RFR(L) 8153224 Monitor deflation prolong safepoints (CR8/v2.08/11-for-jdk14)
David Holmes
david.holmes at oracle.com
Mon Nov 11 10:52:54 UTC 2019
Hi Robbin,
Can you clarify your comments regarding use of Atomic::load and
Atomic::store please. Seems to me you are suggesting using those for
some memory ordering affect not for any atomicity effect per-se.
For example you say:
> 242 jint l_ref_count = ref_count();
> 243 ADIM_guarantee(l_ref_count > 0, "must be positive: l_ref_count=%d,
> ref_count=%d", l_ref_count, ref_count());
> Please use Atomic::load() in ref_count.
But it seems to me that to solve the problem of the compiler not
reissuing the load of ref_count, you should be using
OrderAccess::loadload() in the ref_count() method.
Or are you simply saying that given:
jint l1 = ref_count();
jint l2 = ref_count();
where ref_count simply does "return _ref_count;"
the compiler could treat the above as:
jint l1 = ref_count();
jint l2 = l1;
whereas if we have ref_count defined as "return
Atomic_load(&_ref_count);" then the compiler cannot do that?
I don't like seeing Atomic::load/store being used just to trick the
compiler that way. I thought we already relied on use of volatile to
disallow such optimisations and that this was the accepted way to do it.
Thanks,
David
On 8/11/2019 11:35 pm, Robbin Ehn wrote:
> Hi Dan,
>
> Thanks for looking into this, some comments on v8:
>
> ##################
> src/hotspot/cpu/sparc/globalDefinitions_sparc.hpp
> src/hotspot/cpu/x86/globalDefinitions_x86.hpp
> src/hotspot/share/logging/logTag.hpp
> src/hotspot/share/oops/markWord.hpp
> src/hotspot/share/runtime/basicLock.cpp
> src/hotspot/share/runtime/safepoint.cpp
> src/hotspot/share/runtime/serviceThread.cpp
> src/hotspot/share/runtime/sharedRuntime.cpp
> src/hotspot/share/runtime/synchronizer.hpp
> src/hotspot/share/runtime/vmOperations.cpp
> src/hotspot/share/runtime/vmOperations.hpp
> src/hotspot/share/runtime/vmStructs.cpp
> src/hotspot/share/runtime/vmThread.cpp
> test/hotspot/gtest/oops/test_markWord.cpp
>
> No comments.
>
> ##################
> I don't see the benefit of having the -HandshakeAfterDeflateIdleMonitors
> code paths.
> Removing that option would mean these files can be reverted:
> src/hotspot/cpu/aarch64/globals_aarch64.hpp
> src/hotspot/cpu/arm/globals_arm.hpp
> src/hotspot/cpu/ppc/globals_ppc.hpp
> src/hotspot/cpu/s390/globals_s390.hpp
> src/hotspot/cpu/sparc/globals_sparc.hpp
> src/hotspot/cpu/x86/globals_x86.hpp
> src/hotspot/cpu/x86/macroAssembler_x86.cpp
> src/hotspot/cpu/x86/macroAssembler_x86.hpp
> src/hotspot/cpu/zero/globals_zero.hpp
>
> And one less option here:
> src/hotspot/share/runtime/globals.hpp
>
> ##################
> src/hotspot/share/prims/jvm.cpp
>
> Unclear if this is a good idea.
>
> ##################
> src/hotspot/share/prims/whitebox.cpp
>
> This would assume the test expects the right thing, but that is not
> obvious.
>
> ##################
> src/hotspot/share/prims/jvmtiEnvBase.cpp
>
> The current pending and waiting monitor is only changed by the
> JavaThread itself.
> It only sets it after _contentions is increased.
> It clears it before _contentions is decreased.
> We are depending on safepoint or the thread is suspended, so it can't be
> deflated since _contentions are > 0.
> Plus the thread have already increased the ref count and can't decrease
> it (since at safepoint or suspended).
>
> ##################
> src/hotspot/share/runtime/objectMonitor.cpp
>
> ###1
> You have several these (and in other files):
> 242 jint l_ref_count = ref_count();
> 243 ADIM_guarantee(l_ref_count > 0, "must be positive: l_ref_count=%d,
> ref_count=%d", l_ref_count, ref_count());
> Please use Atomic::load() in ref_count.
> Since this is dependent on ref_count being volatile, otherwise the
> compiler may only do one load.
>
> ###2
> 307 // Prevent deflation. See ObjectSynchronizer::deflate_monitor(),
> ...
> 311 Atomic::add(1, &_contentions);
> In ObjectSynchronizer::deflate_monitor if you would check ref count
> instead of _contetion, we could remove contention.
> Since all waiters also have a ref count it looks like we don't need
> waiters either.
> In ObjectSynchronizer::deflate_monitor:
> if (mid->_contentions != 0 || mid->_waiters != 0) {
> Why not just do:
> if (mid->ref_count()) {
> ?
>
> ##################
> src/hotspot/share/runtime/objectMonitor.hpp
>
> ###1
> 252 intptr_t is_busy() const {
> 253 // TODO-FIXME: assert _owner == null implies _recursions = 0
> 254 // We do not include _ref_count in the is_busy() check because
> 255 // _ref_count is for indicating that the ObjectMonitor* is in
> 256 // use which is orthogonal to whether the ObjectMonitor itself
> 257 // is in use for a locking operation.
>
> But in the non-debug code we always check:
> + if (mid->is_busy() || mid->ref_count() != 0) {
>
> So it seem like you should have a method including ref count.
>
> ##################
> src/hotspot/share/runtime/objectMonitor.inline.hpp
>
> Use Atomic::load for ref count.
>
> ##################
> src/hotspot/share/runtime/synchronizer.cpp
>
> ###1
> 139 static volatile int g_om_free_count = 0; // # on g_free_list
> 140 static volatile int g_om_in_use_count = 0; // # on g_om_in_use_list
> 141 static volatile int g_om_population = 0; // # Extant -- in
> circulation
> 142 static volatile int g_om_wait_count = 0; // # on g_wait_list
> No padding here, aren't they more contended than the fields in the OM?
>
> ###2
> 151 static bool is_next_marked(ObjectMonitor* om) {
>
> Is only used in ObjectSynchronizer::om_flush.
> Here you fetch a OM and read the next field, this do not need LA
> semantics on supported platforms.
> This would only need Atomic::load.
>
> ###3
> 191 static void set_next(ObjectMonitor* om, ObjectMonitor* value) {
>
> In no place you need SR, in the only places it would made a difference:
> 345 OrderAccess::storestore();
> 346 set_next(cur, next); // Unmark the previous list head.
> and
> 1714 OrderAccess::storestore();
> 1715 set_next(in_use_list, next);
>
> You have a storestore already!
>
> This code reads as:
> OrderAccess::storestore();
> OrderAccess::loadstore();
> OrderAccess::storestore();
> om->_next_om = value
>
> So it should be an Atomic::store.
>
> ###4
> 198 static bool mark_list_head(ObjectMonitor* volatile * list_p
>
> Since the mark is an embedded spinlock I think the terminology should be
> changed. (that the spinlock is inside a the next pointer should be
> abstracted away)
> E.g. mark_next_loop would just be lock.
> The load of the list heads should use Atmoic:load.
> It also seem a bit wired to return next for the locking method.
> And output parameter can just be returned, and return NULL if list head
> is NULL.
> E.g.
>
> 198 static ObjectMonitor* get_list_head_locked(ObjectMonitor* volatile
> * list_p) {
> 200 while (true) {
> 201 ObjectMonitor* mid = Atomic::load(list_p);
> 202 if (mid == NULL) {
> 203 return NULL; // The list is empty.
> 204 }
> 205 if (try_lock(mid)) {
> 206 if (Atmoic::load(list_p) != mid) {
> 207 // The list head changed so we have to retry.
> 208 unlock(mid);
> 210 } else {
> return mid;
> }
> 214 }
> // Yield ?
> 215 }
> 216 }
>
> With colleteral changes.
>
> ###5
> 220 static ObjectMonitor* unmarked_next(ObjectMonitor* om)
> Atomic::store is what needed.
>
> ###6
> 333 static void prepend_to_common(
>
> 345 OrderAccess::storestore();
> 346 set_next(cur, next); // Unmark the previous list head.
> Double storestore. (fixed by changing set_next to Atomic::store)
>
> ###7
> 375 static ObjectMonitor* take_from_start_of_common(ObjectMonitor*
> volatile * list_p,
>
> Triple storestore here.
>
> 386 Atomic::dec(count_p);
> 387 // mark_list_head() used cmpxchg() above, switching list head
> can be lazier:
> 388 OrderAccess::storestore();
> 389 // Unmark take, but leave the next value for any lagging list
> 390 // walkers. It will get cleaned up when take is prepended to
> 391 // the in-use list:
> 392 set_next(take, next);
> 393 return take;
>
> Reads:
> count_p--
> OrderAccess::loadstore();
> OrderAccess::storestore();
> OrderAccess::storestore();
> OrderAccess::loadstore();
> OrderAccess::storestore();
> take->_next_om = next;
>
> Fixed by changing set_next to Atomic::store and removing the
> OrderAccess::storestore();
>
> ###8
> ObjectSynchronizer::om_release(
>
> 1591 if (m == mid) {
> 1592 // We found 'm' on the per-thread in-use list so try to
> extract it.
> 1593 if (cur_mid_in_use == NULL) {
> 1594 // mid is the list head and it is marked. Switch the list
> head
> 1595 // to next which unmarks the list head, but leaves mid
> marked:
> 1596 self->om_in_use_list = next;
> 1597 // mark_list_head() used cmpxchg() above, switching list
> head can be lazier:
> 1598 OrderAccess::storestore();
> 1599 } else {
> 1600 // mid and cur_mid_in_use are marked. Switch
> cur_mid_in_use's
> 1601 // next field to next which unmarks cur_mid_in_use, but
> leaves
> 1602 // mid marked:
> 1603 OrderAccess::release_store(&cur_mid_in_use->_next_om, next);
> 1604 }
> 1605 extracted = true;
> 1606 Atomic::dec(&self->om_in_use_count);
> 1607 // Unmark mid, but leave the next value for any lagging list
> 1608 // walkers. It will get cleaned up when mid is prepended to
> 1609 // the thread's free list:
> 1610 set_next(mid, next);
> 1611 break;
> 1612 }
>
> This does not look correct. Before taking this branch we have done a
> cmpxchg in mark_list_head or mark_next_loop.
> This is how it reads:
> OrderAccess::storestore(); // from previous cmpxchg
> OrderAccess::loadstore(); // from previous cmpxchg
> 1591 if (m == mid) {
> 1593 if (cur_mid_in_use == NULL) {
> 1596 self->om_in_use_list = next;
> 1598 OrderAccess::storestore();
> 1599 } else {
> OrderAccess::storestore();
> OrderAccess::loadstore();
> 1603 cur_mid_in_use->_next_om = next;
> 1604 }
> 1605 extracted = true;
> OrderAccess::storestore();
> OrderAccess::fence(); //
> storestore|storeload|loadstore|loadload
> self->om_in_use_count--; // Atomic::dec
> OrderAccess::storestore();
> OrderAccess::loadstore();
> OrderAccess::storestore();
> OrderAccess::loadstore();
> mid->_next_om = next; // Atomic::store
> 1611 break;
> 1612 }
>
> extracted is local variable so you so not need any orderaccess before it
> set.
> Fixed by changing set_next to Atomic::store, removing the
> OrderAccess::storestore() and changing OrderAccess::release_store to
> Atmoic::store();
>
> ###9
> 1653 void ObjectSynchronizer::om_flush(Thread* self) {
>
> 1714 OrderAccess::storestore();
> 1715 set_next(in_use_list, next);
> Fixed by changing set_next to Atomic::store.
>
> ###10
> 1737 self->om_free_list = NULL;
> 1738 OrderAccess::storestore(); // Lazier memory is okay for list
> walkers.
>
> prepend_list_to_g_free_list/prepend_list_to_g_om_in_use_list does first
> thing cmpxchg so there is no need for this storestore.
>
> ###11
> 1797 void ObjectSynchronizer::inflate(ObjectMonitorHandle* omh_p,
> Thread* self,
>
> 1938 // Once ObjectMonitor is configured and the object is associated
> 1939 // with the ObjectMonitor, it is safe to allow async deflation:
> 1940 assert(m->is_new(), "freshly allocated monitor must be new");
> 1941 m->set_allocation_state(ObjectMonitor::Old);
>
> So we use ref count, contention, waiter, owner and allocation state to
> keep OM alive in different scenarios.
> There is not way for me to keep track of that. I don't see why you would
> need more than owner and ref count.
> If you allocate the om with ref count 1 you can remove _allocation_state
> and just decrease ref count here instead.
>
> ###12
> 2079 bool ObjectSynchronizer::deflate_monitor
>
> 2112 if (AsyncDeflateIdleMonitors) {
> 2113 // clear() expects the owner field to be NULL and we won't race
> 2114 // with the simple C2 ObjectMonitor
>
> The macro assambler code is not just executed by C2, so this comment is
> a bit misleading. (there are some more also)
>
> ###13
> 2306 int ObjectSynchronizer::deflate_monitor_list(
>
> Same issue as ObjectSynchronizer::om_release.
> Fixed by changing set_next to Atomic::store, removing the
> OrderAccess::storestore() and changing OrderAccess::release_store to
> Atmoic::store();
>
> ###14
> 2474 if (SafepointSynchronize::is_synchronizing() &&
>
> This is the wrong method to call, it should
> SafepointMechanism::should_block(Thread* thread);
>
> ###15
> 2578 void ObjectSynchronizer::deflate_idle_monitors_using_JT() {
>
> 2616 g_wait_list = NULL;
> 2617 OrderAccess::storestore(); // Lazier memory sync is okay for
> list walkers.
>
> I don't see that g_wait_list is ever simutainously read.
> Either it is accessed by serviceThread outside a safepoint or by
> VMThread inside a safepoint?
>
> It looks like g_wait_list can just be a local in:
> void ObjectSynchronizer::deflate_idle_monitors_using_JT()
>
> (disregarding the debug code that might read it in a safepoint)
>
> ###16
> 2722 assert(SafepointSynchronize::is_synchronizing(), "sanity
> check");
>
> This is the wrong method to call, it should
> SafepointMechanism::should_block(Thread* thread);
>
> ##################
> src/hotspot/share/runtime/vframe.cpp
>
> We are at safepoint or current thread or in a handshake, current pending
> and waiting monitor is already stable.
>
> ##################
> src/hotspot/share/services/threadService.cpp
>
> These changes are only needed for the -HandshakeAfterDeflateIdleMonitors
> path.
>
> ##################
> test/jdk/java/rmi/server/UnicastRemoteObject/unexportObject/UnexportLeak.java
>
>
> Note: if OM had a weak to object instead this would not be needed.
>
> Thanks, Robbin
>
>
> On 11/4/19 10:03 PM, Daniel D. Daugherty wrote:
>> Greetings,
>>
>> I have made changes to the Async Monitor Deflation code in response to
>> the CR7/v2.07/10-for-jdk14 code review cycle. Thanks to David H., Robbin
>> and Erik O. for their comments!
>>
>> JDK14 Rampdown phase one is coming on Dec. 12, 2019 and the Async Monitor
>> Deflation project needs to push before Nov. 12, 2019 in order to allow
>> for sufficient bake time for such a big change. Nov. 12 is _next_ Tuesday
>> so we have 8 days from today to finish this code review cycle and push
>> this code for JDK14.
>>
>> Carsten and Roman! Time for you guys to chime in again on the code
>> reviews.
>>
>> I have attached the change list from CR7 to CR8 instead of putting it in
>> the body of this email. I've also added a link to the CR7-to-CR8-changes
>> file to the webrevs so it should be easy to find.
>>
>> Main bug URL:
>>
>> JDK-8153224 Monitor deflation prolong safepoints
>> https://bugs.openjdk.java.net/browse/JDK-8153224
>>
>> The project is currently baselined on jdk-14+21.
>>
>> Here's the full webrev URL for those folks that want to see all of the
>> current Async Monitor Deflation code in one go (v2.08 full):
>>
>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/11-for-jdk14.v2.08.full
>>
>> Some folks might want to see just what has changed since the last review
>> cycle so here's a webrev for that (v2.08 inc):
>>
>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/11-for-jdk14.v2.08.inc/
>>
>> The OpenJDK wiki did not need any changes for this round:
>>
>> https://wiki.openjdk.java.net/display/HotSpot/Async+Monitor+Deflation
>>
>> The jdk-14+21 based v2.08 version of the patch has been thru Mach5
>> tier[1-8]
>> testing on Oracle's usual set of platforms. It has also been through
>> my usual
>> set of stress testing on Linux-X64, macOSX and Solaris-X64 with the
>> addition
>> of Robbin's "MoCrazy 1024" test running in parallel with the other
>> tests in
>> my lab. Some testing is still running, but so far there are no new
>> regressions.
>>
>> I have not yet done a SPECjbb2015 round on the CR8/v2.08/11-for-jdk14
>> bits.
>>
>> Thanks, in advance, for any questions, comments or suggestions.
>>
>> Dan
>>
>>
>> On 10/17/19 5:50 PM, Daniel D. Daugherty wrote:
>>> Greetings,
>>>
>>> The Async Monitor Deflation project is reaching the end game. I have no
>>> changes planned for the project at this time so all that is left is code
>>> review and any changes that results from those reviews.
>>>
>>> Carsten and Roman! Time for you guys to chime in again on the code
>>> reviews.
>>>
>>> I have attached the list of fixes from CR6 to CR7 instead of putting it
>>> in the main body of this email.
>>>
>>> Main bug URL:
>>>
>>> JDK-8153224 Monitor deflation prolong safepoints
>>> https://bugs.openjdk.java.net/browse/JDK-8153224
>>>
>>> The project is currently baselined on jdk-14+19.
>>>
>>> Here's the full webrev URL for those folks that want to see all of the
>>> current Async Monitor Deflation code in one go (v2.07 full):
>>>
>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/10-for-jdk14.v2.07.full
>>>
>>>
>>> Some folks might want to see just what has changed since the last review
>>> cycle so here's a webrev for that (v2.07 inc):
>>>
>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/10-for-jdk14.v2.07.inc/
>>>
>>>
>>> The OpenJDK wiki has been updated to match the CR7/v2.07/10-for-jdk14
>>> changes:
>>>
>>> https://wiki.openjdk.java.net/display/HotSpot/Async+Monitor+Deflation
>>>
>>> The jdk-14+18 based v2.07 version of the patch has been thru Mach5
>>> tier[1-8]
>>> testing on Oracle's usual set of platforms. It has also been through
>>> my usual
>>> set of stress testing on Linux-X64, macOSX and Solaris-X64 with the
>>> addition
>>> of Robbin's "MoCrazy 1024" test running in parallel with the other
>>> tests in
>>> my lab.
>>>
>>> The jdk-14+19 based v2.07 version of the patch has been thru Mach5
>>> tier[1-3]
>>> test on Oracle's usual set of platforms. Mach5 tier[4-8] are in process.
>>>
>>> I did another round of SPECjbb2015 testing in Oracle's Aurora
>>> Performance lab
>>> using using their tuned SPECjbb2015 Linux-X64 G1 configs:
>>>
>>> - "base" is jdk-14+18
>>> - "v2.07" is the latest version and includes C2
>>> inc_om_ref_count() support
>>> on LP64 X64 and the new HandshakeAfterDeflateIdleMonitors option
>>> - "off" is with -XX:-AsyncDeflateIdleMonitors specified
>>> - "handshake" is with -XX:+HandshakeAfterDeflateIdleMonitors
>>> specified
>>>
>>> hbIR hbIR
>>> (max attempted) (settled) max-jOPS critical-jOPS runtime
>>> --------------- --------- -------- ------------- -------
>>> 34282.00 30635.90 28831.30 20969.20 3841.30 base
>>> 34282.00 30973.00 29345.80 21025.20 3964.10 v2.07
>>> 34282.00 31105.60 29174.30 21074.00 3931.30
>>> v2.07_handshake
>>> 34282.00 30789.70 27151.60 19839.10 3850.20
>>> v2.07_off
>>>
>>> - The Aurora Perf comparison tool reports:
>>>
>>> Comparison max-jOPS critical-jOPS
>>> ---------------------- --------------------
>>> --------------------
>>> base vs 2.07 +1.78% (s, p=0.000) +0.27% (ns,
>>> p=0.790)
>>> base vs 2.07_handshake +1.19% (s, p=0.007) +0.58% (ns,
>>> p=0.536)
>>> base vs 2.07_off -5.83% (ns, p=0.394) -5.39% (ns,
>>> p=0.347)
>>>
>>> (s) - significant (ns) - not-significant
>>>
>>> - For historical comparison, the Aurora Perf comparision tool
>>> reported for v2.06 with a baseline of jdk-13+31:
>>>
>>> Comparison max-jOPS critical-jOPS
>>> ---------------------- --------------------
>>> --------------------
>>> base vs 2.06 -0.32% (ns, p=0.345) +0.71% (ns,
>>> p=0.646)
>>> base vs 2.06_off +0.49% (ns, p=0.292) -1.21% (ns,
>>> p=0.481)
>>>
>>> (s) - significant (ns) - not-significant
>>>
>>> Thanks, in advance, for any questions, comments or suggestions.
>>>
>>> Dan
>>>
>>>
>>> On 8/28/19 5:02 PM, Daniel D. Daugherty wrote:
>>>> Greetings,
>>>>
>>>> The Async Monitor Deflation project has rebased to JDK14 so it's time
>>>> for our first code review in that new context!!
>>>>
>>>> I've been focused on changing the monitor list management code to be
>>>> lock-free in order to make SPECjbb2015 happier. Of course with a change
>>>> like that, it takes a while to chase down all the new and wonderful
>>>> races. At this point, I have the code back to the same stability that
>>>> I had with CR5/v2.05/8-for-jdk13.
>>>>
>>>> To lay the ground work for this round of review, I pushed the following
>>>> two fixes to jdk/jdk earlier today:
>>>>
>>>> JDK-8230184 rename, whitespace, indent and comments changes in
>>>> preparation
>>>> for lock free Monitor lists
>>>> https://bugs.openjdk.java.net/browse/JDK-8230184
>>>>
>>>> JDK-8230317 serviceability/sa/ClhsdbPrintStatics.java fails
>>>> after 8230184
>>>> https://bugs.openjdk.java.net/browse/JDK-8230317
>>>>
>>>> I have attached the list of fixes from CR5 to CR6 instead of putting
>>>> in the main body of this email.
>>>>
>>>> Main bug URL:
>>>>
>>>> JDK-8153224 Monitor deflation prolong safepoints
>>>> https://bugs.openjdk.java.net/browse/JDK-8153224
>>>>
>>>> The project is currently baselined on jdk-14+11 plus the fixes for
>>>> JDK-8230184 and JDK-8230317.
>>>>
>>>> Here's the full webrev URL for those folks that want to see all of the
>>>> current Async Monitor Deflation code in one go (v2.06 full):
>>>>
>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/9-for-jdk14.v2.06.full/
>>>>
>>>>
>>>>
>>>> The primary focus of this review cycle is on the lock-free Monitor List
>>>> management changes so here's a webrev for just that patch (v2.06c):
>>>>
>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/9-for-jdk14.v2.06c.inc/
>>>>
>>>>
>>>> The secondary focus of this review cycle is on the bug fixes that have
>>>> been made since CR5/v2.05/8-for-jdk13 so here's a webrev for just that
>>>> patch (v2.06b):
>>>>
>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/9-for-jdk14.v2.06b.inc/
>>>>
>>>>
>>>> The third and final bucket for this review cycle is the rename,
>>>> whitespace,
>>>> indent and comments changes made in preparation for lock free
>>>> Monitor list
>>>> management. Almost all of that was extracted into JDK-8230184 for the
>>>> baseline so this bucket now has just a few comment changes relative to
>>>> CR5/v2.05/8-for-jdk13. Here's a webrev for the remainder (v2.06a):
>>>>
>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/9-for-jdk14.v2.06a.inc/
>>>>
>>>>
>>>>
>>>> Some folks might want to see just what has changed since the last
>>>> review
>>>> cycle so here's a webrev for that (v2.06 inc):
>>>>
>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/9-for-jdk14.v2.06.inc/
>>>>
>>>>
>>>>
>>>> Last, but not least, some folks might want to see the code before the
>>>> addition of lock-free Monitor List management so here's a webrev for
>>>> that (v2.00 -> v2.05):
>>>>
>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/9-for-jdk14.v2.05.inc/
>>>>
>>>>
>>>> The OpenJDK wiki will need minor updates to match the CR6 changes:
>>>>
>>>> https://wiki.openjdk.java.net/display/HotSpot/Async+Monitor+Deflation
>>>>
>>>> but that should only be changes to describe per-thread list async
>>>> monitor
>>>> deflation being done by the ServiceThread.
>>>>
>>>> (I did update the OpenJDK wiki for the CR5 changes back on 2019.08.14)
>>>>
>>>> This version of the patch has been thru Mach5 tier[1-8] testing on
>>>> Oracle's usual set of platforms. It has also been through my usual set
>>>> of stress testing on Linux-X64, macOSX and Solaris-X64.
>>>>
>>>> I did a bunch of SPECjbb2015 testing in Oracle's Aurora Performance lab
>>>> using using their tuned SPECjbb2015 Linux-X64 G1 configs. This was
>>>> using
>>>> this patch baselined on jdk-13+31 (for stability):
>>>>
>>>> hbIR hbIR
>>>> (max attempted) (settled) max-jOPS critical-jOPS runtime
>>>> --------------- --------- -------- ------------- -------
>>>> 34282.00 28837.20 27905.20 19817.40 3658.10 base
>>>> 34965.70 29798.80 27814.90 19959.00 3514.60 v2.06d
>>>> 34282.00 29100.70 28042.50 19577.00 3701.90
>>>> v2.06d_off
>>>> 34282.00 29218.50 27562.80 19397.30 3657.60
>>>> v2.06d_ocache
>>>> 34965.70 29838.30 26512.40 19170.60 3569.90 v2.05
>>>> 34282.00 28926.10 27734.00 19835.10 3588.40
>>>> v2.05_off
>>>>
>>>> The "off" configs are with -XX:-AsyncDeflateIdleMonitors specified and
>>>> the "ocache" config is with 128 byte cache line sizes instead of 64
>>>> byte
>>>> cache lines sizes. "v2.06d" is the last set of changes that I made
>>>> before
>>>> those changes were distributed into the "v2.06a", "v2.06b" and "v2.06c"
>>>> buckets for this review recycle.
>>>>
>>>>
>>>> Thanks, in advance, for any questions, comments or suggestions.
>>>>
>>>> Dan
>>>>
>>>>
>>>> On 7/11/19 3:49 PM, Daniel D. Daugherty wrote:
>>>>> Greetings,
>>>>>
>>>>> I've been focused on chasing down and fixing the rare test failures
>>>>> that only pop up rarely. So this round is primarily fixes for races
>>>>> with a few additional fixes that came from Karen's review of CR4.
>>>>> Thanks Karen!
>>>>>
>>>>> I have attached the list of fixes from CR4 to CR5 instead of putting
>>>>> in the main body of this email.
>>>>>
>>>>> Main bug URL:
>>>>>
>>>>> JDK-8153224 Monitor deflation prolong safepoints
>>>>> https://bugs.openjdk.java.net/browse/JDK-8153224
>>>>>
>>>>> The project is currently baselined on jdk-13+29. This will likely be
>>>>> the last JDK13 baseline for this project and I'll roll to the JDK14
>>>>> (jdk/jdk) repo soon...
>>>>>
>>>>> Here's the full webrev URL:
>>>>>
>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/8-for-jdk13.full/
>>>>>
>>>>> Here's the incremental webrev URL:
>>>>>
>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/8-for-jdk13.inc/
>>>>>
>>>>> I have not yet checked the OpenJDK wiki to see if it needs any updates
>>>>> to match the CR5 changes:
>>>>>
>>>>> https://wiki.openjdk.java.net/display/HotSpot/Async+Monitor+Deflation
>>>>>
>>>>> (I did update the OpenJDK wiki for the CR4 changes back on 2019.06.26)
>>>>>
>>>>> This version of the patch has been thru Mach5 tier[1-3] testing on
>>>>> Oracle's usual set of platforms. Mach5 tier[4-6] is running now and
>>>>> Mach5 tier[78] will follow. I'll kick off the usual stress testing
>>>>> on Linux-X64, macOSX and Solaris-X64 as those machines become
>>>>> available.
>>>>> Since I haven't made any performance changes in this round, I'll only
>>>>> be running SPECjbb2015 to gather the latest monitorinflation logs.
>>>>>
>>>>> Next up:
>>>>>
>>>>> - We're still seeing 4-5% lower performance with SPECjbb2015 on
>>>>> Linux-X64 and we've determined that some of that comes from
>>>>> contention on the gListLock. So I'm going to investigate removing
>>>>> the gListLock. Yes, another lock free set of changes is coming!
>>>>> - Of course, going lock free often causes new races and new failures
>>>>> so that's a good reason for make those changes isolated in their
>>>>> own round (and not holding up CR5/v2.05/8-for-jdk13 anymore).
>>>>> - I finally have a potential fix for the Win* failure with
>>>>> gc/g1/humongousObjects/TestHumongousClassLoader.java
>>>>> but I haven't run it through Mach5 yet so it'll be in the next
>>>>> round.
>>>>> - Some RTM tests were recently re-enabled in Mach5 and I'm seeing some
>>>>> monitor related failures there. I suspect that I need to go take a
>>>>> look at the C2 RTM macro assembler code and look for things that
>>>>> might
>>>>> conflict if Async Monitor Deflation. If you're interested in that
>>>>> kind
>>>>> of issue, then see the macroAssembler_x86.cpp sanity check that I
>>>>> added in this round!
>>>>>
>>>>> Thanks, in advance, for any questions, comments or suggestions.
>>>>>
>>>>> Dan
>>>>>
>>>>>
>>>>> On 5/26/19 8:30 PM, Daniel D. Daugherty wrote:
>>>>>> Greetings,
>>>>>>
>>>>>> I have a fix for an issue that came up during performance testing.
>>>>>> Many thanks to Robbin for diagnosing the issue in his SPECjbb2015
>>>>>> experiments.
>>>>>>
>>>>>> Here's the list of changes from CR3 to CR4. The list is a bit
>>>>>> verbose due to the complexity of the issue, but the changes
>>>>>> themselves are not that big.
>>>>>>
>>>>>> Functional:
>>>>>> - Change SafepointSynchronize::is_cleanup_needed() from calling
>>>>>> ObjectSynchronizer::is_cleanup_needed() to calling
>>>>>> ObjectSynchronizer::is_safepoint_deflation_needed():
>>>>>> - is_safepoint_deflation_needed() returns the result of
>>>>>> monitors_used_above_threshold() for safepoint based
>>>>>> monitor deflation (!AsyncDeflateIdleMonitors).
>>>>>> - For AsyncDeflateIdleMonitors, it only returns true if
>>>>>> there is a special deflation request, e.g., System.gc()
>>>>>> - This solves a bug where there are a bunch of Cleanup
>>>>>> safepoints that simply request async deflation which
>>>>>> keeps the async JavaThreads from making progress on
>>>>>> their async deflation work.
>>>>>> - Add AsyncDeflationInterval diagnostic option. Description:
>>>>>> Async deflate idle monitors every so many milliseconds when
>>>>>> MonitorUsedDeflationThreshold is exceeded (0 is off).
>>>>>> - Replace ObjectSynchronizer::gOmShouldDeflateIdleMonitors() with
>>>>>> ObjectSynchronizer::is_async_deflation_needed():
>>>>>> - is_async_deflation_needed() returns true when
>>>>>> is_async_cleanup_requested() is true or when
>>>>>> monitors_used_above_threshold() is true (but no more often than
>>>>>> AsyncDeflationInterval).
>>>>>> - if AsyncDeflateIdleMonitors Service_lock->wait() now waits for
>>>>>> at most GuaranteedSafepointInterval millis:
>>>>>> - This allows is_async_deflation_needed() to be checked at
>>>>>> the same interval as GuaranteedSafepointInterval.
>>>>>> (default is 1000 millis/1 second)
>>>>>> - Once is_async_deflation_needed() has returned true, it
>>>>>> generally cannot return true for AsyncDeflationInterval.
>>>>>> This is to prevent async deflation from swamping the
>>>>>> ServiceThread.
>>>>>> - The ServiceThread still handles async deflation of the global
>>>>>> in-use list and now it also marks JavaThreads for async deflation
>>>>>> of their in-use lists.
>>>>>> - The ServiceThread will check for async deflation work every
>>>>>> GuaranteedSafepointInterval.
>>>>>> - A safepoint can still cause the ServiceThread to check for
>>>>>> async deflation work via is_async_deflation_requested.
>>>>>> - Refactor code from ObjectSynchronizer::is_cleanup_needed() into
>>>>>> monitors_used_above_threshold() and remove is_cleanup_needed().
>>>>>> - In addition to System.gc(), the VM_Exit VM op and the final
>>>>>> VMThread safepoint now set the is_special_deflation_requested
>>>>>> flag to reduce the in-use monitor population that is reported by
>>>>>> ObjectSynchronizer::log_in_use_monitor_details() at VM exit.
>>>>>>
>>>>>> Test update:
>>>>>> - test/hotspot/gtest/oops/test_markOop.cpp is updated to work with
>>>>>> AsyncDeflateIdleMonitors.
>>>>>>
>>>>>> Collateral:
>>>>>> - Add/clarify/update some logging messages.
>>>>>>
>>>>>> Cleanup:
>>>>>> - Updated comments based on Karen's code review.
>>>>>> - Change 'special cleanup' -> 'special deflation' and
>>>>>> 'async cleanup' -> 'async deflation'.
>>>>>> - comment and function name changes
>>>>>> - Clarify MonitorUsedDeflationThreshold description;
>>>>>>
>>>>>>
>>>>>> Main bug URL:
>>>>>>
>>>>>> JDK-8153224 Monitor deflation prolong safepoints
>>>>>> https://bugs.openjdk.java.net/browse/JDK-8153224
>>>>>>
>>>>>> The project is currently baselined on jdk-13+22.
>>>>>>
>>>>>> Here's the full webrev URL:
>>>>>>
>>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/7-for-jdk13.full/
>>>>>>
>>>>>> Here's the incremental webrev URL:
>>>>>>
>>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/7-for-jdk13.inc/
>>>>>>
>>>>>> I have not updated the OpenJDK wiki to reflect the CR4 changes:
>>>>>>
>>>>>> https://wiki.openjdk.java.net/display/HotSpot/Async+Monitor+Deflation
>>>>>>
>>>>>> The wiki doesn't say a whole lot about the async deflation invocation
>>>>>> mechanism so I have to figure out how to add that content.
>>>>>>
>>>>>> This version of the patch has been thru Mach5 tier[1-8] testing on
>>>>>> Oracle's usual set of platforms. My Solaris-X64 stress kit run is
>>>>>> running now. Kitchensink8H on product, fastdebug, and slowdebug bits
>>>>>> are running on Linux-X64, MacOSX and Solaris-X64. I still have to run
>>>>>> my stress kit on Linux-X64. I still have to run the SPECjbb2015
>>>>>> baseline and CR4 runs on Linux-X64, MacOSX and Solaris-X64.
>>>>>>
>>>>>> Thanks, in advance, for any questions, comments or suggestions.
>>>>>>
>>>>>> Dan
>>>>>>
>>>>>> On 5/6/19 11:52 AM, Daniel D. Daugherty wrote:
>>>>>>> Greetings,
>>>>>>>
>>>>>>> I had some discussions with Karen about a race that was in the
>>>>>>> ObjectMonitor::enter() code in CR2/v2.02/5-for-jdk13. This race was
>>>>>>> theoretical and I had no test failures due to it. The fix is pretty
>>>>>>> simple: remove the special case code for async deflation in the
>>>>>>> ObjectMonitor::enter() function and rely solely on the ref_count
>>>>>>> for ObjectMonitor::enter() protection.
>>>>>>>
>>>>>>> During those discussions Karen also floated the idea of using the
>>>>>>> ref_count field instead of the contentions field for the Async
>>>>>>> Monitor Deflation protocol. I decided to go ahead and code up that
>>>>>>> change and I have run it through the usual stress and Mach5 testing
>>>>>>> with no issues. It's also known as v2.03 (for those for with the
>>>>>>> patches) and as webrev/6-for-jdk13 (for those with webrev URLs).
>>>>>>> Sorry for all the names...
>>>>>>>
>>>>>>> Main bug URL:
>>>>>>>
>>>>>>> JDK-8153224 Monitor deflation prolong safepoints
>>>>>>> https://bugs.openjdk.java.net/browse/JDK-8153224
>>>>>>>
>>>>>>> The project is currently baselined on jdk-13+18.
>>>>>>>
>>>>>>> Here's the full webrev URL:
>>>>>>>
>>>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/6-for-jdk13.full/
>>>>>>>
>>>>>>> Here's the incremental webrev URL:
>>>>>>>
>>>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/6-for-jdk13.inc/
>>>>>>>
>>>>>>> I have also updated the OpenJDK wiki to reflect the CR3 changes:
>>>>>>>
>>>>>>> https://wiki.openjdk.java.net/display/HotSpot/Async+Monitor+Deflation
>>>>>>>
>>>>>>>
>>>>>>> This version of the patch has been thru Mach5 tier[1-8] testing on
>>>>>>> Oracle's usual set of platforms. My Solaris-X64 stress kit run had
>>>>>>> no issues. Kitchensink8H on product, fastdebug, and slowdebug bits
>>>>>>> had no failures on Linux-X64; MacOSX fastdebug and slowdebug and
>>>>>>> Solaris-X64 release had the usual "Too large time diff" complaints.
>>>>>>> 12 hour Inflate2 runs on product, fastdebug and slowdebug bits on
>>>>>>> Linux-X64, MacOSX and Solaris-X64 had no failures. My Linux-X64
>>>>>>> stress kit is running right now.
>>>>>>>
>>>>>>> I've done the SPECjbb2015 baseline and CR3 runs. I need to gather
>>>>>>> the results and analyze them.
>>>>>>>
>>>>>>> Thanks, in advance, for any questions, comments or suggestions.
>>>>>>>
>>>>>>> Dan
>>>>>>>
>>>>>>>
>>>>>>> On 4/25/19 12:38 PM, Daniel D. Daugherty wrote:
>>>>>>>> Greetings,
>>>>>>>>
>>>>>>>> I have a small but important bug fix for the Async Monitor
>>>>>>>> Deflation
>>>>>>>> project ready to go. It's also known as v2.02 (for those for
>>>>>>>> with the
>>>>>>>> patches) and as webrev/5-for-jdk13 (for those with webrev URLs).
>>>>>>>> Sorry
>>>>>>>> for all the names...
>>>>>>>>
>>>>>>>> JDK-8222295 was pushed to jdk/jdk two days ago so that baseline
>>>>>>>> patch
>>>>>>>> is out of our hair.
>>>>>>>>
>>>>>>>> Main bug URL:
>>>>>>>>
>>>>>>>> JDK-8153224 Monitor deflation prolong safepoints
>>>>>>>> https://bugs.openjdk.java.net/browse/JDK-8153224
>>>>>>>>
>>>>>>>> The project is currently baselined on jdk-13+17.
>>>>>>>>
>>>>>>>> Here's the full webrev URL:
>>>>>>>>
>>>>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/5-for-jdk13.full/
>>>>>>>>
>>>>>>>> Here's the incremental webrev URL (JDK-8153224):
>>>>>>>>
>>>>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/5-for-jdk13.inc/
>>>>>>>>
>>>>>>>> I still have to update the OpenJDK wiki to reflect the CR2 changes:
>>>>>>>>
>>>>>>>> https://wiki.openjdk.java.net/display/HotSpot/Async+Monitor+Deflation
>>>>>>>>
>>>>>>>>
>>>>>>>> This version of the patch has been thru Mach5 tier[1-6] testing on
>>>>>>>> Oracle's usual set of platforms. Mach5 tier[7-8] is running now.
>>>>>>>> My stress kit is running on Solaris-X64 now. Kitchensink8H is
>>>>>>>> running
>>>>>>>> now on product, fastdebug, and slowdebug bits on Linux-X64, MacOSX
>>>>>>>> and Solaris-X64. 12 hour Inflate2 runs are running now on product,
>>>>>>>> fastdebug and slowdebug bits on Linux-X64, MacOSX and Solaris-X64.
>>>>>>>> I'll start my my stress kit on Linux-X64 sometime on Sunday (after
>>>>>>>> my jdk-13+18 stress run is done).
>>>>>>>>
>>>>>>>> I'll do SPECjbb2015 baseline and CR2 runs after all the stress
>>>>>>>> testing is done.
>>>>>>>>
>>>>>>>> Thanks, in advance, for any questions, comments or suggestions.
>>>>>>>>
>>>>>>>> Dan
>>>>>>>>
>>>>>>>>
>>>>>>>> On 4/19/19 11:58 AM, Daniel D. Daugherty wrote:
>>>>>>>>> Greetings,
>>>>>>>>>
>>>>>>>>> I finally have CR1 for the Async Monitor Deflation project
>>>>>>>>> ready to
>>>>>>>>> go. It's also known as v2.01 (for those for with the patches)
>>>>>>>>> and as
>>>>>>>>> webrev/4-for-jdk13 (for those with webrev URLs). Sorry for all the
>>>>>>>>> names...
>>>>>>>>>
>>>>>>>>> Main bug URL:
>>>>>>>>>
>>>>>>>>> JDK-8153224 Monitor deflation prolong safepoints
>>>>>>>>> https://bugs.openjdk.java.net/browse/JDK-8153224
>>>>>>>>>
>>>>>>>>> Baseline bug fixes URL:
>>>>>>>>>
>>>>>>>>> JDK-8222295 more baseline cleanups from Async Monitor
>>>>>>>>> Deflation project
>>>>>>>>> https://bugs.openjdk.java.net/browse/JDK-8222295
>>>>>>>>>
>>>>>>>>> The project is currently baselined on jdk-13+15.
>>>>>>>>>
>>>>>>>>> Here's the webrev for the latest baseline changes (JDK-8222295):
>>>>>>>>>
>>>>>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/4-for-jdk13.8222295
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Here's the full webrev URL (JDK-8153224 only):
>>>>>>>>>
>>>>>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/4-for-jdk13.full/
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Here's the incremental webrev URL (JDK-8153224):
>>>>>>>>>
>>>>>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/4-for-jdk13.inc/
>>>>>>>>>
>>>>>>>>> So I'm looking for reviews for both JDK-8222295 and the latest
>>>>>>>>> version
>>>>>>>>> of JDK-8153224...
>>>>>>>>>
>>>>>>>>> I still have to update the OpenJDK wiki to reflect the CR changes:
>>>>>>>>>
>>>>>>>>> https://wiki.openjdk.java.net/display/HotSpot/Async+Monitor+Deflation
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> This version of the patch has been thru Mach5 tier[1-3] testing on
>>>>>>>>> Oracle's usual set of platforms. Mach5 tier[4-6] is running now
>>>>>>>>> and
>>>>>>>>> Mach5 tier[78] will be run later today. My stress kit on
>>>>>>>>> Solaris-X64
>>>>>>>>> is running now. Linux-X64 stress testing will start on Sunday. I'm
>>>>>>>>> planning to do Kitchensink runs, SPECjbb2015 runs and my monitor
>>>>>>>>> inflation stress tests on Linux-X64, MacOSX and Solaris-X64.
>>>>>>>>>
>>>>>>>>> Thanks, in advance, for any questions, comments or suggestions.
>>>>>>>>>
>>>>>>>>> Dan
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On 3/24/19 9:57 AM, Daniel D. Daugherty wrote:
>>>>>>>>>> Greetings,
>>>>>>>>>>
>>>>>>>>>> Welcome to the OpenJDK review thread for my port of Carsten's
>>>>>>>>>> work on:
>>>>>>>>>>
>>>>>>>>>> JDK-8153224 Monitor deflation prolong safepoints
>>>>>>>>>> https://bugs.openjdk.java.net/browse/JDK-8153224
>>>>>>>>>>
>>>>>>>>>> Here's a link to the OpenJDK wiki that describes my port:
>>>>>>>>>>
>>>>>>>>>> https://wiki.openjdk.java.net/display/HotSpot/Async+Monitor+Deflation
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Here's the webrev URL:
>>>>>>>>>>
>>>>>>>>>> http://cr.openjdk.java.net/~dcubed/8153224-webrev/3-for-jdk13/
>>>>>>>>>>
>>>>>>>>>> Here's a link to Carsten's original webrev:
>>>>>>>>>>
>>>>>>>>>> http://cr.openjdk.java.net/~cvarming/monitor_deflate_conc/0/
>>>>>>>>>>
>>>>>>>>>> Earlier versions of this patch have been through several
>>>>>>>>>> rounds of
>>>>>>>>>> preliminary review. Many thanks to Carsten, Coleen, Robbin, and
>>>>>>>>>> Roman for their preliminary code review comments. A very special
>>>>>>>>>> thanks to Robbin and Roman for building and testing the patch in
>>>>>>>>>> their own environments (including specJBB2015).
>>>>>>>>>>
>>>>>>>>>> This version of the patch has been thru Mach5 tier[1-8]
>>>>>>>>>> testing on
>>>>>>>>>> Oracle's usual set of platforms. Earlier versions have been run
>>>>>>>>>> through my stress kit on my Linux-X64 and Solaris-X64 servers
>>>>>>>>>> (product, fastdebug, slowdebug).Earlier versions have run
>>>>>>>>>> Kitchensink
>>>>>>>>>> for 12 hours on MacOSX, Linux-X64 and Solaris-X64 (product,
>>>>>>>>>> fastdebug
>>>>>>>>>> and slowdebug). Earlier versions have run my monitor inflation
>>>>>>>>>> stress
>>>>>>>>>> tests for 12 hours on MacOSX, Linux-X64 and Solaris-X64 (product,
>>>>>>>>>> fastdebug and slowdebug).
>>>>>>>>>>
>>>>>>>>>> All of the testing done on earlier versions will be redone on the
>>>>>>>>>> latest version of the patch.
>>>>>>>>>>
>>>>>>>>>> Thanks, in advance, for any questions, comments or suggestions.
>>>>>>>>>>
>>>>>>>>>> Dan
>>>>>>>>>>
>>>>>>>>>> P.S.
>>>>>>>>>> One subtest in
>>>>>>>>>> gc/g1/humongousObjects/TestHumongousClassLoader.java
>>>>>>>>>> is currently failing in -Xcomp mode on Win* only. I've been
>>>>>>>>>> trying
>>>>>>>>>> to characterize/analyze this failure for more than a week now. At
>>>>>>>>>> this point I'm convinced that Async Monitor Deflation is
>>>>>>>>>> aggravating
>>>>>>>>>> an existing bug. However, I plan to have a better handle on that
>>>>>>>>>> failure before these bits are pushed to the jdk/jdk repo.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>
More information about the hotspot-runtime-dev
mailing list