Bug 149720 – REGRESSION(r190370): Speedometer spits out an extra console message

• Status: RESOLVED
• Resolution: DUPLICATE
• Component: JavaScriptCore

[Ryosuke Niwa]

run-perf-tests Speedometer fails with the following error message:

Running Speedometer/Full.html (1 of 1)
ERROR: CONSOLE MESSAGE: line 208: Miss the info bar? Run TodoMVC from a server to avoid a cross-origin error.
FAILED
Finished: 30.388546 s

Either r190370 fixed an existing bug or introduced a new regression.

[Ryosuke Niwa]

Created attachment 262280 [details]
Screenshot of Mavericks perf bot builds list

[Ryosuke Niwa]

The regression went away when we rolled out the patch so I'm pretty certain this is the culprit.

[Ryosuke Niwa]

Committed r190535: <http://trac.webkit.org/changeset/190535>

[Ryosuke Niwa]

Temporarily ignored this console message.

[Csaba Osztrogonác]

(In reply to comment #4)
> Temporarily ignored this console message.

I think it is paper overing this serious regression. Why is it 
good to measure the performace if the test is completely broken?

[Csaba Osztrogonác]

*** This bug has been marked as a duplicate of bug 149621 ***

[Ryosuke Niwa]

(In reply to comment #5)
> (In reply to comment #4)
> > Temporarily ignored this console message.
> 
> I think it is paper overing this serious regression. Why is it 
> good to measure the performace if the test is completely broken?

I don't think this is a serious regression. Tail optimization does change the JS behavior, and the benchmark isn't failing on the browser. It's just that we have one extra console message when ran inside WRT.

[Csaba Osztrogonác]

(In reply to comment #7)
> (In reply to comment #5)
> > (In reply to comment #4)
> > > Temporarily ignored this console message.
> > 
> > I think it is paper overing this serious regression. Why is it 
> > good to measure the performace if the test is completely broken?
> 
> I don't think this is a serious regression. Tail optimization does change
> the JS behavior, and the benchmark isn't failing on the browser. It's just
> that we have one extra console message when ran inside WRT.

I don't know the internals of this test. But theoretically I don't think 
if a JS engine should change the behaviour of any JS script anytime.

[Ryosuke Niwa]

Wait, this isn't a duplicate of 149621.

[Radar WebKit Bug Importer]

<rdar://problem/22995488>

[Radar WebKit Bug Importer]

<rdar://problem/22995494>

[Ryosuke Niwa]

Oh, nvm. I misread it.

*** This bug has been marked as a duplicate of bug 149621 ***

[Radar WebKit Bug Importer]

<rdar://problem/22995514>