style: Check transitions per longhand to know which transitions to keep.
This fixes bug https://bugzilla.mozilla.org/show_bug.cgi?id=1406111
This fixes the fishy TransitionProperty mapping which I complained about in my
previous refactor.
Turns out that those properties could only be longhands, and thus the expansion
we did before that (and which I removed) was correct.
This fixes the bug by moving back to the previous correct behavior but using the
correct types.
The optimization to avoid creating a HashSet if we're transitioning all
properties or had no existing transition is removed since now we're creating a
LonghandIdSet, which is cheap, and that was only a performance optimization.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/18763)
<!-- Reviewable:end -->
Change MediaExpressionValue::from_css_value to only accept pixels
MozReview-Commit-ID: Hn3twVa8xLo
<!-- Please describe your changes on the following line: -->
https://bugzilla.mozilla.org/show_bug.cgi?id=1404097https://reviewboard.mozilla.org/r/184746/
---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `__` with appropriate data: -->
- [ ] `./mach build -d` does not report any errors
- [ ] `./mach test-tidy` does not report any errors
- [ ] These changes fix #__ (github issue number if applicable).
<!-- Either: -->
- [ ] There are tests for these changes OR
- [X] These changes do not require tests because existing tests provide coverage.
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/18764)
<!-- Reviewable:end -->
This fixes the fishy TransitionProperty mapping which I complained about in my
previous refactor.
Turns out that those properties could only be longhands, and thus the expansion
we did before that (and which I removed) was correct.
This fixes the bug by moving back to the previous correct behavior but using the
correct types.
The optimization to avoid creating a HashSet if we're transitioning all
properties or had no existing transition is removed since now we're creating a
LonghandIdSet, which is cheap, and that was only a performance optimization.
Don't assume that inputs to compute_style_with_inputs have any rules.
It could be a text style, which never has any rules attached to it.
Fixes https://bugzilla.mozilla.org/show_bug.cgi?id=1406222
<!-- Please describe your changes on the following line: -->
---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `__` with appropriate data: -->
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
- [X] These changes fix https://bugzilla.mozilla.org/show_bug.cgi?id=1406222
<!-- Either: -->
- [ ] There are tests for these changes OR
- [ ] These changes do not require tests because _____
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/18767)
<!-- Reviewable:end -->
Upgrade to rustc 1.22.0-nightly (d7e73e4b1 2017-10-04)
This version fixes large regressions in incremental compilation performance.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/18760)
<!-- Reviewable:end -->
Use raw pointers instead of transmute for UnsafeFlow
This makes the code depend a bit less on rustc implementation details.
r? pcwalton
---
- [x] `./mach build -d` does not report any errors
- [x] `./mach test-tidy` does not report any errors
- [x] These changes do not require tests because they are refactoring only
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/18754)
<!-- Reviewable:end -->
Bump base64 to 0.6
This allows us to update hyper to 0.10.13 (closer to 0.11)
- [x] `./mach build -d` does not report any errors
- [x] `./mach test-tidy` does not report any errors
- [x] These changes do not require tests because they're just a version bump
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/18743)
<!-- Reviewable:end -->
Only generate structs file for the current build
This PR changes build_gecko.rs to only generate the `structs.rs` for the current build (rather than both), depending on whether `gecko_debug` feature is set.
The in-tree files are switched to use the previous release one, because that's what we currently use for stylo test and it is in general what we really care about. For this change, `gecko_debug` mode is removed from Servo CI (in `build-geckolib`) with the assumption that people general do that build locally for stylo development, so it is less likely to be broken than release.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/18753)
<!-- Reviewable:end -->
style: Iterate in the expected order in the custom_properties module.
In #18745, I replaced a few manual iterations over `index` with the iterator,
and it changed the behavior of `layout/style/test/test_variables_order.html`,
since it turns out that the iterator iterates right to left.
I think this is just an accident that happened due to inconsistencies in how we
were iterating over it, and that our behavior was inconsistent (since we
iterated rtl in some cases, but ltr in others seems like it'd be inconsistent
depending on the depth of the tree and different stuff).
This brings back the expected behavior again, and ensures we iterate over a
consistent order every time.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/18761)
<!-- Reviewable:end -->
Update web-platform-tests
This updates web-platform-tests to a46616a5b18e. It also removes our local copy of wptrunner in favour of the upstream one that we can modify and sync more easily. Carried from #17416.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/18756)
<!-- Reviewable:end -->
In #18745, I replaced a few manual iterations over `index` with the iterator,
and it changed the behavior of `layout/style/test/test_variables_order.html`,
since it turns out that the iterator iterates right to left.
I think this is just an accident that happened due to inconsistencies in how we
were iterating over it, and that our behavior was inconsistent (since we
iterated rtl in some cases, but ltr in others seems like it'd be inconsistent
depending on the depth of the tree and different stuff).
This brings back the expected behavior again, and ensures we iterate over a
consistent order every time.
style: Use the flattened tree parent to find the closest non-nac ancestor of an element.
This only matters for document level NAC like canvas custom content, in which
case otherwise we inherit from the document element (which is wrong).
Bug: 1405635
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/18752)
<!-- Reviewable:end -->
This only matters for document level NAC like canvas custom content, in which
case otherwise we inherit from the document element (which is wrong).
Bug: 1405635
If the first test selected by a testrunner thread had a CRASH expectation,
the testrunner believed that it needed to restart and would never actually
get around to running the test.
