Support old-space allocation in generated code (bump block only for now).

Support old-space allocation in generated code (bump block only for now).

First steps towards general pretenuring support.
* Generalize Heap::Top/EndAddress.
* Add testing flag to exercise new code paths.
* Also update the slow-case runtime calls, to ensure a fresh block will be allocated in old.

Next steps are general invalidation of the generated code and adding a policy that doesn't blow up the store buffers), and freelist allocation.

Committed: https://code.google.com/p/dart/source/detail?r=40530

[Ivan Posva, 2014-09-19 18:45:55]

First round of comments.

I guess after the lazy walkability change lands, it will reduce the amount of
changes in this CL as well.

-Ivan

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/assembler_arm.cc
File runtime/vm/assembler_arm.cc (right):

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/assembler_arm...
runtime/vm/assembler_arm.cc:3172: const intptr_t instance_size =
cls.instance_size();
ASSERT(instance_size != 0);

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/heap.cc
File runtime/vm/heap.cc (right):

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/heap.cc#newco...
runtime/vm/heap.cc:401: ASSERT(space == kPretenured);
Please add a TODO to the definition of kPretenured to mark the fact that this is
a temporary solution.

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/intrinsifier_...
File runtime/vm/intrinsifier_x64.cc (right):

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/intrinsifier_...
runtime/vm/intrinsifier_x64.cc:1349:
Class::Handle(isolate->object_store()->one_byte_string_class());
Maybe just define a cid local variable above and avoid going the long way around
here.

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/stub_code_arm.cc
File runtime/vm/stub_code_arm.cc (right):

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/stub_code_arm...
runtime/vm/stub_code_arm.cc:675: __ LoadImmediate(TMP,
RawObject::ClassIdTag::encode(cls.id()));
ditto

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/stub_code_arm...
runtime/vm/stub_code_arm.cc:873: const Class& context_class =
Class::ZoneHandle(Object::context_class());
ditto: Just use the kXYZCid everywhere where we know the exact class being used.

[koda, 2014-09-19 20:48:29]

Addressed all comments and sync'ed after submitting page iterator CL.

PTAL

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/assembler_arm.cc
File runtime/vm/assembler_arm.cc (right):

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/assembler_arm...
runtime/vm/assembler_arm.cc:3172: const intptr_t instance_size =
cls.instance_size();
On 2014/09/19 18:45:55, Ivan Posva wrote:
> ASSERT(instance_size != 0);

Done.

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/heap.cc
File runtime/vm/heap.cc (right):

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/heap.cc#newco...
runtime/vm/heap.cc:401: ASSERT(space == kPretenured);
On 2014/09/19 18:45:55, Ivan Posva wrote:
> Please add a TODO to the definition of kPretenured to mark the fact that this
is
> a temporary solution.

Done.

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/intrinsifier_...
File runtime/vm/intrinsifier_x64.cc (right):

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/intrinsifier_...
runtime/vm/intrinsifier_x64.cc:1349:
Class::Handle(isolate->object_store()->one_byte_string_class());
On 2014/09/19 18:45:55, Ivan Posva wrote:
> Maybe just define a cid local variable above and avoid going the long way
around
> here.

Done.

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/stub_code_arm.cc
File runtime/vm/stub_code_arm.cc (right):

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/stub_code_arm...
runtime/vm/stub_code_arm.cc:675: __ LoadImmediate(TMP,
RawObject::ClassIdTag::encode(cls.id()));
On 2014/09/19 18:45:55, Ivan Posva wrote:
> ditto

Done.

https://codereview.chromium.org/578443003/diff/40001/runtime/vm/stub_code_arm...
runtime/vm/stub_code_arm.cc:873: const Class& context_class =
Class::ZoneHandle(Object::context_class());
On 2014/09/19 18:45:55, Ivan Posva wrote:
> ditto: Just use the kXYZCid everywhere where we know the exact class being
used.

Done.

[Ivan Posva, 2014-09-19 22:24:35]

LGTMwC

-Ivan

https://codereview.chromium.org/578443003/diff/60001/runtime/vm/stub_code_x64.cc
File runtime/vm/stub_code_x64.cc (right):

https://codereview.chromium.org/578443003/diff/60001/runtime/vm/stub_code_x64...
runtime/vm/stub_code_x64.cc:892: intptr_t cid = context_class.id();
?

[koda, 2014-09-19 23:46:36]

Committed patchset #4 (id:60001) manually as r40530 (presubmit successful).

[koda, 2014-09-19 23:46:59]

https://codereview.chromium.org/578443003/diff/60001/runtime/vm/stub_code_x64.cc
File runtime/vm/stub_code_x64.cc (right):

https://codereview.chromium.org/578443003/diff/60001/runtime/vm/stub_code_x64...
runtime/vm/stub_code_x64.cc:892: intptr_t cid = context_class.id();
On 2014/09/19 22:24:35, Ivan Posva wrote:
> ?

Done.