Bug 80906 – _runtime_feature_list isn't supported on Apple Mac DRT

• Component: New Bugs
• Status: RESOLVED FIXED

[Dirk Pranke]

_runtime_feature_list isn't supported on Apple Mac DRT

[Dirk Pranke]

Created attachment 131445 [details]
Patch

[Eric Seidel (no email)]

Comment on attachment 131445 [details]
Patch

I'm not sure what the purpose of adding this stub is?

[Alexey Proskuryakov]

Comment on attachment 131445 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=131445&action=review

Thanks for fixing this!

> Tools/Scripts/webkitpy/layout_tests/port/apple.py:99
> +        # FIXME: Implement this properly?

I think that it's worth commenting that (a) it crashes because the script doesn't set DYLD_FRAMEWORK_PATH, and (b) this feature may be obsolete now.

[Dirk Pranke]

Created attachment 131453 [details]
Patch

[Eric Seidel (no email)]

Comment on attachment 131453 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=131453&action=review

Are we 100% sure this is AppleWin-only?

> Tools/Scripts/webkitpy/layout_tests/port/webkit.py:253
> +        """A port can implement this to detect features only known through compile-time or run-time flags."""

compile-time?

> Tools/Scripts/webkitpy/layout_tests/port/webkit.py:566
> +        import pdb
> +        pdb.set_trace()

I don't think you meant to includ ethis.

[Dirk Pranke]

(In reply to comment #2)
> (From update of attachment 131445 [details])
> I'm not sure what the purpose of adding this stub is?

To override the default, which assumed the port implemented this ... however, as you can see in the second patch, flipping the base class implementation is a more accurate way to model things.

(In reply to comment #5)
> (From update of attachment 131453 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=131453&action=review
> 
> Are we 100% sure this is AppleWin-only?
> 

Seems to be.

> > Tools/Scripts/webkitpy/layout_tests/port/webkit.py:253
> > +        """A port can implement this to detect features only known through compile-time or run-time flags."""
> 
> compile-time?
>

I thought the intent of this was to also detect which features had been compiled in (or out). At least that's the way that I look at it, even though you're actually running something to determine that. I can stick with run-time flags if you prefer.

> > Tools/Scripts/webkitpy/layout_tests/port/webkit.py:566
> > +        import pdb
> > +        pdb.set_trace()
> 
> I don't think you meant to includ ethis.

Indeed. I thought I got rid of that but somehow it seems to have snuck back in.

[Dirk Pranke]

Created attachment 131456 [details]
Patch

[Eric Seidel (no email)]

Comment on attachment 131453 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=131453&action=review

>>> Tools/Scripts/webkitpy/layout_tests/port/webkit.py:253
>>> +        """A port can implement this to detect features only known through compile-time or run-time flags."""
>> 
>> compile-time?
> 
> I thought the intent of this was to also detect which features had been compiled in (or out). At least that's the way that I look at it, even though you're actually running something to determine that. I can stick with run-time flags if you prefer.

Does the compile-time detection work through the same function?  I thought it was a different set of functions...

[Dirk Pranke]

(In reply to comment #8)
> (From update of attachment 131453 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=131453&action=review
> 
> >>> Tools/Scripts/webkitpy/layout_tests/port/webkit.py:253
> >>> +        """A port can implement this to detect features only known through compile-time or run-time flags."""
> >> 
> >> compile-time?
> > 
> > I thought the intent of this was to also detect which features had been compiled in (or out). At least that's the way that I look at it, even though you're actually running something to determine that. I can stick with run-time flags if you prefer.
> 
> Does the compile-time detection work through the same function?  I thought it was a different set of functions...

Ah, true, I forgot about the "grep webcore for symbols" path ... we're way too fancy here.

[Eric Seidel (no email)]

Comment on attachment 131456 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=131456&action=review

> Tools/ChangeLog:8
> +        It looks like this feature is actually only supported on the Win

It looks like I came to the same conclusion when writing http://trac.webkit.org/changeset/99377. :)

[Dirk Pranke]

Created attachment 131458 [details]
rework docstring

[Eric Seidel (no email)]

Comment on attachment 131456 [details]
Patch

It seems like this support was a dream.  A dream which never came to be.  Time to end the dream I guess.

[Dirk Pranke]

Committed r110513: <http://trac.webkit.org/changeset/110513>