ARC: cycle detector (#12823)

* first implementation of the =trace and =dispose hooks for the cycle collector
* a cycle collector for ARC: progress
* manual: the .acyclic pragma is a thing once again
* gcbench: adaptations for --gc:arc
* enable valgrind tests for the strutils tests
* testament: better valgrind support
* ARC refactoring: growable jumpstacks
* ARC cycle detector: non-recursive algorithm
* moved and renamed core/ files back to system/
* refactoring: --gc:arc vs --gc:orc since 'orc' is even more experimental and we want to ship --gc:arc soonish
This commit is contained in:
Andreas Rumpf 2019-12-17 17:37:50 +01:00 • committed by GitHub
commit 83a736a34a
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
45 changed files with 939 additions and 229 deletions

View file

@ -6017,8 +6017,32 @@ The ``noreturn`` pragma is used to mark a proc that never returns.
acyclic pragma
--------------
The ``acyclic`` pragma applies to type declarations. It is deprecated and
ignored.
The ``acyclic`` pragma can be used for object types to mark them as acyclic
even though they seem to be cyclic. This is an **optimization** for the garbage
collector to not consider objects of this type as part of a cycle:
.. code-block:: nim
type
Node = ref NodeObj
NodeObj {.acyclic.} = object
left, right: Node
data: string
Or if we directly use a ref object:
.. code-block:: nim
type
Node {.acyclic.} = ref object
left, right: Node
data: string
In the example a tree structure is declared with the ``Node`` type. Note that
the type definition is recursive and the GC has to assume that objects of
this type may form a cyclic graph. The ``acyclic`` pragma passes the
information that this cannot happen to the GC. If the programmer uses the
``acyclic`` pragma for data types that are in reality cyclic, the memory leaks
can be the result, but memory safety is preserved.
final pragma