Commit graph

1863 commits

Author SHA1 Message Date
Théophane Hufschmitt dc4730ee94
Merge pull request #5720 from tomberek/flake_search
flakes: search up to git or filesystem boundary
2022-01-18 10:59:12 +01:00
Eelco Dolstra fc2443a67c
Merge pull request #5812 from pennae/small-perf-improvements
improve parser performance a bit
2022-01-17 19:49:52 +01:00
Alexander Bantyev e3690ab393
Add more tests for flake upward searching 2022-01-14 17:21:18 +03:00
tomberek 2dead20924
Apply suggestions from code review
Co-authored-by: Théophane Hufschmitt <7226587+thufschmitt@users.noreply.github.com>
2022-01-14 09:16:34 -05:00
pennae 72f42093e7 optimize unescapeStr
mainly to avoid an allocation and a copy of a string that can be
modified in place (ever since EvalState holds on to the buffer, not the
generated parser itself).

 # before

Benchmark 1: nix search --offline nixpkgs hello
  Time (mean ± σ):     571.7 ms ±   2.4 ms    [User: 563.3 ms, System: 8.0 ms]
  Range (min … max):   566.7 ms … 579.7 ms    50 runs

Benchmark 2: nix eval -f ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
  Time (mean ± σ):     376.6 ms ±   1.0 ms    [User: 345.8 ms, System: 30.5 ms]
  Range (min … max):   374.5 ms … 379.1 ms    50 runs

Benchmark 3: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
  Time (mean ± σ):      2.922 s ±  0.006 s    [User: 2.707 s, System: 0.215 s]
  Range (min … max):    2.906 s …  2.934 s    50 runs

 # after

Benchmark 1: nix search --offline nixpkgs hello
  Time (mean ± σ):     570.4 ms ±   2.8 ms    [User: 561.3 ms, System: 8.6 ms]
  Range (min … max):   564.6 ms … 578.1 ms    50 runs

Benchmark 2: nix eval -f ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
  Time (mean ± σ):     375.4 ms ±   1.3 ms    [User: 343.2 ms, System: 31.7 ms]
  Range (min … max):   373.4 ms … 378.2 ms    50 runs

Benchmark 3: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
  Time (mean ± σ):      2.925 s ±  0.006 s    [User: 2.704 s, System: 0.219 s]
  Range (min … max):    2.910 s …  2.942 s    50 runs
2022-01-13 18:06:15 +01:00
pennae 34e3bd10e3 avoid copies of parser input data
when given a string yacc will copy the entire input to a newly allocated
location so that it can add a second terminating NUL byte. since the
parser is a very internal thing to EvalState we can ensure that having
two terminating NUL bytes is always possible without copying, and have
the parser itself merely check that the expected NULs are present.

 # before

Benchmark 1: nix search --offline nixpkgs hello
  Time (mean ± σ):     572.4 ms ±   2.3 ms    [User: 563.4 ms, System: 8.6 ms]
  Range (min … max):   566.9 ms … 579.1 ms    50 runs

Benchmark 2: nix eval -f ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
  Time (mean ± σ):     381.7 ms ±   1.0 ms    [User: 348.3 ms, System: 33.1 ms]
  Range (min … max):   380.2 ms … 387.7 ms    50 runs

Benchmark 3: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
  Time (mean ± σ):      2.936 s ±  0.005 s    [User: 2.715 s, System: 0.221 s]
  Range (min … max):    2.923 s …  2.946 s    50 runs

 # after

Benchmark 1: nix search --offline nixpkgs hello
  Time (mean ± σ):     571.7 ms ±   2.4 ms    [User: 563.3 ms, System: 8.0 ms]
  Range (min … max):   566.7 ms … 579.7 ms    50 runs

Benchmark 2: nix eval -f ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
  Time (mean ± σ):     376.6 ms ±   1.0 ms    [User: 345.8 ms, System: 30.5 ms]
  Range (min … max):   374.5 ms … 379.1 ms    50 runs

Benchmark 3: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
  Time (mean ± σ):      2.922 s ±  0.006 s    [User: 2.707 s, System: 0.215 s]
  Range (min … max):    2.906 s …  2.934 s    50 runs
2022-01-13 18:06:15 +01:00
pennae eee0bcee22 avoid allocations in SymbolTable::create
speeds up parsing by ~3%, system builds by a bit more than 1%

 # before

Benchmark 1: nix search --offline nixpkgs hello
  Time (mean ± σ):     574.7 ms ±   2.8 ms    [User: 566.3 ms, System: 8.0 ms]
  Range (min … max):   569.2 ms … 580.7 ms    50 runs

Benchmark 2: nix eval -f ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
  Time (mean ± σ):     394.4 ms ±   0.8 ms    [User: 361.8 ms, System: 32.3 ms]
  Range (min … max):   392.7 ms … 395.7 ms    50 runs

Benchmark 3: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
  Time (mean ± σ):      2.976 s ±  0.005 s    [User: 2.757 s, System: 0.218 s]
  Range (min … max):    2.966 s …  2.990 s    50 runs

 # after

Benchmark 1: nix search --offline nixpkgs hello
  Time (mean ± σ):     572.4 ms ±   2.3 ms    [User: 563.4 ms, System: 8.6 ms]
  Range (min … max):   566.9 ms … 579.1 ms    50 runs

Benchmark 2: nix eval -f ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
  Time (mean ± σ):     381.7 ms ±   1.0 ms    [User: 348.3 ms, System: 33.1 ms]
  Range (min … max):   380.2 ms … 387.7 ms    50 runs

Benchmark 3: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
  Time (mean ± σ):      2.936 s ±  0.005 s    [User: 2.715 s, System: 0.221 s]
  Range (min … max):    2.923 s …  2.946 s    50 runs
2022-01-13 18:06:15 +01:00
pennae 61a9d16d5c don't strdup tokens in the lexer
every stringy token the lexer returns is turned into a Symbol and not
used further, so we don't have to strdup. using a string_view is
sufficient, but due to limitations of the current parser we have to use
a POD type that holds the same information.

gives ~2% on system build, 6% on search, 8% on parsing alone

 # before

Benchmark 1: nix search --offline nixpkgs hello
  Time (mean ± σ):     610.6 ms ±   2.4 ms    [User: 602.5 ms, System: 7.8 ms]
  Range (min … max):   606.6 ms … 617.3 ms    50 runs

Benchmark 2: nix eval -f hackage-packages.nix
  Time (mean ± σ):     430.1 ms ±   1.4 ms    [User: 393.1 ms, System: 36.7 ms]
  Range (min … max):   428.2 ms … 434.2 ms    50 runs

Benchmark 3: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
  Time (mean ± σ):      3.032 s ±  0.005 s    [User: 2.808 s, System: 0.223 s]
  Range (min … max):    3.023 s …  3.041 s    50 runs

 # after

Benchmark 1: nix search --offline nixpkgs hello
  Time (mean ± σ):     574.7 ms ±   2.8 ms    [User: 566.3 ms, System: 8.0 ms]
  Range (min … max):   569.2 ms … 580.7 ms    50 runs

Benchmark 2: nix eval -f hackage-packages.nix
  Time (mean ± σ):     394.4 ms ±   0.8 ms    [User: 361.8 ms, System: 32.3 ms]
  Range (min … max):   392.7 ms … 395.7 ms    50 runs

Benchmark 3: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
  Time (mean ± σ):      2.976 s ±  0.005 s    [User: 2.757 s, System: 0.218 s]
  Range (min … max):    2.966 s …  2.990 s    50 runs
2022-01-13 18:06:14 +01:00
Théophane Hufschmitt e61c4bc25a
Merge pull request #5887 from pennae/avoid-streams
avoid std::?stream overhead when it's not helpful
2022-01-12 10:52:40 +01:00
pennae 5838354d34 optimize ExprConcatStrings::eval
constructing an ostringstream for non-string concats (like integer addition) is
a small constant cost that we can avoid. for string concats we can keep all the
string temporaries we get from coerceToString and concatenate them in one go,
which saves a lot of intermediate temporaries and copies in ostringstream. we
can also avoid copying the concatenated string again by directly allocating it
in GC memory and moving ownership of the concatenated string into the target
value.

saves about 2% on system eval.

before:

  Benchmark 1: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
    Time (mean ± σ):      2.837 s ±  0.031 s    [User: 2.562 s, System: 0.191 s]
    Range (min … max):    2.796 s …  2.892 s    20 runs

after:

  Benchmark 1: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
    Time (mean ± σ):      2.790 s ±  0.035 s    [User: 2.532 s, System: 0.187 s]
    Range (min … max):    2.722 s …  2.836 s    20 runs
2022-01-12 10:07:21 +01:00
Eelco Dolstra f39d94a55b Merge branch 'readfile' of https://github.com/abbradar/nix 2022-01-10 16:55:28 +01:00
Nikolay Amiantov c66865dff1 builtins.readFile: Propagate path context
Co-authored-by: Shea Levy <shea@shealevy.com>
2022-01-09 13:07:00 +03:00
Eelco Dolstra 3ec02c6743
Merge pull request #5546 from NixOS/allow-paths-in-flake-local-options
allow paths in flake local options
2022-01-06 21:55:20 +01:00
Eelco Dolstra 2b4c944823 Remove EvalState::mkAttrs() 2022-01-04 20:29:17 +01:00
Eelco Dolstra 17daec0b83 Move empty attrset optimisation 2022-01-04 19:23:11 +01:00
Eelco Dolstra ca5baf2392 Turn mkString(Symbol) into a method 2022-01-04 19:09:40 +01:00
Eelco Dolstra ed93aec3c3 Remove non-method mkPath() 2022-01-04 18:45:16 +01:00
Eelco Dolstra 263a8d293c Remove non-method mk<X> functions 2022-01-04 18:40:39 +01:00
Eelco Dolstra cc08364315 Remove non-method mkString() 2022-01-04 18:24:42 +01:00
Eelco Dolstra 6d9a6d2cc3 Ensure that attrsets are sorted
Previously you had to remember to call value->attrs->sort() after
populating value->attrs. Now there is a BindingsBuilder helper that
wraps Bindings and ensures that sort() is called before you can use
it.
2022-01-04 18:00:33 +01:00
pennae 00c993f48b add zipAttrsWith primop
nixpkgs can save a good bit of eval memory with this primop. zipAttrsWith is
used quite a bit around nixpkgs (eg in the form of recursiveUpdate), but the
most costly application for this primop is in the module system. it improves
the implementation of zipAttrsWith from nixpkgs by not checking an attribute
multiple times if it occurs more than once in the input list, allocates less
values and set elements, and just avoids many a temporary object in general.

nixpkgs has a more generic version of this operation, zipAttrsWithNames, but
this version is only used once so isn't suitable for being the base of a new
primop. if it were to be used more we should add a second primop instead.
2022-01-03 21:05:53 +01:00
Eelco Dolstra 96d08fcd66
Merge pull request #5839 from tweag/balsoft/yet-another-follows-bugfix
flake.cc: computeLocks: Only verify overrides when they could change
2022-01-03 20:51:58 +01:00
Sebastian Ullrich d0c8e9254e Fix IFD with chroot store 2021-12-29 19:00:02 +01:00
Lily Foster 92c7d33ee3 Sort attrs from tables in fromTOML
This was dropped in 10a8b5d for the migration from cpptoml to toml11 but
seems to be necessary for the attrsets to work correctly.

Fixes #5833
2021-12-28 10:11:00 -05:00
Alexander Bantyev 2664a216e5
flake.cc: computeLocks: Only verify overrides when they could change
When we check for disappeared overrides, we can get "false positives"
for follows and overrides which are defined in the dependencies of the
flake we are locking, since they are not parsed by
parseFlakeInputs. However, at that point we already know that the
overrides couldn't have possible been changed if the input itself
hasn't changed (since we check that oldLock->originalRef == *input.ref
for the input's parent). So, to prevent this, only perform this check
when it was possible that the flake changed (e.g. the flake we're
locking, or a new input, or the input has changed and mustRefetch ==
true).
2021-12-28 15:30:06 +03:00
Yorick a4ab0a74d9
Fix accidental O(n^2 * log n) performance in NixRepl::addAttrsToScope
Only sort once, after adding all of the attrs first. This reduces my
`nix repl '<nixpkgs>'` loading time from 1.07s to 103ms.

Fixes #5823
2021-12-27 13:18:55 +01:00
tomberek f404189368
Merge branch 'master' into flake_search 2021-12-23 15:22:52 -05:00
Nicolas Mattia b16643b6fc Use int64_t and NixFloat in fromTOML types
This makes sure that values parsed from TOML have a proper size. Using
e.g. `double` caused issues on i686 where the size of `double` (32bit)
was too small to accommodate some values.
2021-12-23 14:12:49 +01:00
regnat dc89dfa7b3 Properly return false on builtins.pathExists /someNonAllowedPath
Follow-up from https://github.com/NixOS/nix/pull/5807 to fix https://github.com/NixOS/nix/pull/5807#issuecomment-1000135394
2021-12-23 10:49:33 +01:00
Eelco Dolstra 89faff93e2
Merge pull request #5807 from NixOS/5805-ca-ifd
Fix IFD with CA derivations
2021-12-21 18:47:34 +01:00
Eelco Dolstra 25cea2b737 Merge branch 'nm-toml11' of https://github.com/nmattia/nix 2021-12-21 14:18:04 +01:00
regnat d90f9d4b99 Fix IFD with CA derivations
Rewrite the string taken by the IFD-like primops to contain the actual
output paths of the derivations rather than the placeholders

Fix #5805
2021-12-21 09:36:50 +01:00
regnat cbbd21ec07 Factor out the path realisation bit of IFD 2021-12-21 09:36:19 +01:00
pennae 09b245690a bulk-allocate Value instances in the evaluator
calling GC_malloc for each value is significantly more expensive than
allocating a bunch of values at once with GC_malloc_many. "a bunch" here
is a GC block size, ie 16KiB or less.

this gives a 1.5% performance boost when evaluating our nixos system.

tested with

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

 # on master

  Time (mean ± σ):      3.335 s ±  0.007 s    [User: 2.774 s, System: 0.293 s]
  Range (min … max):    3.315 s …  3.347 s    50 runs

 # with this change

  Time (mean ± σ):      3.288 s ±  0.006 s    [User: 2.728 s, System: 0.292 s]
  Range (min … max):    3.274 s …  3.307 s    50 runs
2021-12-20 23:01:28 +01:00
Nicolas Mattia 9dea5768ef Clean up toml parsing code 2021-12-20 21:39:37 +01:00
Nicolas Mattia 5d06836b9f Use long in fromTOML 2021-12-17 23:15:26 +01:00
Nicolas Mattia 10a8b5d3ae Update primops/fromTOML.cc to use toml11 2021-12-17 22:13:04 +01:00
regnat ab902521b1 Allow paths in flake local settings
Fix #5505
2021-12-17 09:31:13 +01:00
Théophane Hufschmitt 8868da45a8
Merge pull request #5623 from yorickvP/fix-5621
flakes: fix boolean and int nixConfig values
2021-12-14 10:35:37 +01:00
Eelco Dolstra 32a62b0d25 Re-allow inputs.x.url = "/path/to/file" 2021-12-13 20:59:33 +01:00
Eelco Dolstra cc6406cc59 Merge branch 'better-interpolation-error-location' of https://github.com/greedy/nix 2021-12-13 19:20:48 +01:00
Tom Bereknyei b6cc0a704d flakes: search up to git or filesystem boundary
While parsing a flakeref, upon not finding a flake.nix, search upwards
until git or filesystem boundary.
2021-12-03 10:55:30 -05:00
Silvan Mosberger 90700736c7 Introduce builtins.groupBy primop
This function is very useful in nixpkgs, but its implementation in Nix
itself is rather slow due to it requiring a lot of attribute set and
list appends.
2021-12-02 21:54:51 +01:00
Alexander Bantyev 8cbf862e6f
Flakes: computeLocks: pass correct LockParent when reusing oldLock
Previously, when we were attempting to reuse the old lockfile
information in the computeLocks function, we have passed the parent of
the current input to the next computeLocks call. This was incorrect,
since the follows are resolved relative to the parent. This caused
issues when we tried to reuse oldLock but couldn't for some
reason (read: mustRefetch is true), in that case the follows were
resolved incorrectly.

Fix this by passing the correct parent, and adding some tests to
prevent this particular regression from happening again.

Closes https://github.com/NixOS/nix/issues/5697
2021-12-02 12:02:34 +03:00
Yorick bd628cf3da
flakes: fix boolean and int nixConfig values
Some type confusion was causing ints to be pointers, and bools
to be ints. Fixes #5621
2021-11-29 15:53:22 +01:00
Andreas Rammhold 90d8178009
Don't move the arguments of the primOp
Moving arguments of the primOp into the registration structure makes it
impossible to initialize a second EvalState with the correct primOp
registration. It will end up registering all those "RegisterPrimOp"'s
with an arity of zero on all but the 2nd instance of the EvalState.

Not moving the memory will add a tiny bit of memory overhead during the
eval since we need a copy of all the argument lists of all the primOp's.
The overhead shouldn't be too bad as it is static (based on the amonut
of registered operations) and only occurs once during the interpreter
startup.
2021-11-28 02:06:47 +01:00
Eelco Dolstra 6f46434f32
Merge pull request #5648 from edolstra/list-iter
Support range-based for loop over list values
2021-11-25 17:13:25 +01:00
Eelco Dolstra b6c8e57056 Support range-based for loop over list values 2021-11-25 16:31:39 +01:00
regnat c47027f3a1 Fix the error when accessing a forbidden path in pure eval
If we’re in pure eval mode, then tell that in the error message rather
than (wrongly) speaking about restricted mode.

Fix https://github.com/NixOS/nix/issues/5611
2021-11-25 14:48:01 +01:00
Eelco Dolstra d58f149140
Merge pull request #5631 from Infinisil/list-compare
Make lists be comparable
2021-11-24 15:48:05 +01:00