Re-insert linebreak in dwarf_stack_trace_test

Re-insert linebreak in dwarf_stack_trace_test

This file was auto-formatted 2 months ago.
This - among other things - removed a linebreak which shifted linenumbers
by one, but didn't update the expected linenumbers from a throw accordingly.
Somehow the testserver has claimed that this has passed until now.
I have no idea how that happened.

This CL re-introduces the linebreak by inserting a comment,
which (currently) forces dartfmt to have the wanted linebreak.

BUG=
R=dmitryas@google.com

Committed: https://github.com/dart-lang/sdk/commit/d03a0d798e6769bcb1b302443e5bcfc89e6de551

[jensj, 2017-06-21 11:34:22]

jensj@google.com changed reviewers:
+ dmitryas@google.com, kmillikin@google.com, whesse@google.com

[jensj, 2017-06-21 11:34:22]

Bill: Do you have any idea about how this could happen?
https://uberchromegw.corp.google.com/i/client.dart/builders/precomp-linux-deb...
--- build 722 failed (as it should, it seems to me), but the previous builds
didn't. Build 721 says "standalone/dwarf_stack_trace_test: pass" - though the
test hasn't been updated for 2 months and the status file was last updated 5
days ago.

(Maybe the bot changed?
```
  if (!result.stdout.contains("shared object")) {
    print("Skipping test because we are not running from a dylib");
    return;
  }
```)

[Dmitry Stefantsov, 2017-06-21 11:41:04]

LGTM

I also can't reproduce the bot behavior locally when the test "passes" while
having this formatting issue.  It's indeed strange that bots didn't detect this
issue until recently.

[jensj, 2017-06-21 11:47:43]

Description was changed from

==========
Re-insert linebreak in dwarf_stack_trace_test

This file was auto-formatted 2 months ago.
This - among other things - removed a linebreak which shifted linenumbers
by one, but didn't update the expected linenumbers from a throw accordingly.
Somehow the testserver has claimed that this has passed until now.
I have no idea how that happened.

This CL re-introduces the linebreak by inserting a comment,
which (currently) forces dartfmt to have the wanted linebreak.

BUG=
==========

to

==========
Re-insert linebreak in dwarf_stack_trace_test

This file was auto-formatted 2 months ago.
This - among other things - removed a linebreak which shifted linenumbers
by one, but didn't update the expected linenumbers from a throw accordingly.
Somehow the testserver has claimed that this has passed until now.
I have no idea how that happened.

This CL re-introduces the linebreak by inserting a comment,
which (currently) forces dartfmt to have the wanted linebreak.

BUG=
R=dmitryas@google.com

Committed:
https://github.com/dart-lang/sdk/commit/d03a0d798e6769bcb1b302443e5bcfc89e6de551
==========

[jensj, 2017-06-21 11:47:45]

Committed patchset #1 (id:1) manually as
d03a0d798e6769bcb1b302443e5bcfc89e6de551 (presubmit successful).

[Bill Hesse, 2017-06-21 12:04:54]

On 2017/06/21 11:34:22, jensj wrote:
> Bill: Do you have any idea about how this could happen?
>
https://uberchromegw.corp.google.com/i/client.dart/builders/precomp-linux-deb...
> --- build 722 failed (as it should, it seems to me), but the previous builds
> didn't. Build 721 says "standalone/dwarf_stack_trace_test: pass" - though the
> test hasn't been updated for 2 months and the status file was last updated 5
> days ago.
> 
> (Maybe the bot changed?
> ```
>   if (!result.stdout.contains("shared object")) {
>     print("Skipping test because we are not running from a dylib");
>     return;
>   }
> ```)

No, no idea.  The test was actually passing in the test.py run, so it doesn't
seem to be a bot problem.

[jensj, 2017-06-21 12:41:23]

https://github.com/dart-lang/sdk/issues/29948