Commit graph

2101 commits

Author SHA1 Message Date
Ben Burdette 2acdb90438 remove debug code 2022-05-12 14:20:45 -06:00
Ben Burdette 2d0d1ec99d remove debug code 2022-05-12 14:15:35 -06:00
Ben Burdette 4f48095c66 Merge branch 'debugThrow' into debug-exploratory-PR 2022-05-12 14:11:35 -06:00
Ben Burdette 1ea13084c9 template-ize debugThrow 2022-05-12 13:59:58 -06:00
Ben Burdette 7cd7c7c91a
Merge branch 'master' into debug-exploratory-PR 2022-05-09 09:30:44 -06:00
Jan Tojnar 59d9551c25
libexpr: Fix manual link in error message
It was changed to the old manual in 8895fa70a4
2022-05-08 18:59:00 +02:00
Andreas Rammhold 059ae7f6c4
Add unit tests for libexpr (#5377)
* libexpr: fix builtins.split example

The example was previously indicating that multiple whitespaces would be
collapsed into a single captured whitespace. That isn't true and was
likely a mistake when being documented initially.

* Fix segfault on unitilized list when printing value

Since lists are just chunks of memory the individual elements in the
list might be unitilized when a programming error happens within Nix.

In this case the values are null-initialized (at least with Boehm GC)
and we can avoid a nullptr deref when printing them.

I ran into this issue while ensuring that new expression tests would
show the actual value on an assertion failure.

This is unlikely to cause any runtime performance regressions as
printing values is not really in the hot path (unless the repl is the
primary use case).

* Add operator<< for ValueTypes

* Add libexpr tests

This introduces tests for libexpr that evalulate various trivial Nix
language expressions and primop invocations that should be good smoke
tests wheter or not the implementation is behaving as expected.
2022-05-06 18:05:27 +02:00
Ben Burdette fc66f48812 debugError() 2022-05-06 09:09:49 -06:00
Ben Burdette 2c9fafdc9e trying debugThrow 2022-05-06 08:47:21 -06:00
Ben Burdette dea998b2f2 traceable_allocator 2022-05-05 20:26:10 -06:00
Ben Burdette f400c5466d rename valmap 2022-05-05 15:43:23 -06:00
Ben Burdette 09fcfee925 don't print the 'break' argument 2022-05-05 15:34:59 -06:00
Ben Burdette 0ac121a094
Merge branch 'NixOS:master' into debug-exploratory-PR 2022-05-05 13:54:20 -06:00
Eelco Dolstra 58645a78ab builtins.break: Return argument when debugging is not enabled 2022-05-05 17:17:03 +02:00
Eelco Dolstra dd8b91eebc Style fixes
In particular, use std::make_shared and enumerate(). Also renamed some
fields to fit naming conventions.
2022-05-05 17:17:03 +02:00
Alexander Shpilkin b3ed32d0fd Add forgotten null check 2022-05-04 22:13:49 +03:00
Eelco Dolstra c98648bef0 Merge remote-tracking branch 'origin/master' into debug-exploratory-PR 2022-05-04 14:10:21 +02:00
Alain Zscheile 1385b20078
Get rid of most .at calls (#6393)
Use one of `get` or `getOr` instead which will either return a null-pointer (with a nicer error message) or a default value when the key is missing.
2022-05-04 07:44:32 +02:00
Eelco Dolstra 4a79cba511 Allow selecting derivation outputs using 'installable!outputs'
E.g. 'nixpkgs#glibc^dev,static' or 'nixpkgs#glibc^*'.
2022-05-03 13:43:52 +02:00
Ben Burdette 172a83d22a line endings 2022-04-29 11:24:54 -06:00
Ben Burdette ca6cba8b81 fix 'suggestions' error 2022-04-29 10:51:10 -06:00
Ben Burdette 2a5632c70d incorporate PosIdx changes, symbol changes. 2022-04-29 10:02:17 -06:00
Ben Burdette 6e19947993 Merge branch 'master' into debug-merge-master 2022-04-28 12:32:57 -06:00
Eelco Dolstra 717298c749 Bump eval cache schema version 2022-04-26 17:17:51 +02:00
Eelco Dolstra 1ddabe1a01 nix: Respect meta.outputsToInstall, and use all outputs by default
'nix profile install' will now install all outputs listed in the
package's meta.outputsToInstall attribute, or all outputs if that
attribute doesn't exist. This makes it behave consistently with
nix-env. Fixes #6385.

Furthermore, for consistency, all other 'nix' commands do this as
well. E.g. 'nix build' will build and symlink the outputs in
meta.outputsToInstall, defaulting to all outputs. Previously, it only
built/symlinked the first output. Note that this means that selecting
a specific output using attrpath selection (e.g. 'nix build
nixpkgs#libxml2.dev') no longer works. A subsequent PR will add a way
to specify the desired outputs explicitly.
2022-04-26 17:17:27 +02:00
Eelco Dolstra b12c33510c EvalCache AttrKey: Use Symbol instead of std::string 2022-04-26 14:16:20 +02:00
Eelco Dolstra 474695975d EvalCache: Revert to using symbols in getAttr() 2022-04-26 14:01:42 +02:00
Eelco Dolstra fab731a9d4 Don't pass Symbol by reference
Since Symbol is just an integer, passing it by const reference is
never advantageous.
2022-04-26 13:25:17 +02:00
pennae a385e51a08 rename SymbolIdx -> Symbol, Symbol -> SymbolStr
after #6218 `Symbol` no longer confers a uniqueness invariant on the
string it wraps, it is now possible to create multiple symbols that
compare equal but whose string contents have different addresses. this
guarantee is now only provided by `SymbolIdx`, leaving `Symbol` only as
a string wrapper that knows about the intricacies of how symbols need to
be formatted for output.

this change renames `SymbolIdx` to `Symbol` to restore the previous
semantics of `Symbol` to that name. we also keep the wrapper type and
rename it to `SymbolStr` instead of returning plain strings from lookups
into the symbol table because symbols are formatted for output in many
places. theoretically we do not need `SymbolStr`, only a function that
formats a string for output as a symbol, but having to wrap every symbol
that appears in a message into eg `formatSymbol()` is error-prone and
inconvient.
2022-04-25 15:37:01 +02:00
Théophane Hufschmitt c4ffc8e2f8
Merge pull request #6218 from pennae/pos-symbol-tables
reduce the size of Attr from 3 pointers to 2 on 64 bit machines
2022-04-22 10:28:06 +02:00
Théophane Hufschmitt 7ca6fbc8ca Move ChunkedVector to its own header 2022-04-22 10:01:02 +02:00
Tom Bereknyei f25112d383 fix: builtins.toFile adds path to allowedPaths
The produced path is then allowed be imported or utilized elsewhere:
```
assert (43 == import (builtins.toFile "source" "43")); "good"
```

This will still fail on write-only stores.
2022-04-21 16:41:37 -04:00
pennae 8168a4cf4a shrink Attr by 8 bytes on 64bit machines
with position and symbol tables in place we can now shrink Attr by a full
pointer with some simple field reordering. since Attr is a very hot struct this
has substantial impact on memory use, decreasing GC allocations and heap size by
10-15% each. we also get a ~15% performance improvement due to reduced GC
loading.

pure parsing has taken a hit over the branch base because positions are now
slightly more expensive to create, but overall we get a noticeable improvement.

before (on memory-friendliness):

  Benchmark 1: nix search --no-eval-cache --offline ../nixpkgs hello
    Time (mean ± σ):      6.960 s ±  0.028 s    [User: 5.832 s, System: 0.897 s]
    Range (min … max):    6.886 s …  7.005 s    20 runs

  Benchmark 2: nix eval -f ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
    Time (mean ± σ):     328.1 ms ±   1.7 ms    [User: 295.8 ms, System: 32.2 ms]
    Range (min … max):   324.9 ms … 331.2 ms    20 runs

  Benchmark 3: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
    Time (mean ± σ):      2.688 s ±  0.029 s    [User: 2.365 s, System: 0.238 s]
    Range (min … max):    2.642 s …  2.742 s    20 runs

after:

  Benchmark 1: nix search --no-eval-cache --offline ../nixpkgs hello
    Time (mean ± σ):      6.902 s ±  0.039 s    [User: 5.844 s, System: 0.783 s]
    Range (min … max):    6.820 s …  6.956 s    20 runs

  Benchmark 2: nix eval -f ../nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix
    Time (mean ± σ):     330.7 ms ±   2.2 ms    [User: 300.6 ms, System: 30.0 ms]
    Range (min … max):   327.5 ms … 334.5 ms    20 runs

  Benchmark 3: nix eval --raw --impure --expr 'with import <nixpkgs/nixos> {}; system'
    Time (mean ± σ):      2.330 s ±  0.027 s    [User: 2.040 s, System: 0.234 s]
    Range (min … max):    2.272 s …  2.383 s    20 runs
2022-04-21 21:56:34 +02:00
pennae 8775be3393 store Symbols in a table as well, like positions
this slightly increases the amount of memory used for any given symbol, but this
increase is more than made up for if the symbol is referenced more than once in
the EvalState that holds it. on average every symbol should be referenced at
least twice (once to introduce a binding, once to use it), so we expect no
increase in memory on average.

symbol tables are limited to 2³² entries like position tables, and similar
arguments apply to why overflow is not likely: 2³² symbols would require as many
string instances (at 24 bytes each) and map entries (at 24 bytes or more each,
assuming that the map holds on average at most one item per bucket as the docs
say). a full symbol table would require at least 192GB of memory just for
symbols, which is well out of reach. (an ofborg eval of nixpks today creates
less than a million symbols!)
2022-04-21 21:56:31 +02:00
pennae 00a3280232 don't use Symbol in Pos to represent a path
PosTable deduplicates origin information, so using symbols for paths is no
longer necessary. moving away from path Symbols also reduces the usage of
symbols for things that are not keys in attribute sets, which will become
important in the future when we turn symbols into indices as well.
2022-04-21 21:46:10 +02:00
pennae 6526d1676b replace most Pos objects/ptrs with indexes into a position table
Pos objects are somewhat wasteful as they duplicate the origin file name and
input type for each object. on files that produce more than one Pos when parsed
this a sizeable waste of memory (one pointer per Pos). the same goes for
ptr<Pos> on 64 bit machines: parsing enough source to require 8 bytes to locate
a position would need at least 8GB of input and 64GB of expression memory. it's
not likely that we'll hit that any time soon, so we can use a uint32_t index to
locate positions instead.
2022-04-21 21:46:06 +02:00
pennae 34b72775cf make throw*Error member functions of EvalState
when we introduce position and symbol tables we'll need to do lookups to turn
indices into those tables into actual positions/symbols. having the error
functions as members of EvalState will avoid a lot of churn for adding lookups
into the tables for each caller.
2022-04-21 21:25:18 +02:00
pennae 39df15fb8e don't use full Pos for findPackageFilename/editorFor
only file and line of the returned position were ever used, it wasn't actually
used a position. as such we may as well use a path+int pair for only those two
values and remove a use of Pos that would not work well with a position table.
2022-04-21 21:25:18 +02:00
pennae 38de79fcf7 remove Bindings::need
a future commit will remove the ability to convert the symbol type used in
bindings to strings. since we only have two users we can inline the error check.
2022-04-21 21:25:18 +02:00
pennae ff0fd91ed2 remove Symbol::empty
the only use of this function is to determine whether a lambda has a non-set
formal, but this use is arguably better served by Symbol::set and using a
non-Symbol instead of an empty symbol in the parser when no such formal is present.
2022-04-21 21:25:18 +02:00
pennae 90b5c0a1a6 turn primop names into strings
we don't *need* symbols here. the only advantage they have over strings is
making call-counting slightly faster, but that's a diagnostic feature and thus
needn't be optimized.

this also fixes a move bug that previously didn't show up: PrimOp structs were
accessed after being moved from, which technically invalidates them. previously
the names remained valid because Symbol copies on move, but strings are
invalidated. we now copy the entire primop struct instead of moving since primop
registration happen once and are not performance-sensitive.
2022-04-21 21:25:17 +02:00
Eelco Dolstra 51712bf012
Merge pull request #6128 from ncfavier/fix-completion
Shell completion improvements
2022-04-19 13:45:33 +02:00
Eelco Dolstra d89840b103 Make InstallableFlake::toValue() and toDerivation() behave consistently
In particular, this means that 'nix eval` (which uses toValue()) no
longer auto-calls functions or functors (because
AttrCursor::findAlongAttrPath() doesn't).

Fixes #6152.

Also use ref<> in a few places, and don't return attrpaths from
getCursor() because cursors already have a getAttrPath() method.
2022-04-14 14:07:04 +02:00
Ben Burdette 8b197c492e remove comma 2022-04-09 21:54:41 -06:00
Ben Burdette a61841ac41 don't use std::map merge 2022-04-09 07:45:23 -06:00
Ben Burdette 27d45f9eb3 minor cleanup 2022-04-08 15:46:12 -06:00
Ben Burdette b8b8ec7101 move throw to preverve Error type; turn off debugger for tryEval 2022-04-08 12:34:27 -06:00
Ben Burdette 1a93ac8133 Merge remote-tracking branch 'upstream/master' into upstream-merge 2022-04-07 13:42:01 -06:00
Ben Burdette 50b52d5110 remove debug code 2022-04-07 12:03:18 -06:00
Eelco Dolstra 589f6f267b fetchClosure: Don't allow URL query parameters
Allowing this is a potential security hole, since it allows the user
to specify parameters like 'local-nar-cache'.
2022-04-06 11:52:51 +02:00