RFR (S) : 8022441

David Chase david.r.chase at oracle.com
Fri Aug 9 06:51:26 PDT 2013


Porting this intrinsic to Graal, Doug Simon noticed some places where things were not right
in the interpreted version of the intrinsic, though in ways that currently slip past our existing
tests (I verified this by single-stepping the interpreter right over setting SP to 8, and which
was later corrected -- at least for my run -- by subsequent code).  Three flaws:
1) botched SP restore
2) wrong registers (not conforming to convention) for Windows
3) rare corner case for negative numbers

correct problems

hand-verified by me;
report of correct behavior from Doug Simon;
JPRT focussed on the CRC test;
Local jtreg on the CRC test, interpreted and not.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 163 bytes
Desc: Message signed with OpenPGP using GPGMail
Url : http://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/attachments/20130809/8e88e746/signature.asc 

More information about the hotspot-compiler-dev mailing list