in order to enable arbitrary public-key-based trust polices.
## Derived build traces {#derived}
Implementations that wish to memoize the above may also keep additional *derived* build trace entries that do map unresolved derivations.
But if they do so, they *must* also keep the underlying base entries with resolved derivation keys around.
Firstly, this ensures that the derived entries are merely cache, which could be recomputed from scratch.
Secondly, this ensures the coherence of the derived build trace.
Unlike with base build traces, incoherence with derived build traces is possible.
The key ingredient is that derivation resolution is only deterministic with respect to a fixed base build trace.
Without fixing the base build trace, it inherits the subjectivity of base build traces themselves.
Concretely, suppose there are three derivations \\(a\\), \\(b\\), and \\(c\\).
Let \\(a\\) be a resolved derivation, but let \\(b\\) and \\(c\\) be unresolved and both take as an input an output of \\(a\\).
Now suppose that derived entries are made for \\(b\\) and \\(c\\) based on two different entries of \\(a\\).
(This could happen if \\(a\\) is non-deterministic, \\(a\\) and \\(b\\) are built in one store, \\(a\\) and \\(c\\) are built in another store, and then a third store substitutes from both of the first two stores.)
If trusting the derived build trace entries for \\(b\\) and \\(c\\) requires that each's underlying entry for \\(a\\) be also trusted, the two different mappings for \\(a\\) will be caught.
However, if \\(b\\) and \\(c\\)'s entries can be combined in isolation, there will be nothing to catch the contradiction in their hidden assumptions about \\(a\\)'s output.
[derivation]: ./derivation/index.md
[output]: ./derivation/outputs/index.md
[store object]: @docroot@/store/store-object.md

View File

@@ -8,7 +8,7 @@
- Once this is done, the derivation is *normalized*, replacing each input deriving path with its store path, which we now know from realising the input.
## Builder Execution
## Builder Execution {#builder-execution}
The [`builder`](./derivation/index.md#builder) is executed as follows:

View File

@@ -102,11 +102,11 @@ But rather than somehow scanning all the other fields for inputs, Nix requires t
### System {#system}
The system type on which the [`builder`](#attr-builder) executable is meant to be run.
The system type on which the [`builder`](#builder) executable is meant to be run.
A necessary condition for Nix to schedule a given derivation on some [Nix instance] is for the "system" of that derivation to match that instance's [`system` configuration option] or [`extra-platforms` configuration option].
By putting the `system` in each derivation, Nix allows *heterogenous* build plans, where not all steps can be run on the same machine or same sort of machine.
By putting the `system` in each derivation, Nix allows *heterogeneous* build plans, where not all steps can be run on the same machine or same sort of machine.
Nix can schedule builds such that it automatically builds on other platforms by [forwarding build requests](@docroot@/advanced-topics/distributed-builds.md) to other Nix instances.
[`system` configuration option]: @docroot@/command-ref/conf-file.md#conf-system
@@ -245,7 +245,7 @@ If those other derivations *also* abide by this common case (and likewise for tr
> note the ".drv"
> ```
## Extending the model to be higher-order
## Extending the model to be higher-order {#dynamic}
**Experimental feature**: [`dynamic-derivations`](@docroot@/development/experimental-features.md#xp-feature-dynamic-derivations)

View File

@@ -167,10 +167,10 @@ It is only in the potential for that check to fail that they are different.
>
> In a future world where floating content-addressing is also stable, we in principle no longer need separate [fixed](#fixed) content-addressing.
> Instead, we could always use floating content-addressing, and separately assert the precise value content address of a given store object to be used as an input (of another derivation).
> A stand-alone assertion object of this sort is not yet implemented, but its possible creation is tracked in [Issue #11955](https://github.com/NixOS/nix/issues/11955).
> A stand-alone assertion object of this sort is not yet implemented, but its possible creation is tracked in [issue #11955](https://github.com/NixOS/nix/issues/11955).
>
> In the current version of Nix, fixed outputs which fail their hash check are still registered as valid store objects, just not registered as outputs of the derivation which produced them.
> This is an optimization that means if the wrong output hash is specified in a derivation, and then the derivation is recreated with the right output hash, derivation does not need to be rebuilt --- avoiding downloading potentially large amounts of data twice.
> This is an optimization that means if the wrong output hash is specified in a derivation, and then the derivation is recreated with the right output hash, derivation does not need to be rebuilt &mdash; avoiding downloading potentially large amounts of data twice.
> This optimisation prefigures the design above:
> If the output hash assertion was removed outside the derivation itself, Nix could additionally not only register that outputted store object like today, but could also make note that derivation did in fact successfully download some data.
For example, for the "fetch URL" example above, making such a note is tantamount to recording what data is available at the time of download at the given URL.

View File

@@ -43,7 +43,7 @@ In particular, the specification decides:
- if the content is content-addressed, how is it content addressed
- if the content is content-addressed, [what is its content address](./content-address.md#fixed-content-addressing) (and thus what is its [store path])
- if the content is content-addressed, [what is its content address](./content-address.md#fixed) (and thus what is its [store path])
## Types of derivations

View File

@@ -6,26 +6,221 @@
That is to say, an input-addressed output's store path is a function not of the output itself, but of the derivation that produced it.
Even if two store paths have the same contents, if they are produced in different ways, and one is input-addressed, then they will have different store paths, and thus guaranteed to not be the same store object.
<!---
## Modulo content addressed derivation outputs {#hash-quotient-drv}
### Modulo fixed-output derivations
A naive implementation of an output hash computation for input-addressed outputs would be to hash the derivation hash and output together.
This clearly has the uniqueness properties we want for input-addressed outputs, but suffers from an inefficiency.
Specifically, new builds would be required whenever a change is made to a fixed-output derivation, despite having provably no differences in the inputs to the new derivation compared to what it used to be.
Concretely, this would cause a "mass rebuild" whenever any fetching detail changes, including mirror lists, certificate authority certificates, etc.
**TODO hash derivation modulo.**
To solve this problem, we compute output hashes differently, so that certain output hashes become identical.
We call this concept quotient hashing, in reference to quotient types or sets.
So how do we compute the hash part of the output path of a derivation?
This is done by the function `hashDrv`, shown in Figure 5.10.
It distinguishes between two cases.
If the derivation is a fixed-output derivation, then it computes a hash over just the `outputHash` attributes.
So how do we compute the hash part of the output paths of an input-addressed derivation?
This is done by the function `hashQuotientDerivation`, shown below.
If the derivation is not a fixed-output derivation, we replace each element in the derivations inputDrvs with the result of a call to `hashDrv` for that element.
(The derivation at each store path in `inputDrvs` is converted from its on-disk ATerm representation back to a `StoreDrv` by the function `parseDrv`.) In essence, `hashDrv` partitions store derivations into equivalence classes, and for hashing purpose it replaces each store path in a derivation graph with its equivalence class.
First, a word on inputs.
`hashQuotientDerivation` is only defined on derivations whose [inputs](@docroot@/store/derivation/index.md#inputs) take the first-order form:
```typescript
type ConstantPath = {
path: StorePath;
};
The recursion in Figure 5.10 is inefficient:
it will call itself once for each path by which a subderivation can be reached, i.e., `O(V k)` times for a derivation graph with `V` derivations and with out-degree of at most `k`.
In the actual implementation, memoisation is used to reduce this to `O(V + E)` complexity for a graph with E edges.
type FirstOrderOutputPath = {
drvPath: StorePath;
output: OutputName;
};
-->
type FirstOrderDerivingPath = ConstantPath | FirstOrderOutputPath;
type Inputs = Set<FirstOrderDerivingPath>;
```
For the algorithm below, we adopt a derivation where the two types of (first order) derived paths are partitioned into two sets, as follows:
```typescript
type Derivation = {
// inputs: Set<FirstOrderDerivingPath>; // replaced
inputSrcs: Set<ConstantPath>; // new instead
inputDrvOutputs: Set<FirstOrderOutputPath>; // new instead
// ...other fields...
};
```
In the [currently-experimental][xp-feature-dynamic-derivations] higher-order case where outputs of outputs are allowed as [deriving paths][deriving-path] and thus derivation inputs, derivations using that generalization are not valid arguments to this function.
Those derivations must be (partially) [resolved](@docroot@/store/resolution.md) enough first, to the point where no such higher-order inputs remain.
Then, and only then, can input addresses be assigned.
```
function hashQuotientDerivation(drv) -> Hash:
assert(drv.outputs are input-addressed)
drv ← drv with {
inputDrvOutputs = (
assert(drvPath is store path)
case hashOutputsOrQuotientDerivation(readDrv(drvPath)) of
drvHash : Hash →
(drvHash.toBase16(), output)
outputHashes : Map[String, Hash] →
(outputHashes[output].toBase16(), "out")
| (drvPath, output) ∈ drv.inputDrvOutputs
)
}
return hashSHA256(printDrv(drv))
function hashOutputsOrQuotientDerivation(drv) -> Map[String, Hash] | Hash:
if drv.outputs are content-addressed:
return {
outputName ↦ hashSHA256(
"fixed:out:" + ca.printMethodAlgo() +
":" + ca.hash.toBase16() +
":" + ca.makeFixedOutputPath(drv.name, outputName))
| (outputName ↦ output) ∈ drv.outputs
, ca = output.contentAddress // or get from build trace if floating
}
else: // drv.outputs are input-addressed
return hashQuotientDerivation(drv)
```
### `hashQuotientDerivation`
We replace each element in the derivation's `inputDrvOutputs` using data from a call to `hashOutputsOrQuotientDerivation` on the `drvPath` of that element.
When `hashOutputsOrQuotientDerivation` returns a single drv hash (because the input derivation in question is input-addressing), we simply swap out the `drvPath` for that hash, and keep the same output name.
When `hashOutputsOrQuotientDerivation` returns a map of content addresses per-output, we look up the output in question, and pair it with the output name `out`.
The resulting pseudo-derivation (with hashes instead of store paths in `inputDrvs`) is then printed (in the ["ATerm" format](@docroot@/protocols/derivation-aterm.md)) and hashed, and this becomes the hash of the "quotient derivation".
When calculating output hashes, `hashQuotientDerivation` is called on an almost-complete input-addressing derivation, which is just missing its input-addressed outputs paths.
The derivation hash is then used to calculate output paths for each output.
<!-- TODO describe how this is done. -->
Those output paths can then be substituted into the almost-complete input-addressed derivation to complete it.
> **Note**
>
> There may be an unintentional deviation from specification currently implemented in the `(outputHashes[output].toBase16(), "out")` case.
> This is not fatal because the deviation would only apply for content-addressing derivations with more than one output, and that only occurs in the floating case, which is [experimental][xp-feature-ca-derivations].
> Once this bug is fixed, this note will be removed.
### `hashOutputsOrQuotientDerivation`
How does `hashOutputsOrQuotientDerivation` in turn work?
It consists of two main cases, based on whether the outputs of the derivation are to be input-addressed or content-addressed.
#### Input-addressed outputs case
In the input-addressed case, it just calls `hashQuotientDerivation`, and returns that derivation hash.
This makes `hashQuotientDerivation` and `hashOutputsOrQuotientDerivation` mutually-recursive.
> **Note**
>
> In this case, `hashQuotientDerivation` is being called on a *complete* input-addressing derivation that already has its output paths calculated.
> The `inputDrvs` substitution takes place anyways.
#### Content-addressed outputs case
If the outputs are [content-addressed](./content-address.md), then it computes a hash for each output derived from the content-address of that output.
> **Note**
>
> In the [fixed](./content-address.md#fixed) content-addressing case, the outputs' content addresses are statically specified in advance, so this always just works.
> (The fixed case is what the pseudo-code shows.)
>
> In the [floating](./content-address.md#floating) case, the content addresses are not specified in advance.
> This is what the "or get from [build trace](@docroot@/store/build-trace.md) if floating" comment refers to.
> In this case, the algorithm is *stuck* until the input in question is built, and we know what the actual contents of the output in question is.
>
> That is OK however, because there is no problem with delaying the assigning of input addresses (which, remember, is what `hashQuotientDerivation` is ultimately for) until all inputs are known.
### Performance
The recursion in the algorithm is potentially inefficient:
it could call itself once for each path by which a subderivation can be reached, i.e., `O(V^k)` times for a derivation graph with `V` derivations and with out-degree of at most `k`.
In the actual implementation, [memoisation](https://en.wikipedia.org/wiki/Memoization) is used to reduce this cost to be proportional to the total number of `inputDrvOutputs` encountered.
### Semantic properties
*See [this chapter's appendix](@docroot@/store/math-notation.md) on grammar and metavariable conventions.*
In essence, `hashQuotientDerivation` partitions input-addressing derivations into equivalence classes: every derivation in that equivalence class is mapped to the same derivation hash.
We can characterize this equivalence relation directly, by working bottom up.
We start by defining an equivalence relation on first-order output deriving paths that refer content-addressed derivation outputs. Two such paths are equivalent if they refer to the same store object:
\\[
\\begin{prooftree}
\\AxiomC{$d\_1$ is content-addressing}
\\AxiomC{$d\_2$ is content-addressing}
\\AxiomC{$
{}^\*(\text{path}(d\_1), o\_1)
\=
{}^\*(\text{path}(d\_2), o\_2)
$}
\\TrinaryInfC{$(\text{path}(d\_1), o\_1) \\,\\sim_{\\mathrm{CA}}\\, (d\_2, o\_2)$}
\\end{prooftree}
\\]
where \\({}^*(s, o)\\) denotes the store object that the output deriving path refers to.
We will also need the following construction to lift any equivalence relation on \\(X\\) to an equivalence relation on (finite) sets of \\(X\\) (in short, \\(\\mathcal{P}(X)\\)):
\\[
\\begin{prooftree}
\\AxiomC{$\\forall a \\in A. \\exists b \\in B. a \\,\\sim\_X\\, b$}
\\AxiomC{$\\forall b \\in B. \\exists a \\in A. b \\,\\sim\_X\\, a$}
\\BinaryInfC{$A \\,\\sim_{\\mathcal{P}(X)}\\, B$}
\\end{prooftree}
\\]
Now we can define the equivalence relation \\(\\sim_\\mathrm{IA}\\) on input-addressed derivation outputs. Two input-addressed outputs are equivalent if their derivations are equivalent (via the yet-to-be-defined \\(\\sim_{\\mathrm{IADrv}}\\) relation) and their output names are the same:
\\[
\\begin{prooftree}
\\AxiomC{$d\_1$ is input-addressing}
\\AxiomC{$d\_2$ is input-addressing}
\\AxiomC{$d\_1 \\,\\sim_{\\mathrm{IADrv}}\\, d\_2$}
\\AxiomC{$o\_1 = o\_2$}
\\QuaternaryInfC{$(\text{path}(d\_1), o\_1) \\,\\sim_{\\mathrm{IA}}\\, (\text{path}(d\_2), o\_2)$}
\\end{prooftree}
\\]
And now we can define \\(\\sim_{\\mathrm{IADrv}}\\).
Two input-addressed derivations are equivalent if their content-addressed inputs are equivalent, their input-addressed inputs are also equivalent, and they are otherwise equal:
<!-- cheating a bit with the semantics to get a good layout that fits on the page -->
\\[
\\begin{prooftree}
\\alwaysNoLine
\\AxiomC{$
\\mathrm{caInputs}(d\_1)
\\,\\sim_{\\mathcal{P}(\\mathrm{CA})}\\,
\\mathrm{caInputs}(d\_2)
$}
\\AxiomC{$
\\mathrm{iaInputs}(d\_1)
\\,\\sim_{\\mathcal{P}(\\mathrm{IA})}\\,
\\mathrm{iaInputs}(d\_2)
$}
\\BinaryInfC{$
d\_1\left[\\mathrm{inputDrvOutputs} := \\{\\}\right]
\=
d\_2\left[\\mathrm{inputDrvOutputs} := \\{\\}\right]
$}
\\alwaysSingleLine
\\UnaryInfC{$d\_1 \\,\\sim_{\\mathrm{IADrv}}\\, d\_2$}
\\end{prooftree}
\\]
where \\(\\mathrm{caInputs}(d)\\) returns the content-addressed inputs of \\(d\\) and \\(\\mathrm{iaInputs}(d)\\) returns the input-addressed inputs.
> **Note**
>
> An astute reader might notice that that nowhere does `inputSrcs` enter into these definitions.
> That means that replacing an input derivation with its outputs directly added to `inputSrcs` always results in a derivation in a different equivalence class, despite the resulting input closure (as would be mounted in the store at build time) being the same.
> [Issue #9259](https://github.com/NixOS/nix/issues/9259) is about creating a coarser equivalence relation to address this.
>
> \\(\\sim_\mathrm{Drv}\\) from [derivation resolution](@docroot@/store/resolution.md) is such an equivalence relation.
> It is coarser than this one: any two derivations which are "'hash quotient derivation'-equivalent" (\\(\\sim_\mathrm{IADrv}\\)) are also "resolution-equivalent" (\\(\\sim_\mathrm{Drv}\\)).
> It also relates derivations whose `inputDrvOutputs` have been rewritten into `inputSrcs`.
[deriving-path]: @docroot@/store/derivation/index.md#deriving-path
[xp-feature-dynamic-derivations]: @docroot@/development/experimental-features.md#xp-feature-dynamic-derivations
[xp-feature-ca-derivations]: @docroot@/development/experimental-features.md#xp-feature-ca-derivations
[xp-feature-git-hashing]: @docroot@/development/experimental-features.md#xp-feature-git-hashing
[xp-feature-impure-derivations]: @docroot@/development/experimental-features.md#xp-feature-impure-derivations

View File

@@ -46,7 +46,7 @@ be many different serialisations.
For these reasons, Nix has its very own archive format—the Nix Archive (NAR) format,
which is carefully designed to avoid the problems described above.
The exact specification of the Nix Archive format is in [specified here](../../protocols/nix-archive.md).
The exact specification of the Nix Archive format is in [specified here](../../protocols/nix-archive/index.md).
## Content addressing File System Objects beyond a single serialisation pass

View File

@@ -0,0 +1,16 @@
# Appendix: Math notation
A few times in this manual, formal "proof trees" are used for [natural deduction](https://en.wikipedia.org/wiki/Natural_deduction)-style definition of various [relations](https://en.wikipedia.org/wiki/Relation_(mathematics)).
The following grammar and assignment of metavariables to syntactic categories is used in these sections.
\\begin{align}
s, t &\in \text{store-path} \\\\
o &\in \text{output-name} \\\\
i, p &\in \text{deriving-path} \\\\
d &\in \text{derivation}
\\end{align}
\\begin{align}
\text{deriving-path} \quad p &::= s \mid (p, o)
\\end{align}

View File

@@ -0,0 +1,219 @@
# Derivation Resolution
*See [this chapter's appendix](@docroot@/store/math-notation.md) on grammar and metavariable conventions.*
To *resolve* a derivation is to replace its [inputs] with the simplest inputs &mdash; plain store paths &mdash; that denote the same store objects.
Derivations that only have store paths as inputs are likewise called *resolved derivations*.
(They are called that whether they are in fact the output of derivation resolution, or just made that way without non-store-path inputs to begin with.)
## Input Content Equivalence of Derivations
[Deriving paths][deriving-path] intentionally make it possible to refer to the same [store object] in multiple ways.
This is a consequence of content-addressing, since different derivations can produce the same outputs, and the same data can also be manually added to the store.
This is also a consequence even of input-addressing, as an output can be referred to by derivation and output name, or directly by its [computed](./derivation/outputs/input-address.md) store path.
Since dereferencing deriving paths is thus not injective, it induces an equivalence relation on deriving paths.
Let's call this equivalence relation \\(\\sim\\), where \\(p_1 \\sim p_2\\) means that deriving paths \\(p_1\\) and \\(p_2\\) refer to the same store object.
**Content Equivalence**: Two deriving paths are equivalent if they refer to the same store object:
\\[
\\begin{prooftree}
\\AxiomC{${}^*p_1 = {}^*p_2$}
\\UnaryInfC{$p_1 \\,\\sim_\\mathrm{DP}\\, p_2$}
\\end{prooftree}
\\]
where \\({}^\*p\\) denotes the store object that deriving path \\(p\\) refers to.
This also induces an equivalence relation on sets of deriving paths:
\\[
\\begin{prooftree}
\\AxiomC{$\\{ {}^*p | p \\in P_1 \\} = \\{ {}^*p | p \\in P_2 \\}$}
\\UnaryInfC{$P_1 \\,\\sim_{\\mathcal{P}(\\mathrm{DP})}\\, P_2$}
\\end{prooftree}
\\]
**Input Content Equivalence**: This, in turn, induces an equivalence relation on derivations: two derivations are equivalent if their inputs are equivalent, and they are otherwise equal:
\\[
\\begin{prooftree}
\\AxiomC{$\\mathrm{inputs}(d_1) \\,\\sim_{\\mathcal{P}(\\mathrm{DP})}\\, \\mathrm{inputs}(d_2)$}
\\AxiomC{$
d\_1\left[\\mathrm{inputs} := \\{\\}\right]
\=
d\_2\left[\\mathrm{inputs} := \\{\\}\right]
$}
\\BinaryInfC{$d_1 \\,\\sim_\\mathrm{Drv}\\, d_2$}
\\end{prooftree}
\\]
Derivation resolution always maps derivations to input-content-equivalent derivations.
## Resolution relation
Dereferencing a derived path &mdash; \\({}^\*p\\) above &mdash; was just introduced as a black box.
But actually it is a multi-step process of looking up build results in the [build trace] that itself depends on resolving the lookup keys.
Resolution is thus a recursive multi-step process that is worth diagramming formally.
We can do this with a small-step binary transition relation; let's call it \\(\rightsquigarrow\\).
We can then conclude dereferenced equality like this:
\\[
\\begin{prooftree}
\\AxiomC{$p\_1 \\rightsquigarrow^* p$}
\\AxiomC{$p\_2 \\rightsquigarrow^* p$}
\\BinaryInfC{${}^*p\_1 = {}^*p\_2$}
\\end{prooftree}
\\]
I.e. by showing that both original items resolve (over 0 or more small steps, hence the \\({}^*\\)) to the same exact item.
With this motivation, let's now formalize a [small-step](https://en.wikipedia.org/wiki/Operational_semantics#Small-step_semantics) system of reduction rules for resolution.
### Formal rules
### \\(\text{resolved}\\) unary relation
\\[
\\begin{prooftree}
\\AxiomC{$s \in \text{store-path}$}
\\UnaryInfC{$s$ resolved}
\\end{prooftree}
\\]
\\[
\\begin{prooftree}
\\AxiomC{$\forall i \in \mathrm{inputs}(d). i \text{ resolved}$}
\\UnaryInfC{$d$ resolved}
\\end{prooftree}
\\]
### \\(\rightsquigarrow\\) binary relation
> **Remark**
>
> Actually, to be completely formal we would need to keep track of the build trace we are choosing to resolve against.
>
> We could do that by making \\(\rightsquigarrow\\) a ternary relation, which would pass the build trace to itself until it finally uses it in that one rule.
> This would add clutter more than insight, so we didn't bother to write it.
>
> There are other options too, like saying the whole reduction rule system is parameterized on the build trace, essentially [currying](https://en.wikipedia.org/wiki/Currying) the ternary \\(\rightsquigarrow\\) into a function from build traces to the binary relation written above.
#### Core build trace lookup rule
\\[
\\begin{prooftree}
\\AxiomC{$s \in \text{store-path}$}
\\AxiomC{${}^*s \in \text{derivation}$}
\\AxiomC{${}^*s$ resolved}
\\AxiomC{$\mathrm{build\text{-}trace}[s][o] = t$}
\\QuaternaryInfC{$(s, o) \rightsquigarrow t$}
\\RightLabel{\\scriptsize output path resolution}
\\end{prooftree}
\\]
#### Inductive rules
\\[
\\begin{prooftree}
\\AxiomC{$i \\rightsquigarrow i'$}
\\AxiomC{$i \\in \\mathrm{inputs}(d)$}
\\BinaryInfC{$d \\rightsquigarrow d[i \\mapsto i']$}
\\end{prooftree}
\\]
\\[
\\begin{prooftree}
\\AxiomC{$d \\rightsquigarrow d'$}
\\UnaryInfC{$(\\mathrm{path}(d), o) \\rightsquigarrow (\\mathrm{path}(d'), o)$}
\\end{prooftree}
\\]
\\[
\\begin{prooftree}
\\AxiomC{$p \\rightsquigarrow p'$}
\\UnaryInfC{$(p, o) \\rightsquigarrow (p', o)$}
\\end{prooftree}
\\]
### Properties
Like all well-behaved evaluation relations, partial resolution is [*confluent*](https://en.wikipedia.org/wiki/Confluence_(abstract_rewriting)).
Also, if we take the symmetric closure of \\(\\rightsquigarrow^\*\\), we end up with the equivalence relations of the previous section.
Resolution respects content equivalence for deriving paths, and input content equivalence for derivations.
> **Remark**
>
> We chose to define from scratch an "resolved" unary relation explicitly above.
> But it can also be defined as the normal forms of the \\(\\rightsquigarrow^\*\\) relation:
>
> \\[ a \text{ resolved} \Leftrightarrow \forall b. b \rightsquigarrow^* a \Rightarrow b = a\\]
>
> In prose, resolved terms are terms which \\(\\rightsquigarrow^\*\\) only relates on the left side to the same term on the right side; they are the terms which can be resolved no further.
## Partial versus Complete Resolution
Similar to evaluation, we can also speak of *partial* versus *complete* derivation resolution.
Partial derivation resolution is what we've actually formalized above with \\(\\rightsquigarrow^\*\\).
Complete resolution is resolution ending in a resolved term (deriving path or derivation).
(Which is a normal form of the relation, per the remark above.)
With partial resolution, a derivation is related to equivalent derivations with the same or simpler inputs, but not all those inputs will be plain store paths.
This is useful when the input refers to a floating content addressed output we have not yet built &mdash; we don't know what (content-address) store path will used for that derivation, so we are "stuck" trying to resolve the deriving path in question.
(In the above formalization, this happens when the build trace is missing the keys we wish to look up in it.)
Complete resolution is a *functional* relation, i.e. values on the left are uniquely related with values on the right.
It is not however, a *total* relation (in general, assuming arbitrary build traces).
This is discussed in the next section.
## Termination
For static derivations graphs, complete resolution is indeed total, because it always terminates for all inputs.
(A relation that is both total and functional is a function.)
For [dynamic][xp-feature-dynamic-derivations] derivation graphs, however, this is not the case &mdash; resolution is not guaranteed to terminate.
The issue isn't rewriting deriving paths themselves:
a single rewrite to normalize an output deriving path to a constant one always exists, and always proceeds in one step.
The issue is that dynamic derivations (i.e. those that are filled-in the graph by a previous resolution) may have more transitive dependencies than the original derivation.
> **Example**
>
> Suppose we have this deriving path
> ```json
> {
> "drvPath": {
> "drvPath": "...-foo.drv",
> "output": "bar.drv"
> },
> "output": "baz"
> }
> ```
> and derivation `foo` is already resolved.
> When we resolve deriving path we'll end up with something like.
> ```json
> {
> "drvPath": "...-foo-bar.drv",
> "output": "baz"
> }
> ```
> So far is just an atomic single rewrite, with no termination issues.
> But the derivation `foo-bar` may have its *own* dynamic derivation inputs.
> Resolution must resolve that derivation first before the above deriving path can finally be normalized to a plain `...-foo-bar-baz` store path.
The important thing to notice is that while "build trace" *keys* must be resolved.
The *value* those keys are mapped to have no such constraints.
An arbitrary store object has no notion of being resolved or not.
But, an arbitrary store object can be read back as a derivation (as will in fact be done in case for dynamic derivations / nested output deriving paths).
And those derivations need *not* be resolved.
It is those dynamic non-resolved derivations which are the source of non-termination.
By the same token, they are also the reason why dynamic derivations offer greater expressive power.
[store object]: @docroot@/store/store-object.md
[inputs]: @docroot@/store/derivation/index.md#inputs
[build trace]: @docroot@/store/build-trace.md
[deriving-path]: @docroot@/store/derivation/index.md#deriving-path
[xp-feature-dynamic-derivations]: @docroot@/development/experimental-features.md#xp-feature-dynamic-derivations

View File

@@ -1,7 +1,7 @@
# Content-Addressing Store Objects
Just [like][fso-ca] [File System Objects][File System Object],
[Store Objects][Store Object] can also be [content-addressed](@docroot@/glossary.md#gloss-content-addressed),
[Store Objects][Store Object] can also be [content-addressed](@docroot@/glossary.md#gloss-content-address),
unless they are [input-addressed](@docroot@/glossary.md#gloss-input-addressed-store-object).
For store objects, the content address we produce will take the form of a [Store Path] rather than regular hash.
@@ -107,7 +107,7 @@ References (to other store objects and self-references alike) are supported so l
>
> This method is part of the [`git-hashing`][xp-feature-git-hashing] experimental feature.
This uses the corresponding [Git](../file-system-object/content-address.md#serial-git) method of file system object content addressing.
This uses the corresponding [Git](../file-system-object/content-address.md#git) method of file system object content addressing.
References are not supported.

View File

@@ -6,7 +6,7 @@
>
> A rendered store path
Nix implements references to [store objects](./index.md#store-object) as *store paths*.
Nix implements references to [store objects](./store-object.md) as *store paths*.
Think of a store path as an [opaque], [unique identifier]:
The only way to obtain store path is by adding or building store objects.

15
doc/manual/theme/head.hbs Normal file
View File

@@ -0,0 +1,15 @@
<script>
MathJax = {
loader: {load: ['[tex]/bussproofs']},
tex: {
packages: {'[+]': ['bussproofs']},
// Doesn't seem to work in mathjax 3
//formatError: function(jax, error) {
// console.log(`TeX error in "${jax.latex}": ${error.message}`);
// return jax.formatError(error);
//}
}
};
</script>
<!-- Load a newer versino of MathJax than mdbook does by default, and which in particular has working relative paths for the "bussproofs" extension. -->
<script async src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/3.0.1/es5/tex-mml-chtml.js"></script>

View File

@@ -10,7 +10,7 @@
tag ? "latest",
bundleNixpkgs ? true,
channelName ? "nixpkgs",
channelURL ? "https://nixos.org/channels/nixpkgs-unstable",
channelURL ? "https://channels.nixos.org/nixpkgs-unstable",
extraPkgs ? [ ],
maxLayers ? 70,
nixConf ? { },

View File

@@ -32,7 +32,7 @@
let
inherit (nixpkgs) lib;
officialRelease = true;
officialRelease = false;
linux32BitSystems = [ "i686-linux" ];
linux64BitSystems = [
@@ -413,6 +413,14 @@
supportsCross = false;
};
"nix-json-schema-checks" = {
supportsCross = false;
};
"nix-kaitai-struct-checks" = {
supportsCross = false;
};
"nix-perl-bindings" = {
supportsCross = false;
};
@@ -467,6 +475,27 @@
}
);
apps = forAllSystems (
system:
let
pkgs = nixpkgsFor.${system}.native;
opener = if pkgs.stdenv.isDarwin then "open" else "xdg-open";
in
{
open-manual = {
type = "app";
program = "${pkgs.writeShellScript "open-nix-manual" ''
path="${self.packages.${system}.nix-manual.site}/index.html"
if ! ${opener} "$path"; then
echo "Failed to open manual with ${opener}. Manual is located at:"
echo "$path"
fi
''}";
meta.description = "Open the Nix manual in your browser";
};
}
);
devShells =
let
makeShell = import ./packaging/dev-shell.nix { inherit lib devFlake; };

View File

@@ -142,7 +142,6 @@ release:
$ git pull
$ NEW_VERSION=2.13.0
$ echo $NEW_VERSION > .version
$ ... edit .mergify.yml to add the previous version ...
$ git checkout -b bump-$NEW_VERSION
$ git commit -a -m 'Bump version'
$ git push --set-upstream origin bump-$NEW_VERSION

View File

@@ -60,3 +60,4 @@ if get_option('unit-tests')
subproject('libflake-tests')
endif
subproject('nix-functional-tests')
subproject('json-schema-checks')

View File

@@ -0,0 +1,6 @@
extern "C" [[gnu::retain, gnu::weak]] 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.
return "abort_on_error=1:print_summary=1:detect_leaks=0:detect_odr_violation=0";
}

View File

@@ -1,7 +1,3 @@
asan_test_options_env = {
'ASAN_OPTIONS' : 'abort_on_error=1:print_summary=1:detect_leaks=0',
}
# 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') or 'undefined' in get_option(
@@ -10,3 +6,6 @@ if cxx.get_id() == 'clang' and ('address' in get_option('b_sanitize') or 'undefi
add_project_link_arguments('-shared-libasan', language : 'cpp')
endif
if 'address' in get_option('b_sanitize')
deps_other += declare_dependency(sources : 'asan-options.cc')
endif

View File

@@ -66,3 +66,4 @@ endif
nix_soversion = meson.project_version().split('+')[0].split('pre')[0]
subdir('assert-fail')
subdir('asan-options')

View File

@@ -204,6 +204,25 @@ let
mesonFlags = [ (lib.mesonBool "b_asneeded" false) ] ++ prevAttrs.mesonFlags or [ ];
};
enableSanitizersLayer =
finalAttrs: prevAttrs:
let
sanitizers = lib.optional scope.withASan "address" ++ lib.optional scope.withUBSan "undefined";
in
{
mesonFlags =
(prevAttrs.mesonFlags or [ ])
++ lib.optionals (lib.length sanitizers > 0) (
[
(lib.mesonOption "b_sanitize" (lib.concatStringsSep "," sanitizers))
]
++ (lib.optionals stdenv.cc.isClang [
# https://www.github.com/mesonbuild/meson/issues/764
(lib.mesonBool "b_lundef" false)
])
);
};
nixDefaultsLayer = finalAttrs: prevAttrs: {
strictDeps = prevAttrs.strictDeps or true;
enableParallelBuilding = true;
@@ -246,6 +265,16 @@ in
inherit filesetToSource;
/**
Whether meson components are built with [AddressSanitizer](https://clang.llvm.org/docs/AddressSanitizer.html).
*/
withASan = false;
/**
Whether meson components are built with [UndefinedBehaviorSanitizer](https://clang.llvm.org/docs/UndefinedBehaviorSanitizer.html).
*/
withUBSan = false;
/**
A user-provided extension function to apply to each component derivation.
*/
@@ -332,6 +361,7 @@ in
setVersionLayer
mesonLayer
fixupStaticLayer
enableSanitizersLayer
scope.mesonComponentOverrides
];
mkMesonExecutable = mkPackageBuilder [
@@ -342,6 +372,7 @@ in
mesonLayer
mesonBuildLayer
fixupStaticLayer
enableSanitizersLayer
scope.mesonComponentOverrides
];
mkMesonLibrary = mkPackageBuilder [
@@ -353,6 +384,7 @@ in
mesonBuildLayer
mesonLibraryLayer
fixupStaticLayer
enableSanitizersLayer
scope.mesonComponentOverrides
];
@@ -406,6 +438,16 @@ in
*/
nix-external-api-docs = callPackage ../src/external-api-docs/package.nix { version = fineVersion; };
/**
JSON schema validation checks
*/
nix-json-schema-checks = callPackage ../src/json-schema-checks/package.nix { };
/**
Kaitai struct schema validation checks
*/
nix-kaitai-struct-checks = callPackage ../src/kaitai-struct-checks/package.nix { };
nix-perl-bindings = callPackage ../src/perl/package.nix { };
/**
@@ -458,7 +500,7 @@ in
Example:
```
overrideScope (finalScope: prevScope: { aws-sdk-cpp = null; })
overrideScope (finalScope: prevScope: { aws-crt-cpp = null; })
```
*/
overrideScope = f: (scope.overrideScope f).nix-everything;

View File

@@ -16,21 +16,6 @@ in
scope: {
inherit stdenv;
aws-sdk-cpp =
(pkgs.aws-sdk-cpp.override {
apis = [
"identity-management"
"s3"
"transfer"
];
customMemoryManagement = false;
}).overrideAttrs
{
# only a stripped down version is built, which takes a lot less resources
# to build, so we don't need a "big-parallel" machine.
requiredSystemFeatures = [ ];
};
boehmgc =
(pkgs.boehmgc.override {
enableLargeConfig = true;
@@ -90,37 +75,13 @@ scope: {
installPhase = lib.replaceStrings [ "--without-python" ] [ "" ] old.installPhase;
});
libgit2 =
if lib.versionAtLeast pkgs.libgit2.version "1.9.0" then
pkgs.libgit2
else
pkgs.libgit2.overrideAttrs (attrs: {
# libgit2: Nixpkgs 24.11 has < 1.9.0, which needs our patches
nativeBuildInputs =
attrs.nativeBuildInputs or [ ]
# gitMinimal does not build on Windows. See packbuilder patch.
++ lib.optionals (!stdenv.hostPlatform.isWindows) [
# Needed for `git apply`; see `prePatch`
pkgs.buildPackages.gitMinimal
];
# Only `git apply` can handle git binary patches
prePatch =
attrs.prePatch or ""
+ lib.optionalString (!stdenv.hostPlatform.isWindows) ''
patch() {
git apply
}
'';
patches =
attrs.patches or [ ]
++ [
./patches/libgit2-mempack-thin-packfile.patch
]
# gitMinimal does not build on Windows, but fortunately this patch only
# impacts interruptibility
++ lib.optionals (!stdenv.hostPlatform.isWindows) [
# binary patch; see `prePatch`
./patches/libgit2-packbuilder-callback-interruptible.patch
];
});
libgit2 = pkgs.libgit2.overrideAttrs (attrs: {
cmakeFlags = (attrs.cmakeFlags or [ ]) ++ [
(lib.mesonBool "EXPERIMENTAL_SHA256" true)
];
postInstall = (attrs.postInstall or "") + ''
substituteInPlace $(find $dev/include -type f) --replace-quiet '#include "git2/' '#include "git2-experimental/'
'';
});
}

View File

@@ -70,6 +70,9 @@ pkgs.nixComponents2.nix-util.overrideAttrs (
# We use this shell with the local checkout, not unpackPhase.
src = null;
# Workaround https://sourceware.org/pipermail/gdb-patches/2025-October/221398.html
# Remove when gdb fix is rolled out everywhere.
separateDebugInfo = false;
env = {
# For `make format`, to work without installing pre-commit
@@ -93,38 +96,45 @@ pkgs.nixComponents2.nix-util.overrideAttrs (
++ map (transformFlag "libcmd") (ignoreCrossFile pkgs.nixComponents2.nix-cmd.mesonFlags);
nativeBuildInputs =
attrs.nativeBuildInputs or [ ]
++ pkgs.nixComponents2.nix-util.nativeBuildInputs
++ pkgs.nixComponents2.nix-store.nativeBuildInputs
++ pkgs.nixComponents2.nix-fetchers.nativeBuildInputs
++ pkgs.nixComponents2.nix-expr.nativeBuildInputs
++ lib.optionals havePerl pkgs.nixComponents2.nix-perl-bindings.nativeBuildInputs
++ lib.optionals buildCanExecuteHost pkgs.nixComponents2.nix-manual.externalNativeBuildInputs
++ pkgs.nixComponents2.nix-internal-api-docs.nativeBuildInputs
++ pkgs.nixComponents2.nix-external-api-docs.nativeBuildInputs
++ pkgs.nixComponents2.nix-functional-tests.externalNativeBuildInputs
++ lib.optional (
!buildCanExecuteHost
# Hack around https://github.com/nixos/nixpkgs/commit/bf7ad8cfbfa102a90463433e2c5027573b462479
&& !(stdenv.hostPlatform.isWindows && stdenv.buildPlatform.isDarwin)
&& stdenv.hostPlatform.emulatorAvailable pkgs.buildPackages
&& lib.meta.availableOn stdenv.buildPlatform (stdenv.hostPlatform.emulator pkgs.buildPackages)
) pkgs.buildPackages.mesonEmulatorHook
++ [
pkgs.buildPackages.cmake
pkgs.buildPackages.gnused
pkgs.buildPackages.shellcheck
pkgs.buildPackages.changelog-d
modular.pre-commit.settings.package
(pkgs.writeScriptBin "pre-commit-hooks-install" modular.pre-commit.settings.installationScript)
pkgs.buildPackages.nixfmt-rfc-style
pkgs.buildPackages.shellcheck
pkgs.buildPackages.gdb
]
++ lib.optional (stdenv.cc.isClang && stdenv.hostPlatform == stdenv.buildPlatform) (
lib.hiPrio pkgs.buildPackages.clang-tools
)
++ lib.optional stdenv.hostPlatform.isLinux pkgs.buildPackages.mold-wrapped;
let
inputs =
attrs.nativeBuildInputs or [ ]
++ pkgs.nixComponents2.nix-util.nativeBuildInputs
++ pkgs.nixComponents2.nix-store.nativeBuildInputs
++ pkgs.nixComponents2.nix-fetchers.nativeBuildInputs
++ pkgs.nixComponents2.nix-expr.nativeBuildInputs
++ lib.optionals havePerl pkgs.nixComponents2.nix-perl-bindings.nativeBuildInputs
++ lib.optionals buildCanExecuteHost pkgs.nixComponents2.nix-manual.externalNativeBuildInputs
++ pkgs.nixComponents2.nix-internal-api-docs.nativeBuildInputs
++ pkgs.nixComponents2.nix-external-api-docs.nativeBuildInputs
++ pkgs.nixComponents2.nix-functional-tests.externalNativeBuildInputs
++ pkgs.nixComponents2.nix-json-schema-checks.externalNativeBuildInputs
++ lib.optional (
!buildCanExecuteHost
# Hack around https://github.com/nixos/nixpkgs/commit/bf7ad8cfbfa102a90463433e2c5027573b462479
&& !(stdenv.hostPlatform.isWindows && stdenv.buildPlatform.isDarwin)
&& stdenv.hostPlatform.emulatorAvailable pkgs.buildPackages
&& lib.meta.availableOn stdenv.buildPlatform (stdenv.hostPlatform.emulator pkgs.buildPackages)
) pkgs.buildPackages.mesonEmulatorHook
++ [
pkgs.buildPackages.cmake
pkgs.buildPackages.gnused
pkgs.buildPackages.changelog-d
modular.pre-commit.settings.package
(pkgs.writeScriptBin "pre-commit-hooks-install" modular.pre-commit.settings.installationScript)
pkgs.buildPackages.nixfmt-rfc-style
pkgs.buildPackages.shellcheck
pkgs.buildPackages.include-what-you-use
pkgs.buildPackages.gdb
]
++ lib.optional (stdenv.cc.isClang && stdenv.hostPlatform == stdenv.buildPlatform) (
lib.hiPrio pkgs.buildPackages.clang-tools
)
++ lib.optional stdenv.hostPlatform.isLinux pkgs.buildPackages.mold-wrapped;
in
# FIXME: separateDebugInfo = false doesn't actually prevent -Wa,--compress-debug-sections
# from making its way into NIX_CFLAGS_COMPILE.
lib.filter (p: !lib.hasInfix "separate-debug-info" p) inputs;
buildInputs = [
pkgs.gbenchmark

View File

@@ -62,6 +62,8 @@ let
"nix-cmd"
"nix-cli"
"nix-functional-tests"
"nix-json-schema-checks"
"nix-kaitai-struct-checks"
]
++ lib.optionals enableBindings [
"nix-perl-bindings"
@@ -73,7 +75,7 @@ let
]
);
in
{
rec {
/**
An internal check to make sure our package listing is complete.
*/
@@ -145,13 +147,25 @@ in
)
);
buildNoGc =
# Builds with sanitizers already have GC disabled, so this buildNoGc can just
# point to buildWithSanitizers in order to reduce the load on hydra.
buildNoGc = buildWithSanitizers;
buildWithSanitizers =
let
components = forAllSystems (
system:
nixpkgsFor.${system}.native.nixComponents2.overrideScope (
let
pkgs = nixpkgsFor.${system}.native;
in
pkgs.nixComponents2.overrideScope (
self: super: {
# Boost coroutines fail with ASAN on darwin.
withASan = !pkgs.stdenv.buildPlatform.isDarwin;
withUBSan = true;
nix-expr = super.nix-expr.override { enableGC = false; };
# Unclear how to make Perl bindings work with a dynamically linked ASAN.
nix-perl-bindings = null;
}
)
);

View File

@@ -1,282 +0,0 @@
commit 9bacade4a3ef4b6b26e2c02f549eef0e9eb9eaa2
Author: Robert Hensing <robert@roberthensing.nl>
Date: Sun Aug 18 20:20:36 2024 +0200
Add unoptimized git_mempack_write_thin_pack
diff --git a/include/git2/sys/mempack.h b/include/git2/sys/mempack.h
index 17da590a3..3688bdd50 100644
--- a/include/git2/sys/mempack.h
+++ b/include/git2/sys/mempack.h
@@ -44,6 +44,29 @@ GIT_BEGIN_DECL
*/
GIT_EXTERN(int) git_mempack_new(git_odb_backend **out);
+/**
+ * Write a thin packfile with the objects in the memory store.
+ *
+ * A thin packfile is a packfile that does not contain its transitive closure of
+ * references. This is useful for efficiently distributing additions to a
+ * repository over the network, but also finds use in the efficient bulk
+ * addition of objects to a repository, locally.
+ *
+ * This operation performs the (shallow) insert operations into the
+ * `git_packbuilder`, but does not write the packfile to disk;
+ * see `git_packbuilder_write_buf`.
+ *
+ * It also does not reset the memory store; see `git_mempack_reset`.
+ *
+ * @note This function may or may not write trees and blobs that are not
+ * referenced by commits. Currently everything is written, but this
+ * behavior may change in the future as the packer is optimized.
+ *
+ * @param backend The mempack backend
+ * @param pb The packbuilder to use to write the packfile
+ */
+GIT_EXTERN(int) git_mempack_write_thin_pack(git_odb_backend *backend, git_packbuilder *pb);
+
/**
* Dump all the queued in-memory writes to a packfile.
*
diff --git a/src/libgit2/odb_mempack.c b/src/libgit2/odb_mempack.c
index 6f27f45f8..0b61e2b66 100644
--- a/src/libgit2/odb_mempack.c
+++ b/src/libgit2/odb_mempack.c
@@ -132,6 +132,35 @@ cleanup:
return err;
}
+int git_mempack_write_thin_pack(git_odb_backend *backend, git_packbuilder *pb)
+{
+ struct memory_packer_db *db = (struct memory_packer_db *)backend;
+ const git_oid *oid;
+ size_t iter = 0;
+ int err = -1;
+
+ /* TODO: Implement the recency heuristics.
+ For this it probably makes sense to only write what's referenced
+ through commits, an option I've carved out for you in the docs.
+ wrt heuristics: ask your favorite LLM to translate https://git-scm.com/docs/pack-heuristics/en
+ to actual normal reference documentation. */
+ while (true) {
+ err = git_oidmap_iterate(NULL, db->objects, &iter, &oid);
+ if (err == GIT_ITEROVER) {
+ err = 0;
+ break;
+ }
+ if (err != 0)
+ return err;
+
+ err = git_packbuilder_insert(pb, oid, NULL);
+ if (err != 0)
+ return err;
+ }
+
+ return 0;
+}
+
int git_mempack_dump(
git_buf *pack,
git_repository *repo,
diff --git a/tests/libgit2/mempack/thinpack.c b/tests/libgit2/mempack/thinpack.c
new file mode 100644
index 000000000..604a4dda2
--- /dev/null
+++ b/tests/libgit2/mempack/thinpack.c
@@ -0,0 +1,196 @@
+#include "clar_libgit2.h"
+#include "git2/indexer.h"
+#include "git2/odb_backend.h"
+#include "git2/tree.h"
+#include "git2/types.h"
+#include "git2/sys/mempack.h"
+#include "git2/sys/odb_backend.h"
+#include "util.h"
+
+static git_repository *_repo;
+static git_odb_backend * _mempack_backend;
+
+void test_mempack_thinpack__initialize(void)
+{
+ git_odb *odb;
+
+ _repo = cl_git_sandbox_init_new("mempack_thinpack_repo");
+
+ cl_git_pass(git_mempack_new(&_mempack_backend));
+ cl_git_pass(git_repository_odb(&odb, _repo));
+ cl_git_pass(git_odb_add_backend(odb, _mempack_backend, 999));
+ git_odb_free(odb);
+}
+
+void _mempack_thinpack__cleanup(void)
+{
+ cl_git_sandbox_cleanup();
+}
+
+/*
+ Generating a packfile for an unchanged repo works and produces an empty packfile.
+ Even if we allow this scenario to be detected, it shouldn't misbehave if the
+ application is unaware of it.
+*/
+void test_mempack_thinpack__empty(void)
+{
+ git_packbuilder *pb;
+ int version;
+ int n;
+ git_buf buf = GIT_BUF_INIT;
+
+ git_packbuilder_new(&pb, _repo);
+
+ cl_git_pass(git_mempack_write_thin_pack(_mempack_backend, pb));
+ cl_git_pass(git_packbuilder_write_buf(&buf, pb));
+ cl_assert_in_range(12, buf.size, 1024 /* empty packfile is >0 bytes, but certainly not that big */);
+ cl_assert(buf.ptr[0] == 'P');
+ cl_assert(buf.ptr[1] == 'A');
+ cl_assert(buf.ptr[2] == 'C');
+ cl_assert(buf.ptr[3] == 'K');
+ version = (buf.ptr[4] << 24) | (buf.ptr[5] << 16) | (buf.ptr[6] << 8) | buf.ptr[7];
+ /* Subject to change. https://git-scm.com/docs/pack-format: Git currently accepts version number 2 or 3 but generates version 2 only.*/
+ cl_assert_equal_i(2, version);
+ n = (buf.ptr[8] << 24) | (buf.ptr[9] << 16) | (buf.ptr[10] << 8) | buf.ptr[11];
+ cl_assert_equal_i(0, n);
+ git_buf_dispose(&buf);
+
+ git_packbuilder_free(pb);
+}
+
+#define LIT_LEN(x) x, sizeof(x) - 1
+
+/*
+ Check that git_mempack_write_thin_pack produces a thin packfile.
+*/
+void test_mempack_thinpack__thin(void)
+{
+ /* Outline:
+ - Create tree 1
+ - Flush to packfile A
+ - Create tree 2
+ - Flush to packfile B
+
+ Tree 2 has a new blob and a reference to a blob from tree 1.
+
+ Expectation:
+ - Packfile B is thin and does not contain the objects from packfile A
+ */
+
+
+ git_oid oid_blob_1;
+ git_oid oid_blob_2;
+ git_oid oid_blob_3;
+ git_oid oid_tree_1;
+ git_oid oid_tree_2;
+ git_treebuilder *tb;
+
+ git_packbuilder *pb;
+ git_buf buf = GIT_BUF_INIT;
+ git_indexer *indexer;
+ git_indexer_progress stats;
+ char pack_dir_path[1024];
+
+ char sbuf[1024];
+ const char * repo_path;
+ const char * pack_name_1;
+ const char * pack_name_2;
+ git_str pack_path_1 = GIT_STR_INIT;
+ git_str pack_path_2 = GIT_STR_INIT;
+ git_odb_backend * pack_odb_backend_1;
+ git_odb_backend * pack_odb_backend_2;
+
+
+ cl_assert_in_range(0, snprintf(pack_dir_path, sizeof(pack_dir_path), "%s/objects/pack", git_repository_path(_repo)), sizeof(pack_dir_path));
+
+ /* Create tree 1 */
+
+ cl_git_pass(git_blob_create_from_buffer(&oid_blob_1, _repo, LIT_LEN("thinpack blob 1")));
+ cl_git_pass(git_blob_create_from_buffer(&oid_blob_2, _repo, LIT_LEN("thinpack blob 2")));
+
+
+ cl_git_pass(git_treebuilder_new(&tb, _repo, NULL));
+ cl_git_pass(git_treebuilder_insert(NULL, tb, "blob1", &oid_blob_1, GIT_FILEMODE_BLOB));
+ cl_git_pass(git_treebuilder_insert(NULL, tb, "blob2", &oid_blob_2, GIT_FILEMODE_BLOB));
+ cl_git_pass(git_treebuilder_write(&oid_tree_1, tb));
+
+ /* Flush */
+
+ cl_git_pass(git_packbuilder_new(&pb, _repo));
+ cl_git_pass(git_mempack_write_thin_pack(_mempack_backend, pb));
+ cl_git_pass(git_packbuilder_write_buf(&buf, pb));
+ cl_git_pass(git_indexer_new(&indexer, pack_dir_path, 0, NULL, NULL));
+ cl_git_pass(git_indexer_append(indexer, buf.ptr, buf.size, &stats));
+ cl_git_pass(git_indexer_commit(indexer, &stats));
+ pack_name_1 = strdup(git_indexer_name(indexer));
+ cl_assert(pack_name_1);
+ git_buf_dispose(&buf);
+ git_mempack_reset(_mempack_backend);
+ git_indexer_free(indexer);
+ git_packbuilder_free(pb);
+
+ /* Create tree 2 */
+
+ cl_git_pass(git_treebuilder_clear(tb));
+ /* blob 1 won't be used, but we add it anyway to test that just "declaring" an object doesn't
+ necessarily cause its inclusion in the next thin packfile. It must only be included if new. */
+ cl_git_pass(git_blob_create_from_buffer(&oid_blob_1, _repo, LIT_LEN("thinpack blob 1")));
+ cl_git_pass(git_blob_create_from_buffer(&oid_blob_3, _repo, LIT_LEN("thinpack blob 3")));
+ cl_git_pass(git_treebuilder_insert(NULL, tb, "blob1", &oid_blob_1, GIT_FILEMODE_BLOB));
+ cl_git_pass(git_treebuilder_insert(NULL, tb, "blob3", &oid_blob_3, GIT_FILEMODE_BLOB));
+ cl_git_pass(git_treebuilder_write(&oid_tree_2, tb));
+
+ /* Flush */
+
+ cl_git_pass(git_packbuilder_new(&pb, _repo));
+ cl_git_pass(git_mempack_write_thin_pack(_mempack_backend, pb));
+ cl_git_pass(git_packbuilder_write_buf(&buf, pb));
+ cl_git_pass(git_indexer_new(&indexer, pack_dir_path, 0, NULL, NULL));
+ cl_git_pass(git_indexer_append(indexer, buf.ptr, buf.size, &stats));
+ cl_git_pass(git_indexer_commit(indexer, &stats));
+ pack_name_2 = strdup(git_indexer_name(indexer));
+ cl_assert(pack_name_2);
+ git_buf_dispose(&buf);
+ git_mempack_reset(_mempack_backend);
+ git_indexer_free(indexer);
+ git_packbuilder_free(pb);
+ git_treebuilder_free(tb);
+
+ /* Assertions */
+
+ assert(pack_name_1);
+ assert(pack_name_2);
+
+ repo_path = git_repository_path(_repo);
+
+ snprintf(sbuf, sizeof(sbuf), "objects/pack/pack-%s.pack", pack_name_1);
+ git_str_joinpath(&pack_path_1, repo_path, sbuf);
+ snprintf(sbuf, sizeof(sbuf), "objects/pack/pack-%s.pack", pack_name_2);
+ git_str_joinpath(&pack_path_2, repo_path, sbuf);
+
+ /* If they're the same, something definitely went wrong. */
+ cl_assert(strcmp(pack_name_1, pack_name_2) != 0);
+
+ cl_git_pass(git_odb_backend_one_pack(&pack_odb_backend_1, pack_path_1.ptr));
+ cl_assert(pack_odb_backend_1->exists(pack_odb_backend_1, &oid_blob_1));
+ cl_assert(pack_odb_backend_1->exists(pack_odb_backend_1, &oid_blob_2));
+ cl_assert(!pack_odb_backend_1->exists(pack_odb_backend_1, &oid_blob_3));
+ cl_assert(pack_odb_backend_1->exists(pack_odb_backend_1, &oid_tree_1));
+ cl_assert(!pack_odb_backend_1->exists(pack_odb_backend_1, &oid_tree_2));
+
+ cl_git_pass(git_odb_backend_one_pack(&pack_odb_backend_2, pack_path_2.ptr));
+ /* blob 1 is already in the packfile 1, so packfile 2 must not include it, in order to be _thin_. */
+ cl_assert(!pack_odb_backend_2->exists(pack_odb_backend_2, &oid_blob_1));
+ cl_assert(!pack_odb_backend_2->exists(pack_odb_backend_2, &oid_blob_2));
+ cl_assert(pack_odb_backend_2->exists(pack_odb_backend_2, &oid_blob_3));
+ cl_assert(!pack_odb_backend_2->exists(pack_odb_backend_2, &oid_tree_1));
+ cl_assert(pack_odb_backend_2->exists(pack_odb_backend_2, &oid_tree_2));
+
+ pack_odb_backend_1->free(pack_odb_backend_1);
+ pack_odb_backend_2->free(pack_odb_backend_2);
+ free((void *)pack_name_1);
+ free((void *)pack_name_2);
+ git_str_dispose(&pack_path_1);
+ git_str_dispose(&pack_path_2);
+
+}

View File

@@ -1,930 +0,0 @@
commit e9823c5da4fa977c46bcb97167fbdd0d70adb5ff
Author: Robert Hensing <robert@roberthensing.nl>
Date: Mon Aug 26 20:07:04 2024 +0200
Make packbuilder interruptible using progress callback
Forward errors from packbuilder->progress_cb
This allows the callback to terminate long-running operations when
the application is interrupted.
diff --git a/include/git2/pack.h b/include/git2/pack.h
index 0f6bd2ab9..bee72a6c0 100644
--- a/include/git2/pack.h
+++ b/include/git2/pack.h
@@ -247,6 +247,9 @@ typedef int GIT_CALLBACK(git_packbuilder_progress)(
* @param progress_cb Function to call with progress information during
* pack building. Be aware that this is called inline with pack building
* operations, so performance may be affected.
+ * When progress_cb returns an error, the pack building process will be
+ * aborted and the error will be returned from the invoked function.
+ * `pb` must then be freed.
* @param progress_cb_payload Payload for progress callback.
* @return 0 or an error code
*/
diff --git a/src/libgit2/pack-objects.c b/src/libgit2/pack-objects.c
index b2d80cba9..7c331c2d5 100644
--- a/src/libgit2/pack-objects.c
+++ b/src/libgit2/pack-objects.c
@@ -932,6 +932,9 @@ static int report_delta_progress(
{
int ret;
+ if (pb->failure)
+ return pb->failure;
+
if (pb->progress_cb) {
uint64_t current_time = git_time_monotonic();
uint64_t elapsed = current_time - pb->last_progress_report_time;
@@ -943,8 +946,10 @@ static int report_delta_progress(
GIT_PACKBUILDER_DELTAFICATION,
count, pb->nr_objects, pb->progress_cb_payload);
- if (ret)
+ if (ret) {
+ pb->failure = ret;
return git_error_set_after_callback(ret);
+ }
}
}
@@ -976,7 +981,10 @@ static int find_deltas(git_packbuilder *pb, git_pobject **list,
}
pb->nr_deltified += 1;
- report_delta_progress(pb, pb->nr_deltified, false);
+ if ((error = report_delta_progress(pb, pb->nr_deltified, false)) < 0) {
+ GIT_ASSERT(git_packbuilder__progress_unlock(pb) == 0);
+ goto on_error;
+ }
po = *list++;
(*list_size)--;
@@ -1124,6 +1132,10 @@ struct thread_params {
size_t depth;
size_t working;
size_t data_ready;
+
+ /* A pb->progress_cb can stop the packing process by returning an error.
+ When that happens, all threads observe the error and stop voluntarily. */
+ bool stopped;
};
static void *threaded_find_deltas(void *arg)
@@ -1133,7 +1145,12 @@ static void *threaded_find_deltas(void *arg)
while (me->remaining) {
if (find_deltas(me->pb, me->list, &me->remaining,
me->window, me->depth) < 0) {
- ; /* TODO */
+ me->stopped = true;
+ GIT_ASSERT_WITH_RETVAL(git_packbuilder__progress_lock(me->pb) == 0, NULL);
+ me->working = false;
+ git_cond_signal(&me->pb->progress_cond);
+ GIT_ASSERT_WITH_RETVAL(git_packbuilder__progress_unlock(me->pb) == 0, NULL);
+ return NULL;
}
GIT_ASSERT_WITH_RETVAL(git_packbuilder__progress_lock(me->pb) == 0, NULL);
@@ -1175,8 +1192,7 @@ static int ll_find_deltas(git_packbuilder *pb, git_pobject **list,
pb->nr_threads = git__online_cpus();
if (pb->nr_threads <= 1) {
- find_deltas(pb, list, &list_size, window, depth);
- return 0;
+ return find_deltas(pb, list, &list_size, window, depth);
}
p = git__mallocarray(pb->nr_threads, sizeof(*p));
@@ -1195,6 +1211,7 @@ static int ll_find_deltas(git_packbuilder *pb, git_pobject **list,
p[i].depth = depth;
p[i].working = 1;
p[i].data_ready = 0;
+ p[i].stopped = 0;
/* try to split chunks on "path" boundaries */
while (sub_size && sub_size < list_size &&
@@ -1262,7 +1279,7 @@ static int ll_find_deltas(git_packbuilder *pb, git_pobject **list,
(!victim || victim->remaining < p[i].remaining))
victim = &p[i];
- if (victim) {
+ if (victim && !target->stopped) {
sub_size = victim->remaining / 2;
list = victim->list + victim->list_size - sub_size;
while (sub_size && list[0]->hash &&
@@ -1286,7 +1303,7 @@ static int ll_find_deltas(git_packbuilder *pb, git_pobject **list,
}
target->list_size = sub_size;
target->remaining = sub_size;
- target->working = 1;
+ target->working = 1; /* even when target->stopped, so that we don't process this thread again */
GIT_ASSERT(git_packbuilder__progress_unlock(pb) == 0);
if (git_mutex_lock(&target->mutex)) {
@@ -1299,7 +1316,7 @@ static int ll_find_deltas(git_packbuilder *pb, git_pobject **list,
git_cond_signal(&target->cond);
git_mutex_unlock(&target->mutex);
- if (!sub_size) {
+ if (target->stopped || !sub_size) {
git_thread_join(&target->thread, NULL);
git_cond_free(&target->cond);
git_mutex_free(&target->mutex);
@@ -1308,7 +1325,7 @@ static int ll_find_deltas(git_packbuilder *pb, git_pobject **list,
}
git__free(p);
- return 0;
+ return pb->failure;
}
#else
@@ -1319,6 +1336,7 @@ int git_packbuilder__prepare(git_packbuilder *pb)
{
git_pobject **delta_list;
size_t i, n = 0;
+ int error;
if (pb->nr_objects == 0 || pb->done)
return 0; /* nothing to do */
@@ -1327,8 +1345,10 @@ int git_packbuilder__prepare(git_packbuilder *pb)
* Although we do not report progress during deltafication, we
* at least report that we are in the deltafication stage
*/
- if (pb->progress_cb)
- pb->progress_cb(GIT_PACKBUILDER_DELTAFICATION, 0, pb->nr_objects, pb->progress_cb_payload);
+ if (pb->progress_cb) {
+ if ((error = pb->progress_cb(GIT_PACKBUILDER_DELTAFICATION, 0, pb->nr_objects, pb->progress_cb_payload)) < 0)
+ return git_error_set_after_callback(error);
+ }
delta_list = git__mallocarray(pb->nr_objects, sizeof(*delta_list));
GIT_ERROR_CHECK_ALLOC(delta_list);
@@ -1345,31 +1365,33 @@ int git_packbuilder__prepare(git_packbuilder *pb)
if (n > 1) {
git__tsort((void **)delta_list, n, type_size_sort);
- if (ll_find_deltas(pb, delta_list, n,
+ if ((error = ll_find_deltas(pb, delta_list, n,
GIT_PACK_WINDOW + 1,
- GIT_PACK_DEPTH) < 0) {
+ GIT_PACK_DEPTH)) < 0) {
git__free(delta_list);
- return -1;
+ return error;
}
}
- report_delta_progress(pb, pb->nr_objects, true);
+ error = report_delta_progress(pb, pb->nr_objects, true);
pb->done = true;
git__free(delta_list);
- return 0;
+ return error;
}
-#define PREPARE_PACK if (git_packbuilder__prepare(pb) < 0) { return -1; }
+#define PREPARE_PACK error = git_packbuilder__prepare(pb); if (error < 0) { return error; }
int git_packbuilder_foreach(git_packbuilder *pb, int (*cb)(void *buf, size_t size, void *payload), void *payload)
{
+ int error;
PREPARE_PACK;
return write_pack(pb, cb, payload);
}
int git_packbuilder__write_buf(git_str *buf, git_packbuilder *pb)
{
+ int error;
PREPARE_PACK;
return write_pack(pb, &write_pack_buf, buf);
diff --git a/src/libgit2/pack-objects.h b/src/libgit2/pack-objects.h
index bbc8b9430..380a28ebe 100644
--- a/src/libgit2/pack-objects.h
+++ b/src/libgit2/pack-objects.h
@@ -100,6 +100,10 @@ struct git_packbuilder {
uint64_t last_progress_report_time;
bool done;
+
+ /* A non-zero error code in failure causes all threads to shut themselves
+ down. Some functions will return this error code. */
+ volatile int failure;
};
int git_packbuilder__write_buf(git_str *buf, git_packbuilder *pb);
diff --git a/tests/libgit2/pack/cancel.c b/tests/libgit2/pack/cancel.c
new file mode 100644
index 000000000..a0aa9716a
--- /dev/null
+++ b/tests/libgit2/pack/cancel.c
@@ -0,0 +1,240 @@
+#include "clar_libgit2.h"
+#include "futils.h"
+#include "pack.h"
+#include "hash.h"
+#include "iterator.h"
+#include "vector.h"
+#include "posix.h"
+#include "hash.h"
+#include "pack-objects.h"
+
+static git_repository *_repo;
+static git_revwalk *_revwalker;
+static git_packbuilder *_packbuilder;
+static git_indexer *_indexer;
+static git_vector _commits;
+static int _commits_is_initialized;
+static git_indexer_progress _stats;
+
+extern bool git_disable_pack_keep_file_checks;
+
+static void pack_packbuilder_init(const char *sandbox) {
+ _repo = cl_git_sandbox_init(sandbox);
+ /* cl_git_pass(p_chdir(sandbox)); */
+ cl_git_pass(git_revwalk_new(&_revwalker, _repo));
+ cl_git_pass(git_packbuilder_new(&_packbuilder, _repo));
+ cl_git_pass(git_vector_init(&_commits, 0, NULL));
+ _commits_is_initialized = 1;
+ memset(&_stats, 0, sizeof(_stats));
+ p_fsync__cnt = 0;
+}
+
+void test_pack_cancel__initialize(void)
+{
+ pack_packbuilder_init("small.git");
+}
+
+void test_pack_cancel__cleanup(void)
+{
+ git_oid *o;
+ unsigned int i;
+
+ cl_git_pass(git_libgit2_opts(GIT_OPT_ENABLE_FSYNC_GITDIR, 0));
+ cl_git_pass(git_libgit2_opts(GIT_OPT_DISABLE_PACK_KEEP_FILE_CHECKS, false));
+
+ if (_commits_is_initialized) {
+ _commits_is_initialized = 0;
+ git_vector_foreach(&_commits, i, o) {
+ git__free(o);
+ }
+ git_vector_free(&_commits);
+ }
+
+ git_packbuilder_free(_packbuilder);
+ _packbuilder = NULL;
+
+ git_revwalk_free(_revwalker);
+ _revwalker = NULL;
+
+ git_indexer_free(_indexer);
+ _indexer = NULL;
+
+ /* cl_git_pass(p_chdir("..")); */
+ cl_git_sandbox_cleanup();
+ _repo = NULL;
+}
+
+static int seed_packbuilder(void)
+{
+ int error;
+ git_oid oid, *o;
+ unsigned int i;
+
+ git_revwalk_sorting(_revwalker, GIT_SORT_TIME);
+ cl_git_pass(git_revwalk_push_ref(_revwalker, "HEAD"));
+
+ while (git_revwalk_next(&oid, _revwalker) == 0) {
+ o = git__malloc(sizeof(git_oid));
+ cl_assert(o != NULL);
+ git_oid_cpy(o, &oid);
+ cl_git_pass(git_vector_insert(&_commits, o));
+ }
+
+ git_vector_foreach(&_commits, i, o) {
+ if((error = git_packbuilder_insert(_packbuilder, o, NULL)) < 0)
+ return error;
+ }
+
+ git_vector_foreach(&_commits, i, o) {
+ git_object *obj;
+ cl_git_pass(git_object_lookup(&obj, _repo, o, GIT_OBJECT_COMMIT));
+ error = git_packbuilder_insert_tree(_packbuilder,
+ git_commit_tree_id((git_commit *)obj));
+ git_object_free(obj);
+ if (error < 0)
+ return error;
+ }
+
+ return 0;
+}
+
+static int fail_stage;
+
+static int packbuilder_cancel_after_n_calls_cb(int stage, uint32_t current, uint32_t total, void *payload)
+{
+
+ /* Force the callback to run again on the next opportunity regardless
+ of how fast we're running. */
+ _packbuilder->last_progress_report_time = 0;
+
+ if (stage == fail_stage) {
+ int *calls = (int *)payload;
+ int n = *calls;
+ /* Always decrement, including past zero. This way the error is only
+ triggered once, making sure it is picked up immediately. */
+ --*calls;
+ if (n == 0)
+ return GIT_EUSER;
+ }
+
+ return 0;
+}
+
+static void test_cancel(int n)
+{
+
+ int calls_remaining = n;
+ int err;
+ git_buf buf = GIT_BUF_INIT;
+
+ /* Switch to a small repository, so that `packbuilder_cancel_after_n_calls_cb`
+ can hack the time to call the callback on every opportunity. */
+
+ cl_git_pass(git_packbuilder_set_callbacks(_packbuilder, &packbuilder_cancel_after_n_calls_cb, &calls_remaining));
+ err = seed_packbuilder();
+ if (!err)
+ err = git_packbuilder_write_buf(&buf, _packbuilder);
+
+ cl_assert_equal_i(GIT_EUSER, err);
+}
+void test_pack_cancel__cancel_after_add_0(void)
+{
+ fail_stage = GIT_PACKBUILDER_ADDING_OBJECTS;
+ test_cancel(0);
+}
+
+void test_pack_cancel__cancel_after_add_1(void)
+{
+ cl_skip();
+ fail_stage = GIT_PACKBUILDER_ADDING_OBJECTS;
+ test_cancel(1);
+}
+
+void test_pack_cancel__cancel_after_delta_0(void)
+{
+ fail_stage = GIT_PACKBUILDER_DELTAFICATION;
+ test_cancel(0);
+}
+
+void test_pack_cancel__cancel_after_delta_1(void)
+{
+ fail_stage = GIT_PACKBUILDER_DELTAFICATION;
+ test_cancel(1);
+}
+
+void test_pack_cancel__cancel_after_delta_0_threaded(void)
+{
+#ifdef GIT_THREADS
+ git_packbuilder_set_threads(_packbuilder, 8);
+ fail_stage = GIT_PACKBUILDER_DELTAFICATION;
+ test_cancel(0);
+#else
+ cl_skip();
+#endif
+}
+
+void test_pack_cancel__cancel_after_delta_1_threaded(void)
+{
+#ifdef GIT_THREADS
+ git_packbuilder_set_threads(_packbuilder, 8);
+ fail_stage = GIT_PACKBUILDER_DELTAFICATION;
+ test_cancel(1);
+#else
+ cl_skip();
+#endif
+}
+
+static int foreach_cb(void *buf, size_t len, void *payload)
+{
+ git_indexer *idx = (git_indexer *) payload;
+ cl_git_pass(git_indexer_append(idx, buf, len, &_stats));
+ return 0;
+}
+
+void test_pack_cancel__foreach(void)
+{
+ git_indexer *idx;
+
+ seed_packbuilder();
+
+#ifdef GIT_EXPERIMENTAL_SHA256
+ cl_git_pass(git_indexer_new(&idx, ".", GIT_OID_SHA1, NULL));
+#else
+ cl_git_pass(git_indexer_new(&idx, ".", 0, NULL, NULL));
+#endif
+
+ cl_git_pass(git_packbuilder_foreach(_packbuilder, foreach_cb, idx));
+ cl_git_pass(git_indexer_commit(idx, &_stats));
+ git_indexer_free(idx);
+}
+
+static int foreach_cancel_cb(void *buf, size_t len, void *payload)
+{
+ git_indexer *idx = (git_indexer *)payload;
+ cl_git_pass(git_indexer_append(idx, buf, len, &_stats));
+ return (_stats.total_objects > 2) ? -1111 : 0;
+}
+
+void test_pack_cancel__foreach_with_cancel(void)
+{
+ git_indexer *idx;
+
+ seed_packbuilder();
+
+#ifdef GIT_EXPERIMENTAL_SHA256
+ cl_git_pass(git_indexer_new(&idx, ".", GIT_OID_SHA1, NULL));
+#else
+ cl_git_pass(git_indexer_new(&idx, ".", 0, NULL, NULL));
+#endif
+
+ cl_git_fail_with(
+ git_packbuilder_foreach(_packbuilder, foreach_cancel_cb, idx), -1111);
+ git_indexer_free(idx);
+}
+
+void test_pack_cancel__keep_file_check(void)
+{
+ assert(!git_disable_pack_keep_file_checks);
+ cl_git_pass(git_libgit2_opts(GIT_OPT_DISABLE_PACK_KEEP_FILE_CHECKS, true));
+ assert(git_disable_pack_keep_file_checks);
+}
diff --git a/tests/resources/small.git/HEAD b/tests/resources/small.git/HEAD
new file mode 100644
index 0000000000000000000000000000000000000000..cb089cd89a7d7686d284d8761201649346b5aa1c
GIT binary patch
literal 23
ecmXR)O|w!cN=+-)&qz&7Db~+TEG|hc;sO9;xClW2
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/config b/tests/resources/small.git/config
new file mode 100644
index 0000000000000000000000000000000000000000..07d359d07cf1ed0c0074fdad71ffff5942f0adfa
GIT binary patch
literal 66
zcmaz}&M!)h<>D+#Eyyp<EXgmbOv^9IO)M!(Eh^5;&r`5fFyP`$%gjm5%}+@M@=A(I
MQ@J>k5{uv*03B5png9R*
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/description b/tests/resources/small.git/description
new file mode 100644
index 0000000000000000000000000000000000000000..498b267a8c7812490d6479839c5577eaaec79d62
GIT binary patch
literal 73
zcmWH|%S+5nO;IRHEyyp<EXgmbv{pz>$t+PQ$;d2LNXyJgRZve!Elw`VEGWs$&r??@
Q$yWgB0LrH#Y0~2Y0PnOK(EtDd
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/applypatch-msg.sample b/tests/resources/small.git/hooks/applypatch-msg.sample
new file mode 100755
index 0000000000000000000000000000000000000000..dcbf8167fa503f96ff6a39c68409007eadc9b1f3
GIT binary patch
literal 535
zcmY+AX;Q;542A#a6e8^~FyI8r&I~hf2QJ{GO6(?HuvEG*+#R{4EI%zhfA8r{j%sh$
zHE~E-UtQd8{bq4@*S%jq3@bmxwQDXGv#o!N`o3AHMw3xD)hy0#>&E&zzl%vRffo<B
z)-H|+CWHZ~O*S%cfYx9;02_ohIA<Bg(1SxF-6OCb&_lBkf{t<AM9r;%E(Hf#h{|a@
z9>mqo=v6>_2NRa#TwDdYvTVQyueO*15Nlo%=#DXgC0bhF3vTa`LQGaO9;jeD$OP?~
za$G4Q{z+Q_{5V?5h;a-noM$P{<>Q~j4o7u%#P6^o^16{y*jU=-K8GYD_dUtdj4FSx
zSC0C!DvAnv%S!4d<Yg@O<;m`;oSw)=Fz+hrL<mY{rBr8j4pi^88FX3}jKrYUP)>gk
XB^)11aoGMJPCqWs%IS0YSv(eBT&%T6
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/commit-msg.sample b/tests/resources/small.git/hooks/commit-msg.sample
new file mode 100755
index 0000000000000000000000000000000000000000..f3780f92349638ebe32f6baf24c7c3027675d7c9
GIT binary patch
literal 953
zcmaJ<TW`}a6n<`g#Ya}rZCaYOzy?UGG&Tf#VG`?}7@eHtB=MTqneFUS%75oLS;atz
zm&iUo=ey->y@-{3h^^Cx;#d0zEA@DDc$nY4ez&|=%jTg@_HU*ub=!!y$xW09TSjlj
z(`I@QCsM`!9&80$I98wsQ8yK#)Orb<8re8FjkKh630D$QUDwi~(gkX=RunYm$rDjk
zlp%RUSnzA#6yjdG5?T?2DcYKp+v_lts0ljn&bh3J0bD5@N@1UKZ190O6ZeWr-BuZ^
zWRebCX%(%=Xoj#(xYk1Cjtr!=tyBesf@m6}8zY6Ijbz9i9ziI_jG9Mv<Cz(ymp*>R
zDH*e>^ga9IR?2wrSrAVm;eButj4<aWB@zzNl|1Wp@4;}1O?MUF>Y>7(E2?b~jsu>&
zRKCJ7bp#19sqYh627wD%D9R$8=Ml$TNlumDypl~$jBu*G>5fIR^FB0h0Ex&TGZNr>
zL5hs1_K>taRb!|ThN9ns7^@4MXKP+6aGI_UK)T-M#rcP$;kN(Vcf#P)+5GzWa{l@J
z>-E{`$1iiNVYxq27}<DnwLRXQUG0o_hw&da-s5T#H=`Y9D_8=eTZ?cpWatp#a1vs@
z2BjrO)z@aTuI#g#`)oJcnhM7oYLT@~CHX@CNXv4>j;uo%;)r3kJI2xCFF~Ux;$Q%)
wjbk6JlDCM`jU&P+UVOvg`|iYl<7~9k>HHB4I;pdlQ=I-^$DrHaN$@lH1?P!0U;qFB
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/fsmonitor-watchman.sample b/tests/resources/small.git/hooks/fsmonitor-watchman.sample
new file mode 100755
index 0000000000000000000000000000000000000000..41184ebc318c159f51cd1ebe2290559805df89d8
GIT binary patch
literal 4777
zcmbtYYi}F368$Xwipg4lq(BeHMvzvH-4;n7DGJBPqq#tw3aed8+IU5-m)yvL>;Cqh
z8FFRGj$`9CA8ao<GaSz2oMCnz4Rv-gw9b@j_$0GQT1?XYi|;S??Y`Z6`Z;}C6#27N
z8ShS>J?j^$%==FV``-=rhLcPW`McSytRm~mEO7_&_cAVZrf1fFy*ha@8oe%*-aBYE
zcjzZg>LOkgxuUr-XJnHyD;zmPnRaSc#!k_P*d_BttRdc+J6G7za5#+<HG#rlmbrN~
z8DwU-3}VABEwM=0VLP@^Dy6ERR5_J6cmg|GEh*M1EliqCGwe^ZT-iZ$2Yc`4!I#WZ
z5nGGhn7*jeW=2ydsmfAmm#=8AD<<;TI+#z{Q)kW;yE!%GB6f~7EtEMLdM47Qaz*8=
zIObA(VVj-VG{Ax|66d*hi`+bRG>^Y1nkc2Oowk`ya47uUR3Feu?B<phm31&PQB<lt
zb{W(W4wf#Bab%|Q_tKPS?3^o=5)Z8^Vh(#slNI}pO(f^|{U0GZhLnycSaNd&h?CaC
z0XklU6^<ky6rES9T=na$M8P<_)aKMAMo+UDewAu4wF{#&6diFshiudixAoh|&0<BJ
zR>(w;S{(VYzxh}q-=#zP@uxSx{wbyPUMFU;K(06)$o{07&3yI?q{GqMcQ1c_^M<0<
zF4acAV)Il-V(rCTC1(;bsZ*}bl8dmejAk~yb`B}!^0;g^(o9kGUfZfDOvyp@x4OQt
zSgWh6T|3eq;9MFs8-#z+FDM1h(IjRUP|``PxupgJ7CUHOH90gbgl^2~97`?_X{P))
zB*$r1cDlF-%azKND}?Gv`2K8-9v5e`gQoft=j?T<&a13c^!wY_$D`5z-X1g?ty&6-
zQN50{8?bUk9AI->^W@~~nkOghHIC2YN+<JiT_ob7ttND1oh`HH28Y+OV~HedG&uB`
zy}rA*r_xT#bR`Q7-*)3t*(!Hf?jKzyxk=8hdi3L^d<p<uU9q_<4k&xEr4@YWv_vsW
zp(#32bYtA5u|s#1+}h`0kwpz4kUD&+>AXkLQG_2-{Pq3%{`3KUMeG$iIn%%^6*NYb
zn|_BdV#C)n4565Vcc<EWC-nglECZGy!j9I4&;hUCzr(?6TftH=0^@!mI^E@y5HZw8
ztH&kaSNyg=O6riqR^MPOX6oD__Jz@K;*icS)?m$!p{S$n;*EwlO<L!d7;utu(W9D!
zaF!B~xV^2i?wH0s?Lw%Q)(`aPkajs1XojlPv@Y-d5#YFg#KG+!r7A(dJLnkiJMs`p
zV|_=d!upN{TsxH1?sZNdzxeHsmtzTV`1{pykJ_~+^*>X;uT8&z3vSi!HXGbUj2B!R
zdz~&#<?<tHJql=W&((YpzS06y-Z6Cn^H!*9qH?pOrD~(GV=JJ~z{tpLnGK|xp&C1`
zsbn7O86xjF<~G*S1l*;;Bh%6><Up=oKy99?62P^?M&22U6QFRBXLb&u%=Ur<74wRy
zMRG!f{AvZ>fk#L-&k$fLwo$4?>12g@AXOKFekuo#6EHB%gmpD?1eyh%N8s{2wGoTu
z*@6cEZ^ZW!FAF_|JL`NkV7k}0ow|-2jHwbgH0;c@Dq*o?@&c*HnGdyx6^su8Qk%2{
z*ye(dxO*6-&>qn1+zw}tc6;=sOX{4WB=VqjTS^))y1jlX2Q;=e!qMmFA5lC$#;BxC
z=Y%tRpWxb+_uQAvAw7Q{HGV#R$xb&udLCzZ+HN?kTyB};1EJ8UlQ5!>5eGW@)RX0n
zkjj>EF!3=0Gl^8dzv$B^NMGRxJoqN4A`xq-@wCbrx*u2NmIJ1<fUDQ=*^)h6`vzco
z3F+ro$F!F6pc<B;<;isobIgbVGKUBByoQ4#CO({S7g?<Dh0^!7uJ3gxS=6F;+^gQc
zeKi4`4`Fm3p|BU2v{M|-u!#QGCXiM=k=%np0<ZOPQqEjP_nneyOdgEuG9IH&YYPtp
zKB_dvcYCcyhyT#<uhUEL$o~!TUz;cb&|`uSM{Dkv%&W2lcpYL&kv)tUvVv?&>xZ%H
zh;{|4T3(!E9sY#Ni(wUJYs1MmIc9bl)(4Nl3_wD_BWB>i<1S(LX7m*{Q7PU$muMS*
zM!%0EZx-Vw=Zey;erC?SNxF;pY@^A%-krqzfLV2meBp1vWdyArFYn`DD19T)Hw(?n
z)}{NP(Lk(o*?gl#B@pP7^*r|=;PIDT4|F#{2Hzh-AL0Rv$6uT;<CP7qxbYms@WU7}
z%}TsHJ!V_56ZFF{BfI=8jTBxMEATG376pS6a;u1@c?{~sL<N52`U6fuLkz4P@Mb^b
z2`Q48y&!C0&A+WRCJAdmo3u2#?eI=si9Vm47$|`m**x<wKkM=QR&g?C63@P5X@xP8
zi5QG2b-Fdz%S0%1{kKvL%^RTdpK|SU^VJ7JCmKdwp1`;HRoGM7ef^k_s_}2An=cql
za|{IaxQWpdq<ae&x3BOJq+M5QNIk#C{Nv@u>n|WzE4=slK?on@(fZeGhRgQCu56qB
z{+n81Az96qnQjMY*-*r-KV*7;Z#4Q<xfjbcBx_6JAN-HP@bq+eI%HhAB9&vLyOap{
bw<Ywj(b#kdcCk7dCBY;|DBNpPjIa1F6*dXN
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/post-update.sample b/tests/resources/small.git/hooks/post-update.sample
new file mode 100755
index 0000000000000000000000000000000000000000..9f95e41a39cf974958277c8881ac6cce405ebb20
GIT binary patch
literal 246
zcmXZVO?HDY3<Ti4Poaiwjq^yGw9DKf7mz^&Ly=V5q$H=W^Rt|p_r9s#9Ea7VERo!9
zyT9>uJRJJV$M^KdldiMhj?ImK6~FvwJ*L5a){QoM=L5TYHkGO1$UrO3`a>{?Opw|b
zG(#59NQ#jFL9v~vgOVkM@^^(^A}onOE))yWEwhIlk&{ZyseZ^O0b=w8&O=BK{k<5B
k^Q-B@eG}LeHrquz%(SVEp_N)VhYZikCW__82JXfD17`J9Qvd(}
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/pre-applypatch.sample b/tests/resources/small.git/hooks/pre-applypatch.sample
new file mode 100755
index 0000000000000000000000000000000000000000..625837e25f91421b8809a097f4a3103dd387ef31
GIT binary patch
literal 481
zcmY+ATTa6;5Jms9iouO45IBJXEg&Jm9@v1LPHMM_ZR|;#6tQ<EeSrA%_2`_rGr1_8
z?aM?yVtIc%-@9SGSk&8x=grP-Lf`7!^=$7xgL=|ysZ}!av6zL~ywui}<2##V6L@!k
zy=p^)V7%Wzs-g`9<Y9}^)&uN}BCrXR_T3@Z2$gSJON2`X=mAs+%@7n-2I}ZrP|TFA
zvJJGDl3HPLP<@!Q!}zXQvey#qEE#a#$vs97i4=A0stF@YQ)k_ZajaoS^dVYBc&37_
zVI(L=X<V335r9~7T<;|HfKF+yM}}LB9d96V)Si;sj(;9Rh$#P>h$71hSXq*MxP;V&
zj0cY7SCL=x4`a46sF)C>94Gk%=3q$W2s;j6iHtB2$R0%gix4oK@&T~=ALd_o*CKxt
I-`Pv{1Bpzc>;M1&
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/pre-commit.sample b/tests/resources/small.git/hooks/pre-commit.sample
new file mode 100755
index 0000000000000000000000000000000000000000..10b39b2e26981b8f87ea424e735ef87359066dbb
GIT binary patch
literal 1706
zcmZuxU2ohr5PY_N#pZ0-F<{-<Zkx1j&cMNSQ3FN`GzR*R1O+9oPV`BnOj7q@1pV!u
zrF0Hl^i33(yR)-1d-!H%&2|=|^E~_R{N1zNJ-&Zmt-t?iwXv&i+ZN}Km(TX8Q$H4u
zd7(m`|1iDmF5k@xV`p;C4zojASmLc}yN0QDZbhN=ri&CEt=XGuN1IwjGJ#a#`t-kG
zDqY)}7+Ft|;YKwLYbtg$S(-TBO=x3cP1cd}%f4kB!<6Wu-dCwz-)KDMEuM^_Hh*UC
zC`1)|)T<(U6b`+yOH!6p*Ll}@qastwA*dyjsgOf5C=?LTprfORG6O{5L%@S0wyHpj
zu|_A-=NWnYYR5m7kvm6|&T~GzoJ_OKR3sgFUgw?ifho^NQhvK#{6g0=&Fh)%n}#m0
zk1sNmwb_AMDq};OOGw5|;OyX#?yQMMH6yAk(x$3tjFjHE?c$E2XC_xXav8tnIeIG?
zYMI|~MLEVGkuT*>v&v-X^RA+u>k}E$4d&uD7=g_fA8+pNNV=4s0|iD3p<=DTXClTS
zXV23tJ;ECmN@M0j@zUAKEYW@3bv!SeYZ8ZH`YQNTApFVNc;F|9r5p4TqGs=>8E?6y
zi|gY{iM#PG1nL?UE9YCnWTk72kgZPG*Usqw!~Qd3c?~@w2?%eg@~)+VlSs6N5Yf2^
zz;ow<fjf3n`imj7u5lnzt||q9QFM(H@<3EJCK|l5P!$yDfn~U-(5Vu7s+GqxNKyeI
zP=-Oa(KH&gK`NhUa`cLj3B8%qL};DR7dk!`A^h&3-hFB6p($5Ufy^tGir)3et}qK4
zpkPKYWzC+?=&gw-L6S)S=<lfq)%uLUAa%~81Jf9hk)w~U!DItnoSy`m^}#38M}K-o
z!PqisQkV!&z4E*V01ro~qlVK^0BI`pLk6D+)f~*y!hCvwHq8zY9BGh<2s$@b^A<8G
zRaqk}&qZyyv&|0QDFPA%h4TgF&vdlc|JUq*=>F#K#r^&KMq1A`oqVGFpD&-!Pv|Rc
zO3KSqA@h9nSc%bm`0)Amk6*J}@14J*1-219l%%7D!Pl}UK>|lVi0Dfgu2jN3WC!uL
z0ej??b2iSehVgdnWHmZV4kUo*QL#aiIp}U=9x)IXk}JJ7VQ;CI9Rtn5e0VcjbY<bp
zv{}WlG6L;H!EzFKdY>cVt+`x5D+svCGD<sXw4|)E|JX43I1_3P(sI4{wj87bPSrNG
w!SIr>;Z5hm*<gY+X;)Ryx4=nzaab9m`bwE*^s(%u*E3HbUuOR@+&s_z1=MCi2LJ#7
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/pre-merge-commit.sample b/tests/resources/small.git/hooks/pre-merge-commit.sample
new file mode 100755
index 0000000000000000000000000000000000000000..d8e8ad4bcf8337f23c78a98537ecab4831301202
GIT binary patch
literal 473
zcmaJ-O;f@!5WV+TJd4B0whSt$H%Dh2t`2u6q1!glCNbGU;n%yxdNsG~zR#WA6xIwy
zWEZHoU#u?nykD=Y<HPgeWDkDm^kTof*l(|%^gh!nHrZpo^vhMDjV;E1GD~K7wV*+D
zz9lry9T0cHcm_KhDVXYvQ==FrLTT4u=bEr{U1yl7%thf%wJnv<XQZ`ZbQEezaWdS%
zI;c?h9a)Y!ux<WK8rQd_aA^?61hv_Pf<t7*z1USuL40FxYz<|hybsO?qnN}aMpcuf
z6phFw1%Xx=wUk(m>E$jSEQZ%SQ(}oLgslTvrKK@9Qf#b!hajVFnp9@oIix;NcI9Wk
xjnh0ya!AWet{I7YpD;y6HXyzI*lfSvH=o6*7mJZPkuaYpm>vzZ`wyGEBtOQPo|pgt
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/pre-push.sample b/tests/resources/small.git/hooks/pre-push.sample
new file mode 100755
index 0000000000000000000000000000000000000000..02cbd80c287f959fe33975bb66c56293e3f5b396
GIT binary patch
literal 1431
zcmaJ>U60!~5PUX&#a1@z9B{IIZkjLT0t5kq9#8~D(I5{+8&J~9;#ndUk~-ZT`r|uG
z$#K$$J{TsK<m~Lsu9iP+t-0TZ=sa(K+C6);54X>s*LP1}9!GoZ@4I4myMMG_di|of
z%?llx{O8TS-#^<H#%^V=)RNv>;(OioEmPy%kwWQBA1OMzV{hsQ8XFzS1k!~YQoLa5
zhtP1fA$q6VmMbbAC_9)4I628k*O5J$NR19uHe4QYDK<==I~SQk)Nu%xQ~<Hy8U>KH
z53w=!ke(FGb_PpnZfd*+hnXDTn;2*`u^~;?+5C~cn?bRka7NR%06%e6O91{MAgN6J
zmlO8{Biw4&wr&&(z4p3eln`E}XR9m9bNYZ7Ibrg(4yZIXrfgD7N*AFD7L3YSM#j}%
zo__rOS5fr;@8UM<6cl+cv_$YB$PQ&9dv($eM*))g!_cu!QcSh-mqE9i#QDZT)=o#`
z?8!RtE?w6p?GkGZ-6yt_p~5~4ecu|Sf^)6096%h*q-eNiEA1;Xwg)p~Q&iGSG7-IQ
z9aII&`ps$WOojFA`*bjG<mBv1n0hcYZWN0~(X01-hx(ExqWqaXgSk*@-GMp|K_3!5
z9|O21N3%~izh(4fbp9wzd+!b&7cVwSP5H00)m5ej-(s=Pl#(90UOhn@OA9u+D{i@r
za4*CP0I#<d-)-#xq5q-iY5nIef2s5OuQjcA>kFk|E@sHHuD}W^d`7YJ3YE^zrQnqR
zGoq?;YGKe)93o|_=^f%3U1KYZGPOXRRxK7w`UUbMMa3<86OmVH!EKP$8RCrn9mWX+
zC?9yF!fRVLmud3hF<}x;;sR}f(*r}6Gap3fR6zLHR~kbMgD{98N`L+r&?3p~*0+FX
zcAL%j=(SO}xTJUTvA`&Lf`2mv4koPG9&|<CA~EHbWHMoFPwT(&U=7t0`RoFZPO9Kq
zwwe$i=T|AXY#hD$aZlNMH`wZ%gwilGJ(zeYpOn*F3cy0XKXio^Sj#WXx=PWV`WGaA
B&~*R+
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/pre-rebase.sample b/tests/resources/small.git/hooks/pre-rebase.sample
new file mode 100755
index 0000000000000000000000000000000000000000..193ee3f127387f469894675f0d621b5bb9c275da
GIT binary patch
literal 5006
zcmb7IX>;2+68$XxiXKL@ma;l5d2^5Ba_rPh_DHI-u1#&_upttZXp;no03$20|NFiM
zK#D#xQ>!Z3JkX8T-LDVm!B5j7y_{;JDmmTTef+K1oIiPzeEr+Ai*<2PUgnG4^ZB>p
z_fkAvoR1emuf~ri^K$-px=4#D-v<wZ2Xv&$O_eTJh6d4)=DWL(NBs9G{k<+yMMw0T
z$VH*-+LM)}u&m^`l8~1nt(3Z;R8v(KbY5#i3z+~8h0D}Xvq&3J8BMWDizPNpaeb~9
zBN9bSkthfXzskapf%Zt{*e#}{QaNiaAVZ4{$;;I6<vKNhO@%7P-(;l-x=pPoExHC!
zB(hA#cDdD?s4P=!)=-K{<kHAWKetl-8I8wwO<ihJNs-$dEvr;&S_@6E=mNSJ5;mg#
zyb)MbqKH<one{qrV;ZQ6WL}yLtyi*ekNLf|uC6M!)Cmq7*l?g0d6`MlE49|}>Y9w&
z`bCv#<YfTKtb`!}CyNYd;|(C?vRVQmWOfR9X?FZ#=f$No)^#4>2zVn=YnJyeNey(Y
zRh`9vtLw~A+5zsjp|W0Nsa|29Rm!B>OoG5a+vi;ari8O>KkU!KAWg_fa3btK2x*_@
z0bEc7J;Ubghm}n9bOi(Sv_B66nQ7U)J7f0fO}<cB8i8vG{r39s_>8Wuf*uorcIgEG
zOHc|-V6+HlRhOP}?Cn?@5iwSl43abmBA^2lyL$+cpabCGVES+v^j^FO_}?FIp<qP?
z#_?*YMHIkyZxJxS;h@A)WDJ0bN&+F-#_lFjAf^g_Pb#5YXqYe|dZUpa^zI)VOBXQQ
zAMhT>%En%Ll?Z*7*}TwrZyg5OSZ9rY-`aU~Mc-jjv{Ll)FLMgtB4ujktfQ`Xhqrka
zT=P!A;9w^;Z?PqpLwOLu=cj3L>TdUKw2;DMu)`oVkj}<z_EjO_2uWYuvKG==%Zu?h
zJiMVR^c30RbpW}<+z;jjoNC}YI4f6QC7d-08*4mCB1>#bcDx4tYg=j%D`+i{W~fVM
zVmZ>W9VMyin9c-0KzI_;iZ-g|OyzuG`Yq%(%dvl;ifnVr0;jWE&S`z|rQu=!yHBBO
zx`OJ;oOQ(KKM<$(bC38o>pD0%|HA(E0TRw7qj$fJ_pRN+7Nm>dS<q{AcOz#-;d7_2
zL$z(_nhH{vOzT(}>C(gLg{(`t+5Z=?o+}wXU4tHy+&%F&aRhFebeEhR2R5|<c6J);
zEY(q5F5<n*XP0|=PtPBn$B)V~d$Os-?&8UlaVe_|jdkzoWNsTP-_uyq4$R6o<h`&@
z{loXa{^#TF=NJBYu9qB?hs}x=It_O}ZjX(_wy9@X(lmkRpNi0{8T{O_b_j*JC^_SM
zz3G?1$KCNWF-|`Jbx2cQ-y5LW?Z2eikngTZmsx5C(@({8<l)Ue+gIp##Mosfa~iZN
zZ|NLN9uE6Xaqpwk+@D+f?$tg2JRCY`pwZwbR9OvEn*zYm`ffKIzl4{r{ZgjfpbuX)
z_r0=0y3)T-j$gljPyEJO*6Y<pj7G72aLoqaTpc=#u)*xJcVUm0;k(n;r)^DQNa6Oj
zWvlHEGg~lz`Q_8`yQ9<BZ;ylE1Z}bD<8}<uB9Y5lRB=;JUD0h?_)4H&EhJjvwzJx?
zr<o_vk&75j_5^d$8Z$_Oc1=R-I_DlNZ2@~81oV*J6%o3RuiCz}^VEW>$#Ycbp^w@t
zTl%=f1t=w+WpJzF<|CE@?SCNAz)%9?w33lQ8vrHJqPfH9@}qs*QXOG71W=ylx;wOB
zcx!Bj^)Yy6WX$a^vBkBJ5Cob<oubBW+a#9bX{0bkMTX_2sIrs`HMk@$q{dK5h2-g}
z({`~#gm#G?c#>qlaDx_B0c<3b+8)f84LCrt;e;qxc+7>VbwVK{skNv!wvBiTa^9Iu
zkwP;VK)jH$WJ{`MRwAA9fal!y0dtV;FWg8PTkWU>CwnqD>1ZX2B@;$DlX%C5MI+}{
z9xQVnffR*~v2KAUj*hCd<gRCjR7~6Q(vF%VT7j97iv3j4Z0p%mU`7t061~l7b$!#t
z3*Pveii}aQ?QD9aiX>gul~`bk#mk`o>zk9)<2Uc8?hUZAEvd!`9em)~$Z)zev>w^8
zyAgCP_$&Y)7HSQ84`xG}OeTavaEswwF|8Xpi5iZzZa@hCiv(J-%bfFC&)HLlO+Rhw
zG6g?9eL5&A!SuJnQ6}LxG%tU+@vZ`i+!+Rz6iYvsTdhnPo7lW{m-}{hya@viX4)XZ
zngaw+j;gloB#|UwI@8sOmQpc`h+bicQJnQIB5eifIMQNgD2+oai33m!34~xU|0Azj
zhu$8z+T5^;Pxx@d{N)pzOJLSa^e;aDf$W%N5XcOf!mGC9l9j$Ev2h6N+6ZQC+CJzl
zaM7?S!SrFLS2DASjj(h6y1WN3N?|bmqmyzm!&nLoE|`rKBOc_yDF$a#FsUn!IQf(t
zdC&Us(kQz*7mv<H12p@UD8XaLXdK{>H^j*^MC@>wTDb}g%~sx*ng#>{@lR=XG-Z5_
z#<9*Oh0joMzt;nS)ObAp)347`D=}r-;nV!TbIq&xrGRGsF6fZg+!VkfUei@_&l-M&
zPqQ+Dw)RV}+)I8RuqAxa`Pv8e&!_gXS=e2-un>=Ktn}-;%lLZxaVn?Q>yZCb2R3Wk
z77zr%;Rq&h|2ncqyKYmFI0148JVY7Q$V5p=dWj<MQ<8r+@QLq+UROk&%quICq^O2C
zp(17882jXI)_GaqzAY4AjoB_nh8k*r1mJ>+Qqpu%i|xp2<qF`Tw6&3`h654ceBjZ7
z4>C=WaOb2Wudn^h0EcD%$p9YVU1fnoRV9`(cy(vv6K>FXS!2jY>1GnU--7)4usH&K
zao*&P^@9~YmUe|ZdLW@C>H;!*<TIU}-!Tw#3oqZA*99}b3&uHiGO<{I6!pMnAiQdS
P!fA@Yk4s_DjDP<F98V`a
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/pre-receive.sample b/tests/resources/small.git/hooks/pre-receive.sample
new file mode 100755
index 0000000000000000000000000000000000000000..32b597f8c32399017c3d7d51134d1f6490ad9575
GIT binary patch
literal 601
zcmZ`#+iu%141Kn~f>Vt3>Nw4M*;=?j(TBD#O@XCv0|MEhA;z}kTFRv@`tPHhp=&Yh
zg%Zhg4i7o_k{a5i&f5;tZ==%}^Sn4aD_6%qs<o-wO_Prn;}`SPs_*$C$(7T|$#C3`
zPt%-C8gelZ1GqAP8`ZQmg0{8-S9H{R@D>_XAuJt&EumdH4Yu`UjT<s+s_~uXh}qA8
zg|_HG)%7Pdc&$7*uR0HF@)~vmFjqyD?XZwCbLen^h5t)sm9<90Oa!@Y%8!~rF8G?W
zkzmCF8kMtuuelMHIAlqqnm?72LeGM1J4`w(kX9&%LQn}ForlDLjBoCyvxmo@x3kH^
z^loxLyPiDWPo-cBMnsg2M6}kuPGHEGBqVkC{D&9Kt%xFAsTw4QC1$_=fwBl=3dI+e
zaSxI}JS}=Z(Ec80eF`!Zq3mqapXI|UN!a)t;@4hcu%Eq2xcoW}%!><-+XHTuHss+b
YOmM2;hq8Egm*4=7_P9T{21QBYH*F=mfB*mh
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/prepare-commit-msg.sample b/tests/resources/small.git/hooks/prepare-commit-msg.sample
new file mode 100755
index 0000000000000000000000000000000000000000..b1970da1b6d3f42f00069fd17c325de72cda812e
GIT binary patch
literal 1702
zcmb_cTW{Mo6n>t6#i?x6xmZ$SFLf{QfG*3r0L?Pg?px55l8$UTGO3bO;spKi{V3XX
z))weX0X>M9bNMcZ-6yG%>(n}JI2|25dr<ew@wmMG{l(3lx~bQz>}WZBP@ih?JX^+@
zu#5O48P>yRX(m<b*PU*sORp92TCD1dX`%HE+1$w5k<(Ngu7zQ83#MGJR?<<W=d@yL
z#heqwo{FmCg0g#x<~R+PBD#}q(MBn;V$x;%UrJPP3*l%XtlvTWChI2SfJ$9e`YvSj
zRSOQ?NUgSMLI`3vL48YBHzwzVXoe7v0ef|0YHgV$N@?N(-R)rPqRDrK$n(%+OF$`P
zWdjC5N~`#RjV9}aYwQ4_yF5O-$h2`>fDIhYP)doc1&TADZa@ZGpusJ$6G+e$ZMcmC
zoOosDQPS}l{H?YPsq(4;0SGkATa9eeqAaDcj<jNAU+LTSmM1jo(ti~T0B7acJnnTX
zTarYy;Husdxal0!S<ba8=uu&a*SNYtr7|d7$g-q3_JHER2*oBsVW~i~XXdMx%LW~0
zT*960LF<qZ6K&FZ;vGmtyywBU3^Q>q8n2wALbFwU@2i@FAaRV!=uw-nwx1gKn2SvY
z>Ff>;2sg!+Hxfkwv1lsiii=p6WenF=5)6LZc<a$zDCD$GRuwvG4Fr+B#h?#9gTbio
zk#MdxC@WY%zSGN#i}Ts_#q`bf-{)`7CcWeB*7WlIyHjioJJWw&A5VItPUq3^9!r}S
zbykelFV-VFvcr>QaZ=aS_}+-4Y&?!@HWh|<^gJ21!|T@+%On#w6azxPHV}XsRbe*w
zR_TZ2XEsQa1lPK~biYqg@0-RW@5J1@=<87cFzEUABdCoFH2CZo?}l(Z*!OFqUxo>K
z_d`l#4d9|H6;VPT{X?^{VJ>oL|D7K{BJwwqB>`YcPoGk+9hbvHnoQ{EM|kPgD_`wk
zKm4#2xu;-y`RAm!=L_BnLvJ8$AZm8@?)v<%vwvsw8AF2x6!mTT;c72A_~U9nIq0ST
zv)N0!I!^1p=g8-RQfx5)E_Mb_4I2vtQpI30XZ&t<!9D6nI|;V7YR3)l6=S~QhuwLQ
g$e&^kTY-M99*<-Iw@(78*M5W!4}U}|8YyMx3->-9h5!Hn
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/push-to-checkout.sample b/tests/resources/small.git/hooks/push-to-checkout.sample
new file mode 100755
index 0000000000000000000000000000000000000000..a80611e18896f212c390d845e49a3f6d5693b41d
GIT binary patch
literal 2840
zcmai0U31$u5PXh)#YOS7cE^-rw@uolNhe9&aUS|HtvhX>G$45tVUYj>fRdF?|9kfU
zNR~aG=E)WbEbeyq7JTw}ZuHIE2kUtL<<n;$&G!2F^Je|kx2ug=4L5!H^!ogx`7o$&
z%Il(3zAe6<oe$^F=A|}s`8}CDp*M#3M)gC-)LOeDUpYMl3YNy9R)I-T)pE7sy09aj
zJ7%&5PnSB-F#2{jc><WLR{I2izuK%VHc+{hRfXe<^_q)8RjcE(6WX+F2)iAtDtI{x
ztAHVBq)eSp_E^xcV^i_5KLIHA$g{zEjh?rsacu+(Eyvve2~Kmw%;n3g(kWB56j~Js
z<yE5tYUsAR&PY0wgRvM8x!zgLX8SI!eVY&}YZ|>AoeCNptd-NM1aZLhESzC;I`+Ns
zfmNNjdAp^W8#Q*}l>CT7RB9F5(BbI8ly2l~+E};JW|>&d1)=epZ-8vm8ppkbEVn#R
zt30a5A-c(YQR8eM5%;|UAnO>rt!&@x@G@yp+92%w-}%(5P_+P&Wf_zb$f-Qrl5(7z
z2ah(bkE;!DK(&aAMuQ%1TS>ai?wSXCOCSj=_}8x4IbCx^$}9q)<W{Y<9o<WqZo^oV
z3bR;Qa%VT-M~kU@2n{=+=)C!MD`12;@<F*EtPfV3K#g^1%&ggH?4{m<R$WEKcUI4%
zl6{ik6Bo46pmNjdXg5@?hn;SjG$}rr3Gy#-BGgVD$8oD)OcK(oqca)L_kk)UBZ_&6
z*ourb#Yc8l3J+uSda_Y$GY--5Zp3QK9w^?P%E0xb57?fY+Q#+KU4)+R>whwv)SBt|
zg#MX4;;Oau`m=MI9(^&zPbueY@~>3*ixX%mvR5m_1&nAg@ZKvY1E$O}&EtLiG;mhV
z1xhMIm~fGjmf_#{62f`y;09?I7M1W2tWQvz<}i9lR>OpQyUJi45_&*pQus&EkwY<>
zI|ZAx=*3i9a-)g)hXkvO7>UJ5MNgL(Z+-wpXVcgbSgpmFmbf1~DPA(OVGI&FNLeIE
zNH!_aiH$vsif$_j7=T2{cS(!DOI`~bn@)vSd-0d7xL=DF;UNP|tW}4i<qWTSNCe|y
zg9kV&L9g;FhC@tvsVu#V-brqShHy2qZpA!gg{ZTY>h>DvHtu9tY_pbJ6x(6E*hxgC
zzNDao%qlr-IE%YGbS4hF!n!on7#W3$bX-_hbZAaws^nHu#)Dx=WzdbJ>AKzAy@T$x
zSWE^x9+|TEHVEPyaPYa0DOChp?AeHSBBDbZNokQpAY{lE!7geZI=jV)G^2@<iI(N4
zJLJjy@Y<VI$yq;3bVpJICW3D?YDMDl4Oe5pDf{d`i1_3Qx%4uX`$dOz<9e}jm2B-u
z8-?%!7XuiNF2O&MAdl-iw{drG+$F^zT2Z7WRV#c6;IRZEf>l)&91Zb1+`T+oq9wWF
zRV~kGTGce0O~p^6mj{kT5kL(pv>r;Lvd7VDX*P>A^Th`$3cWO<svk?_?FeP@458*2
zA1PqUOdd%VP5&4~ocLK16{1GLll64c=mU81RMFstpnMoLuI7hNIE4N)U%h*#<Fz{C
z9#>0<l7lRrls|W(8=W1O80P~6+U7<4BqC}N4-4GR3w#{O7YmH?JxwJfru2d?e){$5
z@5R+`7Z;1)FW;OkE-(HPisCS;5F4O^Q>L81p4Ysdo3ZP1(SrR-peEdTo;-@bkB((G
zPHYQXUL!@Q$e(OQ;R9r%@Afz+50I7>*^^c&&|E*r-jN)LH=pM4AqMwWxSv|nqjddE
Z4{_hwv8!W(<d3>T<BC%y-6O5i(|^OS%`gA}
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/sendemail-validate.sample b/tests/resources/small.git/hooks/sendemail-validate.sample
new file mode 100755
index 0000000000000000000000000000000000000000..effa6f54f2a5c1f0cebe258bf4805904be4b4852
GIT binary patch
literal 2365
zcmb_dU2oe)5PUX&#g-Nv2{5JDX_MA~3%Iq?8g*kpRgpYZIFU=~BI=I5J6e{T{`bz^
zk+$48h#x9Ika!=nv$M0y{clD}-j1x(hDWbnzP?l2k8j?TH{brS+Nf21yPm)Nczma>
zYw`X3V>TCdnSD1ru8&`j=2DIPbCT@SnIgUw>$+lEYP}+x8(BMYnr=iT3*ndq)xzaV
z>I+qjv}vC#8_9M+b1p#uNS0M0)q<p>8!3p_LRQ0MA3M`!2foxzRUjbFY@}O~(ki=S
zqscnq8cU*dY)D$$cqE}n)V0yIk>CNKHCrndOtSP*HbOb;nbwAHSb;R+gs^?^Dve%)
zoW}t(*D}$>O3ab0TS^-;J|u&sb-PkZzo#kn*#xYt(;<xzKW(YtQZ$u23?yV#kyh1~
z@+Idh;EG5jXx8@%<;Y_W8SA=|T;MPQ)TB!!<QcbU)YR4)79eeeg4|vp-8jm%Dl3^I
zRS7+i3>FGuwzSb^g&RDiGcOz9TB;Hu`nJh)$W=C=XCSm2AY=$w3G3P-V#Oo+N*;#2
z4ijJ-pBZ=;T(RTgp_HYrD!uW-dTMfkuqY5jwOy)~gM;#=P^i{!l7`pXTS^s(&^{RU
zydaw}OpS#^D1cXM8?FW+fh`t7D(g;yr6|}fdaNtZBx3hlK~IpkTu3!Qq%R+zAo#<L
zU&m+XIFB0>t}Bs8^3$vHD+-TGT@`F>H1Cc#WAVW;&$S6%fE2d6@kLS0g&ihIM{}0z
z8#XhD>b>3{(BH|Px7}&lJ4%y1v<t$W+!F|W@<gaU4;MqSHKR(wdML<XnCv;zaPrSi
zxVCvek26-bf#Q!H+uAgy_{e_1UZCq>(CihZJx@8MPoGdl*BJGD;usf*iS7%;{Joe;
zNFuBa>*~o&qETDPo~u&~$FxE1xb^x&(CbE`Y3GfsibL2rl+L;>P6j&Y3U>K$mkp*6
zd`Q{<^+^&;GskGjwD-%!boR&i-TC<Uvy02w+l$Nb?B}aL-%ZDpluqd=K_@}z*fyuV
zzAs1Hf?3v$k!X7GTc8Q=r`WJ_Uu=>A9UOR|@=GYb5x#<f&WSMH%mCJU<#=7=qFdL6
zG?Wz&6z&J<@I(B>+dhd7fkaVIR^pol`Mv+rUbmZ43dVL6^S7g3{NsPiG$iy$5EDB%
z6KIgnb$H(n&t3e4E6d4V7w^B?JS}JkG)PM6+X3Co`SQs($O*AA+MG~{S7RJ=cy-l&
z>~%3y`tjfx2>uOu<lDGWewcb=oL@}B@B6FCZ?oxSJWldrm%UfOduahk%C0H>tB_^s
ziwG=e=ch|FQ0IkN91US7rhdQkXhwwt$gU0WEVDjo=IPb+?6PC=s8}J*ua(Ms))`UL
fi$|vMHn?H<rred|1&u#kOoJ`%vx)-*g-ZSfgGvdP
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/hooks/update.sample b/tests/resources/small.git/hooks/update.sample
new file mode 100755
index 0000000000000000000000000000000000000000..c3430f6fe27e572a1cebdb2f2bec70add0642a60
GIT binary patch
literal 3707
zcmb_fU2oeq6n(aS#jRq*br9K3y0z;^QgkUc^kKlTrB6js&@yE)mPu8l<0$L?`wmIT
zrsXba))@gJi@e|G+<SfSXe`CeSQ}OG@sr8ZTUlQ{dzM}Q@O-hBi}GeUom`#X%FiYH
zX?m4Rna-0RN2lfK)A3ZuvHcz$L<jUn62D=~vfz{}wIH2VqBLX_O$(JSXeF7H$}q!c
zWY}C&R;eX%X?P{%d;|>_tSE3ettp-hLlsZCxaLX8(nU;bVRB;Ce6@s#eu2|WvLz>-
zvy(&>Gyfp@+BtKnpqWkKi^+v{4jn_pNw_zeuxE<mRXKx8G3;9pl+45&4~hHW!G@wo
za7?X(0B}HbX*ExkDmas*xzV)FxygC8AL?2Z1x-0QJqS@qn8sD7r{bm30@<%eL_gOw
z;~85O$Xw2AS}Qp)5ViRUe3|ir8;&&I<B7Y6^!kkNyYXF4EY(b8_5DsTYn_&?wkdEz
z0y$tADo<&}nGs5kg2-J=0KlEGPb(%<An(pXY{K`qIZCuwiT|2{8JD&5o_~`o6<;dD
zi@J#zCE4={8j%<uy|iutvHu1QKo5Tno-BAF2FwD%%O#UDDum=w!;!PNe-cOFNX4)5
zd>TifiGO|)w}OANj2n2D^K=o3j6P6uOL70#cbA{uzWXDlk1wr9GV1X(2W{RuTvjXV
zCmd<W?kH^?PXjkbF`XZtwu1B+%4@ZvHIwGpJ*8@8`MWAhq^B|Hj1lzj3R8bVuMpNb
zz4Gzk!3T3bY;WEGIww&kq9BYW6EP*q$K|EB-@TH(Fjtz*`HMTO?i+3E;5tdSah&xZ
z+t!x4K7)dpy5wiJhlJz~8qF|r8a&-SV7^I3C@_q=P`yt@_x_F-;PQR)fzP<zNN>8u
zH%V`94=q3)Dk)PHNrnFC(T1)Om6f{Usj;u1R->&XoCYVK2V3ZlgZuF?N}1+33<P7e
z<0yVF?Qoa{l#7q(3&jv=Ab)gpM8A7`p%3InNzSxy)ZER2U0C#9zKpnLY0I?>OER*x
z*9Z=L=zI8CN>A_^jYjt0F$psO$sL=38q5q|SG)qCN6{^>RFh5E&l5GZ$pEahnF&d+
z5c>64t}uJPkf~_!VUj#&N%nC-gUMj%=@B=!V>&}xtj2%@-mOm#rQUSJ3(ccmc+fza
znZ#uxF>N?QN5UrIEd!5RgHEf<eGg}P45aAs(Xs6u!XW9r1I*E6XJ^1movX@xYLuPz
z|7xBN4z@b}#x>W#;(nKYF+D<*rdshJ$X-z2OZ2X;)nn@KSVdVhaA?}@3;6gZxb4<W
z`9sa`0lR_azMX|=t_(FvG@%w2);9P}SG0u&K1(*oX3};~=<<!N*F$UTSxD{V&6mgL
ztw9Ntc2eOF@c!OJytNC4T^#)Ien7-`dI{6s#co~0f^E3J<0Ty)l3xq2u@Y8DXTK>v
zozoWSr{{+!h}zGpumG3H`=AvWpm^9kW;J$Jp^Xl*?8ckr`fqN%c|Z;VC0|cM4vSrk
zH_O8Yvh85nvJp^;``wo8=z0f`FWg?`>gO#y1hjX1{}rTlg9rwIKia8eyGexA3GnuR
z`Rg~XZoW;0pA)vI8=p5!+6sIn#C^FCvR>ffv39h6SCNi9v);%WD;WZ`of_MgwyRWy
z-yY%n*Y>X8<Sf+RyB|Sr2YG@1w~%U$o`(5EDkJ|3$u=eMZA&_wxEsy<Xxh2k^tP?4
RGx&ZHQs^8zuIpu!=pQhfp8Eg*
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/info/exclude b/tests/resources/small.git/info/exclude
new file mode 100644
index 0000000000000000000000000000000000000000..a5196d1be8fb59edf8062bef36d3a602e0812139
GIT binary patch
literal 240
zcmXYqK?=e!6h!x)VxWtv*mf_t5?px$aS_{}Hj?C*<cHXeXE&AZhT*-L3ZoI&*l1%Z
zqG?zr3TvQGZ__}H4(u*%p*rI=cU!%ya5ugfGATh66$IJHgu1Gs0-<N;$V+SsdE)?u
zIq;i$f#WE4f$_MWicZjMEob9LWKMR#iwZq54~QgST^6=i%u0lUkJu-_J**QBMq}ZG
Wth_)NDbl|`oQr&HAFQ5h`0jqAIZ*BZ
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/objects/88/f9a6eaa2cf008b9bc92847178621f21fa99f3e b/tests/resources/small.git/objects/88/f9a6eaa2cf008b9bc92847178621f21fa99f3e
new file mode 100644
index 0000000000000000000000000000000000000000..7d80b8d78e0dc55669d831a6638f48ec9fed0982
GIT binary patch
literal 50
zcmV-20L}k+0V^p=O;s>9W-v4`Ff%bx$Vkn}$!Ay}rnY6F$m-Kg*KD_+;Lx#g4|^&N
I02NaX#p`nv=Kufz
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/objects/af/5626b4a114abcb82d63db7c8082c3c4756e51b b/tests/resources/small.git/objects/af/5626b4a114abcb82d63db7c8082c3c4756e51b
new file mode 100644
index 0000000000000000000000000000000000000000..822bc151862ec3763cf2d3fa2372b93bbd3a4b65
GIT binary patch
literal 30
mcmb<m^geacKghr&@q@?NlP9kSYMj?U<r(;diNWtH+YSKNt_|)0
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/objects/c6/97d4f7a6eac8d7b131673c340bd3cc5bac14d4 b/tests/resources/small.git/objects/c6/97d4f7a6eac8d7b131673c340bd3cc5bac14d4
new file mode 100644
index 0000000000000000000000000000000000000000..5adfa88cc6b00bb19f2031b8ab61f6d07f9ccdb8
GIT binary patch
literal 130
zcmV-|0Db>>0i}&W3IZ_@1U=^!a~EV1casc=c+{&un1qQN*i9hD|0|m(2n|iwp*q%W
z%N;b$hu%cM`$TMo*~EnC1BFP&Pfj~;jZVKXQ96s_PhV<-XAROi+@-v8dBLUa`!;GB
k^i<X>XlEv8$>R)1G>9th&t3j;s7J{?^9n<zzF|~BaA?ar-~a#s
literal 0
HcmV?d00001
diff --git a/tests/resources/small.git/refs/heads/master b/tests/resources/small.git/refs/heads/master
new file mode 100644
index 0000000000000000000000000000000000000000..4eb36d22298f060fd324155ab854d9d6486fc498
GIT binary patch
literal 41
ucmV~$!4Uu;2m`Rc)5uXl$AMP&AHjriQg~T$i(A>|7U^`%mXoWC24Q^m!3%@{
literal 0
HcmV?d00001

View File

@@ -714,7 +714,7 @@ EOF
place_channel_configuration() {
if [ -z "${NIX_INSTALLER_NO_CHANNEL_ADD:-}" ]; then
echo "https://nixos.org/channels/nixpkgs-unstable nixpkgs" > "$SCRATCH/.nix-channels"
echo "https://channels.nixos.org/nixpkgs-unstable nixpkgs" > "$SCRATCH/.nix-channels"
_sudo "to set up the default system channel (part 1)" \
install -m 0644 "$SCRATCH/.nix-channels" "$ROOT_HOME/.nix-channels"
fi

View File

@@ -213,7 +213,7 @@ fi
# Subscribe the user to the Nixpkgs channel and fetch it.
if [ -z "$NIX_INSTALLER_NO_CHANNEL_ADD" ]; then
if ! "$nix/bin/nix-channel" --list | grep -q "^nixpkgs "; then
"$nix/bin/nix-channel" --add https://nixos.org/channels/nixpkgs-unstable
"$nix/bin/nix-channel" --add https://channels.nixos.org/nixpkgs-unstable
fi
if [ -z "$_NIX_INSTALLER_TEST" ]; then
if ! "$nix/bin/nix-channel" --update nixpkgs; then

View File

@@ -15,7 +15,7 @@ programmatically:
1. Embedding the evaluator
2. Writing language plug-ins
Embedding means you link the Nix C libraries in your program and use them from
Embedding means you link the Nix C API libraries in your program and use them from
there. Adding a plug-in means you make a library that gets loaded by the Nix
language evaluator, specified through a configuration option.

View File

@@ -0,0 +1 @@
../../.version

View File

@@ -0,0 +1 @@
../../src/libstore-tests/data/build-result

View File

@@ -0,0 +1 @@
../../src/libstore-tests/data/realisation

Some files were not shown because too many files have changed in this diff Show More