Set ScopeInfo when compiling for unit tests.

Set ScopeInfo when compiling for unit tests.

This is required to correctly treat eval in unit tests.

BUG=
R=mstarzinger@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=23193

[sigurds, 2014-08-08 11:53:44]

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/functio...
File test/cctest/compiler/function-tester.h (right):

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/functio...
test/cctest/compiler/function-tester.h:50: if (!info.shared_info().is_null()) {
Please advice if you think the null check is necessary. (Copied from
compiler.cc)

[Michael Starzinger, 2014-08-14 08:48:08]

LGTM with comments.

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/functio...
File test/cctest/compiler/function-tester.h (right):

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/functio...
test/cctest/compiler/function-tester.h:50: if (!info.shared_info().is_null()) {
On 2014/08/08 11:53:44, sigurds wrote:
> Please advice if you think the null check is necessary. (Copied from
> compiler.cc)

Since we initialize the CompilationInfo from a JSFunction, the
SharedFunctionInfo should be initialized. I would be fine with dropping the
check.

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-ch...
File test/cctest/compiler/test-changes-lowering.cc (right):

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-ch...
test/cctest/compiler/test-changes-lowering.cc:64: if
(!info.shared_info().is_null()) {
Likewise.

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-co...
File test/cctest/compiler/test-codegen-deopt.cc (right):

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-co...
test/cctest/compiler/test-codegen-deopt.cc:54: if
(!info.shared_info().is_null()) {
Likewise.

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-pi...
File test/cctest/compiler/test-pipeline.cc (right):

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-pi...
test/cctest/compiler/test-pipeline.cc:32: if (!info.shared_info().is_null()) {
As discussed offline: Let's drop this as this test case only tests the minimal
pipeline.

[sigurds, 2014-08-14 09:53:34]

Will land once nopush is gone.

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/functio...
File test/cctest/compiler/function-tester.h (right):

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/functio...
test/cctest/compiler/function-tester.h:50: if (!info.shared_info().is_null()) {
On 2014/08/14 08:48:07, Michael Starzinger wrote:
> On 2014/08/08 11:53:44, sigurds wrote:
> > Please advice if you think the null check is necessary. (Copied from
> > compiler.cc)
> 
> Since we initialize the CompilationInfo from a JSFunction, the
> SharedFunctionInfo should be initialized. I would be fine with dropping the
> check.

Done.

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-ch...
File test/cctest/compiler/test-changes-lowering.cc (right):

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-ch...
test/cctest/compiler/test-changes-lowering.cc:64: if
(!info.shared_info().is_null()) {
On 2014/08/14 08:48:07, Michael Starzinger wrote:
> Likewise.

Done.

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-co...
File test/cctest/compiler/test-codegen-deopt.cc (right):

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-co...
test/cctest/compiler/test-codegen-deopt.cc:54: if
(!info.shared_info().is_null()) {
On 2014/08/14 08:48:07, Michael Starzinger wrote:
> Likewise.

Done.

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-pi...
File test/cctest/compiler/test-pipeline.cc (right):

https://codereview.chromium.org/453953002/diff/1/test/cctest/compiler/test-pi...
test/cctest/compiler/test-pipeline.cc:32: if (!info.shared_info().is_null()) {
On 2014/08/14 08:48:07, Michael Starzinger wrote:
> As discussed offline: Let's drop this as this test case only tests the minimal
> pipeline.

Done.

[sigurds, 2014-08-19 11:24:32]

Committed patchset #2 manually as 23193 (presubmit successful).