Bug 196849 – WebThread should run at a higher priority than user initiated

• Status: RESOLVED
• Resolution: FIXED
• Component: New Bugs

[Ryosuke Niwa]

WebThread should run at a higher priority than user initiated

[Ryosuke Niwa]

<rdar://problem/46851062>

[Ryosuke Niwa]

Created attachment 367282 [details]
Fixes the bug

[Ryosuke Niwa]

<rdar://problem/46851062>

[Ryosuke Niwa]

Created attachment 367283 [details]
Fixes the bug

[Geoffrey Garen]

Can you include a description of the workload you used to qualify this change, and the improvement you saw?

[Geoffrey Garen]

Comment on attachment 367283 [details]
Fixes the bug

View in context: https://bugs.webkit.org/attachment.cgi?id=367283&action=review

r=me

> Source/WebCore/platform/ios/wak/WebCoreThread.mm:670
> -    pthread_attr_setstacksize(&tattr, 200 * 4096);
> +    pthread_attr_setstacksize(&tattr, 800 * KB);

I'm pretty sure this change is harmless -- but it doesn't really seem related to this bug.

[Ryosuke Niwa]

(In reply to Geoffrey Garen from comment #6)
> Comment on attachment 367283 [details]
> Fixes the bug
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=367283&action=review
> 
> r=me
> 
> > Source/WebCore/platform/ios/wak/WebCoreThread.mm:670
> > -    pthread_attr_setstacksize(&tattr, 200 * 4096);
> > +    pthread_attr_setstacksize(&tattr, 800 * KB);
> 
> I'm pretty sure this change is harmless -- but it doesn't really seem
> related to this bug.

Oh ya, I forgot to mention in the change log. I'm just replaying 200 * 4096 by 200 * 4 * 1024 -> 800 * 1024 -> 800 KB. I think the code comes from when page size was 4KB in iOS.

[Ryosuke Niwa]

(In reply to Geoffrey Garen from comment #5)
> Can you include a description of the workload you used to qualify this
> change, and the improvement you saw?

Added: "This improves the blocked time in StartWebThread from 2~3ms to 250μs while cold launching iBooks to an opened book."

[Ryosuke Niwa]

Committed r244229: <https://trac.webkit.org/changeset/244229>