Compile and simulate first dart function on arm generated from ast.

Compile and simulate first dart function on arm generated from ast.
Committed: https://code.google.com/p/dart/source/detail?r=19152

[regis, 2013-02-26 19:30:25]

There is a enough to review for everyone :-)

[srdjan, 2013-02-26 23:23:09]

LGTM with comments & question.

https://codereview.chromium.org/12335102/diff/1/runtime/vm/code_generator_tes...
File runtime/vm/code_generator_test.cc (right):

https://codereview.chromium.org/12335102/diff/1/runtime/vm/code_generator_tes...
runtime/vm/code_generator_test.cc:568: #endif
Add comment to both endif-s what they are (similar to removed comment)

https://codereview.chromium.org/12335102/diff/1/runtime/vm/compiler.cc
File runtime/vm/compiler.cc (right):

https://codereview.chromium.org/12335102/diff/1/runtime/vm/compiler.cc#newcod...
runtime/vm/compiler.cc:553: DisassembleCode(parsed_function.function(), false);
Why this extra disassemble?

https://codereview.chromium.org/12335102/diff/1/runtime/vm/flow_graph_allocat...
File runtime/vm/flow_graph_allocator.cc (right):

https://codereview.chromium.org/12335102/diff/1/runtime/vm/flow_graph_allocat...
runtime/vm/flow_graph_allocator.cc:80: 
You don't need to set to false, blocked_cpu_registers_() is initializing the
array to false already.

[regis, 2013-02-27 00:19:26]

Thanks Srdjan. See my replies/question.
I'll upload after others have reviewed.

-- Regis

https://codereview.chromium.org/12335102/diff/1/runtime/vm/code_generator_tes...
File runtime/vm/code_generator_test.cc (right):

https://codereview.chromium.org/12335102/diff/1/runtime/vm/code_generator_tes...
runtime/vm/code_generator_test.cc:568: #endif
On 2013/02/26 23:23:09, srdjan wrote:
> Add comment to both endif-s what they are (similar to removed comment)

Done.

https://codereview.chromium.org/12335102/diff/1/runtime/vm/compiler.cc
File runtime/vm/compiler.cc (right):

https://codereview.chromium.org/12335102/diff/1/runtime/vm/compiler.cc#newcod...
runtime/vm/compiler.cc:553: DisassembleCode(parsed_function.function(), false);
On 2013/02/26 23:23:09, srdjan wrote:
> Why this extra disassemble?

It is not an extra one. The helper CompileParsedFunctionHelper does not
disassemble (CompileFunctionHelper does).

Codegen tests (using CompileParsedFunctionHelper directly) did not get
disassembled.

We could move the disassemble functionality from CompileFunctionHelper into
CompileParsedFunctionHelper, but I did not want to restructure code and possibly
change the order in which things are printed out and expected by your team.

https://codereview.chromium.org/12335102/diff/1/runtime/vm/flow_graph_allocat...
File runtime/vm/flow_graph_allocator.cc (right):

https://codereview.chromium.org/12335102/diff/1/runtime/vm/flow_graph_allocat...
runtime/vm/flow_graph_allocator.cc:80: 
On 2013/02/26 23:23:09, srdjan wrote:
> You don't need to set to false, blocked_cpu_registers_() is initializing the
> array to false already.

Where does this happen? Does blocked_cpu_registers_() in the initializer list of
the constructor initialize the array?

[srdjan, 2013-02-27 00:27:24]

https://codereview.chromium.org/12335102/diff/1/runtime/vm/flow_graph_allocat...
File runtime/vm/flow_graph_allocator.cc (right):

https://codereview.chromium.org/12335102/diff/1/runtime/vm/flow_graph_allocat...
runtime/vm/flow_graph_allocator.cc:80: 
On 2013/02/27 00:19:26, regis wrote:
> On 2013/02/26 23:23:09, srdjan wrote:
> > You don't need to set to false, blocked_cpu_registers_() is initializing the
> > array to false already.
> 
> Where does this happen? Does blocked_cpu_registers_() in the initializer list
of
> the constructor initialize the array?

Yes.

[regis, 2013-02-27 17:14:05]

Committed patchset #2 manually as r19152 (presubmit successful).