Commit graph

9597 commits

Author SHA1 Message Date
Qyriad 2cd1ef2201 Merge "libfetchers: fix URL logging" into main 2024-05-27 14:20:15 +00:00
Qyriad ec5039653d Merge "Revert "bump the extra --version info from info to notice, -vv -> -v"" into main 2024-05-26 21:51:05 +00:00
eldritch horrors ed6b3165ea worker: process timeouts first, and exclusively
check goals for timeouts first, and their activity fds only if no
timeout has occurred. checking for timeouts *after* activity sets
us up for assertion failures by running multiple build completion
notifiers, the first of which will kill/reap the the goal process
and consuming the Pid instance. when the second notifier attempts
to do the same it will core dump with an assertion failure in Pid
and take down not only the single goal, but the entire daemon and
all goals it was building. luckily this is rare in practice since
it requires a build to both finish and time out at the same time.

writing a test for this is not feasible due to how much it relies
on scheduling to actually trigger the underlying bug, but on idle
machines it can usually be triggered by running multiple sleeping
builds with timeout set to the sleep duration and `--keep-going`:

    nix-build --timeout 10 --builders '' --keep-going -E '
      with import <nixpkgs> {};
      builtins.genList
        (i: runCommand "foo-${toString i}" {} "sleep 10")
        100
    '

Change-Id: I394d36b2e5ffb909cf8a19977d569bbdb71cb67b
2024-05-26 20:57:09 +00:00
Qyriad 742c62a6eb docstrings: NixRepl::getDerivationPath: exceptions directly thrown
getDerivationPath() directly throws nix::Error for invalid derivations

Change-Id: I81ead950060b789794fa683b61c6349fece1690d
2024-05-26 17:19:26 +00:00
Qyriad 9f16a20f3d Revert "bump the extra --version info from info to notice, -vv -> -v"
This reverts commit d003dcd7f4.

d0390b5cf¹ bumped the default verbosity of nix3 commands, making this
change obsolete.

[1]: d0390b5cf2

Change-Id: I1ddc36ea4514cd8d2855ff334accf6e98c210879
2024-05-26 11:15:10 -06:00
Maximilian Bosch 5986a720d4 Merge "git putFile: support flake maximalists" into main 2024-05-26 10:16:40 +00:00
Qyriad ebd00b2d0b libfetchers: fix URL logging
8c06b7b43¹ made libfetchers log the URL being fetched just before the
actual fetch, particularly in case something freezes. This used the base
URL, to not include query parameters, as the Nixpkgs lib tests assume
that stderr logs will be equal across shallow and non-shallow git
fetches (and shallow fetches have the ?shallow=1 query parameter).

8c06b7b43 assumed that the `base` field of ParsedURL would be populated,
as the comment simply says "URL without query/fragment"... but
apparently it is not populated when the URL being fetched is *already*
fetched, which caused libfetchers to log things like

fetching gitlab input ''

which is. silly. but you know, busted lix be busted.

Anyway, with this commit we just remove the query params before printing
instead, which seems to do the right thing

[1]: 8c06b7b431

Change-Id: I9b9988992029aa6abef786f20b66e68c2ebb97d4
2024-05-25 19:37:38 -06:00
Artemis Tosini 53e2b0740c
libstore: remove duplicate builder variable
The `builder` local variable and duplicate `args.push_back` are no
longer required since the Darwin sandbox stopped using `sandbox-exec`.
The `drv->isBuiltin` check is not required either, as args are not
accessed when the builder is builtin.

Change-Id: I80b939bbd6f727b01793809921810ff09b579d54
2024-05-25 22:14:55 +00:00
jade dd53bce476 Merge "libstore: parse the buildMode instead of unchecked cast" into main 2024-05-25 17:42:09 +00:00
alois31 ddfe379a6b Merge "libstore/build: always enable seccomp filtering and no-new-privileges" into main 2024-05-25 04:21:53 +00:00
jade 2a7a824d83 libstore: parse the buildMode instead of unchecked cast
Change-Id: Icf6af7935e8f139bef36b40ad475e973aa48855c
2024-05-24 20:45:05 -06:00
Qyriad 5b4b216fac Merge "nix3-upgrade-nix: fix when there are differing pnames" into main 2024-05-25 02:20:24 +00:00
Qyriad 1c0f3c540e Merge changes I3ab84cc5,Iba34ad42 into main
* changes:
  nix3: always use the same verbosity default (info)
  libfetchers: log fetches by URL just before they happen
2024-05-25 01:52:44 +00:00
Qyriad f3f68fcfac nix3-upgrade-nix: fix when there are differing pnames
Change-Id: I19c7e24a4d46137127e76b7bb133e0184d73d1b6
2024-05-25 01:50:32 +00:00
Qyriad 076c19e0d1 change "evaluating file" logs to debug
I can't imagine wanting this unless you are debugging something (in
which case it's very useful)

Change-Id: I90c6f182c18486e9f6b15a59379bbb8e88fb8e7f
2024-05-24 23:03:08 +00:00
alois31 f047e4357b libstore/build: always enable seccomp filtering and no-new-privileges
Seccomp filtering and the no-new-privileges functionality improve the security
of the sandbox, and have been enabled by default for a long time. In
lix-project/lix#265 it was decided that they
should be enabled unconditionally. Accordingly, remove the allow-new-privileges
(which had weird behavior anyway) and filter-syscall settings, and force the
security features on. Syscall filtering can still be enabled at build time to
support building on architectures libseccomp doesn't support.

Change-Id: Iedbfa18d720ae557dee07a24f69b2520f30119cb
2024-05-24 21:19:29 +00:00
Qyriad d0390b5cf2 nix3: always use the same verbosity default (info)
Change-Id: I3ab84cc583e3e8b1c05a8ae1a7a087204f513d03
2024-05-24 15:15:42 -06:00
Qyriad 8c06b7b431 libfetchers: log fetches by URL just before they happen
Addresses but does not close #305, as we still need an indicator for
frozen fetches.

Change-Id: Iba34ad42dc1c8772f7da249b90fe794b041bbf73
2024-05-24 15:15:42 -06:00
Qyriad 00bf2b105d Merge changes I462a8cf0,I3b0bcea3,I2acd56e7,Ifc149764,I9e2ef170 into main
* changes:
  docs: linkify nix3-build mention in nix-build.md
  build: make internal-api-docs PHONY
  cleanup lookupFileArg
  add docstring to lookupFileArg
  add libcmd test for lookupFileArg
2024-05-23 23:28:42 +00:00
jade 9530b7f2b2 Merge "packaging: rename nixexpr -> lixexpr and so on" into main 2024-05-23 23:19:28 +00:00
jade c97e17144e packaging: rename nixexpr -> lixexpr and so on
This breaks downstreams linking to us on purpose to make sure that if
someone is linking to Lix they're doing it on purpose and crucially not
mixing up Nix and Lix versions in compatibility code.

We still need to fix the internal includes to follow the same schema so
we can drop the single-level include system entirely. However, this
requires a little more effort.

This adds pkg-config for libfetchers and config.h.

Migration path:
expr.hh      -> lix/libexpr/expr.hh
nix/config.h -> lix/config.h

To apply this migration automatically, remove all `<nix/>` from
includes, so: `#include <nix/expr.hh>` -> `#include <expr.hh>`. Then,
the correct paths will be resolved from the tangled mess, and the
clang-tidy automated fix will work.

Then run the following for out of tree projects:

```
lix_root=$HOME/lix
(cd $lix_root/clang-tidy && nix develop -c 'meson setup build && ninja -C build')
run-clang-tidy -checks='-*,lix-fixincludes' -load=$lix_root/clang-tidy/build/liblix-clang-tidy.so -p build/ -fix src
```

Related: lix-project/nix-eval-jobs#5
Fixes: lix-project/lix#279
Change-Id: I7498e903afa6850a731ef8ce77a70da6b2b46966
2024-05-23 16:45:23 -06:00
Qyriad a0dcfbb084 cleanup lookupFileArg
Change-Id: I2acd56e7a542b12138f43c95af78fdd50e944619
2024-05-23 14:11:10 -06:00
Qyriad 3a597f1d0b add docstring to lookupFileArg
Change-Id: Ifc149764f5a15725d3d630677c6da29def4b0f3e
2024-05-23 14:11:10 -06:00
Qyriad 677cf75473 repl: log errors writing to history file
These errors are now logged and explicitly ignored, rather than
implicitly ignored.

Change-Id: Ia26015466a17f2b11952df5317a4d150d79dc184
2024-05-23 18:10:23 +00:00
Pierre Bourdon 79121e0c44 Merge "repl: do not crash when tab-completing import errors" into main 2024-05-23 10:48:38 +00:00
eldritch horrors f281727e70 Merge "libstore: remove unused copyPath function" into main 2024-05-23 02:31:50 +00:00
Pierre Bourdon d8bc3bfb6d
repl: do not crash when tab-completing import errors
File not found while importing causes a SysError, not an EvalError,
which is not currently caught by the tab-completion handler. Ignoring
all SysErrors might seem "dangerous" but this is the tab-completion
handler, any exception being bubbled up from there causes unexpected
behavior (causes the whole repl to exit).

Fixes #340.

Change-Id: I643048a47935e77f582decc539d9e51bdb96c890
2024-05-23 02:52:54 +02:00
jade d05e0b9f1f fix: enlarge envSize by several times to not be close to nixpkgs size
nixpkgs has 23000 attributes, and our previous limit would be hit if you
have more than one nixpkgs in the environment, for example, because
`repl-overlays` will load the new stuff from the environment on top of
the existing environment.

This is not really testable since if we did write such a test, it would
just be testing this constant tbh...

Fixes: lix-project/lix#337
Change-Id: I49197bfb4db55b082f914f0d70e84f5f5f110954
2024-05-22 17:08:37 -07:00
Qyriad 06c1375e52 Merge "libfetchers: fallback to memory SQLite if fs IO fails" into main 2024-05-22 21:17:46 +00:00
jade dcc7ea5498 release notes: add a bunch of them
Also fix typos introduced by the commits I read.

I have run the addDrvOutputDependencies release note past Ericson since
I was confused by what the heck it was doing, and he was saying it was
reasonable.

Change-Id: Id015353b00938682f7faae7de43df7f991a5237e
2024-05-22 21:13:56 +02:00
eldritch horrors 47523944c5 libstore: remove unused copyPath function
Change-Id: Ibfb0848d8b337bd38947a745c240838865cea401
2024-05-22 18:26:01 +02:00
Qyriad 6881476232 libfetchers: fallback to memory SQLite if fs IO fails
nix::fetchers::CacheImpl uses $XDG_CACHE_HOME, or its default based on
$HOME, to store its SQLite database. If the current process can't write
to that directory for whatever reason, though, any eval-time fetching
would fail just initializing the cache.

With this change, IO errors initializing the fetcher cache are logged
but ignored, and nix::fetchers::CacheImpl falls back to an in-memory¹
database instead.

Notably, this will fix any uses eval fetching while Lix itself is being
run in a derivation builder (such as during tests), as the derivation
builder does not set $XDG_CACHE_HOME, and sets $HOME to the non-existent
directory /homeless-shelter.

Before:

$ env -u XDG_CACHE_HOME HOME=/homeless-shelter nix -Lv eval --impure -E 'fetchTarball "https://git.lix.systems/lix-project/lix/archive/main.tar.gz"'
error:
       … while calling the 'fetchTarball' builtin
         at «string»:1:1:
            1| fetchTarball "https://git.lix.systems/lix-project/lix/archive/main.tar.gz"
             | ^

       error: creating directory '/homeless-shelter': Permission denied

After:

$ env -u XDG_CACHE_HOME HOME=/homeless-shelter nix -Lv eval --impure -E 'fetchTarball "https://git.lix.systems/lix-project/lix/archive/main.tar.gz"'
warning: ignoring error initializing Lix fetcher cache: error: creating directory '/homeless-shelter': Permission denied
"/nix/store/s9lxdnn0awp37n560bg4fgr497ah4hvw-source"

¹: https://www.sqlite.org/inmemorydb.html

Change-Id: I15c38c9baaf215fc6e192b8a4c70b9692a69bc22
2024-05-21 09:30:25 -06:00
Qyriad 20981461d4 print type and value in "flake attr is not a derivation" errors
This turns errors like:

error: flake output attribute 'hydraJobs' is not a derivation or path

into errors like:

error: expected flake output attribute 'hydraJobs' to be a derivation or
path but found a set: { binaryTarball = «thunk»; build = «thunk»; etc> }

This change affects all InstallableFlake commands.

Change-Id: I899757af418b6f98201006ec6ee13a448c07077c
2024-05-21 05:55:13 -06:00
Artemis Tosini 3de77e6dbd Merge "libutil: Create chmodPath function" into main 2024-05-20 15:13:53 +00:00
Graham Christensen 8220da8a53
git putFile: support flake maximalists
Passing the commit message as an argument causes update failures on repositories with
lots of flake inputs. In some cases, the commit message is over 250,000 bytes.

Upstream PR: https://github.com/NixOS/nix/pull/10686

(cherry picked from commit 8b5e8f4fba5728f2b3e90fcd1ab15df77e3ea0e8)
Change-Id: I2c196a21cc9bedc24d57a828a0c5b9467e072f76
2024-05-20 13:51:29 +02:00
Qyriad 589953e832 Merge "fix -Wdeprecated-copy on clang (BaseError copy assignment)" into main 2024-05-20 00:11:12 +00:00
puck bfb91db4f6 repl-interacter: save history after entering every line
Fixes: lix-project/lix#328
Change-Id: Iedd79ff5f72e84766ebd234c63856170afc624f0
2024-05-19 22:47:45 +00:00
Artemis Tosini 5411fbf204
libutil: Create chmodPath function
Move the identical static `chmod_` functions in libstore to
libutil. the function is called `chmodPath` instead of `chmod`
as otherwise it will shadow the standard library chmod in the nix
namespace, which is somewhat confusing.

Change-Id: I7b5ce379c6c602e3d3a1bbc49dbb70b1ae8f7bad
2024-05-19 22:07:58 +00:00
Qyriad 4eb6779ea8 fix -Wdeprecated-copy on clang (BaseError copy assignment)
2bbe3efd1¹ added the -Wdeprecated-copy warning, and fixed the instances
of it which GCC warned about, in HintFmt and ref<T>. However, when
building with Clang, there is an additional deprecated-copy warning in
BaseError. This commit explicitly defaults the copy assignment operator
for BaseError and silences this warning.

1: 2bbe3efd16
Change-Id: I50aa4a7ab1a7aae5d7b31f765994abd3db06379d
2024-05-19 12:32:13 -06:00
eldritch horrors 774c56094f libstore: fix old RemoteStore::addToStore serializer
having the serializer write into `*conn` is not legal because we are
in a sinkToSource that will be drained by the remote we're connected
to. writing into `*conn` directly can break the framing protocol. it
is unlikely this code was ever run: to protocol it caters to is from
2016(!) and thoroughly untested in-tree, and since it's been present
since nix 2.17 and the 1.18 protocol broken here is nix 2.0 we might
safely assume that daemons older than nix 2.1 are no longer used now

see also #325 (though that wants <2.3 gone, this is sadly only <2.1)

Change-Id: I9d674c18f6d802f61c5d85dfd9608587b73e70a5
2024-05-19 11:57:55 +00:00
Alyssa Ross 139d31f876 Improve nix-store --delete failure message
On several occasions I've found myself confused when trying to delete
a store path, because I am told it's still alive, but
nix-store --query --roots doesn't show anything.  Let's save future
users this confusion by mentioning that a path might be alive due to
having referrers, not just roots.

(cherry picked from commit 979a019014569eee7d0071605f6ff500b544f6ac)

Upstream-PR: https://github.com/NixOS/nix/pull/10733
Change-Id: I54ae839a85f3de3393493fba27fd40d7d3af0516
2024-05-18 14:49:40 -06:00
puck 1fe58bd8a7 nix cat/dump-path/key: stop progress bar before writeFull
These commands outputs data that may not end with a newline. This
causes problems when the progress bar redraws, as that completely
wipes the last line of output. As nix key generate-secret outputs
a single line of text with no output, it shows up entirely blank,
making it look like nothing happened.

Fixes: lix-project/lix#320
Change-Id: I5ac706d71d839b6dfa760b60a351414cd96297cf
2024-05-18 17:51:16 +00:00
Pierre Bourdon d1c8fd3b09 Merge "derived-path: refuse built derived path with a non-derivation base" into main 2024-05-18 07:26:26 +00:00
julia 7a3745b076
Deprecate the online flake registries and vendor the default registry
Fixes #183, #110, #116.

The default flake-registry option becomes 'vendored', and refers
to a vendored flake-registry.json file in the install path.

Vendored copy of the flake-registry is from github:NixOS/flake-registry
at commit 9c69f7bd2363e71fe5cd7f608113290c7614dcdd.

Change-Id: I752b81c85ebeaab4e582ac01c239d69d65580f37
2024-05-18 12:27:23 +10:00
puck 23c92f0815 Merge "primops: change to std::function, allowing the passing of user data" into main 2024-05-17 21:37:41 +00:00
puck 92e1df23b3 Merge "Loosen constness on listElems() result" into main 2024-05-17 21:37:35 +00:00
Qyriad 93b7edfd07 Merge "docs: mention importNative/exec in allow-unsafe-native-code-during-evaluation" into main 2024-05-17 18:17:05 +00:00
julia 005ee33a9a Merge "Allow enabling core dumps from builds for nix & child processes" into main 2024-05-17 01:10:44 +00:00
Qyriad 5ff076d8ad docs: mention importNative/exec in allow-unsafe-native-code-during-evaluation
Both of these still needs their own actual documentation, but they are
at least now mentioned that they exist and what they're enabled by.

Change-Id: I235b9e8e627e04ed06611423c8e67a8eca233120
2024-05-17 00:41:35 +00:00
Pierre Bourdon 5a1824ebe1
derived-path: refuse built derived path with a non-derivation base
Example: /nix/store/dr53sp25hyfsnzjpm8mh3r3y36vrw3ng-neovim-0.9.5^out

This is nonsensical since selecting outputs can only be done for a
buildable derivation, not for a realised store path. The build worker
side of things ends up crashing with an assertion when trying to handle
such malformed paths.

Change-Id: Ia3587c71fe3da5bea45d4e506e1be4dd62291ddf
2024-05-17 02:16:15 +02:00