Batch sync, default and continuous lanes

[tyao1]

Summary

This is the other approach for unifying default and sync lane #25524.
The approach in that PR is to merge default and continuous lane into the sync lane, and use a new field to track the priority. But there are a couple places that field will be needed, and it is difficult to correctly reset the field when there is no sync lane.

In this PR we take the other approach that doesn't remove any lane, but batch them to get the behavior we want.

How did you test this change?

yarn test

[sizebot]

Comparing: fabef7a...1cba37e

Critical size changes

Includes critical production bundles, as well as any change greater than 2%:

Name	+/-	Base	Current	+/- gzip	Base gzip	Current gzip
oss-stable/react-dom/cjs/react-dom.production.min.js	=	154.75 kB	154.77 kB	+0.02%	49.08 kB	49.08 kB
oss-experimental/react-dom/cjs/react-dom.production.min.js	+0.02%	156.70 kB	156.73 kB	+0.01%	49.73 kB	49.73 kB
facebook-www/ReactDOM-prod.classic.js	+0.07%	534.49 kB	534.84 kB	+0.12%	95.15 kB	95.26 kB
facebook-www/ReactDOM-prod.modern.js	+0.07%	519.59 kB	519.94 kB	+0.13%	92.96 kB	93.08 kB

Significant size changes

Includes any change greater than 0.2%:

Expand to show

Name	+/-	Base	Current	+/- gzip	Base gzip	Current gzip
facebook-www/ReactIs-dev.modern.js	+0.72%	9.56 kB	9.63 kB	+0.59%	2.53 kB	2.55 kB
facebook-www/ReactIs-dev.classic.js	+0.72%	9.56 kB	9.63 kB	+0.55%	2.54 kB	2.55 kB

Generated by 🚫 dangerJS against 1cba37e

[Andarist]

Is this going to fix those two?
#24831
#25191

[acdlite]

@Andarist That's one of the motivations, yeah.

[acdlite]

Overall looks great! Just one more question. Could you please look before merging?

[acdlite]

Can you remind me why you changed this?

The desired behavior is that DefaultLane should only interrupt a transition if there's also a pending SyncLane that's entangled with. But if they are entangled, then the nextLane >= wipLane check above would have already evaluated to true.

So I don't think this condition needs to change.

Either way, we should probably have a test for this.

[tyao1]

ah makes sense. I wasn't clear about the desired behavior.

[acdlite]

Use an else instead:

let lane;
if (enableUnifiedSyncLane && (renderLane & SyncUpdateLanes) !== NoLane) {
  lane = SyncHydrationLane;
} else  {
  ...
}

[Siegrift]

Based on #25700 (comment)

Should these two issues be closed?
#24831
#25191