The style candidate cache had regressed a few times (see #12534), and my
intuition is that being able to disable all style sharing with a single rule in
the page is really unfortunate.
This commit redesigns the style sharing cache in order to be a optimistic cache,
but then reject candidates if they match different sibling-affecting selectors
in the page, for example.
So far the numbers have improved, but not so much as I'd wanted (~10%/20% of
non-incremental restyling time in general). The current implementation is really
dumb though (we recompute and re-match a lot of stuff), so we should be able to
optimise it quite a bit.
I have different ideas for improving it (that may or may not work), apart of the
low-hanging fruit like don't re-matching candidates all the time but I have to
measure the real impact.
Also, I need to verify it against try.
Fix a cached style cascade bug that only manifested in sequential mode
When copying cached styles, keep the `writing_mode` up to date.
---
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
- [X] These changes fix#11818 (github issue number if applicable).
- [X] There are tests for these changes
EDIT: The test is now working. I ran it with the first commit (the actual fix) reverted and it failed.
<!-- 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/12839)
<!-- Reviewable:end -->
Add `--version` flag
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy --faster` does not report any errors
- [X] These changes fix#11241 (github issue number if applicable).
Either:
- [ ] There are tests for these changes OR
- [ ] These changes do not require tests because _____
Not 100% sure of a good way to test this, so I'm submitting as is for feedback. Manually testing it appears to work fine.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="35" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/11262)
<!-- Reviewable:end -->
This commit adds the `--profiler-trace-path` flag. When combined with `-p` to
enable profiling, it dumps a profile as a self-contained HTML file to the given
path. The profile visualizes the traced operations as a gant-chart style
timeline.
In rust-url 1.0 the `Url` struct is going to have private fields, and there
is no way to to create an aribitrary one without going through the parser.
The plugin never had a clear demonstrated performance benefit,
it was made mostly because it was possible and relatively easy at the time.
Added --soft-fail option.
At the moment, wptrunner always runs the WPT tests with `servo --hard-fail`. To test hardening, we need to switch off `--hard-fail`. This PR introduces a `--soft-fail` option to do that.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="35" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/10420)
<!-- Reviewable:end -->
Skip printing the backtrace for RecvError/SendError
We currently get tons of useless backtraces clogging up the output when we have a panic cascade. This adds a handler that outputs a single line when a thread panics due to a sender or receiver hanging up, since this is almost always due to a panic cascade.
We could add a commandline arg that gets us back the old behavior, though I'm not sure if this is necessary.
r? @asajeffrey
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="35" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/10329)
<!-- Reviewable:end -->
Load prefs.json from profile-dir if --profile-dir is specified at launch
In response to #10098
Tries to load `prefs.json` from the profile-dir and merge them into the preferences if `--profile-dir` is specified at launch. The profile-dir preferences take precedence over the default preferences, but command line preferences still take precedence over everything.
Also adds some tests for `prefs.rs`. These rely on the contents of `resources/prefs.json` (at least `test_get_set_reset_extend()` does), so they may need to be re-worked a bit.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="35" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/10114)
<!-- Reviewable:end -->
Allow setting preferences to false in WPT tests
First patch to servo - apologies if I did something stupid :)
This is a fix for #10161. I have squashed the commits into one.
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="35" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/10204)
<!-- Reviewable:end -->
Parallel display list construction hasn't been shown to give any
performance gains. It is also incompatible with the current flat display
list implementation. Once flat display lists have landed, we can explore
possible benefits of parallel construction once again.
WebRender is an experimental GPU accelerated rendering backend for Servo.
The WebRender backend can be specified by running Servo with the -w option (otherwise the default rendering backend will be used).
WebRender has many bugs, and missing features - but it is usable to browse most websites - please report any WebRender specific rendering bugs you encounter!
