Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Thu, 10 Mar 2022 18:11:58 GMT, Mikhailo Seledtsov
wrote:
>> A trivial fix to solve a memory leak/memory pinning for long runs of
>> suspendthrd003.
>>
>> See the bug report for the gory analysis details.
>>
>> This fix was included in my jdk-19+12 stress runs so the updated test was
>> executed
>> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
>> configs. The
>> test was also executed in parallel for 3 runs with my
>> StressWrapper_StopAtExit config
>> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the
>> updated test was
>> tested with Mach5 Tier[1567] with no failures on any platform.
>
> Marked as reviewed by mseledtsov (Committer).
@mseledts - Thanks for the review.
-
PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Wed, 9 Mar 2022 20:41:37 GMT, Daniel D. Daugherty wrote:
> A trivial fix to solve a memory leak/memory pinning for long runs of
> suspendthrd003.
>
> See the bug report for the gory analysis details.
>
> This fix was included in my jdk-19+12 stress runs so the updated test was
> executed
> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
> configs. The
> test was also executed in parallel for 3 runs with my
> StressWrapper_StopAtExit config
> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the updated
> test was
> tested with Mach5 Tier[1567] with no failures on any platform.
Marked as reviewed by mseledtsov (Committer).
-
PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Thu, 10 Mar 2022 16:17:22 GMT, Leonid Mesnik wrote:
>> A trivial fix to solve a memory leak/memory pinning for long runs of
>> suspendthrd003.
>>
>> See the bug report for the gory analysis details.
>>
>> This fix was included in my jdk-19+12 stress runs so the updated test was
>> executed
>> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
>> configs. The
>> test was also executed in parallel for 3 runs with my
>> StressWrapper_StopAtExit config
>> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the
>> updated test was
>> tested with Mach5 Tier[1567] with no failures on any platform.
>
> Marked as reviewed by lmesnik (Reviewer).
@lmesnik - Thanks for the review.
-
PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Wed, 9 Mar 2022 20:41:37 GMT, Daniel D. Daugherty wrote:
> A trivial fix to solve a memory leak/memory pinning for long runs of
> suspendthrd003.
>
> See the bug report for the gory analysis details.
>
> This fix was included in my jdk-19+12 stress runs so the updated test was
> executed
> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
> configs. The
> test was also executed in parallel for 3 runs with my
> StressWrapper_StopAtExit config
> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the updated
> test was
> tested with Mach5 Tier[1567] with no failures on any platform.
Marked as reviewed by lmesnik (Reviewer).
-
PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Wed, 9 Mar 2022 20:41:37 GMT, Daniel D. Daugherty wrote:
> A trivial fix to solve a memory leak/memory pinning for long runs of
> suspendthrd003.
>
> See the bug report for the gory analysis details.
>
> This fix was included in my jdk-19+12 stress runs so the updated test was
> executed
> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
> configs. The
> test was also executed in parallel for 3 runs with my
> StressWrapper_StopAtExit config
> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the updated
> test was
> tested with Mach5 Tier[1567] with no failures on any platform.
@lmesnik and @mseledts - Can you guys take a look when you get the chance?
I'm making a minor addition to the NSK Log class.
-
PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Thu, 10 Mar 2022 04:52:46 GMT, David Holmes wrote: >> The fix looks fine, but what about StopAtExit? That test seemed to use a lot >> more memory than suspendthrd003. > > @plummercj - see https://bugs.openjdk.java.net/browse/JDK-8282704 for > StopAtExit. @dholmes-ora, @plummercj and @alexmenkov - Thanks for the reviews! One clarification: > and only print them if an error actually occurs Log will also print them if the mode is set to verbose. I manually verified my fix by setting the log to verbose at the end of the test run and I verified that I only had logging lines for the last loop rather than all gazillion of them... - PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Wed, 9 Mar 2022 20:41:37 GMT, Daniel D. Daugherty wrote:
> A trivial fix to solve a memory leak/memory pinning for long runs of
> suspendthrd003.
>
> See the bug report for the gory analysis details.
>
> This fix was included in my jdk-19+12 stress runs so the updated test was
> executed
> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
> configs. The
> test was also executed in parallel for 3 runs with my
> StressWrapper_StopAtExit config
> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the updated
> test was
> tested with Mach5 Tier[1567] with no failures on any platform.
Marked as reviewed by amenkov (Reviewer).
-
PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Thu, 10 Mar 2022 03:55:01 GMT, Chris Plummer wrote:
>> A trivial fix to solve a memory leak/memory pinning for long runs of
>> suspendthrd003.
>>
>> See the bug report for the gory analysis details.
>>
>> This fix was included in my jdk-19+12 stress runs so the updated test was
>> executed
>> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
>> configs. The
>> test was also executed in parallel for 3 runs with my
>> StressWrapper_StopAtExit config
>> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the
>> updated test was
>> tested with Mach5 Tier[1567] with no failures on any platform.
>
> The fix looks fine, but what about StopAtExit? That test seemed to use a lot
> more memory than suspendthrd003.
> @plummercj - see https://bugs.openjdk.java.net/browse/JDK-8282704 for
> StopAtExit.
Thanks!
-
PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Wed, 9 Mar 2022 20:41:37 GMT, Daniel D. Daugherty wrote:
> A trivial fix to solve a memory leak/memory pinning for long runs of
> suspendthrd003.
>
> See the bug report for the gory analysis details.
>
> This fix was included in my jdk-19+12 stress runs so the updated test was
> executed
> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
> configs. The
> test was also executed in parallel for 3 runs with my
> StressWrapper_StopAtExit config
> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the updated
> test was
> tested with Mach5 Tier[1567] with no failures on any platform.
Marked as reviewed by cjplummer (Reviewer).
-
PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Thu, 10 Mar 2022 03:55:01 GMT, Chris Plummer wrote:
>> A trivial fix to solve a memory leak/memory pinning for long runs of
>> suspendthrd003.
>>
>> See the bug report for the gory analysis details.
>>
>> This fix was included in my jdk-19+12 stress runs so the updated test was
>> executed
>> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
>> configs. The
>> test was also executed in parallel for 3 runs with my
>> StressWrapper_StopAtExit config
>> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the
>> updated test was
>> tested with Mach5 Tier[1567] with no failures on any platform.
>
> The fix looks fine, but what about StopAtExit? That test seemed to use a lot
> more memory than suspendthrd003.
@plummercj - see https://bugs.openjdk.java.net/browse/JDK-8282704 for
StopAtExit.
-
PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Wed, 9 Mar 2022 20:41:37 GMT, Daniel D. Daugherty wrote:
> A trivial fix to solve a memory leak/memory pinning for long runs of
> suspendthrd003.
>
> See the bug report for the gory analysis details.
>
> This fix was included in my jdk-19+12 stress runs so the updated test was
> executed
> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
> configs. The
> test was also executed in parallel for 3 runs with my
> StressWrapper_StopAtExit config
> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the updated
> test was
> tested with Mach5 Tier[1567] with no failures on any platform.
The fix looks fine, but what about StopAtExit? That test seemed to use a lot
more memory than suspendthrd003.
-
PR: https://git.openjdk.java.net/jdk/pull/7764
Re: RFR: 8282314: nsk/jvmti/SuspendThread/suspendthrd003 may leak memory
On Wed, 9 Mar 2022 20:41:37 GMT, Daniel D. Daugherty wrote:
> A trivial fix to solve a memory leak/memory pinning for long runs of
> suspendthrd003.
>
> See the bug report for the gory analysis details.
>
> This fix was included in my jdk-19+12 stress runs so the updated test was
> executed
> in the NSK JVM/TI testsuite for 3 runs in {fastdebug,release,slowdebug}
> configs. The
> test was also executed in parallel for 3 runs with my
> StressWrapper_StopAtExit config
> for 101 minutes in {fastdebug,release,slowdebug} configs. Lastly, the updated
> test was
> tested with Mach5 Tier[1567] with no failures on any platform.
Hi Dan,
This seems okay. This Log utility, by default, will keep all non error messages
in the log and only print them if an error actually occurs so that you can see
the full history. The log just grows without bound. Clearing the log
periodically, when you know any previous messages are not needed, seems a
reasonable thing to do.
Thanks,
David
-
Marked as reviewed by dholmes (Reviewer).
PR: https://git.openjdk.java.net/jdk/pull/7764
