RFR: 8300926: Several startup regressions ~6-70% in 21-b6 all platforms

[Daniel D. Daugherty]

On Thu, 16 Feb 2023 16:04:11 GMT, Erik Österlund <eosterlund at openjdk.org> wrote:

>> Hi all, please consider.
>> 
>> The original issue was when thread 1 asked to deopt nmethod set X and thread 2 asked for the same or a subset of X.
>> All method will already be marked, but the actual deoptimizing, not entrant, patching PC on stacks and patching post call nops, was not done yet. Which meant thread 2 could 'pass' thread 1.
>> Most places did deopt under Compile_lock, thus this is not an issue, but WB and clearCallSiteContext do not.
>> 
>> Since a handshakes may take long before completion and Compile_lock is used for so much more than deopts.
>> The fix in https://bugs.openjdk.org/browse/JDK-8299074 instead always emitted a handshake even when everything was already marked. (instead of adding Compile_lock to all places)
>> 
>> This turnout to be problematic in the startup, for example the number of deopt handshakes in jetty dry run startup went from 5 to 39 handshakes.
>> 
>> This fix first adds a barrier for which you do not pass until the requested deopts have happened and it coalesces the handshakes.
>> Secondly it moves handshakes part out of the Compile_lock where it is possible.
>> 
>> Which means we fix the performance bug and we reduce the contention on Compile_lock, meaning higher throughput in compiler and things such as class-loading.
>> 
>> It passes t1-t7 with flying colours! t8 still not completed and I'm redoing some testing due to last minute simplifications.
>> 
>> Thanks, Robbin
>
> src/hotspot/share/runtime/deoptimization.cpp line 110:
> 
>> 108:   MutexLocker ml(CompiledMethod_lock, Mutex::_no_safepoint_check_flag);
>> 109:   // If there is nothing to deopt _gen is the same as comitted.
>> 110:   _gen = DeoptimizationScope::_committed_deopt_gen;
> 
> Could we find a better name for the "_gen" member? If I'm reading this right, it's some kind of _required_gen, right? As in we need to have committed up until this number, or it isn't safe to continue.

The comment from deoptimization.hpp is this:

  // The highest gen we need to execute/wait for
  uint64_t _gen;

so better names might be: `_highest_gen` or `_highest_required_gen` or Erik's
`_required_gen`. In any case, if you rename this field, please update the comment
in deoptimization.hpp to match...

-------------

PR: https://git.openjdk.org/jdk/pull/12585