Commit graph

934 commits

Author SHA1 Message Date
Valentin Gagarin
ac730622e8
document where the value of builtins.nixPath comes from (#9113)
* document default values for `nix-path`

also note how it's overridden and note the effect of `restrict-eval`
2024-03-09 17:57:57 +00:00
pennae
5d9fdab3de use byte indexed locations for PosIdx
we now keep not a table of all positions, but a table of all origins and
their sizes. position indices are now direct pointers into the virtual
concatenation of all parsed contents. this slightly reduces memory usage
and time spent in the parser, at the cost of not being able to report
positions if the total input size exceeds 4GiB. this limit is not unique
to nix though, rustc and clang also limit their input to 4GiB (although
at least clang refuses to process inputs that are larger, we will not).

this new 4GiB limit probably will not cause any problems for quite a
while, all of nixpkgs together is less than 100MiB in size and already
needs over 700MiB of memory and multiple seconds just to parse. 4GiB
worth of input will easily take multiple minutes and over 30GiB of
memory without even evaluating anything. if problems *do* arise we can
probably recover the old table-based system by adding some tracking to
Pos::Origin (or increasing the size of PosIdx outright), but for time
being this looks like more complexity than it's worth.

since we now need to read the entire input again to determine the
line/column of a position we'll make unsafeGetAttrPos slightly lazy:
mostly the set it returns is only used to determine the file of origin
of an attribute, not its exact location. the thunks do not add
measurable runtime overhead.

notably this change is necessary to allow changing the parser since
apparently nothing supports nix's very idiosyncratic line ending choice
of "anything goes", making it very hard to calculate line/column
positions in the parser (while byte offsets are very easy).
2024-03-06 23:48:42 +01:00
John Ericson
d4ad1fcf30 Avoid creating temporary store object for git over the wire
Instead, serialize as NAR and send that over, then rehash sever side.
This is alorithmically simpler, but comes at the cost of a newer
parameter to `Store::addToStoreFromDump`.

Co-authored-by: Eelco Dolstra <edolstra@gmail.com>
2024-02-27 11:27:34 -05:00
John Ericson
201551c937 Add Git object hashing to the store layer
Part of RFC 133

Extracted from our old IPFS branches.

Co-Authored-By: Matthew Bauer <mjbauer95@gmail.com>
Co-Authored-By: Carlo Nucera <carlo.nucera@protonmail.com>
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
Co-authored-by: Florian Klink <flokli@flokli.de>
2024-02-27 11:27:34 -05:00
Johannes Kirschbauer
d83008c3a7
documentation: clarify genericClosure (#10003)
* doc: clarify genericClosure documentation

Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
2024-02-24 12:34:53 +00:00
Rebecca Turner
f05c13ecc2
Remove the concept of "skipped frames" 2024-02-22 17:14:55 -08:00
Théophane Hufschmitt
774e7ca584
Merge pull request #9914 from 9999years/debugger-on-trace
Enter debugger on `builtins.trace` with an option
2024-02-21 18:32:56 +01:00
Rebecca Turner
b111fba8cd
Add documentation, rename to debugger-on-trace 2024-02-21 09:07:39 -08:00
Rebecca Turner
2a8fe9a938
:quit in the debugger should quit the whole program 2024-02-20 10:01:13 -08:00
Théophane Hufschmitt
6f4bb1b584
Merge pull request #10043 from edolstra/fix-readonly-fetchToStore
fetchToStore(): Don't always respect settings.readOnlyMode
2024-02-20 12:07:48 +01:00
Eelco Dolstra
7cb4d0c5b7 fetchToStore(): Don't always respect settings.readOnlyMode
It's now up to the caller whether readOnlyMode should be applied. In
some contexts (like InputScheme::fetch()), we always need to fetch.
2024-02-20 11:46:49 +01:00
Alois Wohlschlager
89e21ab4bd
Restore builtins.pathExists behavior on broken symlinks
Commit 83c067c0fa changed `builtins.pathExists`
to resolve symlinks before checking for existence. Consequently, if the path
refers to a symlink itself, existence of the target of the symlink (instead of
the symlink itself) was checked. Restore the previous behavior by skipping
symlink resolution in the last component.
2024-02-13 18:09:55 +01:00
Rebecca Turner
8b7eb7400b
Enter debugger on builtins.trace with an option 2024-02-08 12:07:37 -08:00
Rebecca Turner
c0e7f50c1a
Rename hintfmt to HintFmt 2024-02-08 11:58:25 -08:00
Théophane Hufschmitt
1ba9780cf5
Merge pull request #9834 from 9999years/structured-errors
Towards structured error classes
2024-02-08 20:00:25 +01:00
Eelco Dolstra
a6737b7e17 CanonPath, SourcePath: Change operator + to /
This is less confusing and makes it more similar to std::filesystem::path.
2024-02-05 15:17:39 +01:00
Rebecca Turner
87dc4bc7d1
Attach positions to errors in derivationStrict 2024-02-01 16:39:38 -08:00
Rebecca Turner
c6a89c1a16
libexpr: Support structured error classes
While preparing PRs like #9753, I've had to change error messages in
dozens of code paths. It would be nice if instead of

    EvalError("expected 'boolean' but found '%1%'", showType(v))

we could write

    TypeError(v, "boolean")

or similar. Then, changing the error message could be a mechanical
refactor with the compiler pointing out places the constructor needs to
be changed, rather than the error-prone process of grepping through the
codebase. Structured errors would also help prevent the "same" error
from having multiple slightly different messages, and could be a first
step towards error codes / an error index.

This PR reworks the exception infrastructure in `libexpr` to
support exception types with different constructor signatures than
`BaseError`. Actually refactoring the exceptions to use structured data
will come in a future PR (this one is big enough already, as it has to
touch every exception in `libexpr`).

The core design is in `eval-error.hh`. Generally, errors like this:

    state.error("'%s' is not a string", getAttrPathStr())
      .debugThrow<TypeError>()

are transformed like this:

    state.error<TypeError>("'%s' is not a string", getAttrPathStr())
      .debugThrow()

The type annotation has moved from `ErrorBuilder::debugThrow` to
`EvalState::error`.
2024-02-01 16:39:38 -08:00
John Ericson
caea7dcb7e Change an allowPath call to take a store path again
This looks like a revert of #5844, but is not.

That one was needed because
d90f9d4b99 (diff-0f59bb6f197822ef9f19ceae9624989499d170c84dfdc1f486a8959bb4588cafR85)
changed the type of the argument to `allowPath` from a `StorePath` to a
`Path`. But since
caabc4f648 (diff-0f59bb6f197822ef9f19ceae9624989499d170c84dfdc1f486a8959bb4588cafL100-R92),
it is a `StorePath` again.

I think this is worth changing because we want to be very careful about
`toRealPath` and the evaluator --- ideally the choice of real path does
not affect evaluation at all. So using it fewer times is better.
2024-01-31 11:44:06 -05:00
Eelco Dolstra
b36ff47e7c Resolve symlinks in a few more places
Fixes #9882.
2024-01-30 15:35:31 +01:00
lexi
08f38a3a40
Fix typo in primops.cc (and therefore Nix docs)
This also fixes the typo in the Nix docs at https://nixos.org/manual/nix/unstable/language/builtins.
2024-01-25 15:30:51 +01:00
Rebecca Turner
cb7fbd4d83
Print value on type error
Adds the failing value to `value is <TYPE> while a <TYPE> is expected`
error messages.
2024-01-22 08:56:02 -08:00
Mel Zuser
25c889baac Fix performance of builtins.substring for empty substrings
When returning a 0-length substring, avoid calling coerceToString,
since it returns a string_view with the string's length, which is
expensive to compute for large strings.
2024-01-12 08:37:04 -08:00
Rebecca Turner
4feb7d9f71
Combine AbstractPos, PosAdapter, and Pos
Also move `SourcePath` into `libutil`.

These changes allow `error.hh` and `error.cc` to access source path and
position information, which we can use to produce better error messages
(for example, we could consider omitting filenames when two or more
consecutive stack frames originate from the same file).
2024-01-08 10:59:41 -08:00
Robert Hensing
83f5622545
Merge pull request #9658 from pennae/env-diet
reduce the size of Env by one pointer
2023-12-31 13:57:16 +01:00
pennae
1fe66852ff reduce the size of Env by one pointer
since `up` and `values` are both pointer-aligned the type field will
also be pointer-aligned, wasting 48 bits of space on most machines. we
can get away with removing the type field altogether by encoding some
information into the `with` expr that created the env to begin with,
reducing the GC load for the absolutely massive amount of single-entry
envs we create for lambdas. this reduces memory usage of system eval by
quite a bit (reducing heap size of our system eval from 8.4GB to 8.23GB)
and gives similar savings in eval time.

running `nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'`

before:

  Time (mean ± σ):      5.576 s ±  0.003 s    [User: 5.197 s, System: 0.378 s]
  Range (min … max):    5.572 s …  5.581 s    10 runs

after:

  Time (mean ± σ):      5.408 s ±  0.002 s    [User: 5.019 s, System: 0.388 s]
  Range (min … max):    5.405 s …  5.411 s    10 runs
2023-12-30 18:55:13 +01:00
Shea Levy
c3942ef85f
Build IFD in the build store when using eval-store.
Previously, IFDs would be built within the eval store, even though one
is typically using `--eval-store` precisely to *avoid* local builds.

Because the resulting Nix expression must be copied back to the eval
store in order to be imported, this requires the eval store to trust
the build store's signatures.
2023-12-23 21:33:56 -05:00
Robert Hensing
ee439734e9
Merge pull request #9582 from pennae/misc-opts
a packet of small optimizations
2023-12-22 17:00:59 +01:00
Silvan Mosberger
ea454d8687 Undeprecate isNull
There's no good reason to deprecate it:
- For consistency reasons it should continue to exist, such that all
primitive types have a corresponding `builtins.is*` primop.
- There's no implementation cost to continuing to have this function
- It costs users time to try to migrate away from it, e.g.
  https://github.com/NixOS/nixpkgs/pull/219747 and https://github.com/NixOS/nixpkgs/pull/275548
- Using it can give easier-to-read code like `all isNull list`

Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
2023-12-20 17:55:19 +01:00
pennae
2b0e95e7aa use singleton expr to generate black hole errors
this also reduces forceValue code size and removes the need for
hideInDiagnostics. coopting thunk forcing like this has the additional
benefit of clarifying how these errors can happen in the first place.
2023-12-19 19:32:16 +01:00
pennae
74c134914c compare string values with strcmp
string_view()ification calls strlen() first, which we don't need here.
2023-12-19 19:32:16 +01:00
pennae
78353deb02 encode black holes as tApp values
checking for isBlackhole in the forceValue hot path is rather more
expensive than necessary, and with a little bit of trickery we can move
such handling into the isApp case. small performance benefit, but under
some circumstances we've seen 2% improvement as well.

〉 nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'

before:

  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

after:

  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
2023-12-19 19:32:16 +01:00
John Ericson
ed26b186fb Remove now-redundant text-hashing store methods
`addTextToStore` and `computeStorePathFromDump` are now redundant.

Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
2023-12-18 10:44:10 -05:00
John Ericson
dfc876531f Organize content addressing, use SourceAccessor with Store::addToStore
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
2023-12-18 10:41:54 -05:00
Matthew Bauer
bcbdb09ccf Add eval-system option
`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>
2023-12-14 19:04:00 -05:00
Robert Hensing
d4f6b1d38b
Merge pull request #9497 from edolstra/move-access-control
Move restricted/pure-eval access control out of the evaluator and into the accessor
2023-12-08 22:21:50 +01:00
Peter Kolloch
d38ec12855 Update src/libexpr/primops.cc
Co-authored-by: Théophane Hufschmitt <7226587+thufschmitt@users.noreply.github.com>
2023-12-06 23:43:42 +01:00
Peter Kolloch
8afeaf05c4 Add docs/rl-notes for nix hash convert / builtins.convertHash
https://github.com/NixOS/nix/issues/8876
2023-12-06 23:43:42 +01:00
Peter Kolloch
5334c9c792 HashType: Rename to HashAlgorithm
To be consistent with CLI, nix API
and many other references.

As part of this, we also converted it to a scoped enum.

https://github.com/NixOS/nix/issues/8876
2023-12-06 23:43:42 +01:00
Eelco Dolstra
83c067c0fa PosixSourceAccessor: Don't follow any symlinks
All path components must not be symlinks now (so the user needs to
call `resolveSymlinks()` when needed).
2023-12-05 23:02:59 +01:00
Valentin Gagarin
24b781773f fix random docs errors
remove link to the contributing guide from user documentation.
it doesn't help here, and the target at first glance shows redundant
information.
2023-12-02 03:02:59 +01:00
Robert Hensing
188c803ddb
Merge pull request #9508 from infinisil/revert-7762-boost-regex
Revert "Switch from std::regex to boost::regex"
2023-12-01 02:08:58 +01:00
Valentin Gagarin
4781e7fa70 Document each store type on its own page
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
2023-12-01 01:27:52 +01:00
Silvan Mosberger
908a011a4a
Revert "Switch from std::regex to boost::regex" 2023-12-01 00:50:20 +01:00
Eelco Dolstra
305939655a Remove superfluous use of hasAccessControl() 2023-11-30 16:28:33 +01:00
Eelco Dolstra
ea95327e72 Move restricted/pure-eval access control out of the evaluator and into the accessor 2023-11-30 16:16:17 +01:00
Eelco Dolstra
cb7f25869d
Merge pull request #9430 from hercules-ci/remove-vlas
Fix stack overflow in `filter`
2023-11-30 15:31:44 +01:00
Yorick van Pelt
d2f5e263e3
Switch from std::regex to boost::regex 2023-11-25 15:14:18 +01:00
Philipp Schuster
5be0e6b314 doc: primops: fix typo 2023-11-24 10:50:01 +01:00
Philipp Schuster
2ce8c9650b
doc: primops: add more info for foldl (#9254)
* doc: primops: add more info for foldl

From the existing doc it is not obvious whether the first or the
second argument is the accumulator. This is however relevant to
know, as for certain scenarios, this might change the behavior.

Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
2023-11-23 21:02:20 +00:00