[Webkit-unassigned] [Bug 188742] [GLIB] Expose JavaScriptCore options in GLib public API

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Thu Jan 24 03:07:30 PST 2019


--- Comment #13 from Carlos Garcia Campos <cgarcia at igalia.com> ---
(In reply to Michael Catanzaro from comment #12)
> Comment on attachment 359149 [details]
> Patch
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=359149&action=review
> Sorry it took me much longer than usual to get to this.

No problem.

> The options are global, not specific to a particular VM?


> JSC_OPTIONS_USE_FTL, and JSC_OPTIONS_USE_LLINT, it would be nice to have a
> little bit more documentation. The only one library users are likely to
> understand is the plain JIT, and even then we should probabyl explain it a
> little bit. DGF, FTL, and LLINT are probably just alphabet soup even to
> experienced developers, unless you're familiar with WebKit internals.

I expect users of those options to know what they are.

> Lastly... use-case. You used to require that new APIs be used by a real
> application. Do you have plans to use it somewhere? Surely not adding API
> just for the sake of it?
> > Source/JavaScriptCore/API/glib/JSCOptions.cpp:33
> > + * JavaScript options allow to change the behavior of the JavaScript engine.
> allow changing
> > Source/JavaScriptCore/API/glib/JSCOptions.cpp:37
> > + * Only a few of them are documented, you can use the undocumented options at
> documented;
> > Source/JavaScriptCore/API/glib/JSCOptions.cpp:38
> > + * your own risk (you can find the list of options in the WebKit source code).
> your own risk. (You can find the list of options in the WebKit source code.)
> > Source/JavaScriptCore/API/glib/JSCOptions.cpp:52
> > +typedef int32_t int32;
> > +typedef size_t size;
> typedefs are fine, but a more modern/natural way would be:
> using int32 = int32_t;
> using size = size_t;
> I guess these are for use by the FOR_EACH options macros?


> > Tools/TestWebKitAPI/Tests/JavaScriptCore/glib/TestJSC.cpp:3376
> > +        "--jsc-configFile=/tmp/bar",
> Does the config file actually get created by the tests? I assume not, since
> you're not actually running any JS. But if so, it should be cleaned.

It's not created.

You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20190124/54102754/attachment-0001.html>

More information about the webkit-unassigned mailing list