This bug was filed from the Socorro interface and is report bp-a84b5e92-c691-4a26-a4b1-b6b5b2130925. ============================================================= 0 libGLESv2.dll rx::VertexDataConverter<float,rx::WidenRule<0,4>,rx::ConversionRule<5126,0,0>,rx::DefaultVertexValues<float,0> >::convertArray(float const *,unsigned int,unsigned int,float *) gfx/angle/src/libglesv2/renderer/vertexconversion.h 1 libGLESv2.dll rx::VertexBuffer9::storeVertexAttributes(gl::VertexAttribute const &,int,int,int,unsigned int) gfx/angle/src/libglesv2/renderer/VertexBuffer9.cpp 2 libGLESv2.dll rx::VertexBufferInterface::storeVertexAttributes(gl::VertexAttribute const &,int,int,int,unsigned int *) gfx/angle/src/libglesv2/renderer/VertexBuffer.cpp 3 libGLESv2.dll rx::VertexDataManager::prepareVertexData(gl::VertexAttribute const * const,gl::ProgramBinary *,int,int,rx::TranslatedAttribute *,int) gfx/angle/src/libglesv2/renderer/VertexDataManager.cpp 4 libGLESv2.dll rx::Renderer9::applyVertexBuffer(gl::ProgramBinary *,gl::VertexAttribute * const,int,int,int) gfx/angle/src/libglesv2/renderer/Renderer9.cpp 5 libGLESv2.dll gl::Context::drawArrays(unsigned int,int,int,int) gfx/angle/src/libglesv2/Context.cpp 6 libGLESv2.dll glDrawArraysInstancedANGLE gfx/angle/src/libglesv2/libGLESv2.cpp 7 xul.dll mozilla::gl::GLContext::fDrawArraysInstanced(unsigned int,int,int,int) obj-firefox/dist/include/GLContext.h 8 xul.dll mozilla::WebGLContext::DrawArraysInstanced(unsigned int,int,int,int) content/canvas/src/WebGLContextVertices.cpp 9 xul.dll mozilla::dom::WebGLExtensionInstancedArraysBinding::drawArraysInstancedANGLE obj-firefox/dom/bindings/WebGLRenderingContextBinding.cpp 10 @0x8b3ec04 Frame #10 is some weird address in all the cases I looked at, more reports at https://crash-stats.mozilla.com/report/list?signature=rx%3A%3AVertexDataConverter%3Cfloat%2C%20rx%3A%3AWidenRule%3Cint%2C%204%3E%2C%20rx%3A%3AConversionRule%3Cint%2C%200%2C%20int%3E%2C%20rx%3A%3ADefaultVertexValues%3Cfloat%2C%20int%3E%20%3E%3A%3AconvertArray%28float%20const%2A%2C%20unsigned%20int%2C%20unsigned%20int%2C%20float%2A%29 This started spiking on Aurora with the uplift last week, apparently. It's #5 in Aurora 26 data from builds of the last 3 days right now.

26 has a new version of Angle. Jeff and Benoit, thoughts?

This should just be bug 916816, which is awaiting uplift approval for Aurora.

(In reply to Jeff Gilbert [:jgilbert] from comment #2) > This should just be bug 916816, which is awaiting uplift approval for Aurora. Makes sense, I was already wondering why the crash isn't also present on Nightly in the last few days.

Thanks!

This crash still gets reported but it seems to be very rare. We currently have 1 report in 50 Beta, 2 reports in 49 Release, and 6 reports in 48 Release. Considering the original crash volume I suspect the original issue is fixed but there remains an edge case or a different bug under the same signature. Based on this I am going to close this bug but please reopen it if you can provide a reproducible testcase.