Comment #2 on issue 3392 by [email protected]: Miscompilation of numeric
comparison
http://code.google.com/p/v8/issues/detail?id=3392
For background, we've observed that part of one of our massive GWT webapps
has been broken in recent versions of Chrome, which was finally tied down
and reduced to the example in the OP today (and then further reduced to the
test case in 328343004). Based on our investigations of which reductions
caused the bug to disappear, escape analysis seemed a plausible culprit (as
replacing the single-field object "d" with its field resulted in things
working), as did number representations (as on x86, the bug disappeared
when -4611686018427387904 was replaced with any value less than -(2^^30),
and on x64, it didn't seem to reproduce with -(2^^30), hence the value of
-(2^^62)).
You have no idea how pleased I am to see these suspicions confirmed and to
see a fix proposed - thankyou.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings
--
--
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/d/optout.
