| Summary: | REGRESSION (r171239): Much more time spent taking snapshots during the PLT | ||||||
|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Tim Horton <thorton> | ||||
| Component: | WebKit2 | Assignee: | Tim Horton <thorton> | ||||
| Status: | RESOLVED FIXED | ||||||
| Severity: | Normal | CC: | commit-queue, mitz, rniwa, sam, simon.fraser | ||||
| Priority: | P2 | Keywords: | InRadar | ||||
| Version: | 528+ (Nightly build) | ||||||
| Hardware: | Unspecified | ||||||
| OS: | Unspecified | ||||||
| Bug Depends on: | 135190 | ||||||
| Bug Blocks: | |||||||
| Attachments: |
|
||||||
|
Description
Tim Horton
2014-07-22 14:37:35 PDT
Created attachment 235316 [details]
patch
This change broke three API tests, so I'm going to roll out. Tests that failed: WKUserContentController.ScriptMessageHandlerWithNavigation WebKit2.RestoreSessionStateContainingFormData Tests that timed out: DeviceScaleFactorOnBack.WebKit2 Re-opened since this is blocked by bug 135190 What! How is that possible? I added this message very recently and the *only* thing it does is take a snapshot. Will investigate. The UI process is stuck in a HashTable lock adding a string to the AtomicString table under decidePolicyForResponse. Still unclear how I could have influenced this. Given http://trac.webkit.org/changeset/171367 I wonder if ap was off-by-one somehow. Testing now. Rolling out my patch changes nothing, rolling out Joe's fixes all three tests. gg. (In reply to comment #8) > Rolling out my patch changes nothing, rolling out Joe's fixes all three tests. gg. Rolled this one back in in http://trac.webkit.org/changeset/171471, will go roll Joe out. |