RFR(M) 7199175: JSR 292: C1 needs patching when invokedynamic/invokehandle call site is not linked
vladimir.kozlov at oracle.com
Fri May 31 09:20:26 PDT 2013
On 5/31/13 3:28 AM, Roland Westrelin wrote:
> Thanks for reviewing this, Vladimir.
>>>> + develop(bool, C1PatchInvokeDynamic, true, \
>>>> Do we really want to turn if off? During development I'm sure it was helpful but after that?
>>> That's something John asked for.
>> I agree with Christian that this flag is weird. Why you would want to bailout even for testing?
> It's not a trivial change. So if tests that used to work start failing and this change is suspected, it will be as easy as running with -XX:-C1PatchInvokeDynamic to verify this change is the cause. I think it's a good idea to have this flag.
So you are saying that you are fixing performance issue only. Okay, may make it diagnostic then to be able use it with
>> Could you split some checks into 2 lines in c1_Runtime1.cpp? They are too long.
> Here is a new webrev in case you want to check I split the right checks.
This looks good.
More information about the hotspot-compiler-dev