Open Bug 851078 Opened 11 years ago Updated 2 years ago

Various D2D reftests fail on new test slave hardware

Categories

(Core :: Graphics, defect)

x86_64
Windows 7
defect

Tracking

()

People

(Reporter: jimm, Unassigned)

References

Details

Spin off from bug 848936. We annotated most of these with min/max pixel differences, but a few were simply disabled due to hard failures. All of these fail due to rounding errors in color calculations. Bas felt this was the result of flawed drivers.

skip-if(B2G) fuzzy-if(d2d,1,10000) == dynamic-filter-contents-01a.svg dynamic-filter-contents-01-ref.svg
skip-if(B2G) fuzzy-if(d2d,1,10000) == dynamic-filter-contents-01b.svg dynamic-filter-contents-01-ref.svg
skip-if(d2d) == opacity-and-gradient-02.svg opacity-and-gradient-02-ref.svg

Note these currently fail on Win8 only because that is the only platform we are running on new test hardware. The plan is to migrate all Win tests to the new hardware in time so they will likely fail on other os as well when that happens.
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.