Sputnik tests - forward to r94.

Sputnik tests - forward to r94.


BUG=
TEST=

[Martin Maly, 2011-02-22 20:36:36]

This brings Sputnik tests to current head.
Recently Sputnik introduced a flag that runs all/some tests in strict mode. I am
keeping it off for the moment but have todo to turn it on as the strict mode
implementation progresses.

Thank you!
Martin

[Lasse Reichstein, 2011-02-23 06:12:32]

LGTM

http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status
File test/sputnik/sputnik.status (right):

http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
test/sputnik/sputnik.status:193: # with undefined/null. Test expects otherwise.
Object.prototype.toString has special cases for null and undefined as the two
first steps of the algorithm. After that, it does ToObject on the remaining
values.

http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
test/sputnik/sputnik.status:199: # V8 should not wrap this when calling builtin
functions
We should have avoided that already. Are we wrapping this now?

http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
test/sputnik/sputnik.status:223: S15.3_A3_T1: FAIL
Agree.

http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
test/sputnik/sputnik.status:225: S15.3_A3_T3: FAIL
And agree. 
The tests did depend on a variable not being present before the assignment to
the global property.

[Martin Maly, 2011-02-23 18:12:10]

Thanks, Lasse, for reviewing.

Rico, if you are ok with the change, feel free to commit it and update the build
bots.

Thank you!
Martin

http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status
File test/sputnik/sputnik.status (right):

http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
test/sputnik/sputnik.status:193: # with undefined/null. Test expects otherwise.
On 2011/02/23 06:12:32, Lasse Reichstein wrote:
> Object.prototype.toString has special cases for null and undefined as the two
> first steps of the algorithm. After that, it does ToObject on the remaining
> values.

Cleaned up. These 2 failures are captured below as well as a V8 bug.

http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
test/sputnik/sputnik.status:199: # V8 should not wrap this when calling builtin
functions
On 2011/02/23 06:12:32, Lasse Reichstein wrote:
> We should have avoided that already. Are we wrapping this now?

not wrapping for builtins is on my todo list still. Some builtins may already
work but not the builtins implemented in javascript themselves. Once that's
done, all these tests will start passing.

http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
test/sputnik/sputnik.status:223: S15.3_A3_T1: FAIL
On 2011/02/23 06:12:32, Lasse Reichstein wrote:
> Agree.

Done.

http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
test/sputnik/sputnik.status:225: S15.3_A3_T3: FAIL
On 2011/02/23 06:12:32, Lasse Reichstein wrote:
> And agree. 
> The tests did depend on a variable not being present before the assignment to
> the global property.

Done.

[Rico, 2011-02-24 06:39:42]

Landed in revision 6918

On 2011/02/23 18:12:10, Martin Maly wrote:
> Thanks, Lasse, for reviewing.
> 
> Rico, if you are ok with the change, feel free to commit it and update the
build
> bots.
> 
> Thank you!
> Martin
> 
> http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status
> File test/sputnik/sputnik.status (right):
> 
>
http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
> test/sputnik/sputnik.status:193: # with undefined/null. Test expects
otherwise.
> On 2011/02/23 06:12:32, Lasse Reichstein wrote:
> > Object.prototype.toString has special cases for null and undefined as the
two
> > first steps of the algorithm. After that, it does ToObject on the remaining
> > values.
> 
> Cleaned up. These 2 failures are captured below as well as a V8 bug.
> 
>
http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
> test/sputnik/sputnik.status:199: # V8 should not wrap this when calling
builtin
> functions
> On 2011/02/23 06:12:32, Lasse Reichstein wrote:
> > We should have avoided that already. Are we wrapping this now?
> 
> not wrapping for builtins is on my todo list still. Some builtins may already
> work but not the builtins implemented in javascript themselves. Once that's
> done, all these tests will start passing.
> 
>
http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
> test/sputnik/sputnik.status:223: S15.3_A3_T1: FAIL
> On 2011/02/23 06:12:32, Lasse Reichstein wrote:
> > Agree.
> 
> Done.
> 
>
http://codereview.chromium.org/6542066/diff/1/test/sputnik/sputnik.status#new...
> test/sputnik/sputnik.status:225: S15.3_A3_T3: FAIL
> On 2011/02/23 06:12:32, Lasse Reichstein wrote:
> > And agree. 
> > The tests did depend on a variable not being present before the assignment
to
> > the global property.
> 
> Done.