Summary: | Optimizing loads when creating new pages | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Jiewen Tan <jiewen_tan> | ||||||||
Component: | WebKit Misc. | Assignee: | Jiewen Tan <jiewen_tan> | ||||||||
Status: | RESOLVED FIXED | ||||||||||
Severity: | Normal | CC: | berto, cgarcia, commit-queue, darin, ews-watchlist, ggaren, gustavo, jiewen_tan, mcatanzaro, tsavell, webkit-bug-importer, youennf | ||||||||
Priority: | P2 | Keywords: | InRadar | ||||||||
Version: | WebKit Nightly Build | ||||||||||
Hardware: | Unspecified | ||||||||||
OS: | Unspecified | ||||||||||
Attachments: |
|
Description
Jiewen Tan
2019-03-09 01:17:31 PST
Created attachment 364123 [details]
Patch
The internal counterpart is within: <rdar://problem/46325687>. Thanks for the patch. If this patch contains new public API please make sure it follows the guidelines for new WebKit2 GTK+ API. See http://trac.webkit.org/wiki/WebKitGTK/AddingNewWebKit2API Created attachment 364175 [details]
Patch
Created attachment 364178 [details]
Patch
Comment on attachment 364178 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=364178&action=review > Source/WebKit/UIProcess/WebPageProxy.h:2191 > +#if HAVE(LOAD_OPTIMIZER) > +WEBPAGEPROXY_LOADOPTIMIZER_ADDITIONS_2 > +#endif How do these hooks work? It's strange that we would have non-open source code tightly coupled to open source like this. It's one thing to depend on an internal library, or to provide APIs for use by internal code, but just #including secret code into WebKit internals like this seems... unusual. What gets optimized? Comment on attachment 364178 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=364178&action=review r+ so Jiewen can keep making progress >> Source/WebKit/UIProcess/WebPageProxy.h:2191 >> +#endif > > How do these hooks work? > > It's strange that we would have non-open source code tightly coupled to open source like this. It's one thing to depend on an internal library, or to provide APIs for use by internal code, but just #including secret code into WebKit internals like this seems... unusual. > > What gets optimized? I understand your unease. But this has happened multiple times before before when an engineer at Apple started development on something that Apple wasn't ready to reveal. I think it’s OK for now; we can discuss it in more detail when someone from Apple upstreams everything. And it’s all inside an #if. Comment on attachment 364178 [details]
Patch
Thanks Darin for r+ this patch.
Comment on attachment 364178 [details] Patch Clearing flags on attachment: 364178 Committed r242697: <https://trac.webkit.org/changeset/242697> All reviewed patches have been landed. Closing bug. Reverted r242697 for reason: Broke internal builders. Committed r242705: <https://trac.webkit.org/changeset/242705> (In reply to Truitt Savell from comment #12) > Reverted r242697 for reason: > > Broke internal builders. > > Committed r242705: <https://trac.webkit.org/changeset/242705> Thanks for the heads up. I forgot that I have to land the OpenSource and Internal changes together. Will hold the same patch after the Internal one is ready to land. Committed r242760: <https://trac.webkit.org/changeset/242760> |