Closed Bug 1467688 Opened 6 years ago Closed 6 years ago

Style changes to perspective get their change hints ignored if the frame isn't also transformed.

Categories

(Core :: CSS Parsing and Computation, defect)

defect
Not set
normal

Tracking

()

RESOLVED FIXED
mozilla63
Tracking Status
firefox62 --- fixed
firefox63 --- fixed

People

(Reporter: mattwoodrow, Assigned: mattwoodrow)

References

(Blocks 1 open bug)

Details

Attachments

(2 files)

nsStyleDisplay::CalcDifference set 'hint |= nsChangeHint_NeutralChange;' if HasTransformStyle returns false (which is the case for just perspective), and so ignores any change hints we generated for a new/different perspective value.

Looks like this bug has existed since I first implemented perspective, the change hint checks were initially in an if (HasTransformStyle()) block.
Bug 1466954 has a crashtest that triggers this (and hits an RDL assertion).
See Also: → 1466954
FWIW, this is a crash report caused by the same reason. bp-16522335-6c5b-495b-928e-97fe40180613
Actually, the original description isn't quite right. The check for a changed value of HasPerspectiveStyle() sets the value of 'hint' and not 'transformHint', so it always applied.

The bug here is that we only generate nsChangeHint_UpdateContainingBlock in that case (since the check for changes to mChildPerspective *is* scoped to transformed frame) and not nsChangeHint_RepaintFrame.

The broken testcases are applying the new perspective to the document element (which is already a containing block for fixed elements), so it ends up being a no-op, and nothing invalidates for the newly added perspective.

We do have an existing testcase for the normal case of this (layout/reftests/w3c-css/submitted/transforms/perspective-containing-block-dynamic-1b.html) and that passes already.
Comment on attachment 8986079 [details]
Bug 1467688 - Make sure we invalidate for perspective changes even if the frame isn't otherwise transformed.

https://reviewboard.mozilla.org/r/251528/#review257800

::: layout/style/nsStyleStruct.cpp:3879
(Diff revision 1)
> +  if (HasPerspectiveStyle() != aNewData.HasPerspectiveStyle()) {
> +    // A change from/to being a containing block for position:fixed.
> +    hint |= nsChangeHint_UpdateContainingBlock |
> +            nsChangeHint_UpdateOverflow |
> +            nsChangeHint_RepaintFrame;
> +  } else if (mChildPerspective != aNewData.mChildPerspective) {
> +    hint |= nsChangeHint_UpdateOverflow |
> +            nsChangeHint_RepaintFrame;
> +  }
> +

I did now mostly same fix for this bug. :)

I think we can rewrite this if-else block like this;

```c++
  if (mChildPerspective != anewData.mChildPerspective) {
    hint |= nsChangeHint_UpdateOverflow |
            nsChangeHint_RepaintFrame;
    if (HasPerspectiveStyle()! = aNewData.HasPerspectiveStyle()) {
      hint |= nsChangeHint_UpdateContainingBlock;
    }
  }
```
Assignee: nobody → matt.woodrow
Blocks: RDLbugs
Comment on attachment 8986079 [details]
Bug 1467688 - Make sure we invalidate for perspective changes even if the frame isn't otherwise transformed.

https://reviewboard.mozilla.org/r/251528/#review258936

Seems like it might be good to add a reftest rather than just a crashtest?
Attachment #8986079 - Flags: review?(dbaron) → review+
We're sorry, Autoland could not rebase your commits for you automatically. Please manually rebase your commits and try again.

hg error in cmd: hg rebase -s 09fe2c0b64bbf5f7856e3216d1746ffd971f62f1 -d 5a560404cf9e: rebasing 470889:09fe2c0b64bb "Bug 1467688 - Make sure we invalidate for perspective changes even if the frame isn't otherwise transformed. r=dbaron" (tip)
merging layout/base/crashtests/crashtests.list
merging layout/style/nsStyleStruct.cpp
warning: conflicts while merging layout/base/crashtests/crashtests.list! (edit, then use 'hg resolve --mark')
unresolved conflicts (see hg resolve, then hg rebase --continue)
Pushed by mwoodrow@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/b976829f826a
Make sure we invalidate for perspective changes even if the frame isn't otherwise transformed. r=dbaron
Backout by aciure@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/3446bd708e60
Backed out 1 changesets for reftest failures crashtests/1467688.html CLOSED TREE
Pushed by mwoodrow@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/08ad3e81d9af
Make sure we invalidate for perspective changes even if the frame isn't otherwise transformed. r=dbaron
https://hg.mozilla.org/mozilla-central/rev/08ad3e81d9af
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla63
Comment on attachment 8986079 [details]
Bug 1467688 - Make sure we invalidate for perspective changes even if the frame isn't otherwise transformed.

Approval Request Comment
[Feature/Bug causing the regression]: RDL
[User impact if declined]: Assertion crashes on Nightly/Dev Edition, potentially incorrect z-ordering on release (though usually temporary).
[Is this code covered by automated tests?]: Crashtest added.
[Has the fix been verified in Nightly?]: By me, yes
[Needs manual test from QE? If yes, steps to reproduce]: No
[List of other uplifts needed for the feature/fix]: None
[Is the change risky?]: No
[Why is the change risky/not risky?]: Just adds extra style change hints to make sure we invalidate. Should never be a problem to hint too many changes.
[String changes made/needed]: None.
Flags: needinfo?(matt.woodrow)
Attachment #8986079 - Flags: approval-mozilla-beta?
Comment on attachment 8986079 [details]
Bug 1467688 - Make sure we invalidate for perspective changes even if the frame isn't otherwise transformed.

Crash fix, adds new tests, let's uplift for beta 8.
Attachment #8986079 - Flags: approval-mozilla-beta? → approval-mozilla-beta+
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: