New GCC build warning introduced in the last few days: { In file included from /mozilla-central/js/src/jsobjinlines.h:23:0, from /mozilla-central/js/src/vm/Shape-inl.h:29, from /mozilla-central/js/src/jsscriptinlines.h:21, from /mozilla-central/js/src/vm/ScopeObject-inl.h:12, from /mozilla-central/js/src/ion/BaselineFrame-inl.h:14, from /mozilla-central/js/src/vm/Stack-inl.h:19, from /mozilla-central/js/src/jsinferinlines.h:24, from /mozilla-central/js/src/ion/MIR.h:16, from /mozilla-central/js/src/ion/IonBuilder.h:13, from /mozilla-central/js/src/jsworkers.cpp:13: /mozilla-central/js/src/jstypedarray.h:277:28: warning: inline function 'static uint32_t js::TypedArray::type(JSObject*)' used but never defined [enabled by default] static inline uint32_t type(JSObject *obj); } hg bisect says this was introduced by this cset: changeset: 130152:fc9427895561 user: Brian Hackett <bhackett1024@gmail.com> date: Sat Apr 27 15:02:42 2013 -0600 summary: Bug 864214 - Generate MIR nodes for accessing statically known typed arrays on x86, r=luke,jandem.

Though I included the output for jsworkers.cpp in comment 0, this new warning gets spammed for numerous .cpp files in /js. (with a backtrace of the #include chain)

I actually get 506 lines of build warning spam from this, in GCC 4.8 (including the diagnostic #include paths and the printout of the line of the used-but-not-defined declaration). Attaching the full warning output shortly.

Fixed by Bug 889146 Part 11...? https://hg.mozilla.org/mozilla-central/diff/6f7ee0fd72e8/js/src/jstypedarray.h

Yeah, it looks like that definitely would've fixed this (by providing type()'s definition with the declaration). Resolving as FIXED by that bug. Thanks for closing the loop on this!