Request for review: 8005895: Inefficient InstanceKlass field packing wasts memory
aleksey.shipilev at oracle.com
Thu Jan 10 10:52:12 PST 2013
I think we need to see if this should be fixed in lambda/lambda first.
Otherwise I'm OK with the either plan, preferring to solve this on the spot.
On 10.01.2013, at 22:46, Jiangli Zhou <jiangli.zhou at oracle.com> wrote:
> Hi Aleksey,
> Good question! I don't know why _has_default_methods was not in the enumerated flags in the first place. Looks like it can be handled that way. Although packing _has_default_methods into the _misc_flags would not yield any memory saving currently, but it may help in the future. Please let me know if you prefer packing _has_default_methods into _misc_flags as part of this change. Or, I can file a enhancement bug so we can keep track of it.
> On 01/10/2013 10:11 AM, Aleksey Shipilev wrote:
>> Why is _has_default_methods not in the enumerated flags there in the first place?
>> On 10.01.2013, at 21:57, Jiangli Zhou<jiangli.zhou at oracle.com> wrote:
>>> Please review the following trivial change for repacking the InstanceKlass fields to improve memory efficiency.
>>> The 'bool' typed variables, '_is_marked_dependent' and '_has_default_methods' were grouped with u2 typed fields. For both, there were 1-byte padding added on 32bit machine. We can pack those two boolean variables together to avoid the padding, then move one of the u2 field to be together with '_init_state' and '_reference_type'. That saves 4 bytes without any real code change.
More information about the hotspot-runtime-dev