Always use --leg_only option.

Always use --leg_only option.

Committed: https://code.google.com/p/dart/source/detail?r=2567

[ahe, 2011-12-15 15:51:09]

@Bill: please take a look at dart/tools/testing/architecture.py and let me know
if that is going to be a problem for test.dart.

[kasperl, 2011-12-19 05:38:47]

LGTM.

[ahe, 2011-12-19 14:34:31]

Hi Mads,

Please take a look at this and let me know if it is a problem for test.dart.

Cheers,
Peter

[Bill Hesse, 2011-12-19 14:37:30]

I didn't see any problem, since you are just adding an option to the command
line.  Since you had gotten an LGTM and committed, I didn't bother to respond. 
Is there something I am missing, or is this just adding an option to the
command, and parsing an extra status file?  How could any of that be a problem?

[Mads Ager (google), 2011-12-19 14:38:27]

I'll update test.dart. Yes, this needs updating in test.dart. We need to pass
the option and we need to take the new status files into account.

[ahe, 2011-12-19 14:40:31]

On 2011/12/19 14:37:30, Bill Hesse wrote:
> I didn't see any problem, since you are just adding an option to the command
> line.  Since you had gotten an LGTM and committed, I didn't bother to respond.

> Is there something I am missing, or is this just adding an option to the
> command, and parsing an extra status file?  How could any of that be a
problem?

I'd like your opinion on parsing an extra status file and whether or not it is
going to be a problem for test.dart. There I two reason for me asking: 1. I
don't know if it will impact test.dart (I assume it doesn't as you must already
handle it for the co19 tests). 2. I want to make sure you're aware that I made
this change.

[ngeoffray, 2011-12-20 10:30:32]

LGTM!

http://codereview.chromium.org/8957008/diff/3002/dart/frog/tests/leg/leg.status
File dart/frog/tests/leg/leg.status (right):

http://codereview.chromium.org/8957008/diff/3002/dart/frog/tests/leg/leg.stat...
dart/frog/tests/leg/leg.status:17: *: Skip # Leg is not self-hosted (yet).
I don't think that's the reason. I believe that's because we cannot compile the
code yet.

http://codereview.chromium.org/8957008/diff/3002/dart/tools/testing/architect...
File dart/tools/testing/architecture.py (right):

http://codereview.chromium.org/8957008/diff/3002/dart/tools/testing/architect...
dart/tools/testing/architecture.py:539: return ['--enable_leg', '--leg_only']
leg_only does not need enable_leg.

[ahe, 2011-12-20 11:21:08]

Hi Kasper, Bill, Mads, and Nicolas,

Thanks for taking a look!

Cheers,
Peter

http://codereview.chromium.org/8957008/diff/3002/dart/frog/tests/leg/leg.status
File dart/frog/tests/leg/leg.status (right):

http://codereview.chromium.org/8957008/diff/3002/dart/frog/tests/leg/leg.stat...
dart/frog/tests/leg/leg.status:17: *: Skip # Leg is not self-hosted (yet).
On 2011/12/20 10:30:32, ngeoffray wrote:
> I don't think that's the reason. I believe that's because we cannot compile
the
> code yet.

Well, when you say "the code", you refer to Leg itself. That  Leg cannot compile
Leg means that Leg is not self-hosted :-)

http://codereview.chromium.org/8957008/diff/3002/dart/tools/testing/architect...
File dart/tools/testing/architecture.py (right):

http://codereview.chromium.org/8957008/diff/3002/dart/tools/testing/architect...
dart/tools/testing/architecture.py:539: return ['--enable_leg', '--leg_only']
On 2011/12/20 10:30:32, ngeoffray wrote:
> leg_only does not need enable_leg.

Well, since there is a bug in scripts/bootstrap/frog_wrapper.py, it does. I'll
send out a separate CL.