| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |\
| |
| | |
Paulper stack reduction
|
| | |
| |
| |
| |
| | |
One fewer Int to be whizzing around the parser hoping to
be confused with other Ints.
|
| | |
| |
| |
| |
| | |
- anonymize unused pattern binder
- avoid negations
|
| |/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- ThisType(some.package) need not be erased, we can let
that prefix through unchanged and avoid churning through
allocations.
- Sharpen the condition in `rebindInnerClass`. The answer
to the deleted comment in the code, is "anonymous and
local classes", which are specifically excluded in the
new formulation.
- Finally, detect if erasure of the TypeRef is an identity and reuse
the original.
Waste not, want not.
To expand on the first point, here is what used to happen during
erasure:
scala> val scalaPack = typeOf[Predef.type].prefix
scalaPack: $r.intp.global.Type = scala.type
scala> typeDeconstruct.show(scalaPack)
res19: String = ThisType(package scala)
scala> typeDeconstruct.show(erasure.scalaErasure(scalaPack))
res20: String = TypeRef(TypeSymbol(final class scala extends ))
Showing one step of the erasure type map:
scala> typeDeconstruct.show(scalaPack.asInstanceOf[SubType].underlying.typeOfThis)
res21: String = SingleType(pre = ThisType(package <root>), package scala)
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It we can only safely use vals in Definitions for top-level symbols.
Otherwise, when the IDE switches to loading the symbol from source,
we can hold on to a stale symbol, which in turn impedes implicit
materialization of TypeTags.
This commit moves (most) of the accessors for member symbols
into RunDefinitions, and changes calling code accordingly.
This is a win for presentation compiler correctness, and
might even shave of a few cycles.
In a few cases, I have had to leave a `def` to a member symbol
in Definitions so we can get to it from the SymbolTable cake,
which doesn't see RunDefinitions.
The macro FastTrack facility now correctly recreates the mapping
from Symbol to macro implementation each run, using a new facility
in perRunCaches to create a run-indexed cache.
The enclosed test recreates the situation reported in the ticket,
in which TypeTags.scala is loaded from source.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We are all used to calls to `definitions.PredefModule`, or
`defintions.Predef_???` to grab the symbol of some well known
entity. But you'll notice that some of these are lazy vals,
and others are defs.
Why is this so? In the presentation compiler, a member like
`Predef.???` will be assigned a new symbol after the user
browses into `Predef.scala`.
Mistakenly using vals in definitions leads to subtle
IDE bugs like SI-7678. We are able to trigger these
situations in tests, as noted in the comments of that issue.
Changing the vals to defs, on the other hand, has a performance
penalty. Some schemes to workaround this have shown up:
cache them per-implicit search, or compare method names and owners
rather than symbols on hot paths in the type checker.
This commit introduces a facility to cache these sort of symbols
per-run, and uses it to check for `Predef.conforms` and
and for the class/type tag materializers.
A followup pull request (WIP: https://github.com/retronym/scala/compare/ticket/7678-2)
will expand the use of to address the widespread and unsafe caching
of member symbols that I found while investigating SI-7678.
|
| |\
| |
| | |
Delay delambdafication and put the lambda's body into the containing class
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This commit puts a real body on the Delambdafy phase.
From a lambda, Delambdafy will create
1) a static forwarder at the top level of the class that contained
the lambda
2) a new top level class that
a) has fields and a constructor taking the captured environment
(including possbily the "this" reference)
b) an apply method that calls the static forwarder
c) if needed a bridge method for the apply method
3) an instantiation of the newly created class which replaces the
lambda
Trees.scala is modified to add two more convenient factories
for templates and classdefs.
A few basic tests are included to verify that it works as expected.
Further commits will have additional tests.
|
| | |
| |
| |
| |
| |
| |
| | |
This commit is purely a refactor. It pulls code needed to adapt a tree
of one type into a tree of another type (by casting, boxing, coercing,
etc) out of Erasure and into common locations that will be usable
from the Delambdafy phase.
|
| |\ \
| |/
|/| |
Microoptimization in implicit search
|
| | |
| |
| |
| | |
Avoid creating a throwaway list of parameter types.
|
| |/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As Paul noted in the comments to SI-6260 (from which I mined some
test cases) "there is no possible basis for conflict here":
scala> class C[A](val a: Any) extends AnyVal
defined class C
scala> class B { def x[A](ca: C[A]) = () }
defined class B
scala> class D extends B { override def x[A](ca: C[A]) = () }
<console>:8: error: bridge generated for member method x: [A](ca: C[A])Unit in class D
which overrides method x: [A](ca: C[A])Unit in class B
clashes with definition of the member itself;
both have erased type (ca: Object)Unit
class D extends B { override def x[A](ca: C[A]) = () }
^
What was happening?
Bridge computation compares `B#x` and `D#x` exitingErasure, which
results in comparing:
ErasedValueType(C[A(in B#x)]) =:= ErasedValueType(C[A(in D#x)])
These types were considered distinct (on the grounds of the unique
type hash consing), even though they have the same erasure and
involve the same value class.
That triggered creation of an bridge. After post-erasure eliminates
the `ErasedValuedType`s, we find that this marvel of enginineering is
bridges `(Object)Unit` right back onto itself. The previous
resolution of SI-6385 (d435f72e5fb7fe) was a test case that confirmed
that we detected the zero-length bridge and reported it nicely, which
happened after related work in SI-6260. But we can simply avoid
creating in it in the first place.
That's what this commit does. It does so by reducing the amount
of information carried in `ErasedValueType` to the bare minimum needed
during the erasure -> posterasure transition.
We need to know:
1. which value class wraps the value, so we can box and unbox
as needed
2. the erasure of the underlying value, which will replace this
type in post-erasure.
This construction means that the bridge above computation now
compares:
ErasedValueType(C, Any) =:= ErasedValueType(C, Any])
I have included a test to show that:
- we don't incur any linkage or other runtime errors in the
reported case (run/t6385.scala)
- a similar case compiles when the signatures align
(pos/t6260a.scala), but does *not* compile when the just
erasures align (neg/t6260c.scala)
- polymorphic value classes continue to erase to the instantiated
type of the unbox: (run/t6260b.scala)
- other cases in SI-6260 remains unsolved and indeed unsolvable
without an overhaul of value classes: (neg/t6260b.scala)
In my travels I spotted a bug in corner case of null, asInstanceOf
and value classes, which I have described in a pending test.
|
| |\
| |
| | |
Favour module accessors symbols in rebind
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The Refchecks tree transformer transforms a nested modules that
overrides a method into a pair of symbols: the module itself, and
an module accessor that matches the overridden symbol.
[[syntax trees at end of typer]] // test1.scala
package <empty> {
abstract trait C2 extends scala.AnyRef {
def O1: Any
};
class C1 extends AnyRef with C2 {
object O1 extends scala.AnyRef
}
}
[[syntax trees at end of refchecks]] // test1.scala
package <empty> {
abstract trait C2 extends scala.AnyRef {
def O1: Any
};
class C1 extends AnyRef with C2 {
object O1 extends scala.AnyRef
@volatile <synthetic> private[this] var O1$module: C1.this.O1.type = _;
<stable> def O1: C1.this.O1.type = {
C1.this.O1$module = new C1.this.O1.type();
C1.this.O1$module
}
}
}
When constructing a TypeRef or SingleType with a prefix and and a symbol,
the factory methods internally use `rebind` to see if the provided symbol
should be replaced with an overriding symbol that is available in that prefix.
Trying this out in the REPL is a bit misleading, because even if you change
phase to `refchecks`, you won't get the desired results because the transform
is not done in an InfoTransformer.
scala> val O1 = typeOf[C1].decl(TermName("O1"))
O1: $r.intp.global.Symbol = object O1
scala> typeRef(typeOf[C2], O1, Nil)
res13: $r.intp.global.Type = C2#O1
scala> res13.asInstanceOf[TypeRef].sym.owner
res14: $r.intp.global.Symbol = class C1
But debugging the test case, we get into `rebind` during an AsSeenFrom
which is where we crashed when `suchThat` encountered the overloaded
module and module accessor symbols:
typeOf[OuterObject.Inner.type].memberType(symbolOf[InnerTrait.Collection])
...
singleTypeAsSeen(OuterTrait.this.Inner.type)
val SingleType(pre, sym) = tp
// pre = OuterTrait.this.type
// sym = OuterTrait.Inner
val pre1 = this(pre) // OuterObject.type
singleType(pre1, sym)
rebind(pre1, sym) // was crashing, now OuterObject.Inner
}
This commit excludes the module symbol from symbol lookup in the prefix in rebind.
|
| |\ \
| | |
| | | |
SI-6840 fixes weird typing of quasiquote arguments
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Previously quasiquote arguments were type checked against Any
which caused weird inference that made splicing of complex expressions
unusable:
val l1 = List(q"foo")
val l2 = List(q"bar")
q"f(..${l1 ++ l2})" // argument type checked as Any instead of List[Tree]
This is fixed by forcing compiler to type check against type
variable which itself isn't used in any other way.
|
| |\ \ \
| |_|/
|/| | |
fixes handling of fancy nested classes in runtime reflection
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Replaces the `jclazz.isMemberClass` check for whether we have an
inner/nested class with `jclazz.getEnclosingClass != null`, because there
exist classes produced by javac (see the attached jar file and the test log)
which have the following properties:
* They are nested within a parent class
* getEnclosingClass returns a non-null value
* isMemberClass returns false
Previously such classes were incorrectly treated as non-nested, were
incorrectly put into an enclosing package rather than an enclosing class,
and had their names trimmed in the process, leading to situations when
a package has multiple declarations with the same name. This is now fixed.
When changing the check, we need to be careful with interpretation of
what Class.getEnclosingXXX methods return. If getEnclosingClass produces
a non-null result, this doesn't mean that the class is inner or nested,
because getEnclosingClass is also not null for local classes (the ones
with getEnclosingMethod != null || getEnclosingConstructor != null).
This is expressed in the order of pattern match clauses in `sOwner`.
Now when the bug is fixed, I also revert b18a2f8798b2, restoring a very
important integrity check in runtime reflection, which I had to disable
a couple hours ago to fix a master breakage. More details at scala-internals:
https://groups.google.com/forum/#!topic/scala-internals/hcnUFk75MgQ
|
| |\ \ \
| |_|/
|/| | |
deprecates raw tree manipulation facilities in macros.Context
|
| | | | |
|
| |\ \ \
| |_|/
|/| | |
Add support for packages into quasiquotes and toolbox, improve handling of fresh names, unhardcode quasiquote expansion logic
|
| | | |
| | |
| | |
| | |
| | | |
This should ensure that concurrent access to the
fresh name creator is properly synchronized.
|
| | | | |
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This commit extracts out freshTermName and freshTypeName to the top-level
with implicit fresh name creator argument. This will let to refactor out
more methods out of tree builder into treegen that are dependent on fresh
name generator. We also save quite a bit of boilerplate by not having to
redefined fresh functions all over the place.
|
| | | | |
|
| | | |
| | |
| | |
| | | |
For sake of consistency with noSelfType.
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
During parsing some names are generated artificially using freshTermName & freshTypeName (e.g. `x$1`). Such names should be reified in a different way because they are assumed to be always fresh and non-overlapping with the environment. So `x$1` should reify down to equivalent of `freshTermName("x$")` rather than `TermName("x$1")`.
But this is not enough. One name can be used more than once in a tree. E.g. `q"_ + 1"` desugars into `q"x$1 => x$1 + 1"`. So we need to ensure that every place where `x$1` is used gets the same fresh name. Hence the need for `withFreshTermName` that lets q"_ + 1" quasiquote desugare into equivalent of `withFreshTermName("x$") { freshx => q"$freshx => $freshx + 1" }`.
For pattern quasiquotes it's a bit different. Due to the fact that end-result must be a pattern we need to represent fresh names as patterns too. A natural way to express that something is fresh is to represent it as a free variable (e.g. any name will do in that place). But due to possible use of the same name in multiple places we need to make sure that all such places have the same values by adding a sequence of guards to the pattern.
Previously such names were reified naively and it could have caused name collision problems and inability to properly much on trees that contain such names.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | | |
Previously due to limited support for expansion in apply position
quasiquotes had to use a compiler hook for deconstruction. Now with
recent changes in pattern matcher it's possible to remove that special
case.
|
| | | |
| | |
| | |
| | |
| | |
| | | |
In order to implement this a new parser entry point
`parseStatsOrPackages` that augments current parseStats with ability
to parse "package name { ... }" syntax.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Recent runtime reflection sync pull request, which among other things,
introduced more restricted package completion rules, led to a nightmarish
regression on java7, which prevented the entire scalac from starting up.
This commit temporarily disables the introduced assert in order to hotfix
runtime reflection. In the hours to come I'll be looking into the root of
the problem, preparing a pull request that reinstates the assert, which
indicates a bug in our code that's been there since 2.10.0-M3, and fixes
the said bug.
Details: https://groups.google.com/forum/#!topic/scala-internals/hcnUFk75MgQ
|
| | | |
| | |
| | |
| | | |
https://github.com/scala/scala/pull/3029
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
We do need lazy to be robust, because initialization sequence might
trigger synced operations in unexpected order, leading to NPE's.
Even if this is optimizable by removing some of the lazies or by
carefully reordering cake layers, there's no guarantee that all this
effort won't break after another reflection refactoring.
|
| | | |
| | |
| | |
| | | |
Fixes this glaring omission of synchronization for a core part of Symbol.info.
|
| | | |
| | |
| | |
| | | |
This is another optimization we discussed with Roland.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This is one of the changes suggested by Roland in order to reduce contention
caused by reflection GIL.
Locks optimized away here are indirectly used for such fundamental operations
as subtyping tests, so the optimization looks quite important.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
In the current synchronization scheme multiple threads can enter the
missingHook trying to materialize a package, which hasn't been created.
That's fine, because makeScalaPackage, which creates and enters package
symbols is synchronized and checks whether the creation is necessary
before commencing. Therefore even if makeScalaPackage is called multiple
times in rapid succession, the calls will be serialized and all calls
except the first one won't do anything.
|
| | | |
| | |
| | |
| | |
| | | |
Called from isVolatile, which is called from isStable, which is a part
of the public reflection API.
|
| | | |
| | |
| | |
| | |
| | |
| | | |
selfType joins the happy family of flags, annotations and privateWithin,
which automatically trigger initialization, when used within runtime
reflection.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
First of all, GIL should only apply to runtime reflection, because noone
is going to run toolboxes in multiple threads: a) that's impossible, b/c
the compiler isn't thread safe, b) ToolBox api prevents that.
Secondly, the only things in symbols which require synchronization are:
1) info/validTo (completers aren't thread-safe),
2) rawInfo and its dependencies (it shares a mutable field with info)
3) non-trivial caches like in typeAsMemberOfLock
If you think about it, other things like sourceModule or associatedFile
don't need synchronization, because they are either set up when a symbol
is created or cloned or when it's completed. The former is obviously safe,
while the latter is safe as well, because before acquiring init-dependent
state of symbols, the compiler calls `initialize`, which is synchronized.
We can say that symbols can be in four possible states: 1) being created,
2) created, but not yet initialized, 3) initializing, 4) initialized.
Of those only #3 is dangerous and needs protection, which is what this
commit does.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | | |
On a serious note, I feel really uncomfortable about having to juggle
this slew of locks. Despite that I can't immediately find a deadlock,
I'm 100% sure there is one hiding in the shadows. Hence, I'm abandoning
all runtime reflection locks in favor of a single per-universe one.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The code to do so is curated with the help of a generator. Because this
needs to inspect code post-typer, the code generator is run during partest
as a code-validator. We could concievably do the same with a macro, but
this approach might be a better starting point which macros continue to
stabilize.
Removes Definitions.AnyRefModule and an already deprecated alias, as
these have been throwing exceptions for more than a year since 6bb5975289.
They used to be used by AnyRef specialization.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
At first I just tried to remove syntheticCoreClasses from missingHook
and put them into the initializer of freshly created mirrors in order to
reduce the non-determinism in mutations of the global symbol table.
And then it didn't work, crashing on me claiming that AnyRef is missing.
Apparently we still need AnyRefClass in missingHook, just because it's
impossible to initialize (i.e. unpickle) ScalaPackageClass without it.
And then it still didn't work, whining about multiple overloaded defs
of some synthetic symbols. That was really tricky, but I figured it out
as well by initializing ScalaPackageClass first before forcing any
synthetic symbols (see the details in comments).
And then it worked, but stopped working half a year later when
Jason and I came to revisit this old pull request. The final twist
was pre-initializing ObjectClass, because it's a dependency of AnyRefClass,
which is a critical dependency of ScalaPackageClass (full information
can be found in comments).
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Mentioned methods mutate the global `atPhaseStack` variable, which can
easily lead to imbalances and, ultimately, to the empty stack error.
Luckily for us, there's only one dummy phase, SomePhase, which is used
by runtime reflection, so there is absolutely zero need to invoke atPhase
in non-compiler reflexive universes.
The cleanest solution would be to override `atPhase` for runtime reflection,
but it's @inline final, so I didn't want to pay performance penalties for
something that's used three times in runtime reflection (during unpickling, in
reflection-specific completers and in `Symbol.typeParams/unsafeTypeParams`).
Therefore I added overrideable analogues of `atPhase` and `atPhaseNotLaterThan`
which are called from the aforementioned code shared between the compiler and
runtime reflection. I also had to duplicate the code of `Symbol.XXXtypeParams`
(only in SynchronizedSymbols, not in normal Symbols) again due to those
methods being very performance-sensitive.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Synchronization via decoration would be neat if it actually worked.
Unfortunately, root symbols never got decorated, therefore their children
also never got decorated and all the way down to the very turtles.
This commit fixes this sad issue by turning root symbols from objects
to lazy vals. Yes, this is going to induce a performance penalty, which
will hopefully not be high enough to invalidate this cornerstone of our
synchronization strategy.
Now when root symbols are lazy vals, they can be overridden in the runtime
reflexive universe and decorated with SynchronizedSymbol, which makes their
children sync and sound.
|
| | |/
|/|
| |
| |
| |
| |
| | |
Too bad I didn't notice that before. That will free up quite a bit of
memory, removing an extraneous field in every single Symbol, namely the:
private volatile Symbols.Symbol.SymbolKind$ SymbolKind$module
|
| |\ \
| | |
| | | |
Fix AsSeenFrom of ThisType from TypeVar prefix
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Restores behaviour for the AsSeenFrom before the refactoring
in b457b6c477.
This commit uniformly considered that a `TypeVar` prefix should not
`matchesPrefixAndClass`; a condition that formerly was only applied
if the type being viewed was a `TypeRef`. This condition was
originally added in cc9e8eda3364d as a backstop for pos/t2797.scala.
This commit leaves that backstop in place where it was, although it
expresses it more directly by checking if `pre baseType clazz` is
`NoType`, which was the case that cropped up in SI-2797:
scala> type T = bc._1.type forSome { val bc: (AnyRef, AnyRef) }
warning: there were 1 feature warning(s); re-run with -feature for details
defined type alias T
scala> val et = typeOf[T].dealias.asInstanceOf[ExistentialType]
et: $r.intp.global.ExistentialType = bc._1.type forSome { val bc: (AnyRef, AnyRef) }
scala> et.withTypeVars( { x =>
| println(x.prefix.typeSymbol)
| println(x.prefix.typeSymbol.isSubClass(typeOf[Tuple2[_, _]].typeSymbol))
| println(x.prefix.baseType(typeOf[Tuple2[_, _]].typeSymbol))
| true
| } , reflect.internal.Depth(0))
type bc.type
true
<notype>
res98: Boolean = true
|
| |\ \ \
| | | |
| | | | |
Traverser and Pickler improvements.
|
