1337612 - stylo: non-fatal assertion why should we have flushed style again?: 'mPresShell && currentGeneration == mPresShell->GetPresContext()->GetRestyleGeneration()'

Status: RESOLVED WORKSFORME

(Core :: CSS Parsing and Computation, defect, P3)

Description

[Xidorn Quan [:xidorn] UTC+10]

There are several mochitests trigger this assertion with the following stack:
###!!! ASSERTION: why should we have flushed style again?: 'mPresShell && currentGeneration == mPresShell->GetPresContext()->GetRestyleGeneration()', file c:/mozilla-source/stylo/layout/style/nsComputedDOMStyle.cpp, line 898
#01: nsComputedDOMStyle::GetPropertyCSSValue (c:\mozilla-source\stylo\layout\style\nscomputeddomstyle.cpp:976)
#02: nsComputedDOMStyle::GetPropertyValue (c:\mozilla-source\stylo\layout\style\nscomputeddomstyle.cpp:402)
#03: mozilla::dom::CSSStyleDeclarationBinding::getPropertyValue (c:\mozilla-source\stylo\obj-firefox-stylo\dom\bindings\cssstyledeclarationbinding.cpp:165)
#04: mozilla::dom::GenericBindingMethod (c:\mozilla-source\stylo\dom\bindings\bindingutils.cpp:2959)
#05: js::CallJSNative (c:\mozilla-source\stylo\js\src\jscntxtinlines.h:239)
#06: js::InternalCallOrConstruct (c:\mozilla-source\stylo\js\src\vm\interpreter.cpp:460)
#07: js::jit::DoCallFallback (c:\mozilla-source\stylo\js\src\jit\baselineic.cpp:3690)
#08: ??? (???:???)
#09: ??? (???:???)

Looks like it happens when getting property value from computed style.

One example is "./mach mochitest --disable-e10s layout/style/test/test_bug437915.html"

Comment 1

[Xidorn Quan [:xidorn] UTC+10]

layout/style/test/test_bug1203766.html
layout/style/test/test_bug437915.html
layout/style/test/test_computed_style.html
layout/style/test/test_inherit_computation.html