Request for reviews (XL): 6829187: compiler optimizations required for JSR 292
Christian.Thalinger at Sun.COM
Thu Dec 3 07:04:03 PST 2009
On Wed, 2009-12-02 at 09:54 -0800, Tom Rodriguez wrote:
> Shouldn't the code in lcm.cpp be guarded by a check for whether the
> call is an invokedynamic? You're making it appear to always be killed
> which isn't true.
> It could also use a comment explaining what it's doing.
> MachCallJavaNode::in_RegMask could also use a comment.
> Otherwise this looks good.
Here is the updated webrev:
More information about the hotspot-compiler-dev