[wpt-sync] Sync PR 20571 - Fix XRFrame.getPose matrix math
Categories
(Testing :: web-platform-tests, task, P4)
Tracking
(firefox73 fixed)
Tracking | Status | |
---|---|---|
firefox73 | --- | fixed |
People
(Reporter: wpt-sync, Unassigned)
References
()
Details
(Whiteboard: [wptsync downstream])
Sync web-platform-tests PR 20571 into mozilla-central (this bug is closed when the sync is complete).
PR: https://github.com/web-platform-tests/wpt/pull/20571
Details from upstream follow.
Alex Cooper <alcooper@chromium.org> wrote:
Fix XRFrame.getPose matrix math
There were a few issues here. Most of them either cancelled each other
out or (due to overrides) weren't hit.
- The comment above XRFrame::getPose was incorrect
- This has now been updated, and the parameters were renamed to better
match the spec, and hopefully prevent further confusion- XRSpace::getPose was returning other_from_base and not base_from_other
- TargetRaySpace and GripSpace actually have overrides that do the
right thing in this case, and so were unaffected.- XRReferenceSpace did call up to this, but mostly cancelled out this
inversion due to the next issue.- XRReferenceSpace::MojoFromSpace was returning SpaceFromMojo
- This caused the error above to not be as noticed.
As a result, the xrFrame_getPose test has had it's matrix updated and a
new test has been added which should hopefully help catch this.Note that if a grip or target space is supplied as the first parameter
to getPose (the common case), the math is being done correctly.
Also note that the two inversions for references spaces also mostly
cancels each other right.
The main issue that was affected is if a reference space with an origin
offset was supplied as the first parameter to xrFrame.getPoseFixed:1030049
Change-Id: I0184607dbd0fef991c382289e1c764f9a1b336a6
Reviewed-on: https://chromium-review.googlesource.com/1947629
WPT-Export-Revision: d9e65007fa7f7e752bec49ff3c9c27d2830af3df
Assignee | ||
Comment 1•6 years ago
|
||
Assignee | ||
Comment 2•6 years ago
|
||
GitHub CI Results
wpt.fyi PR Results Base Results
Ran 1 tests and 2 subtests
Firefox
FAIL : 1
ERROR : 1
NOTRUN: 1
Chrome
FAIL : 1
ERROR : 1
NOTRUN: 1
Safari
FAIL : 1
ERROR : 1
NOTRUN: 1
Comment 4•6 years ago
|
||
bugherder |
Description
•