jjgriego

(^ rebase and fix typo in comment)

For the benefit of reviewers, the differences between this and the previous PR of this patch are visible as the two commits ending here: https://github.com/jjgriego/WebKit/commit/1ba3b813b5b370ee7e202f9601d652c2851fa032

Thanks for an initial review, Saam. Looking forward to talking more about the approach--I've kept things high level for now but agreed that the approach esp. w/ 64-bit arithmetic is...

Failing armv7 wasm test is also broken on trunk, but we're running in more modes now, so EWS is complaining

a few notes after some looking at the places we talked about --- (re: discussion below about stubbing missing instructions out in e.g. B3Special) I've taken a look and actually...

OK, after much agonizing: rebased (including over the GC arrays support) + this revision includes the split AirIRGenerator, which should hopefully make both reviewing easier and tamps down substantially on...

Hi! Thanks for taking a look: > This pass is the lowest-tier allocator, which has the goal of reducing latency, correct? So if this doesn't make this code slower to...

test failures are upstream wasm failures happening in new modes thanks to the new tier support; should be fixed by https://github.com/WebKit/WebKit/pull/6467 when that lands