Because it is inlined, it doesn't show up directly in crash reports and can require manually looking at the minidump to see what actually crashed, for instance:

https://bugs.chromium.org/p/chromium/issues/detail?id=897789#c5
https://bugs.chromium.org/p/chromium/issues/detail?id=893971#c16
https://bugs.chromium.org/p/chromium/issues/detail?id=919230#c3

We can fix the first problem by making sure there is an unoptimizable frame name in the callstack when triggering a crash.

The other issue is it would be nice to have more information on MOJO_RESULT_RESOURCE_EXHAUSTED failures, which has also been coming up with data pipe creation failures. If these are due to exceeding maximum handle counts, perhaps we should be dumping handle counts broken up by type, to get a better sense of where we are "leaking" handles?