The navigation command is always executed on the top-level
browsing context. As such the check for a possible load
event should not be done for the currently selected frame.
At the same time the switch to frame call can be moved
to the parent process.
Differential Revision: https://phabricator.services.mozilla.com/D80619
get() is used a lot within Marionette for getters,
and as such makes it very hard to find the actual
"WebDriver:NavigateTo" implementation.
Differential Revision: https://phabricator.services.mozilla.com/D80617
Copy ensure_subprocess_env from mozbuild.util (not currently accessible from the
automation mozharness environment) to xpcshell harness, and use it.
Differential Revision: https://phabricator.services.mozilla.com/D84020
The navigation command is always executed on the top-level
browsing context. As such the check for a possible load
event should not be done for the currently selected frame.
At the same time the switch to frame call can be moved
to the parent process.
Differential Revision: https://phabricator.services.mozilla.com/D80619
get() is used a lot within Marionette for getters,
and as such makes it very hard to find the actual
"WebDriver:NavigateTo" implementation.
Differential Revision: https://phabricator.services.mozilla.com/D80617
SKIP_BMO_CHECK
The change was introduced in #24481, which assumed request.GET.get would
return a single value (though not documented explicitly, it's a very
reasonable assumption), but unfortunately it actually returns a list.
This caused widespread timeout in many referrer-policy tests, which were
not caught as affected tests.
This change fixes the immediate issue. #24614 will fix MultiDict to match
the intuition.
wpt-commit: 4ba6bcd3466f36dd4ca1a0c59d558d0555b009ac
wpt-pr: 24615
Automatic update from web-platform-tests
Revert "[LayoutNG] Move relative position calculation into NG."
This reverts commit 27880cb4012f7a84f7921c1bc3f023211eb032ea.
Reason for revert:
Suspect of blink_web_tests failures in a lot of bots, like this one:
https://ci.chromium.org/p/chromium/builders/ci/WebKit%20Win10/70261
Unexpected Failures:
* fast/css/relative-positioned-block-nested-with-inline-parent-dynamic.html
* fast/css/relative-positioned-block-nested-with-inline-parent.html
* paint/invalidation/compositing/composited-float-under-composited-inline-individual.html
* paint/invalidation/compositing/composited-float-under-composited-inline.html
* paint/invalidation/compositing/stacked-float-under-composited-inline.html
Original change's description:
> [LayoutNG] Move relative position calculation into NG.
>
> This patch moves the relative position calculation into NG. Previously
> we needed to manually account for the relative position of an object
> in many different places, overflow calculation, geometry APIs etc.
>
> This patch directly updates the child's offset (relative to its parent)
> to directly incorporate this additional offset.
>
> This calculation occurs within the NGBoxFragmentBuilder, and if an
> object is *placed* by LayoutNG...
> ...LayoutBoxModelObject::RelativePositionOffset returns an empty offset.
>
> This fixes a few small issues where previously we didn't resolve
> percentages against the correct available size when contained within
> an anonymous block.
>
> Bug: 1070822, 1058690
> Change-Id: I2c623b02cc5a4605d310596f5c60353a21236bff
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2214608
> Commit-Queue: Ian Kilpatrick <ikilpatrick@chromium.org>
> Reviewed-by: Xianzhu Wang <wangxianzhu@chromium.org>
> Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
> Reviewed-by: Koji Ishii <kojii@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#787991}
TBR=yosin@chromium.org,wangxianzhu@chromium.org,kojii@chromium.org,ikilpatrick@chromium.org,yosin@google.com,mstensho@chromium.org
Change-Id: I5ca67701222509693d3020fa068c279cef39ecbb
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 1070822, 1058690
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2296628
Reviewed-by: Jian Li <jianli@chromium.org>
Commit-Queue: Jian Li <jianli@chromium.org>
Cr-Commit-Position: refs/heads/master@{#788228}
--
wpt-commits: 65ace390982bb405ee10fbea9ba1c92a22199774
wpt-pr: 24592
Automatic update from web-platform-tests
Origin isolation: reorganize WPTs
This creates subfolders for certain broad classes of web platform tests,
and establishes a more uniform naming scheme for the tests.
Bug: 1042415
Change-Id: Ifc5b07f56880bec2e2a0963e0b6807d8de28988c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2290697
Commit-Queue: Domenic Denicola <domenic@chromium.org>
Reviewed-by: James MacLean <wjmaclean@chromium.org>
Cr-Commit-Position: refs/heads/master@{#788258}
--
wpt-commits: 2cc84007df785cc39ccd11f033b7a35a5efd2f03
wpt-pr: 24544
Automatic update from web-platform-tests
Test that MessagePort owners are current, not incumbent
Follows https://github.com/whatwg/html/pull/5728.
--
wpt-commits: 299084220865342716773d5116880a63711339d7
wpt-pr: 24566
Automatic update from web-platform-tests
Test that BroadcastChannel settings object is current, not incumbent
Follows https://github.com/whatwg/html/pull/5729.
--
wpt-commits: 74b3f3fcff249affad48bb1bfadf8877154a0f18
wpt-pr: 24568
Automatic update from web-platform-tests
Python 3: port tests in common/security-features [part 2] (#24481)
* Python 3: port tests in common/security-features [part 2]
* Encode path and improve header value type convert function
--
wpt-commits: 3c8965b973ca10f3cf5a6a817c87e32939e4cb6e
wpt-pr: 24481
Automatic update from web-platform-tests
Python 3: Use bytes for status (#24585)
--
wpt-commits: 0d905012224721a378c987778245bdc871644be9
wpt-pr: 24585
Automatic update from web-platform-tests
Don't skip already loaded custom fonts in FontFallbackIterator
When FontFallbackIterator iterates through a custom font, it checks
RangeSetContributesForHint() and if false, skips the custom font.
While the check was to introduced to prevent unnecesary loading of
fonts that don't load in new characters, it incorrectly blocks already
loaded custom fonts which can be used right away.
Therefore, this patch revises RangeSetContributesForHint() so that we
skip a font only when it's also a pending custom font.
Bug: 1101483
Change-Id: I04ed909d154b137b2bf45e712da199e6a5ddce7b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2278648
Commit-Queue: Xiaocheng Hu <xiaochengh@chromium.org>
Reviewed-by: Dominik Röttsches <drott@chromium.org>
Cr-Commit-Position: refs/heads/master@{#788038}
--
wpt-commits: 9f039866171222befc3c77beaae994829554bdf5
wpt-pr: 24479
Automatic update from web-platform-tests
[LayoutNG] Move relative position calculation into NG.
This patch moves the relative position calculation into NG. Previously
we needed to manually account for the relative position of an object
in many different places, overflow calculation, geometry APIs etc.
This patch directly updates the child's offset (relative to its parent)
to directly incorporate this additional offset.
This calculation occurs within the NGBoxFragmentBuilder, and if an
object is *placed* by LayoutNG...
...LayoutBoxModelObject::RelativePositionOffset returns an empty offset.
This fixes a few small issues where previously we didn't resolve
percentages against the correct available size when contained within
an anonymous block.
Bug: 1070822, 1058690
Change-Id: I2c623b02cc5a4605d310596f5c60353a21236bff
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2214608
Commit-Queue: Ian Kilpatrick <ikilpatrick@chromium.org>
Reviewed-by: Xianzhu Wang <wangxianzhu@chromium.org>
Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
Reviewed-by: Koji Ishii <kojii@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787991}
--
wpt-commits: 4b32cd3ad3d262854efc7628daacb8bc2db9f06f
wpt-pr: 24105
Automatic update from web-platform-tests
Clipboard API: Implement the HTML clipboard reader and writer.
This CL adds two subclasses: ClipboardHTMLReader and
ClipboardHTMLWriter. When HTML is read and written it is "sanitized"
which means that all script objects are removed from the html. They
are both entirely synchronous because it is impossible to use HTML
DOM nodes off of the main thread.
Design Doc: https://docs.google.com/document/d/1tzjfckWUKAAxEbKcUjg21uz7LFDMQ7ZDzfTTj3Lp93Y/edit
Bug: 931839
Change-Id: I29b26f1c721102005bcbb5911549a9546aed3006
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2243532
Commit-Queue: Dylan Sleeper <dsleeps@google.com>
Reviewed-by: Jarryd Goodman <jarrydg@chromium.org>
Reviewed-by: Darwin Huang <huangdarwin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787862}
--
wpt-commits: f1d19687599024a2219d35a6ab0fecb4128891ad
wpt-pr: 24175
Automatic update from web-platform-tests
Update test threshold for Mac 10.11
The threshold for the -100 Hz sine wave is apparently just a little
too tight for MacOS 10.11. Adjust the threshold slightly.
Bug: 1104535
Change-Id: I7102093e2641a748d07ca63c05e9e8e63c9f8dce
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2295211
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787847}
--
wpt-commits: 025147852df6e1e548866c99ff3ff66effb20c1a
wpt-pr: 24580
Automatic update from web-platform-tests
Implement putAll and add web platform test
This adds a simple putAll implementation that loops put. A basic web
platform test is added as well.
Change-Id: I8fcc25ffacb64d15855da921eb14f3770df28694
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2274217
Commit-Queue: Daniel Murphy <dmurph@chromium.org>
Reviewed-by: Victor Costan <pwnall@chromium.org>
Reviewed-by: Daniel Murphy <dmurph@chromium.org>
Reviewed-by: enne <enne@chromium.org>
Reviewed-by: Olivier Yiptong <oyiptong@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787815}
--
wpt-commits: 05c2f1c6fbb22e5c62cd6c83b0ae90880588ce2e
wpt-pr: 24493
Automatic update from web-platform-tests
TC: don't run print tests for WebKitGTK and Servo (#24576)
* print-reftests are currently not supported for this two browsers.
So we need to avoid triggering those runs on taskcluster for them to
avoid failing the whole run. This fixes: #24360 and #24257.
* This patch removes the print-reftest suite from the general block
and adds a new block for it and for the browsers that support this
tests on taskcluster (Chrome and Firefox).
--
wpt-commits: 149ce86ecb408e9df28e1f55c236ae4388e77cc0
wpt-pr: 24576
Automatic update from web-platform-tests
[tools] Fix some deprecation warnings (#24567)
Mostly from forgetting to use raw string literals and the
collections.abc move in Python 3
--
wpt-commits: 5418b374cc363c16267515c521524689c9b2d540
wpt-pr: 24567
Automatic update from web-platform-tests
Remove redundant XrInputSourceArray iterable test
XRInputSourceArray_Iterable was failing on chromium, since as a ".html"
and not a ".https.html" file, it was not loaded in a secure context, and
thus WebXr wasn't exposed (and the test API doesn't attempt to polyfill
itself in if navigator.xr doesn't exist). However, on closer inspection,
this is a redundant test case that is already covered by the idlharness.
Fixed: 1104347
Change-Id: Ica4d52578571627e18b9a007e5bb461cf3e34ceb
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2293227
Commit-Queue: Alexander Cooper <alcooper@chromium.org>
Commit-Queue: Brandon Jones <bajones@chromium.org>
Auto-Submit: Alexander Cooper <alcooper@chromium.org>
Reviewed-by: Brandon Jones <bajones@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787443}
--
wpt-commits: 64d6c9611cba2925370d643167f7b3e93b27fa15
wpt-pr: 24569
Automatic update from web-platform-tests
Report device pixel content box for SVG elements correctly
Currently the device pixel content box for SVG elements is reported as
the same as content box and border box. We change this to report
the device pixel content box size as the bounding box in device pixels.
Bug: 1088497
Change-Id: Iae78c49deb214b516792cefd751a09d57202f9c3
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2240033
Commit-Queue: Daniel Libby <dlibby@microsoft.com>
Reviewed-by: Philip Rogers <pdr@chromium.org>
Reviewed-by: Daniel Libby <dlibby@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#787411}
--
wpt-commits: b9d0a6d56c80282ac7e878e0a249e7ed5a56eb50
wpt-pr: 24096
Automatic update from web-platform-tests
Rename some manual dynamic import tests
This properly follows the filename pattern at https://web-platform-tests.org/writing-tests/manual.html to allow them to be identified as manual.
--
wpt-commits: 859527a3513029d2d529cd42fb5c9f65afe42d58
wpt-pr: 24565
Automatic update from web-platform-tests
Add missing test path for manifest during WPT syncing.
--
Merge loading and updating manifest states when syncing.
--
wpt-commits: 1075231e6c12d2f971fb4321d57db42a62f02216, 04210fcb5e8e84d0c04797a225d94c3a0fac1935
wpt-pr: 24559
Automatic update from web-platform-tests
Fix tools/ unit tests (#24557)
* [lint] Fix a type annotation
This was introduced in #24520 accidentally.
* Skip quic in pytest and coverage
quic/ contains Python 3-only code that leads to warnings when running
`pytest --cov` in Python 2.
* Mark a test as expected to fail on Windows+py3
See #24561 for details.
* Drive-by: change docker-worker schema URL
--
wpt-commits: 8cb1c14b2bac8725dad6dedfb2c8d4af9d1ffbc8
wpt-pr: 24557
Automatic update from web-platform-tests
Add some basic Oscillator tests
Test the output of the OscillatorNode for a sine wave at various
frequencies (including a negative frequency) against the exact
mathematical output. The low frequencies are intended to test
Chrome's interpolation algorithm.
Since we use sine waves or custom waves with a very few known
coefficients (without normalization), we know what the exact output
should be.
Bug: 1013118
Change-Id: Ie68ef5a8e6979aab70d0a4d4c24ee31810599f56
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2290693
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787297}
--
wpt-commits: 980dc0c7625921634185a5369c05506196102b53
wpt-pr: 24541
Automatic update from web-platform-tests
[scroll-animations] Integrate @scroll-timeline with animations
Until now it was possible to parse @scroll-timeline rules, but they
did not actually do anything. This CL performs the necessary
integration into StyleEngine and CSSAnimations (partially) to make
something happen.
The applicable StyleRuleScrollTimeline objects are maintained on
StyleEngine. When calculating the animation update, we now take
the animation-timeline property into account, and look for a
rule on StyleEngine corresponding to the specified timeline name.
This rule (if found) then causes a new ScrollTimeline to be created.
This is by no means a fully functional implementation, but it's a
start. The most notable omission is updating timelines, which is
currently not possible (crbug.com/1097053).
Other omissions, flaws, or postponed work are either tracked as
bugs (with TODOs that reference them) or as failing expectations.
This ensures that we don't miss anything before shipping.
Bug: 1074052
Change-Id: I04e75be1dc970ed2f542e60457f8db25bbb0fe7a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2253752
Commit-Queue: Anders Hartvoll Ruud <andruud@chromium.org>
Reviewed-by: Majid Valipour <majidvp@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787258}
--
wpt-commits: 5fff9da6c050f3d6a9a3d8bc9d9eff6308826e12
wpt-pr: 24252
Automatic update from web-platform-tests
Clipboard API: Implement clipboard-read and clipboard-write Feature Policy.
Complete and expose previously incomplete feature policy integration.
Also, use separate clipboard-read and clipboard-write feature policies,
as opposed to the previous single clipboard feature policy.
Previously, https://crrev.com/c/1913466 aimed to add Feature Policy
integration to the Clipboard API. Unfortunately, this integration wasn't
complete, so sites didn't have the option to allow cross-origin iframes
to access the clipboard at all after this change.
Complete the integration to allow sites to enable this feature policy.
Also, as feature and permission policies are being merged, use
clipboard-read and clipboard-write as feature policy names, so that the
feature policy and permission names match.
Verified by via automated WPTs, and also manually tested by running this
example[1], which fails without this CL, and succeeds with the CL.
Intent to prototype and ship: https://groups.google.com/a/chromium.org/d/msg/blink-dev/IDAapW6w4to/iSH1MXRFCgAJ
[1]: https://smart-flicker-baroness.glitch.me/
Bug: 1074489
Change-Id: I09125d74510ed6c868d84936df595d02326619b7
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2164248
Commit-Queue: Darwin Huang <huangdarwin@chromium.org>
Reviewed-by: Alex Russell <slightlyoff@chromium.org>
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Reviewed-by: Ian Clelland <iclelland@chromium.org>
Reviewed-by: Balazs Engedy <engedy@chromium.org>
Reviewed-by: Victor Costan <pwnall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786914}
--
wpt-commits: e2ebd58e4c8071639d939f6dba26ff5e8e7a8084
wpt-pr: 23727
Automatic update from web-platform-tests
Cleanup newline-removal in manifest_build.py (#24540)
This is a small follow-up to d7423b4ef835ac579b2aa6cbd0e88e47ed1c82cd,
which was an attempt at a hotfix that turned out to be the problem. To
tidy up the code, trim the sha at retrieval time and don't trim the tag
(which turned out not to have a newline in it).
--
wpt-commits: 7b2f44248d9955dd11663479a469926969c8884e
wpt-pr: 24540
Automatic update from web-platform-tests
dom: Generate iterator symbol for interfaces with indexed getters.
--
wpt-commits: c114ff9b2f781a671261d55ca977075d084a7ebf
wpt-pr: 24548
Automatic update from web-platform-tests
Construct StyleResolver in StyleEngine constructor.
Years ago, the StyleResolver was deleted and re-created for certain
style sheet changes. The current lifetime of the StyleResolver is that it
is created once on first EnsureStyleResolver and only freed when the
StyleEngine is disposed. Instead of checking if it needs to be
constructed every time we access it, create it as part of constructing
the StyleEngine.
We don't create it for documents without a Frame for which we should not
use the StyleResolver.
Change-Id: I1f1d2ef322d2599737f584191550f769dcf6828e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2270209
Commit-Queue: Rune Lillesveen <futhark@chromium.org>
Reviewed-by: Robert Flack <flackr@chromium.org>
Reviewed-by: Majid Valipour <majidvp@chromium.org>
Reviewed-by: Xiaocheng Hu <xiaochengh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787158}
--
wpt-commits: bfe92da84ad0ab55f1ee333f64bd5860e6da5ff5
wpt-pr: 24454
Automatic update from web-platform-tests
Python 3: port webappapis and cross-origin tests in html (#24409)
* Python 3: port webappapis and cross-origin tests in html
* Correct use of type on name argument for uuid5
--
wpt-commits: c44d7e9b7dbdecfe831e0feca8bd5ffcd02d424a
wpt-pr: 24409
Automatic update from web-platform-tests
[FlexNG] Fix % max-height definiteness
A replaced element's % max-height was being ignored when it was the
child of a flex item that acquired its definite height by stretching.
This was flaky though, it only broke when the image was loaded before
the flexbox had laid out. If the flexbox was laid out, _then_ the image
loaded and hence caused a relayout, the image would obey the
definiteness that was cached from the flex item's first layout.
Change-Id: I2b429aed111a1b800311287337f951c15fcc72f2
Fixed: 1092934
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2287370
Commit-Queue: David Grogan <dgrogan@chromium.org>
Reviewed-by: Christian Biesinger <cbiesinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787095}
--
wpt-commits: 8e1b09f5bdecd4e9709e2377d4344acef5f9002c
wpt-pr: 24527
Automatic update from web-platform-tests
[NativeFS] Update some tests to getFooHandle.
http://crrev.com/c/2285670 renamed getFile/getDirectory to getFileHandle
and getDirectoryHandle, this updates some tests that were missed in that
change.
Bug: 1103022
Change-Id: Id4b4a60677424cae59af7626e57acc1f6be4bd29
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2291252
Reviewed-by: Giovanni Ortuño Urquidi <ortuno@chromium.org>
Commit-Queue: Marijn Kruisselbrink <mek@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787072}
--
wpt-commits: 319817182b307f3a1637dfddc25d44eb6262a71c
wpt-pr: 24547
Automatic update from web-platform-tests
Re-add paint_invalidation_container in PrePaintTreeWalk.
This is needed in order to make sure to call
SetNeedsCheckRasterInvalidation() on all GraphicsLayers that
need raster invalidation after an invalidation of paint. To do
this involves finding out into which GraphicsLayer a LayoutObject paints.
We'll be able to remove it in favor of a dirty bit on
CompositingLayerAssigner when pre-paint is moved before compositing.
Bug: 1102733
Change-Id: I506d7e574e802502d9414ad67f14f0d4c072b184
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2290547
Commit-Queue: Chris Harrelson <chrishtr@chromium.org>
Commit-Queue: Xianzhu Wang <wangxianzhu@chromium.org>
Auto-Submit: Chris Harrelson <chrishtr@chromium.org>
Reviewed-by: Xianzhu Wang <wangxianzhu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#787058}
--
wpt-commits: e67f139c7c1bc14c85c7def4db56e3a1964624d1
wpt-pr: 24550
Automatic update from web-platform-tests
WebXR - anchors - add WPTs for more complete coverage of the API
Change-Id: I0a0858835a040161b352f465d68c068a8d4847bd
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2284094
Commit-Queue: Piotr Bialecki <bialpio@chromium.org>
Reviewed-by: Alexander Cooper <alcooper@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786984}
--
wpt-commits: ce7bfaec86b1c6c926bdb0f2150e33a3a1bf1f54
wpt-pr: 24496
Automatic update from web-platform-tests
[LayoutNG] Spanner break token and fragmentation improvements.
We were creating incorrect outgoing break tokens from column content
that was interrupted by two or more adjacent spanners. The outgoing
break token would point to the second spanner, rather than any column
content that would follow. To fix this, make sure that we don't
interrupt column layout until we have looked past all adjacent spanners,
to find the right place to resume column layout right away.
This gets especially more fun if we're nested inside another
fragmentation context, because then we may not be able to resume column
content layout in the same outer fragmentation. Therefore we need to
keep proper track of all the spanner break tokens.
Introduce a walker to streamline processing of the multicol parts
(spanners and/or column content), and thus make LayoutChildren() easier
to follow. We're now also better at handling parallel flows established
inside spanners.
Bug: 1066617, 829028
Change-Id: I7ad3f8f19f7258e2f8a588034c0002249efdaf20
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2288705
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786940}
--
wpt-commits: 0b58efec76d3f460dbb65d577e46ba4daf781427
wpt-pr: 24539
Automatic update from web-platform-tests
[LayoutNG] Handle spanner margins correctly across fragments.
Use the same AdjustMarginsForFragmentation() utility function as in
regular block layout, so that we handle block-end margins correctly,
i.e. apply them to only one fragment (which may not necessarily be the
last one, if something inside the spanner establishes a parallel flow).
Bug: 829028
Change-Id: I307037a23f960dc4ed4db2c126e41c59de8ae38e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2289134
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786948}
--
wpt-commits: 0dc51c3f662dcb23da4c60255abe0bd333c8e37e
wpt-pr: 24538
Automatic update from web-platform-tests
Origin isolation: add various navigation WPTs
This adds web platform tests covering various navigation scenarios for
origin isolation. The intention is to cover at least all the non-popup
navigation cases in isolated_origin_browsertest.cc. We also want to
approach the same coverage for navigation as we have for two-iframe
scenarios (i.e. the various parent-child1-child2 tests), but we avoid
doubling the size of the test suite by focusing on the more important
or tricky scenarios instead of just copying all of those.
Bug: 1042415
Change-Id: Ifb937f183af2610f348e4fea036d96aa0504ddab
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2285258
Commit-Queue: Domenic Denicola <domenic@chromium.org>
Reviewed-by: James MacLean <wjmaclean@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786945}
--
wpt-commits: 9d71a04aef33467e378c662d8b58ab1df6ff7719
wpt-pr: 24498
Automatic update from web-platform-tests
Add comment to border-width-011 to make it immediately obvious (#24542)
--
wpt-commits: fd21e757f91e0858a1a2b3870a3c1fb3a59dbd14
wpt-pr: 24542
Automatic update from web-platform-tests
Change document policy parser to use structured header dictionary
This CL is based on the spec change on
https://github.com/w3c/webappsec-feature-policy/pull/388/files?short_path=1a56774.
Since the parameter names are removed in this change, feature names
will have to change so that they remain meaningful. This will be
addressed in follow-up CLs.
Change-Id: Ic87dcb5bf26e440dd5041ae31d7b248e7d85d93a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2281104
Reviewed-by: Dmitry Gozman <dgozman@chromium.org>
Reviewed-by: Ian Clelland <iclelland@chromium.org>
Commit-Queue: Charlie Hu <chenleihu@google.com>
Cr-Commit-Position: refs/heads/master@{#786929}
--
wpt-commits: 78d5125d6c9e0e48accc6040499db5477d5c311b
wpt-pr: 24460
Automatic update from web-platform-tests
Reland: Abstract portal activation handling to avoid hack of making a fake promise.
This is somewhat verbose, but allows us to avoid the unhandled rejection
being visible to script without having to make this hack even worse.
Instead of actually calling activate(), this mirrors the spec structure
by simply calling to underlying logic, passing a resolver where it makes
sense and not (in this case, passing a substitute object) where it
doesn't.
The web test is added to SlowTests because it uses timeouts to check
that things don't happen, and the additional test makes the timeouts 6
seconds total -- which is the test runner timeout. Slow tests get 5x
longer before they are aborted. The alternative would have been to
shorten the timeouts in this test, but that may have made the test more
flaky.
Bug: 1102081
Change-Id: Ied9d26e32ee426bf0a135279a9063f104d0e6460
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2290231
Commit-Queue: Jeremy Roman <jbroman@chromium.org>
Commit-Queue: Adithya Srinivasan <adithyas@chromium.org>
Reviewed-by: Adithya Srinivasan <adithyas@chromium.org>
Auto-Submit: Jeremy Roman <jbroman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786912}
--
wpt-commits: 12cef047ffeaec190bbe1e6ff363dcd70ecdc176
wpt-pr: 24536
Automatic update from web-platform-tests
[LayoutNG] Only spanner siblings are considered adjacent.
The spec isn't crystal clear, but unconditionally considering spanners
that are siblings of another spanner's ancestors as adjacent is
definitely wrong. We need to resume column content layout inside the
parent block of the spanner, since it may have a fixed block-size,
block-end padding / border, or anything else that we'd just lose
otherwise.
This breaks one existing (non-WPT) test, but I'm not convinced that the
test is valid. Filed crbug.com/1100849 for it.
It also fixes 3 WPT tests. They were all quite complex, so I added a
couple of simple ones, that demonstrate this fix more to the point.
Bug: 829028
Change-Id: I9d55153f859dd84a5019e7d4dd09453f7f9ee1e4
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2290150
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Morten Stenshorne <mstensho@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786841}
--
wpt-commits: 90c4db31dd07db9c6009cdbc0358f460099a15fb
wpt-pr: 24534
Automatic update from web-platform-tests
[Taskcluster] Only schedule the sink-task for pull requests (#24511)
We don't need the sink-task for master pushes/etc, and this is a quick
workaround for the fact that Taskcluster doesn't support >100 data
dependencies (and we now schedule more than that for daily epoch runs!)
In order to reliably test this behaviour, we now also ensure the ordering
of tasks is deterministic.
Co-authored-by: Robert Ma <robertma@chromium.org>
--
wpt-commits: 821f177ada8d57dcec1f07327245c0a870c5758f
wpt-pr: 24511
Automatic update from web-platform-tests
[LayoutNG] Only apply block-end margins where the block actually ends.
If we reached the end of a block in an earlier fragmentainer, but
there's overflowing content, we'll revisit this block in subsequent
fragmentainers as long as child content needs it. Repeating the
block-end margins in these fragmentainers would be wrong, though.
Use AdjustForFragmentation() only for margins, and rename it to
AdjustMarginsForFragmentation().
The fieldset layout algorithm was using this one for borders/padding,
while block and multicol layout called
AdjustBorderScrollbarPaddingForFragmentation() in container_builder_.
Become more uniform by calling that method for all in
SetupFragmentBuilderForFragmentation() (via the base layout algorithm
constructor). Since this depends on the initial fragment geometry being
set, also do that in the base layout algorithm constructor.
Bug: 829028
Change-Id: I6808d023e4e79ab489bdb3a48e3c6c47cf9db6c6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2287313
Reviewed-by: Alison Maher <almaher@microsoft.com>
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Ian Kilpatrick <ikilpatrick@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786791}
--
wpt-commits: 931e5b7e2fe9818a7e15caa6378fabb00dbbca6a
wpt-pr: 24522
Automatic update from web-platform-tests
[WPT] Introduce a test-only-api helper script
including a helper function to load Mojo JS bindings. This will enable
us to have better control over loading of *.mojom.js (notably disable
automatic dependency loading) and reduce code duplication.
Refactor WebXR tests to use this helper script, and add missing
dependencies that were previously auto loaded. (WebUSB & WebBluetooth
changes to follow.)
Note that upstream WPT still does not have *.mojom.js available so tests
will continue to fail there (but with a clearer error). Test results on
Chromium waterfall should not change.
Bug: 1094512
Change-Id: If660c4788c185bc7baf9ce6edbb691333e509d4a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2276652
Reviewed-by: Stephen McGruer <smcgruer@chromium.org>
Reviewed-by: Alexander Cooper <alcooper@chromium.org>
Commit-Queue: Robert Ma <robertma@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786746}
--
wpt-commits: 8d7104310d9fdb1b2eb6995219c31489d25010b5
wpt-pr: 24400
Automatic update from web-platform-tests
Actually remove six from Dockerfile
With #24484 removing the dependency on system six, we are now able to
remove it from the Docker image, which was introduced in #23733.
Removing this has the benefit of avoiding introducing accidental
dependency.
--
wpt-commits: dea212bda34384982165966b6085dc9f4ec20858
wpt-pr: 24514
Automatic update from web-platform-tests
Remove inaccurate assert (#24530)
The test and its title are right, but the assert doesn't describe it
well. Dropping it.
--
wpt-commits: 40ef02d00e6e4209a8c48d59f56952ad7654731c
wpt-pr: 24530
Automatic update from web-platform-tests
[css-text-3] Add letter-spacing tests written last October (#24516)
r=frivoal
--
wpt-commits: 3e6a569c8126d6782193ffec32fd501aa038c408
wpt-pr: 24516
Automatic update from web-platform-tests
Fix CSS typed OM CSSSkewX and CSSSkewY toMatrix method to use degrees instead of radians.
The toMatrix method of a CSSTransformComponent should return a DOMMatrix
representing the object transformation.
https://www.w3.org/TR/css-typed-om-1/#dom-csstransformcomponent-tomatrix
The skewXSelf and skewYSelf methods used to apply a transformation on a
matrix must be provided with an angle parameter expressed in degrees.
https://www.w3.org/TR/geometry-1/#dom-dommatrix-skewxselfhttps://www.w3.org/TR/geometry-1/#dom-dommatrix-skewyself
Since the angles passed to the skewXSelf and skewYSelf were in radians,
DOMMatrix returned by the toMatrix method of those two components was incorrect.
R=alancutter@chromium.org
Change-Id: I5edc1cdf1493184584154723508c6e20c212ac92
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2242008
Commit-Queue: Alan Cutter <alancutter@chromium.org>
Reviewed-by: Alan Cutter <alancutter@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786632}
--
wpt-commits: f8d07d1a6788da6efee2913d5768528de28c9b57
wpt-pr: 24418
Automatic update from web-platform-tests
Revert "Do not expose media devices before a successful getUserMedia()"
This reverts commit d64e736aa8c7976b12b33f92472f7950bf0215b9.
Reason for revert: Caused chrome_all_tast_tests / camera.CCAUISanity.fake failures; e.g. https://ci.chromium.org/p/chrome/builders/ci/chromeos-betty-pi-arc-chrome/7680
Original change's description:
> Do not expose media devices before a successful getUserMedia()
>
> This is a spec-compliance change.
>
> This CL also adds some improvements to the blink MediaDevices unit test.
>
> Bug: 1101860
> Change-Id: If2e7824df3b3dbd3cdb80f4bcf6b5c417d6f0ef2
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2279808
> Commit-Queue: Guido Urdaneta <guidou@chromium.org>
> Reviewed-by: Harald Alvestrand <hta@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#786253}
TBR=hta@chromium.org,guidou@chromium.org
Change-Id: I41b676c7d3e437c06d23f2b46270453f1cc81670
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 1101860
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2288282
Reviewed-by: Mohsen Izadi <mohsen@chromium.org>
Commit-Queue: Mohsen Izadi <mohsen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786615}
--
wpt-commits: 66a61640ed3a93fa0ddaf2aabe07193a17066b82
wpt-pr: 24528
Automatic update from web-platform-tests
Revert "Abstract portal activation handling to avoid hack of making a fake promise."
This reverts commit fdc228c7acab87083507a1276f9d8f786966f5d6.
Reason for revert: portal-activate-default.html test timeout failure
https://ci.chromium.org/p/chromium/builders/ci/Mac10.10%20Tests/55803
Original change's description:
> Abstract portal activation handling to avoid hack of making a fake promise.
>
> This is somewhat verbose, but allows us to avoid the unhandled rejection
> being visible to script without having to make this hack even worse.
>
> Instead of actually calling activate(), this mirrors the spec structure
> by simply calling to underlying logic, passing a resolver where it makes
> sense and not (in this case, passing a substitute object) where it
> doesn't.
>
> Bug: 1102081
> Change-Id: I0eecf253928bfce39afcddccf54e9bc6f6c0d549
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2285340
> Reviewed-by: Adithya Srinivasan <adithyas@chromium.org>
> Commit-Queue: Jeremy Roman <jbroman@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#786480}
TBR=jbroman@chromium.org,adithyas@chromium.org
Change-Id: I00da9eee8625ca5da3a09efb7078e22a277dbfd8
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 1102081
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2288712
Reviewed-by: Keishi Hattori <keishi@chromium.org>
Commit-Queue: Keishi Hattori <keishi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#786596}
--
wpt-commits: 3c4df6d40bcb03edae052d0a37dc14291706c6a7
wpt-pr: 24526