RFR (S) JDK-8139651 JDK-8139651, ConcurrentG1Refine uses ints for many of its members that should be unsigned types
Tom Benson
tom.benson at oracle.com
Fri Feb 19 18:24:22 UTC 2016
Correction:
>> Also, I notice you didn't change the max_jint I mentioned in
dirtyCardQueue.cpp:
>>
>> 329 _max_completed_queue = max_jint;
>>
>> Was that by design or by accident? 8^)
> accident! I'll fix it. Should it be SIZE_MAX?
I just noticed that this conflicts with Kim's advice that
_max_completed_queue can't be unsigned. So it will need to stay max_jint.
Tom
On 2/19/2016 1:16 PM, Tom Benson wrote:
> Hi Joe,
>
> On 2/19/2016 11:48 AM, Joseph Provino wrote:
>>
>>
>> On 2/19/2016 11:25 AM, Tom Benson wrote:
>>> Hi Joe,
>>> In g1_globals.hpp, you changed G1ConcRefinementServiceIntervalMillis
>>> to size_t, along with the zone sizes and threshold. I think that
>>> one should be uintx, since it's definitely a count, not size/length.
>> I'll fix that.
>
> Well, not to nit-pick 8^), but your new webrev reverted this to
> intx, but I think it would make a nice uintx , as I said. As long as
> you're there. Or leave it, your call.
>
>>>
>>> Also, I notice you didn't change the max_jint I mentioned in
>>> dirtyCardQueue.cpp:
>>>
>>> 329 _max_completed_queue = max_jint;
>>>
>>> Was that by design or by accident? 8^)
>> accident! I'll fix it. Should it be SIZE_MAX?
>>
>
> I think so, and I see that's what you did.
>
> Tom
>
>> joe
>>
>>>
>>> Aside from that, looks OK to me.
>>> Tom
>>>
>>>
>>> On 2/19/2016 11:05 AM, Joseph Provino wrote:
>>>> New webrev is here:
>>>> http://cr.openjdk.java.net/~jprovino/8139651/webrev.01
>>>>
>>>> thanks.
>>>>
>>>> joe
>>>>
>>>> On 2/18/2016 1:44 PM, Joseph Provino wrote:
>>>>>
>>>>>
>>>>> On 2/12/2016 6:49 PM, Kim Barrett wrote:
>>>>>>> On Feb 11, 2016, at 6:58 PM, Joseph Provino
>>>>>>> <joseph.provino at oracle.com> wrote:
>>>>>>>
>>>>>>> Please review this small change from "int" to "size_t".
>>>>>>>
>>>>>>> CR: https://bugs.openjdk.java.net/browse/JDK-8139651
>>>>>>>
>>>>>>> webrev: http://cr.openjdk.java.net/~jprovino/8139651/webrev.00
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>>
>>>>>> Copyrights need updating.
>>>>> Okay.
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>>
>>>>>> I think the types of the command line options should be changed
>>>>>> to be
>>>>>> consistent with their associated concurrent refinement values, e.g.
>>>>>> G1ConcRefinementGreenZone should have the same type as
>>>>>> ConcurrentG1Refine::_green_zone and the associated accessor
>>>>>> functions.
>>>>>>
>>>>>> This would eliminate the need for some explicit type specifications
>>>>>> for MIN2/MAX2 calls and such.
>>>>> Sounds good.
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>>
>>>>>> I think some of these numbers are semantically not so much
>>>>>> "sizes" as
>>>>>> "counts", and wonder if such might be better typed as uint or uintx
>>>>>> rather than size_t.
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>> src/share/vm/gc/g1/concurrentG1RefineThread.cpp
>>>>>> 70 _threshold = MIN2<size_t>(cg1r()->thread_threshold_step()
>>>>>> * (_worker_id + 1) + cg1r()->green_zone(),
>>>>>> 73 _deactivation_threshold = MAX2<size_t>(_threshold -
>>>>>> cg1r()->thread_threshold_step(), cg1r()->green_zone());
>>>>>>
>>>>>> Are the explicit types for MIN2/MAX2 still needed here? I think all
>>>>>> the argument types are the same.
>>>>> I'll fix that.
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>> src/share/vm/gc/g1/concurrentG1RefineThread.cpp
>>>>>> 138 size_t curr_buffer_num =
>>>>>> (size_t)dcqs.completed_buffers_num();
>>>>>>
>>>>>> Identity cast.
>>>>> Will fix.
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>> src/share/vm/gc/g1/dirtyCardQueue.cpp
>>>>>> 214 if ((size_t)_n_completed_buffers <= stop_at) {
>>>>>>
>>>>>> Identity cast.
>>>>> Will fix.
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>> src/share/vm/gc/g1/dirtyCardQueue.cpp
>>>>>> 222 if (_completed_buffers_head == NULL)
>>>>>> 223 _completed_buffers_tail = NULL;
>>>>>>
>>>>>> [preexisting]
>>>>>>
>>>>>> Please add braces around the "then" clause.
>>>>> okay.
>>>>>>
>>>>>> Perhaps better would be to change
>>>>>>
>>>>>> 222 if (_completed_buffers_head == NULL)
>>>>>> 223 _completed_buffers_tail = NULL;
>>>>>> 224 assert(_n_completed_buffers > 0, "Invariant");
>>>>>> 225 _n_completed_buffers--;
>>>>>>
>>>>>> =>
>>>>>>
>>>>>> assert(_n_completed_buffers > 0, "Invariant");
>>>>>> _n_completed_buffers--;
>>>>>> if (_completed_buffers_head == NULL) {
>>>>>> assert(_n_completed_buffers == 0, "Invariant");
>>>>>> _completed_buffers_tail = NULL;
>>>>>> }
>>>>> Okay.
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>> src/share/vm/gc/g1/g1RemSet.cpp
>>>>>> 294 size_t into_cset_n_buffers =
>>>>>> into_cset_dcqs.completed_buffers_num();
>>>>>>
>>>>>> Unused variable can just be removed.
>>>>> okay.
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>> src/share/vm/gc/g1/ptrQueue.hpp
>>>>>> 220 size_t _process_completed_threshold;
>>>>>> 242 size_t _max_completed_queue;
>>>>>>
>>>>>> These variables cannot presently be changed to an unsigned type. A
>>>>>> negative value is treated specially for both of them, as a flag that
>>>>>> turns off associated behavior. A negative _max_completed_queue
>>>>>> means
>>>>>> there is no maximum. A negative _process_completed_threshold means
>>>>>> never trigger concurrent processing of completed buffers. I think
>>>>>> the
>>>>>> -1 special values (converted to size_t) end up working and providing
>>>>>> the desired behavior by accident, which would explain why these
>>>>>> didn't
>>>>>> lead to any test failures.
>>>>>>
>>>>>> This is a bit of a mess, and I've got a fix for it in progress.
>>>>>>
>>>>>> Unfortunately, undoing the type changes here might make a bit of a
>>>>>> mess elsewhere, requiring additional ugly casts.
>>>>>>
>>>>>> For examples of the use of -1 for these, see the constructor for
>>>>>> SATBMarkQueue (_max_completed_queue) and the initialize call for the
>>>>>> non-Java-thread dirty card queue set in G1CollectedHeap::initialize
>>>>>> (both are -1).
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>> src/share/vm/gc/g1/ptrQueue.cpp
>>>>>> 261 size_t PtrQueueSet::completed_buffers_list_length() {
>>>>>> 262 int n = 0;
>>>>>> 263 BufferNode* cbn = _completed_buffers_head;
>>>>>> 264 while (cbn != NULL) {
>>>>>> 265 n++;
>>>>>> 266 cbn = cbn->next();
>>>>>> 267 }
>>>>>> 268 return n;
>>>>>> 269 }
>>>>>>
>>>>>> The type of n should also be changed to size_t.
>>>>> okay.
>>>>>>
>>>>>> ------------------------------------------------------------------------------
>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>
>
More information about the hotspot-gc-dev
mailing list