RFR(S): 8064703: crash running specjvm98's javac following 8060252
roland.westrelin at oracle.com
Tue Dec 2 09:24:41 UTC 2014
Thanks for the reviews Vladimir & Vladimir.
> On Dec 1, 2014, at 7:31 PM, Vladimir Kozlov <vladimir.kozlov at oracle.com> wrote:
> On 12/1/14 10:05 AM, Roland Westrelin wrote:
>>> null_check() will "confuse" tightly_coupled_allocation(). Why you moved it?
>> I moved tightly_coupled_allocation() back where it was in 8 and before. So tightly_coupled_allocation() only succeeds if the null_check is a noop.
> Got it.
> Looks good.
>>> Vladimir k
>>> On 12/1/14 4:28 AM, Roland Westrelin wrote:
>>>> If an arraycopy is used to initialize a just allocated array, when possible we skip the array initialization and let arraycopy do the initialization. If an uncommon trap is triggered after the allocation and before the arraycopy, then an uninitialized array can escape the compiled method to the interpreter. This is fixed by making the uncommon trap resume execution at the allocation rather than the arraycopy and redo the allocation (Vladimir’s suggestion).
More information about the hotspot-compiler-dev