i.e. #ifndef VM_WHATEVER -> #ifndef RUNTIME_VM_WHATEVER
This lets us remove a hack from the PRESUBMIT.py script that existed
for reasons that are no longer valid, and sets us up to add some
presubmit checks for the GN build.
R=asiva@google.com, rmacnak@google.com
Review URL: https://codereview.chromium.org/2450713004 .
- use attribute 'no msan' on NativeEntry::ReturnValueIsError function
as msan doesn't seem to track the return slot being set by a native
function
- vsnprintf seems to have issues with msan so unpoison the allocated
memory buffer everytime vsnprintf is used to suppress the error.
BUG=
R=fschneider@google.com
Review URL: https://codereview.chromium.org/2383293003 .
Previously if one marker marked the weak property and another marker marked its key then the first marker might stop marking before it sees that the key was marked.
Here is a possible concurrent execution, assuming P is a WeakProperty and K is P.key:
(Marker A) (Marker B)
| |
[ mark property P ] |
| |
[ drain marking stack ] |
[ no more work to do ] |
| [ mark K ]
| [ draing marking stack ]
| [ no more work to do ]
| |
| |
... ...
| |
[Finalize] [Finalize]
|
[Clear P]
In this execution we end up clearing P even though P.key is marked.
To fix this issue without reintroducing central WeakProperty processing
we change the marking phase loop in such a way that markers consider the
marking done only if all of them agree that they have no more weak properties
with marked keys.
Essentially the marking loop now has a separate phase where all markers check
their pending weak properties. The decision to proceed to the next marking phase
(weak handles processing) is done in lock step using barrier - either all threads
advance to the next stage or one of the threads finds a weak property
with marked key and unmarked value and all threads resume marking.
BUG=
R=asiva@google.com
Review URL: https://codereview.chromium.org/2041413005 .
This CL removes optimized access for scalarlist, and only the new TypedData classes
are optimized. I changed the runtime libraries core and math to use typedData
instead of scalarlist (Uint16List is used in StringBuffer, Uint32List by Math.random).
Instead of using LoadIndexed for internal and external arrays,
split external loads into a load of the backing store and a load
of the element.
v3 <- LoadIndexed(v1, index)
becomes
v2 <- LoadUntagged(v1, ExternalTypedData::data_offset)
v3 <- LoadIndexed(v2, index);
For this I introduce two new representations in the IL:
kUntagged (for values that hold a untagged pointer) and
kNoRepresentation (for instructions accept any input
representation)
Deoptimization does not need to know about kUntagged
since these values can never occur in the environment.
Also with this change:
* fix COMPILE_ASSERT and use it in one place.
* Cleanup IL printer output of deopt ids.
Review URL: https://codereview.chromium.org//12871010
git-svn-id: https://dart.googlecode.com/svn/branches/bleeding_edge/dart@20198 260f80e4-7a28-3924-810f-c04153c831b5
replace it with the beginnings of a local (same isolate) IsolateMirror
implementation.
Removed old mirror tests and added two new mirror tests.
Even though mirrors.cc is part of the vm, I chose to implement most of
it using the dart embedding interface instead of our internal
interfaces because the embedding interface was more convenient.
mirrors.cc is basically all new in this CL -- don't pay any attention
to diffs for that file.
Added dart embedding functions required for the functionality in this
CL: Dart_DebugName, Dart_GetNativeInstanceFieldCount,
Dart_RootLibrary, Dart_RegisteredLibraryUrls, and Dart_LibraryName.
Extended or modified some existing dart api functions, primarily to
make them propagate error handles properly.
Added tests for new dart embedding api functionality.
Added the ability to determine if a port is local to the current isolate.
Extended NotImplementedException to accept an optional string
argument. I wanted to give more descriptive error messages.
Review URL: https://chromiumcodereview.appspot.com//10416050
git-svn-id: https://dart.googlecode.com/svn/branches/bleeding_edge/dart@8117 260f80e4-7a28-3924-810f-c04153c831b5
The C message reader can now read lists created like this:
new List()
new List<int>()
new List<String>()
new List<double>()
new List<bool>()
The backward references are now resolved and already allocated
Dart_CObject objects are reused when there is a backward reference.
The reuse of the Dart_CObject objects poses the issue of which objects
where allocated with the supplied allocator and which where not.
Currently this will work best with a zone allocator. This will be added
to the tests in a subsequent change.
R=asiva@google.com, turnidge@google.com
BUG=
TEST=
Review URL: https://chromiumcodereview.appspot.com//9303031
git-svn-id: https://dart.googlecode.com/svn/branches/bleeding_edge/dart@3831 260f80e4-7a28-3924-810f-c04153c831b5
The purpose of this change is twofold:
1. Source in the bin directory can now use the same assertions as
source in the vm directory. The ASSERT macro used by the code
in runtime/bin was just defined to use assert from the standard
C library.
2. Moving other implementation parts from runtime/vm to
runtime/platform (e.g. classes Monitor and Mutex) for sharing
between runtime/bin and runtime/vm will be easier as these
implementations rely on these assertion macros.
Created two gypi files for the platform directory. One for the
headers and one for the source. The source one is only included
when building the VM library and will be present in libdart.a
when the dart executable is linked.
All the code for asserts is still in the dart namespace.
Also re-arranged the order of includes to be alphabetically in
the files touched.
R=ager@google.com, iposva@google.com
BUG=
TEST=
Review URL: http://codereview.chromium.org//9189003
git-svn-id: https://dart.googlecode.com/svn/branches/bleeding_edge/dart@3335 260f80e4-7a28-3924-810f-c04153c831b5
Ryan Macnak