This sets up infrastructure in libutil to allow for signing other than
by a secret key in memory. #9076 uses this to implement remote signing.
(Split from that PR to allow reviewing in smaller chunks.)
Co-Authored-By: Raito Bezarius <masterancpp@gmail.com>
many paths need not be heap-allocated, and derivation env name/valye
pairs can be moved into the map.
before:
Benchmark 1: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
Time (mean ± σ): 6.883 s ± 0.016 s [User: 5.250 s, System: 1.424 s]
Range (min … max): 6.860 s … 6.905 s 10 runs
after:
Benchmark 1: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
Time (mean ± σ): 6.868 s ± 0.027 s [User: 5.194 s, System: 1.466 s]
Range (min … max): 6.828 s … 6.913 s 10 runs
the table is very small compared to cache sizes and a single indexed
load is faster than three comparisons.
before:
Benchmark 1: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
Time (mean ± σ): 6.907 s ± 0.012 s [User: 5.272 s, System: 1.429 s]
Range (min … max): 6.893 s … 6.926 s 10 runs
after:
Benchmark 1: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
Time (mean ± σ): 6.883 s ± 0.016 s [User: 5.250 s, System: 1.424 s]
Range (min … max): 6.860 s … 6.905 s 10 runs
a bunch of derivation strings contain no escape sequences. we can
optimize for this fact by first scanning for the end of a derivation
string and simply returning the contents unmodified if no escape
sequences were found. to make this even more efficient we can also use
BackedStringViews to avoid copies, avoiding heap allocations for
transient data.
before:
Benchmark 1: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
Time (mean ± σ): 6.952 s ± 0.015 s [User: 5.294 s, System: 1.452 s]
Range (min … max): 6.926 s … 6.974 s 10 runs
after:
Benchmark 1: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
Time (mean ± σ): 6.907 s ± 0.012 s [User: 5.272 s, System: 1.429 s]
Range (min … max): 6.893 s … 6.926 s 10 runs
istream sentry objects are very expensive for single-character
operations, and since we don't configure exception masks for the
istreams used here they don't even do anything. all we need is
end-of-string checks and an advancing position in an immutable memory
buffer, both of which can be had for much cheaper than istreams allow.
the effect of this change is most apparent on empty stores.
before:
Benchmark 1: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
Time (mean ± σ): 7.167 s ± 0.013 s [User: 5.528 s, System: 1.431 s]
Range (min … max): 7.147 s … 7.182 s 10 runs
after:
Benchmark 1: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
Time (mean ± σ): 6.963 s ± 0.011 s [User: 5.330 s, System: 1.421 s]
Range (min … max): 6.943 s … 6.974 s 10 runs
as written the comparisons generate copies, even though it looks as
though they shouldn't.
before:
Time (mean ± σ): 4.396 s ± 0.002 s [User: 3.894 s, System: 0.501 s]
Range (min … max): 4.393 s … 4.399 s 10 runs
after:
Time (mean ± σ): 4.260 s ± 0.003 s [User: 3.754 s, System: 0.505 s]
Range (min … max): 4.257 s … 4.266 s 10 runs
resizing a std::string clears the newly added bytes, which is not
necessary here and comes with a ~1.4% slowdown on our test nixos config.
〉 nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
before:
Time (mean ± σ): 4.486 s ± 0.003 s [User: 3.978 s, System: 0.507 s]
Range (min … max): 4.482 s … 4.492 s 10 runs
after:
Time (mean ± σ): 4.429 s ± 0.002 s [User: 3.929 s, System: 0.500 s]
Range (min … max): 4.427 s … 4.433 s 10 runs
`eval-system` option overrides just the value of `builtins.currentSystem`.
This is more useful than overriding `system` since you can build these
derivations on remote builders which can work on the given system.
Co-authored-by: John Ericson <John.Ericson@Obsidian.Systems>
Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
I wrote the `configure.ac` wrong, and so we just got no builds
supporting ACLs.
Also, it needs to be more precise because Darwin puts other stuff in
that same header, evidently.
I don't love the way this code looks. There are two larger problems:
- eval, build/scratch, destination stores (#5025) should have different
types to reflect the fact that they are used for different purposes
and those purposes correspond to different operations. It should be
impossible to "use the wrong store" in my cases.
- Since drvs can end up in both the eval and build/scratch store, we
should have some sort of union/layered store (not on the file sytem
level, just conceptual level) that allows accessing both. This would
get rid of the ugly "check both" boilerplate in this PR.
Still, it might be better to land this now / soon after minimal cleanup,
so we have a concrete idea of what problem better abstractions are
supposed to solve.
Below the comment added by this commit is a much longer comment
followed by a trust check, both of which have confused me on at
least two occasions. I figured it out once, forgot it, then had to
ask @Ericson2314 to explain it, at which point I understood it
again. I think this might confuse other people too, or maybe I will
just forget it a third time. So let's add a comment.
Farther down in the function is the following check:
```
if (!(drvType.isCA() || trusted))
throw Error("you are not privileged to build input-addressed derivations");
```
This seems really strange at first. A key property of Nix is that
you can compute the outpath of a derivation using the derivation
(and its references-closure) without trusting anybody!
The missing insight is that at this point in the code the builder
doesn't necessarily have the references-closure of the derivation
being built, and therefore needs to trust that the derivation's
outPath is honest. It's incredibly easy to overlook this, because
the only difference between these two cases is which of these
identically-named functions we used:
- `readDerivation(Source,Store)`
- `Store::readDerivation()`
These functions have different trust models (except in the special
case where the first function is used on the local store). We
should call the reader's attention to this fact.
Co-authored-by: Cole Helbling <cole.e.helbling@outlook.com>
In https://github.com/NixOS/nix/pull/6134#issuecomment-1079199888,
@thuffschmitt proposed exposing `LegacySSHStore` in Nix for
deduplication with Hydra, at least temporarily. I think that is a good
idea.
Note that the diff will look bad unless one ignores whitespace! Also try
this locally:
```shell-session
git diff --ignore-all-space HEAD^:src/libstore/legacy-ssh-store.cc HEAD:src/libstore/legacy-ssh-store.cc
git diff --ignore-all-space HEAD^:src/libstore/legacy-ssh-store.cc HEAD:src/libstore/legacy-ssh-store.hh
```
This is needed for building CA deriations with a src store / dest store
split. In particular it is needed for Hydra.
https://github.com/NixOS/hydra/issues/838 currently puts realizations,
and thus build outputs, in the local store, but it should not.
In the process, partially undo e89b5bd0bf
in that the ancient < 2.4 version is now supported again by the
serializer again. `LegacySSHStore`, instead of also asserting that the
version is at least 4, just checks that `narHash` is set.
This allows us to better test the serializer in isolation for both
versions (< 4 and >= 4).
- remove prose for the default value, which is shown programmatically
- add note on how this relates to `cores`
- add link to mentioned derivation attribute
Today, with the tests inside a `tests` intermingled with the
corresponding library's source code, we have a few problems:
- We have to be careful that wildcards don't end up with tests being
built as part of Nix proper, or test headers being installed as part
of Nix proper.
- Tests in libraries but not executables is not right:
- It means each executable runs the previous unit tests again, because
it needs the libraries.
- It doesn't work right on Windows, which doesn't want you to load a
DLL just for the side global variable . It could be made to work
with the dlopen equivalent, but that's gross!
This reorg solves these problems.
There is a remaining problem which is that sibbling headers (like
`hash.hh` the test header vs `hash.hh` the main `libnixutil` header) end
up shadowing each other. This PR doesn't solve that. That is left as
future work for a future PR.
Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
This makes for more useful manual table of contents, that displays the
information at a glance.
The `nix help-stores` command is kept as-is, even though it will show up
in the manual with the same information as these pages due to the way it
is written as a "`--help`-style" command. Deciding what to do with that
command is left for a later PR.
This change also lists all store types at the top of the respective overview page.
Co-authored-by: John Ericson <John.Ericson@Obsidian.Systems
The problem was that f880469173 forgot
that the `#include <sys/xattr.h>` was guarded by an `#ifdef __linux__`.
However, the build failure was only on FreeBSD --- turns out other
platforms have this header too!
The fix therefore uses a new configure check so we properly clear ACLs
on more platforms.
`installcheck` doesn't yet work, but the rest of the build can now
happen mostly inside a separate build directory.
Progress on #9342
Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
It is not inherently tied to `LocalStore`, it could probably even go in
`libnixutil`. Functions not attached to `LocalStore` should not be
declared in `local-store.hh`.
I am moving it to facilitate experimenting for #9344. If
canonicalisation should be done client-side in client-side builds, there
wouldn't be a `LocalStore` at all so having to include that header to
get this freestanding function is cumbersome and wrong.
Perhaps canonicalisation should still be done server-side for security
reasons --- I don't mean to make that judgement call now --- but even if
so, this freestanding function still isn't connected to `LocalStore` so
while less urgent it is still better to move out of this header.
This avoids repeated copying of the same source tree between Nix
invocations. It requires the accessor to have a "fingerprint" (e.g. a
Git revision) that uniquely determines its contents.
Closes#9343
See that issue for motivation.
Installing these is disabled by default, but we enable it (and the
additional output we want isntall these too so as not to clutter the
existing ones) to use in cross builds and dev shells.
Before it returned a list of JSON objects with store object information,
including the path in each object. Now, it maps the paths to JSON
objects with the metadata sans path.
This matches how `nix derivation show` works.
Quite hillariously, none of our existing functional tests caught this
change to `path-info --json` though they did use it. So just new
functional tests need to be added.
`Store::pathInfoToJSON` was a rather baroque functions, being full of
parameters to support both parsed derivations and `nix path-info`. The
common core of each, a simple `dValidPathInfo::toJSON` function, is
factored out, but the rest of the logic is just duplicated and then
specialized to its use-case (at which point it is no longer that
duplicated).
This keeps the human oriented CLI logic (which is currently unstable)
and the core domain logic (export reference graphs with structured
attrs, which is stable), separate, which I think is better.
Spent a while debugging why `nix-copy-closure` wasn't working anymore
and it was my shell RC printing something I added for debug.
Hopefully this can save someone else some time.
All OS and IO operations should be moved out, leaving only some misc
portable pure functions.
This is useful to avoid copious CPP when doing things like Windows and
Emscripten ports.
Newly exposed functions to break cycles:
- `restoreSignals`
- `updateWindowSize`
When doing local builds, we get phase reporting lines in the log file,
they look like '@nix {"action":"setPhase","phase":"unpackPhase"}'.
With the ssh-ng protocol, we do have access to these messages, but since we
are only including messages of type resBuildLogLine in the logs, the phase
information does not end up in the log file.
The phase reporting could probably be improved altoghether (it looks like it
is kind of accidental that these JSON messages for phase reporting show up
but others don't, just because they are actually emitted by nixpkgs' stdenv),
but as a first step I propose to make ssh-ng behave in the same way as local builds do.
Deduplicating code moreover enforcing the pattern means:
- It is easier to write new characterization tests because less boilerplate
- It is harder to mess up new tests because there are fewer places to
make mistakes.
Co-authored-by: Jacek Galowicz <jacek@galowicz.de>
I wouldn't call it *good* yet, but this will do for now.
- `RetrieveRegularNARSink` renamed to `RegularFileSink` and moved
accordingly because it actually has nothing to do with NARs in
particular.
- its `fd` field is also marked private
- `copyRecursive` introduced to dump a `SourceAccessor` into a
`ParseSink`.
- `NullParseSink` made so `ParseSink` no longer has sketchy default
methods.
This was done while updating #8918 to work with the new
`SourceAccessor`.
The brings a number of advantages, including:
- Easier to update test data if design changes (and I do think our
derivation JSON is not yet complaint with the guidelines).
- Easier to reuse test data in other implementations, inching closer to
compliance tests for Nix *the concept* rather than any one
implementation.
Before the change builder ID exhaustion printed the following message:
[0/1 built] waiting for UID to build '/nix/store/hiy9136x0iyib4ssh3w3r5m8pxjnad50-python3.11-breathe-4.35.0.drv'
After the change it should be:
[0/1 built] waiting for a free build user ID for '/nix/store/hiy9136x0iyib4ssh3w3r5m8pxjnad50-python3.11-breathe-4.35.0.drv'
Rather than having a misc tutorial page in the grab-bag "package management" section, this information should just be part of the S3 store docs.
---------
Co-authored-by: John Ericson <John.Ericson@Obsidian.Systems>
It does not belong with the data type itself.
This also materializes the fact that `copyPath` does not do any version
negotiation just just hard-codes "16".
The non-standard interface of these serializers makes it harder to test,
but this is fixed in the next commit which then adds those tests.
Worker Protocol:
Note that the worker protocol already had a serialization for
`BuildResult`; this was added in
a4604f1928. It didn't have any versioning
support because at that time reusable seralizers were not away for the protocol
version. It could thus only be used for new messages also introduced in
that commit.
Now that we do support versioning in reusable serializers, we can expand
it to support all known versions and use it in many more places.
The exist test data becomes the version 1.29 tests: note that those
files' contents are unchanged. 1.28 and 1.27 tests are added to cover
the older code-paths.
The keyered build result test only has 1.29 because the keying was also
added in a4604f19284254ac98f19a13ff7c2216de7fe176; the older
serializations are always used unkeyed.
Serve Protocol:
Conversely, no attempt was made to factor out such a serializer for the
serve protocol, so our work there in this commit for that protocol
proceeds from scratch.
It was some ad-hoc functions to account for versions, while the already
factored-out serializer just supported the latest version.
Now, we can fold that version-specific logic into the factored out one,
and so we do.
It is dead code. It was added in
8e0946e8df as part of the repeated /
enforce-determinism feature, but that was removed in
8fdd156a65.
It is not good because it skips many fields. For testing purposes we
will soon want to add a new one that doesn't skip fields, but we want to
make sure making == sensitive to those fields won't change how Nix
works. Proving in this commit that the old version is dead code achieves
that.
hashBase is ambiguous, since it's not about the digital bases, but about
the format of hashes. Base16, Base32 and Base64 are all character maps
for binary encoding.
Rename the enum Base to HashFormat.
Rename variables of type HashFormat from [hash]Base to hashFormat,
including CmdHashBase::hashFormat and CmdToBase::hashFormat.
This will allow us to factor out logic, which is currently scattered
inline, into several reusable instances
The tests are also updated to support versioning. Currently all Worker
and Serve protocol tests are using the minimum version, since no
version-specific serialisers have been created yet. But in subsequent
commits when that changes, we will test individual versions to ensure
complete coverage.
16591eb3cc (diff-19f999107b609d37cfb22c58e7f0bc1cf76edf1180e238dd6389e03cc279b604) (2013) added support for files to doBind
This is work towards allowing users to change the location of chrootRootDir, to, for example, a tmpfs.
inspired by trofi on matrix
> It looks like build sandbox created by nix-daemon runs on the same filesystem, as /nix/store including things like /tmp which makes all small temporary files hit the disk. Is it intentional? If it is is there an easy way to redirect chroot's root to be tmpfs?
dirsInChroot -> pathsInChroot
Add a new experimental `impure-env` setting that is a key-value list of
environment variables to inject into FOD derivations that specify the
corresponding `impureEnvVars`.
This allows clients to make use of this feature (without having to change the
environment of the daemon itself) and might eventually deprecate the current
behaviour (pick whatever is in the environment of the daemon) as it's more
principled and might prevent information leakage.
To start, it is just a clone of the common protocol. But now that we
have the separate protocol implementations, we can add versioning
information without the versions of one protocol leaking into another.
Using the infrastructure from the previous commit, we don't have to
duplicate code for shared behavior.
Motivation: No more perverse incentives. [0] did some awkward things
because the serialisers did not store the version. I don't want anyone
making changes to be pushed towards keeping the serialization logic with
the core data types just because it's easier or the alternative is
tedious.
The actual versioning of the Worker and Serve protocol serialisers
(Common remains unversioned as the underlying mini-protocols are not
versioned) will happen in subsequent commits / PRs.
[0]: fe1f34fa60
Copy the relevant tests to ensure the new interfaces added in the last
commit are tested.
Perhaps I should try to deduplicat these tests some more. However its
not clear how to do that outside of a big ugly C++ macro.
https://github.com/google/googletest/blob/main/docs/advanced.md has some
stuff but it is cumbersome and I didn't figure it out yet.
This is done in a separate commit in order to be sure that the first
commit really didn't change any behavior; if we changed the
implementation and the tests at once, it would be harder to tell whether
or not some behavioral changes slipped in what is supposed to be a "pure
refactor".
Co-Authored-By: Valentin Gagarin <valentin.gagarin@tweag.io>
This introduces some shared infrastructure for our notion of protocols.
We can then define multiple protocols in terms of that notion.
We an also express how particular protocols depend on each other.
For example, we can define a common protocol and a worker protocol,
where the second depends on the first in terms of the data types it can
read and write.
The "serve" protocol can just use the common one for now, but will
eventually need its own machinary just like the worker protocol for
version-aware serialisers
I think it is bad for these reasons when `tests/` contains a mix of
functional and integration tests
- Concepts is harder to understand, the documentation makes a good
unit vs functional vs integration distinction, but when the
integration tests are just two subdirs within `tests/` this is not
clear.
- Source filtering in the `flake.nix` is more complex. We need to
filter out some of the dirs from `tests/`, rather than simply pick
the dirs we want and take all of them. This is a good sign the
structure of what we are trying to do is not matching the structure
of the files.
With this change we have a clean:
```shell-session
$ git show 'HEAD:tests'
tree HEAD:tests
functional/
installer/
nixos/
```
This has been the behaviour before Nix 2.4. It was dropped in a rewrite
in 759947bf72, allowing the creation of
store paths that aren't considered valid by older Nix versions or other
Nix tooling.
Nix 2.4 didn't ship in NixOS until 22.05, and stdenv.mkDerivation in
nixpkgs drops leading periods since April 2022, so it's unlikely anyone
is relying on the current lax behaviour.
Closes#9091.
Change-Id: I4a57bd9899e1b0dba56870ae5a1b680918a18ce9
This was somewhat of a false alarm. The problem was not that the
protocol implementation actually failed to round trip, but that two of
the fields were ignored entirely --- not serialized and deserialized at
all.
For reference, those fields were added in
fa68eb367e.
This reverts commit 5e3986f59c. This
un-implements RFC 92 but fixes the critical bug #9052 which many people
are hitting. This is a decent stop-gap until a minimal reproduction of
that bug is found and a proper fix can be made.
Mostly fixed#9052, but I would like to leave that issue open until we
have a regression test, so I can then properly fix the bug (unbreaking
RFC 92) later.
Detected by `gcc` as:
CXX src/libstore/profiles.o
src/libstore/profiles.cc: In function 'void nix::deleteGenerationsGreaterThan(const Path&, GenerationNumber, bool)':
src/libstore/profiles.cc:186:50: warning: comparison of integer expressions of different signedness: 'int' and 'nix::GenerationNumber' {aka 'long unsigned int'} [-Wsign-compare]
186 | for (auto keep = 0; i != gens.rend() && keep < max; ++i, ++keep);
| ~~~~~^~~~~
Without this change, nix build processes will not drop the locks for derivation goals
which have already been built by another process when the current process gets
round to building them. This means the locks are held until the process
terminates.
If there are other nix build processes in a similar state, they will also try to
acquire the same locks when they try to build the same derivation, and so will
wait until the lock holder terminates (which might be a very long time if it has
a lot to build). In some pathological cases, those processes might be holding
their own locks on other derivations due to the same issue, and this can lead to
deadlock.
Resolves#6468
The Derivation parser and old ATerm unfortunately leaves few ways to get
nice errors when an old version of Nix encounters a new version of the
format. The most likely scenario for this to occur is with a new client
making a derivation that the old daemon it is communicating with cannot
understand.
The extensions we just created for dynamic derivation deps will add a
version field, solving the problem going forward, but there is still the
issue of what to do about old versions of Nix up to now.
The solution here is to carefully catch the bad error from the daemon
that is likely to indicate this problem, and add some extra context to
it.
There is another "Ugly backwards compatibility hack" in
`remote-store.cc` that also works by transforming an error.
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
We use the same nested map representation we used for goals, again in
order to save space. We might someday want to combine with `inputDrvs`,
by doing `V = bool` instead of `V = std::set<OutputName>`, but we are
not doing that yet for sake of a smaller diff.
The ATerm format for Derivations also needs to be extended, in addition
to the in-memory format. To accomodate this, we added a new basic
versioning scheme, so old versions of Nix will get nice errors. (And
going forward, if the ATerm format changes again the errors will be even
better.)
`parsedStrings`, an internal function used as part of parsing
derivations in A-Term format, used to consume the final `]` but expect
the initial `[` to already be consumed. This made for what looked like
unbalanced brackets at callsites, which was confusing. Now it consumes
both which is hopefully less confusing.
As part of testing, we also created a unit test for the A-Term format for
regular non-experimental derivations too.
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
Apply suggestions from code review
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
- Don't assert: Derivation ATerms are not necessarily produced by Nix,
and parsers should always throw graceful errors
- Improve error message from `static void except(..)`, shows both what
we expected and what we actually got.
The intention is that we backport it, and then hopefully a few people
might get slightly better errors if they try out new experimental drv
files (for RFC 92) with an old version of Nix.
Continue with the characterization testing idioms begun in
c70484454f, but this time for unit tests.
Co-authored-by: Andreas Rammhold <andreas@rammhold.de>
To avoid dealing with an optional `drvPath` (because we might not know
it yet) everywhere, make an `CreateDerivationAndRealiseGoal`. This goal
just builds/substitutes the derivation file, and then kicks of a build
for that obtained derivation; in other words it does the chaining of
goals when the drv file is missing (as can already be the case) or
computed (new case).
This also means the `getDerivation` state can be removed from
`DerivationGoal`, which makes the `BasicDerivation` / in memory case and
`Derivation` / drv file file case closer together.
The map type is factored out for clarity, and because we will soon hvae
a second use for it (`Derivation` itself).
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
We're about to split up `DerivationGoal` a bit. At that point
`makeDerivationGoal` will mean something more specific than it does
today. (Perhaps a future rename will make this clearer.)
On the other hand, the more public `Worker::makeGoal` function will
continue to work exactly as before. So by moving some call sites to use
that instead, we preemptively avoid issues in the next step.
Types converted:
- `NixStringContextElem`
- `OutputsSpec`
- `ExtendedOutputsSpec`
- `DerivationOutput`
- `DerivationType`
Existing ones mostly conforming the pattern cleaned up:
- `ContentAddressMethod`
- `ContentAddressWithReferences`
The `DerivationGoal::derivationType` field had a bogus initialization,
now caught, so I made it `std::optional`. I think #8829 can make it
non-optional again because it will ensure we always have the derivation
when we construct a `DerivationGoal`.
See that issue (#7479) for details on the general goal.
`git grep 'Raw::Raw'` indicates the two types I didn't yet convert
`DerivedPath` and `BuiltPath` (and their `Single` variants) . This is
because @roberth and I (can't find issue right now...) plan on reworking
them somewhat, so I didn't want to churn them more just yet.
Co-authored-by: Eelco Dolstra <edolstra@gmail.com>
`EvalState::mkSingleDerivedPathString` previously contained its own
inverse (printing, rather than parsing) in order to validate what was
parsed. Now that is pulled out into its own separate function:
`EvalState::coerceToSingleDerivedPath`.
In additional that pulled out logic is deduplicated with
`EvalState::mkOutputString` via `EvalState::mkOutputStringRaw`, which is
itself deduplicated (and generalized) with
`DownstreamPlaceholder::mkOutputStringRaw`.
All these changes make the unit tests simpler.
(We would ideally write more unit tests for `mkSingleDerivedPathString`
`coerceToSingleDerivedPath` directly, but we cannot yet do that because
the IO in reading the store path won't work when the dummy store cannot
hold anything. Someday we'll have a proper in-memory store which will
work for this.)
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
std::move(state->data) and data.empty() were called in a loop, and
could run with no other threads intervening. Accessing moved objects
is undefined behavior, and could cause a crash.
Virtual methods are no longer valid once the derived destructor has
run. This means the compiler is free to optimize them to be
non-virtual.
Found using clang-tidy
We want to be able to write down `foo.drv^bar.drv^baz`:
`foo.drv^bar.drv` is the dynamic derivation (since it is itself a
derivation output, `bar.drv` from `foo.drv`).
To that end, we create `Single{Derivation,BuiltPath}` types, that are
very similar except instead of having multiple outputs (in a set or
map), they have a single one. This is for everything to the left of the
rightmost `^`.
`NixStringContextElem` has an analogous change, and now can reuse
`SingleDerivedPath` at the top level. In fact, if we ever get rid of
`DrvDeep`, `NixStringContextElem` could be replaced with
`SingleDerivedPath` entirely!
Important note: some JSON formats have changed.
We already can *produce* dynamic derivations, but we can't refer to them
directly. Today, we can merely express building or example at the top
imperatively over time by building `foo.drv^bar.drv`, and then with a
second nix invocation doing `<result-from-first>^baz`, but this is not
declarative. The ethos of Nix of being able to write down the full plan
everything you want to do, and then execute than plan with a single
command, and for that we need the new inductive form of these types.
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
This enables nix to correctly report what will be fetched in the case
that everything is a cache hit.
Note however that if an intermediate build of something which is not
cached could still cause products to end up being substituted if the
intermediate build results in a CA path which is in the cache.
Fixes#8615.
Signed-off-by: Peter Waller <p@pwaller.net>
When receiving a stream of NARs through the ssh-ng protocol, an already
existing path would cause the NAR archive to not be read in the stream,
resulting in trying to parse the NAR as a ValidPathInfo. This results in
the error message:
error: not an absolute path: 'nix-archive-1'
Fixes#6253
Usually this problem is avoided by running QueryValidPaths before
AddMultipleToStore, but can arise when two parallel nix processes gets
the same response from QueryValidPaths. This makes the problem more
prominent when running builds in parallel.
When loading a derivation from a JSON, malformed input would trigger
cryptic "assertion failed" errors. Simply replacing calls to `operator []`
with calls to `.at()` was not enough, as this would cause json.execptions
to be printed verbatim.
Display nice error messages instead and give some indication where the
error happened.
*Before:*
```
$ echo 4 | nix derivation add
error: [json.exception.type_error.305] cannot use operator[] with a string argument with number
$ nix derivation show nixpkgs#hello | nix derivation add
Assertion failed: (it != m_value.object->end()), function operator[], file /nix/store/8h9pxgq1776ns6qi5arx08ifgnhmgl22-nlohmann_json-3.11.2/include/nlohmann/json.hpp, line 2135.
$ nix derivation show nixpkgs#hello | jq '.[] | .name = 5' | nix derivation add
error: [json.exception.type_error.302] type must be string, but is object
$ nix derivation show nixpkgs#hello | jq '.[] | .outputs = { out: "/nix/store/8j3f8j-hello" }' | nix derivation add
error: [json.exception.type_error.302] type must be object, but is string
```
*After:*
```
$ echo 4 | nix derivation add
error: Expected JSON of derivation to be of type 'object', but it is of type 'number'
$ nix derivation show nixpkgs#hello | nix derivation add
error: Expected JSON object to contain key 'name' but it doesn't
$ nix derivation show nixpkgs#hello | jq '.[] | .name = 5' | nix derivation add
error: Expected JSON value to be of type 'string' but it is of type 'number'
$ nix derivation show nixpkgs#hello | jq '.[] | .outputs = { out: "/nix/store/8j3f8j-hello" }' | nix derivation add
error:
… while reading key 'outputs'
error: Expected JSON value to be of type 'object' but it is of type 'string'
```
These docs explain the implementation relative to the local store
originals. The original declaration of virtual methods can still be
consulted for proper interface-level documentation.
As an optimisation for LocalStore, we read all the store directory entries into
a set. Checking for membership of this set is much faster than a stat syscall.
However for LocalOverlayStore, the lower store directory is expected to contain
a vast number of entries and reading them all can take a very long time.
So instead of enumerating them all upfront, we call pathExists as needed. This
means making stat syscalls for each store path, but the upper layer is expected
to be relatively small compared to the lower store so that should be okay.
It was initially unclear to me which of these are temporary state for
the verify paths computation, and which of these are the results of that
computation to be used in the rest of the function. Now, it is clear,
and enforced.
We don't care about non-store-paths in there (things like `.links`, are,
in fact, allowed). So let's just skip them up front and be more strongly
typed.
Will need to do subclass-specific implementations in the next commit.
This isn't because there will be multiple variations of the daemon
protocol (whew!) but because different clients pick and choose different
parts to use.
This makes it more useful. In general, the derivation will be in one
store, and the realisation info is in another.
This also helps us avoid duplication. See how `resolveDerivedPath` is
now simpler because it uses `queryPartialDerivationOutputMap`. In #8369
we get more flavors of derived path, and need more code to resolve them
all, and this problem only gets worse.
The fact that we need a new method to deal with the multiple dispatch is
unfortunate, but this generally relates to the fact that `Store` is a
sub-par interface, too bulky/unwieldy and conflating separate concerns.
Solving that is out of scope of this PR.
This is part of the RFC 92 work. See tracking issue #6316
We were bedeviled by sandboxing issues when working on the layered
store. The problem ended up being that when we have nested nix builds,
and the inner store is inside the build dir (e.g. store is
`/build/nix-test/$name/store`, build dir is `/build`) bind mounts
clobber each other and store paths cannot be found.
After thoroughly cleaning up `local-derivation-goal.cc`, we might be
able to make that work. But that is a lot of work. For now, we just fail
earlier with a proper error message.
Finally, test this: nested sandboxing without the problematic store dir
should work, and with should fail with the expected error message.
Co-authored-by: Dylan Green <67574902+cidkidnix@users.noreply.github.com>
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
Whereas `ContentAddressWithReferences` is a sum type complex because different
varieties support different notions of reference, and
`ContentAddressMethod` is a nested enum to support that,
`ContentAddress` can be a simple pair of a method and hash.
`ContentAddress` does not need to be a sum type on the outside because
the choice of method doesn't effect what type of hashes we can use.
Co-Authored-By: Cale Gibbard <cgibbard@gmail.com>
- Improved API docs from comment
- Exit codes are for `nix-build`, not just `nix-store --release`
- Make note in tests so the magic numbers are not surprising
Picking up where #8387 left off.
Previously it was not possible to open a local store when its database is on a read-only filesystem. Obviously a store on a read-only filesystem cannot be modified, but it would still be useful to be able to query it.
This change adds a new read-only setting to LocalStore. When set to true, Nix will skip operations that fail when the database is on a read-only filesystem (acquiring big-lock, schema migration, etc), and the store database will be opened in immutable mode.
Co-authored-by: Ben Radford <benradf@users.noreply.github.com>
Co-authored-by: cidkidnix <cidkidnix@protonmail.com>
Co-authored-by: Dylan Green <67574902+cidkidnix@users.noreply.github.com>
Co-authored-by: John Ericson <git@JohnEricson.me>
Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
If the garbage collector has acquired the global GC lock, but hasn't
created the GC socket yet, then a client attempting to connect would
get ENOENT. Note that this only happens when the GC runs for the first
time on a machine. Subsequently clients will get ECONNREFUSED which
was already handled.
Fixes#7370.
Pass this around instead of `Source &` and `Sink &` directly. This will
give us something to put the protocol version on once the time comes.
To do this ergonomically, we need to expose `RemoteStore::Connection`,
so do that too. Give it some more API docs while we are at it.
The motivation is exactly the same as for the last commit. In addition,
this anticipates us formally defining separate serialisers for the serve
protocol.
See API docs on that struct for why. The pasing as as template argument
doesn't yet happen in that commit, but will instead happen in later
commit.
Also make `WorkerOp` (now `Op`) and enum struct. This led us to catch
that two operations were not handled!
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
This is generally a fine practice: Putting implementations in headers
makes them harder to read and slows compilation. Unfortunately it is
necessary for templates, but we can ameliorate that by putting them in a
separate header. Only files which need to instantiate those templates
will need to include the header with the implementation; the rest can
just include the declaration.
This is now documenting in the contributing guide.
Also, it just happens that these polymorphic serializers are the
protocol agnostic ones. (Worker and serve protocol have the same logic
for these container types.) This means by doing this general template
cleanup, we are also getting a head start on better indicating which
code is protocol-specific and which code is shared between protocols.
- Greatly expand API docs
- Clean up code in misc ways
- Instead of a complicated single loop on generations, do different
operations in successive subsequent steps.
- Avoid `ref` in one place where `&` is fine
- Just return path instead of mutating an argument in `makeName`
Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
Rather than doing `allowEmpty` as boolean, have separate types and use
`std::optional`. This makes it harder to forget the possibility of an
empty path.
The `build-hook` setting was categorized as a `PathSetting`, but
actually it was split into arguments. No good! Now, it is
`Setting<Strings>` which actually reflects what it means and how it is
used.
Because of the subtyping, we now also have support for
`Setting<std::optional<String>>` in general. I imagine this can be used
to clean up many more settings also.
A library shouldn't require changes to the caller's argument handling,
especially if it doesn't have to, and indeed we don't have to.
This changes the lookup order to prioritize the hardcoded path to nix
if it exists. The static executable still finds itself through /proc
and the like.
Introduce what substituters "are" in the configuration option entry.
Remove arbitrary line breaks for easier editing in the future.
Link glossary some more.
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
Co-authored-by: John Ericson <git@JohnEricson.me>
Previously, for tarball flakes, we recorded the original URL of the
tarball flake, rather than the URL to which it ultimately
redirects. Thus, a flake URL like
http://example.org/patchelf-latest.tar that redirects to
http://example.org/patchelf-<revision>.tar was not really usable. We
couldn't record the redirected URL, because sites like GitHub redirect
to CDN URLs that we can't rely on to be stable.
So now we use the redirected URL only if the server returns the
`x-nix-is-immutable` or `x-amz-meta-nix-is-immutable` headers in its
response.
* Document manual migration for use-xdg-base-directories
As there's currently no automatic migration for use-xdg-base-directories
option, add instructions for manual migration to the option's
description.
Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
When encountering a build error, Nix moves the output paths out of the
chroot into their final location (for “easier debugging of build
failures”). However this was broken for chroot stores as it was moving
it to the _logical_ location, not the _physical_ one.
Fix it by moving to the physical (_real_) location.
Fix https://github.com/NixOS/nix/issues/8395
The `hashed-mirrors` option did use to have this default value,
but it was removed and re-added with an empty default value.
As the autogenerated docs show the (actual) default values from code,
remove this incorrect reference from the docs.
I was updating my nix.conf settings after a few years and noticed this.
This adds a new configuration option to Nix, `always-allow-substitutes`,
whose effect is simple: it causes the `allowSubstitutes` attribute in
derivations to be ignored, and for substituters to always be used.
This is extremely valuable for users of Nix in CI, where usually
`nix-build-uncached` is used. There, derivations which disallow
substitutes cause headaches as the inputs for building already-cached
derivations need to be fetched to spuriously rebuild some simple text
file.
This option should be a good middle-ground, since it doesn't imply
rebuilding the world, such as the approach I took in
https://github.com/NixOS/nixpkgs/pull/221048
Using abstract types like can help cut down on compilation time, both
from scratch, and especially incremental builds during development. The
idea is that `worker-protocol.hh` can declare all the (de)serializers, but
only again abstract types; when code needs to use some (de)serializers, it can
include headers just for the data types it needs to (de)serialize.
`store-api.hh` in particular is a bit of a sledgehammer, and the data
types we want to serialize have their own headers.
This is the more typically way to do [Argument-dependent
lookup](https://en.cppreference.com/w/cpp/language/adl)-leveraging
generic serializers in C++. It makes the relationship between the `read`
and `write` methods more clear and rigorous, and also looks more
familiar to users coming from other languages that do not have C++'s
libertine ad-hoc overloading.
I am returning to this because during the review in
https://github.com/NixOS/nix/pull/6223, it came up as something that
would make the code easier to read --- easier today hopefully already,
but definitely easier if we were have multiple codified protocols with
code sharing between them as that PR seeks to accomplish.
If I recall correctly, the main criticism of this the first time around
(in 2020) was that having to specify the type when writing, e.g.
`WorkerProto<MyType>::write`, was too verbose and cumbersome. This is
now addressed with the `workerProtoWrite` wrapper function.
This method is also the way `nlohmann::json`, which we have used for a
number of years now, does its serializers, for what its worth.
This reverts commit 45a0ed82f0. That
commit in turn reverted 9ab07e99f5.
This is good in general, but in particular ensures when we heavily
refactor it in the next commit there is less likelihood for an
unintentional change in behavior to sneak in.
These items are not templates, and they declared in
`worker-protocol.hh`; therefore they should live in a
`worker-protocol.cc`.
Anything else needlessly diverges from convention. After all, it is not
like this code is only used in `remote-store.cc`; it is also used in
`daemon.cc`. There is no good reason to place it with the client
implementation or the server implementation when it used equally by
both.
This requires switching on SQLITE_OPEN_URI because there is no open flag to
make the database immutable. Without immutable, sqlite will still attempt to
create journal and wal files, even when the database is opened read-only.
https://www.sqlite.org/c3ref/open.html
The immutable parameter is a boolean query parameter that indicates that the
database file is stored on read-only media. When immutable is set, SQLite
assumes that the database file cannot be changed, even by a process with higher
privilege, and so the database is opened read-only and all locking and change
detection is disabled.
Nix does not manage the overlayfs mount point itself, but the correct
functioning of the overlay store does depend on this mount point being set up
correctly. Rather than just assume this is the case, check that the lowerdir
and upperdir options are what we expect them to be. This check is on by
default, but can be disabled if needed.
They were improperly added in 8a93b5a551.
They were not `.gitignore`d because they were stale in that commit --
build artifacts no longer used that name by then and so `.gitignore` was
updated accordingly.
Previously, we relied on the `shutdown()` function to terminate `accept()`
calls on a listening socket. However, this approach did not work on macOS as
the waiting `accept()` call is not considered a connected socket, resulting in
an `ENOTCONN` error. Instead, we now close the listening socket to terminate
the `accept()` call.
Additionally, we fixed a resource management issue where we set the
`daemonSocket` variable to -1, triggering resource cleanup and causing the
`stopDaemon` function to be called twice. This resulted in errors as the socket
was already closed by the time the second `stopDaemon` call was made. Instead of
setting `daemonSocket` to -1, we now release the socket using the `release()`
method on a unique pointer. This properly transfers ownership and allows for
correct resource cleanup.
These changes ensure proper behavior and resource management for the
recursive-nix feature on macOS.
In other words, use a plain `ContentAddress` not
`ContentAddressWithReferences` for `DerivationOutput::CAFixed`.
Supporting fixed output derivations with (fixed) references would be a
cool feature, but it is out of scope at this moment.
Recently, I encountered the "NAR info file 'xxxx' is corrupt" error
with my binary cache. The message is not helpful in determining, which
kind of corruption happened. The file, fetched with curl, looked
reasonably.
This commit adds more information to the error message, which should
allow debugging and hopefully fixing the problem.
We finally test the status quo of remote build trust in a number of
ways. We create a new experimental feature on `nix-daemon` to do so.
PR #3921, which improves the situation with trustless remote building,
will build upon these changes. This code / tests was pull out of there
to make this, so everything is easier to review, and in particular we
test before and after so the new behavior in that PR is readily apparent
from the testsuite diff alone.
Issues:
1. Features gated on disabled experimental settings should warn and be
ignored, not silently succeed.
2. Experimental settings in the same config "batch" (file or env var)
as the enabling of the experimental feature should work.
3. For (2), the order should not matter.
These are analogous to the issues @roberth caught with my changes for
arg handling, but they are instead for config handling.
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
In many cases we are dealing with a collection of realisations, they are
all outputs of the same derivation. In that case, we don't need
"derivation hashes modulos" to be part of our map key, because the
output names alone will be unique. Those hashes are still part of the
realisation proper, so we aren't loosing any information, we're just
"normalizing our schema" by narrowing the "primary key".
Besides making our data model a bit "tighter" this allows us to avoid a
double `for` loop in `DerivationGoal::waiteeDone`. The inner `for` loop
was previously just to select the output we cared about without knowing
its hash. Now we can just select the output by name directly.
Note that neither protocol is changed as part of this: we are still
transferring `DrvOutputs` over the wire for `BuildResult`s. I would only
consider revising this once #6223 is merged, and we can mention protocol
versions inside factored-out serialization logic. Until then it is
better not change anything because it would come a the cost of code
reuse.
If my memory is correct, @edolstra objected to modifying `wantedOutputs`
upon falling back to doing a build (as we did before), because we should
only modify it in response to new requests --- *actual* wants --- and
not because we are "incidentally" building all the outptus beyond what
may have been requested.
That's a fair point, and the alternative is to replace the boolean soup
with proper enums: Instead of modifying `wantedOuputs` som more, we'll
modify `needsRestart` to indicate we are passed the need.
In https://github.com/NixOS/nix/pull/6311#discussion_r834863823, I
realized since derivation goals' wanted outputs can "grow" due to
overlapping dependencies (See `DerivationGoal::addWantedOutputs`, called
by `Worker::makeDerivationGoalCommon`), the previous bug fix had an
unfortunate side effect of causing more pointless rebuilds.
In paticular, we have this situation:
1. Goal made from `DerivedPath::Built { foo, {a} }`.
2. Goal gives on on substituting, starts building.
3. Goal made from `DerivedPath::Built { foo, {b} }`, in fact is just
modified original goal.
4. Though the goal had gotten as far as building, so all outputs were
going to be produced, `addWantedOutputs` no longer knows that and so
the goal is flagged to be restarted.
This might sound far-fetched with input-addressed drvs, where we usually
basically have all our goals "planned out" before we start doing
anything, but with CA derivation goals and especially RFC 92, where *drv
resolution* means goals are created after some building is completed, it
is more likely to happen.
So the first thing to do was restore the clearing of `wantedOutputs` we
used to do, and then filter the outputs in `buildPathsWithResults` to
only get the ones we care about.
But fix also has its own side effect in that the `DerivedPath` in the
`BuildResult` in `DerivationGoal` cannot be trusted; it is merely the
*first* `DerivedPath` for which this goal was originally created.
To remedy this, I made `BuildResult` be like it was before, and instead
made `KeyedBuildResult` be a subclass wit the path. Only
`buildPathsWithResults` returns `KeyedBuildResult`s, everything else
just becomes like it was before, where the "key" is unambiguous from
context.
I think separating the "primary key" field(s) from the other fields is
good practical in general anyways. (I would like to do the same thing
for `ValidPathInfo`.) Among other things, it allows constructions like
`std::map<Key, ThingWithKey>` where doesn't contain duplicate keys and
just precludes the possibility of those duplicate keys being out of
sync.
We might leverage the above someday to overload `buildPathsWithResults`
to take a *set* of return a *map* per the above.
-----
Unfortunately, we need to avoid C++20 strictness on designated
initializers.
(BTW
https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2021/p2287r1.html
this offers some new syntax for this use-case. Hopefully this will be
adopted and we can eventually use it.)
No having that yet, maybe it would be better to not make
`KeyedBuildResult` a subclass to just avoid this.
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
As requested by @roberth, it is good to call out the specific instances
we care about, which is `!` for the RPC protocols, and `^` for humans.
This doesn't take advantage of parametricity as much, but since the
human and computer interfaces are good to decouple anyways (we don't
care if they drift further apart over time in the slightest) some
separation and slight duplication is fine.
Also, unit test both round trips.
More progress on issue #5729
The method trivially generalizes to be store-implementation-agnostic, in
fact.
However, we force it to continue to be unimplemented with `RemoteStore`
and `LegacySSHStore` because the implementation we'd get via the
generalization is probably not the one users expect. This keeps our
hands untied to do it right going forward.
For more about the tension between the scheduler logic being
store-type-agnostic and remote stores doing their own scheduling, see
issues #5025 and #5056.
The warning message should produce an installable name that can be
passed to `nix build`, `nix path-info`, etc. again. Since the CLI
expects that the .drv path and the output names are separated by
a caret, the warning message must also separate the .drv path and output
names with a caret.
However, `DerivedPath::Built.to_string()` uses an exclamation point as
the separator instead. This commit adds a `separator` argument to the
to_string method.
This changes the warning message from:
If this command is now failing try again with '/nix/store/foo.drv!*'
to:
If this command is now failing try again with '/nix/store/foo.drv^*'
More progress on issue #5729.
Instead of having it by the default method in `Store` itself, have it be
the implementation in `DummyStore` and `LegacySSHStore`. Then just the
implementations which fail to provide the method pay the "penalty" of
dealing with the icky `unimplemented` function for non-compliance.
Combined with my other recent PRs, this finally makes `Store` have no
`unsupported` calls!
Getting the occasional SQLITE_BUSY is expected when the database is being
accessed concurrently. The retry will likely succeed so it is pointless to warn
immediately. Instead we track how long each retrySQLite block has been running,
and only begin warning after a second has elapsed (and then every 10 seconds
subsequently).
libutil is a dependency of libstore, so it should always be
initialized as such.
libutil is also a dependency of libmain. Being explicit about this
dependency might be good, but not worth the slight code complexity
until the library structure gets more advanced.
Part of an effort to make it easier to initialize the right things,
by moving code into the appropriate libraries.
This code is bad. We shouldn't unset variables in programs whose
children may need them. Fixing one issue at a time, so postponing.
See https://github.com/NixOS/nix/issues/7731
Part of an effort to make it easier to initialize the right things,
by moving code into the appropriate libraries.
