Need to implement WorkerContextProxy::create for Chromium platform.
Created attachment 27945 [details] Proposed Patch
Created attachment 27947 [details] Proposed Patch
Comment on attachment 27947 [details] Proposed Patch I don't think that checking for PLATFORM(CHROMIUM) when the check is actually for multi-process implementation is great. Is this what we're always doing in such cases?
(In reply to comment #3) > (From update of attachment 27947 [details] [review]) > I don't think that checking for PLATFORM(CHROMIUM) when the check is actually > for multi-process implementation is great. Is this what we're always doing in > such cases? > This check is to say: create WorkerMessagingProxy for all non-Chromium platform For Chromium, we have our own logic to decide what to create. That is, either create our own version of cross-process proxies or use WorkerMessagingProxy directly if it is for subworker or worker from the same origin.
We do not need additional changes other than placing WorkerContextProxy::create in #if. So change the summary to reflect this.
Created attachment 27986 [details] Proposed Patch
Comment on attachment 27986 [details] Proposed Patch > This check is to say: create WorkerMessagingProxy for all non-Chromium platform > For Chromium, we have our own logic to decide what to create. What I was saying was that non-Chromium platforms may want to use their own logic for create() if they use separate processes instead of threads. But looks like we don't have a platform define for that yet. r=me
Reassign to levin for landing.
Committed in r41291.