> [task 2017-08-25T15:21:54.809074Z] 15:21:54 INFO - /home/worker/workspace/build/src/obj-firefox/dist/include/mozilla/dom/BindingUtils.h:254:1: error: inlining failed in call to always_inline 'nsresult mozilla::dom::binding_detail::UnwrapObjectInternal(V&, U&, mozilla::dom::prototypes::ID, uint32_t) [with T = nsGlobalWindow; bool mayBeWrapper = false; U = nsGlobalWindow*; V = JSObject*]': function not considered for inlining > [task 2017-08-25T15:21:54.809454Z] 15:21:54 INFO - UnwrapObjectInternal(V& obj, U& value, prototypes::ID protoID, > [task 2017-08-25T15:21:54.809839Z] 15:21:54 INFO - ^~~~~~~~~~~~~~~~~~~~ > [task 2017-08-25T15:21:52.352383Z] 15:21:52 INFO - /home/worker/workspace/build/src/obj-firefox/dist/include/mozilla/dom/BindingUtils.h:254:1: error: inlining failed in call to always_inline 'nsresult mozilla::dom::binding_detail::UnwrapObjectInternal(V&, U&, mozilla::dom::prototypes::ID, uint32_t) [with T = mozilla::dom::Exception; bool mayBeWrapper = false; U = mozilla::dom::Exception*; V = JSObject*]': function not considered for inlining > [task 2017-08-25T15:21:52.352441Z] 15:21:52 INFO - UnwrapObjectInternal(V& obj, U& value, prototypes::ID protoID, > [task 2017-08-25T15:21:52.352480Z] 15:21:52 INFO - ^~~~~~~~~~~~~~~~~~~~

build-gcc hashes for 6.4 #4715f02413f8a91d02d967521c084990c99ce1a671b8a450a80fbd4245f4b728 gcc-6.4.0.tar.gz #8ceebbf4d9a81afa2b4449113cee4b7cb14a687d7a549a963deb5e2a41458b6b isl-0.15.tar.bz2

I feel like Boris would like to know about this.

Sure. We want this function to be inlined. We are telling GCC so. It sounds like it decides it shouldn't inline it for some reason (why?) and then fails the build instead of silently not inlining it (which is good). But the thing that needs to be fixed is the "decides it shouldn't inline it" bit...

I have a successful build with 6.4 and this did not need fixing, so I'm going to close it. https://treeherder.mozilla.org/#/jobs?repo=try&revision=9a7aca7e2f2cec7aea6e6b28a797e20276bf2979