A quick update from my debug. We definitely have one nsRefreshDriver for the OOP iframe and another one for its parent document, but: * If the animation is hidden, we are using `mozilla::InactiveRefreshDriverTimer::TickOne()` to tick it, so only update once and no other animation update. So there is no ready and no finished promise resolved. * If the animation is not hidden, we are using `mozilla::RefreshDriverTimer::Tick()`, so everything looks normal. Note: If we disable fission, we only have one nsRefreshDriver, and looks like it always runs? Still have to check why we use an inactive refresh driver for hidden animations. Perhaps it's a way for optimization or it's just a bug.
Bug 1578309 Comment 32 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
A quick update from my debug. We definitely have one nsRefreshDriver for the OOP iframe and another one for its parent document, but: * If the animation is hidden, we are using `mozilla::InactiveRefreshDriverTimer::TickOne()` to tick it, so only update once and no other animation update. So there is no ready and no finished promise resolved. * If the animation is not hidden, we are using `mozilla::RefreshDriverTimer::Tick()`, so everything looks normal. Note: If we disable fission, we only have one nsRefreshDriver of the parent document. Still have to check why we use an inactive refresh driver for hidden animations. Perhaps it's a way for optimization or it's just a bug.
A quick update from my debug. We definitely have one nsRefreshDriver for the OOP iframe and another one for its parent document, but: * If the animation is hidden, we are using `mozilla::InactiveRefreshDriverTimer::TickOne()` to tick it, so only update once and no other animation update. So there is no ready and no finished promise resolved. * If the animation is not hidden, we are using `mozilla::RefreshDriverTimer::Tick()`, so everything looks normal. Note: If we disable fission, we only have one nsRefreshDriver for the parent document. Still have to check why we use an inactive refresh driver for hidden animations. Perhaps it's a way for optimization or it's just a bug.