Not LGTM, adding Ben as a reviewer.
Ben spent several hours making sure that OSR code is _not_ installed in
JSFunction objects. This change completely defeats and breaks that
invariant, by
bypassing the choking point that was introduced. I agree that we might want
to
cache OSR code, but we don't want to install it in JSFunction objects.
https://codereview.chromium.org/100613004/diff/20001/src/code-stubs-hydrogen.cc
File src/code-stubs-hydrogen.cc (right):
https://codereview.chromium.org/100613004/diff/20001/src/code-stubs-hydrogen.cc#newcode1195
src/code-stubs-hydrogen.cc:1195: // case, it's probably a good idea to
just reuse that OSR code right now.
I strongly disagree that this is a good idea! Once Ben's work is done,
it will be incorrect to do that. This will just take whatever code is in
the optimized_map and smash it into the JSFunction. OSR code should not
be installed into JSFunctions at all.
https://codereview.chromium.org/100613004/
--
--
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
---
You received this message because you are subscribed to the Google Groups "v8-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.
