ParallelStreamsLoadTest_special_J9_24 hang/timeout at 100% #11904
Labels
Milestone
Comments
|
Looking at the java stacks, this doesn't look like a hang - the load threads are running: The second thread is likely running in JITted code, so the stack can't be walked. The other two cores show both load threads in JITted code, so they seem to have progressed since the first core. |
|
@pshipton Do you have a download link for the SDK that was in use? |
|
and the debug image if that helps https://140-211-168-230-openstack.osuosl.org/artifactory/ci-eclipse-openj9/Build_JDK8_x86-32_windows_Personal/739//debug-image.tar.gz |
|
In java.exe_210207_165625.dmp, the native stacks of the two load threads are: Which indicate that OSR is occurring (as it should be in the given test mode). |
|
Stacks from java.exe_210207_165655.dmp: |
|
Looking at thread 25 in the two cores, the thread has clearly progressed (it does not appear to be looping in the decompiler): |
|
So from what I can see, this is not a hang as such - it may just be running very slowly, or perhaps running on a sick machine. |
|
Snap trace for thread 25 extracted from core 625 shows lots of decompile going on: |
Not at the moment. I've raised this issue: adoptium/STF#103. @pshipton - do you know which jcmd commands result in the creation of javacores and system dumps? |
|
@lumpfish |
|
https://ci.eclipse.org/openj9/job/Test_openjdk11_j9_special.system_x86-64_windows_mixed_Personal_testList_0/3/ |
pshipton
changed the title
|
jcmd should now be used to generate dumps on Windows: see adoptium/STF#103 |
|
https://ci.eclipse.org/openj9/job/Test_openjdk11_j9_special.system_x86-64_windows_mixed_Personal_testList_0/4/ |
|
I note this test appears to complete within 15min on other platforms. |
|
I guess this test isn't converted to use time based testing. I see some modes passing quickly, and other modes taking a longer period of time, 30min+. |
|
@Mesbah-Alam @lumpfish is it in progress to convert this test to time based? |
|
As explained in adoptium/aqa-systemtest#379 - the test is self limiting to two concurrent instances - so there is no point in running it one more than two threads. Any additional threads just terminate immediately upon seeing that there are already two threads running. As you have observed, it (usually?) takes > 5mins to run. If it were changed to time based, then when one instance ends another would replace it until the time limit is reached. In reality, also as explained in adoptium/aqa-systemtest#379, historically the tests have only been running one instance because they were being deemed to have timed out immediately because of a bug in stf (now fixed). So at the moment, the tests are being run exactly once down each of two threads. If they became '5 minute' tests they would mostly still run the same way. On some fast machines the test might be invoked more than once on those two threads. They would also still take longer than five minutes where they currently take longer than five minutes because running the test once takes more than 5 minutes. What is your concern - that they are not running long enough, that they do not take the same amount of time to run across platforms and modes, or something else? |
@lumpfish , I guess the answer to this is what Pete said in one of his previous comments -
|
|
But as I said, they take >30 mins to run once in those modes |
|
My primary concern is this failure, it appears the tests are often trying to run for longer than the timeout. Secondary, although this mode and some other modes can take 30+, there are modes being used where the testing takes less than 5 minutes. As previously discussed re time based testing, the shorter run time may not be long enough to warm up the JIT and find all the bugs. |
|
Looking at the diagnostics we have for the most recent failure #11904 (comment), I was hoping to have javacore files, but I only see core files created via procdump. |
|
I've gathered timing information and discovered that OSRG modes run slow only on Windows, but not on other platforms. Assigning to JIT to investigate why these modes are slow only on Windows. variation: (24) Mode107-OSRG variation: (25) Mode110-OSRG variation: (26) Mode610-OSRG variation: (27) Mode612-OSRG |
|
@nbhuiyan : could you have a look at this please? |
smlambert
pushed a commit
to adoptium/aqa-tests
that referenced
this issue
Mar 1, 2021
The OSRG modes run slow only on Windows, and contribute to Windows special.system tests timing out. The Mode Mode107-OSRG (24) can timeout the 60min limit and fail. Issue eclipse-openj9/openj9#11904 variation: (24) Mode107-OSRG JVM_OPTIONS: -Xgcpolicy:optthruput -Xdebug -Xrunjdwp:transport=dt_socket,address=8888,server=y,onthrow=no.pkg.foo,launch=echo -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 67min aix 8: 9min, 11: 4min osx 8: 3min, 11: 5min plinux 8: 2min, 11: 4min xlinux 8: 4min, 11: 10min zlinux 8: 4min, 11: 6min variation: (25) Mode110-OSRG JVM_OPTIONS: -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 32min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 4min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min variation: (26) Mode610-OSRG JVM_OPTIONS: -Xcompressedrefs -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 25min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 3min, 11: 4min xlinux 8: 5min, 11: 6min zlinux 8: 2min, 11: 2min variation: (27) Mode612-OSRG JVM_OPTIONS: -Xcompressedrefs -Xgcpolicy:gencon -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 33min aix 8: 4min, 11: 5min osx 8: 2min, 11: 4min plinux 8: 3min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min Signed-off-by: Peter Shipton <[email protected]>
pshipton
added a commit
to pshipton/openjdk-tests
that referenced
this issue
Mar 2, 2021
Fixes adoptium#2316 (comment) The OSRG modes run slow only on Windows, and contribute to Windows special.system tests timing out. The Mode Mode107-OSRG (24) can timeout the 60min limit and fail. Issue eclipse-openj9/openj9#11904 variation: (24) Mode107-OSRG JVM_OPTIONS: -Xgcpolicy:optthruput -Xdebug -Xrunjdwp:transport=dt_socket,address=8888,server=y,onthrow=no.pkg.foo,launch=echo -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 67min aix 8: 9min, 11: 4min osx 8: 3min, 11: 5min plinux 8: 2min, 11: 4min xlinux 8: 4min, 11: 10min zlinux 8: 4min, 11: 6min variation: (25) Mode110-OSRG JVM_OPTIONS: -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 32min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 4min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min variation: (26) Mode610-OSRG JVM_OPTIONS: -Xcompressedrefs -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 25min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 3min, 11: 4min xlinux 8: 5min, 11: 6min zlinux 8: 2min, 11: 2min variation: (27) Mode612-OSRG JVM_OPTIONS: -Xcompressedrefs -Xgcpolicy:gencon -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 33min aix 8: 4min, 11: 5min osx 8: 2min, 11: 4min plinux 8: 3min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min Signed-off-by: Peter Shipton <[email protected]>
pshipton
added a commit
to pshipton/openjdk-tests
that referenced
this issue
Mar 2, 2021
Fixes adoptium#2316 (comment) The OSRG modes run slow only on Windows, and contribute to Windows special.system tests timing out. The Mode Mode107-OSRG (24) can timeout the 60min limit and fail. Issue eclipse-openj9/openj9#11904 variation: (24) Mode107-OSRG JVM_OPTIONS: -Xgcpolicy:optthruput -Xdebug -Xrunjdwp:transport=dt_socket,address=8888,server=y,onthrow=no.pkg.foo,launch=echo -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 67min aix 8: 9min, 11: 4min osx 8: 3min, 11: 5min plinux 8: 2min, 11: 4min xlinux 8: 4min, 11: 10min zlinux 8: 4min, 11: 6min variation: (25) Mode110-OSRG JVM_OPTIONS: -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 32min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 4min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min variation: (26) Mode610-OSRG JVM_OPTIONS: -Xcompressedrefs -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 25min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 3min, 11: 4min xlinux 8: 5min, 11: 6min zlinux 8: 2min, 11: 2min variation: (27) Mode612-OSRG JVM_OPTIONS: -Xcompressedrefs -Xgcpolicy:gencon -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 33min aix 8: 4min, 11: 5min osx 8: 2min, 11: 4min plinux 8: 3min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min Signed-off-by: Peter Shipton <[email protected]>
smlambert
pushed a commit
to adoptium/aqa-tests
that referenced
this issue
Mar 2, 2021
Fixes #2316 (comment) The OSRG modes run slow only on Windows, and contribute to Windows special.system tests timing out. The Mode Mode107-OSRG (24) can timeout the 60min limit and fail. Issue eclipse-openj9/openj9#11904 variation: (24) Mode107-OSRG JVM_OPTIONS: -Xgcpolicy:optthruput -Xdebug -Xrunjdwp:transport=dt_socket,address=8888,server=y,onthrow=no.pkg.foo,launch=echo -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 67min aix 8: 9min, 11: 4min osx 8: 3min, 11: 5min plinux 8: 2min, 11: 4min xlinux 8: 4min, 11: 10min zlinux 8: 4min, 11: 6min variation: (25) Mode110-OSRG JVM_OPTIONS: -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 32min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 4min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min variation: (26) Mode610-OSRG JVM_OPTIONS: -Xcompressedrefs -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 25min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 3min, 11: 4min xlinux 8: 5min, 11: 6min zlinux 8: 2min, 11: 2min variation: (27) Mode612-OSRG JVM_OPTIONS: -Xcompressedrefs -Xgcpolicy:gencon -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 33min aix 8: 4min, 11: 5min osx 8: 2min, 11: 4min plinux 8: 3min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min Signed-off-by: Peter Shipton <[email protected]>
patkarns
added a commit
to patkarns/openjdk-tests
that referenced
this issue
Mar 4, 2021
* Add add-exports to HCRLateAttachWorkload_previewEnabled (adoptium#2208) - Add add-exports to HCRLateAttachWorkload_previewEnabled Signed-off-by: Longyu Zhang <[email protected]> * Convert iteration based STF based load tests to time-based tests (adoptium#2203) Signed-off-by: [email protected] <[email protected]> * AUTO: auto exclude test jdk_foreign_native impl=openj9 ver=16+ (adoptium#2219) - related: adoptium#2218 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: smlambert <[email protected]> * get.sh: stop mandating GNU tar --strip when extracting debug images (adoptium#2184) Signed-off-by: Stewart X Addison <[email protected]> * Remove references to eclipse-openj9/openj9#11135 (adoptium#2215) eclipse-openj9/openj9#11135 refers to an uber issue, which will be closed soon. So, the references to eclipse-openj9/openj9#11135 have been replaced with stand-alone trackers for the respective issues. Also, a duplicate for SuperMethodTest.java has been removed. Signed-off-by: Babneet Singh <[email protected]> * Exclude ChaCha20/ChaCha20Poly1305ParamTest.java (adoptium#2221) Issue eclipse-openj9/openj9#11689 Signed-off-by: Peter Shipton <[email protected]> * Update OpenJ9 JDK17 exclude list (adoptium#2217) * Apply adoptium#2188 to OpenJ9 JDK17 exclude list Apply adoptium#2188 to OpenJ9 JDK17 exclude list. Signed-off-by: Babneet Singh <[email protected]> * Apply adoptium#2192 to OpenJ9 JDK17 exclude list Apply adoptium#2192 to OpenJ9 JDK17 exclude list. Signed-off-by: Babneet Singh <[email protected]> * Apply adoptium#2215 to OpenJ9 JDK17 exclude list Apply adoptium#2215 to OpenJ9 JDK17 exclude list. Signed-off-by: Babneet Singh <[email protected]> * Fix charsets,i18n,language_tag and locale_matching for 16 and Create new 4 test cases (adoptium#2214) Co-authored-by: Kentaro Kuramochi <[email protected]> * Allow users to set lightweight option (adoptium#2216) The benefit of having lightweight checkout is to save space on Jenkins master. However, Jenkins does not support using Repository URL and Branch build parameters with lightweight checkout together (see https://issues.jenkins.io/browse/JENKINS-48431). In this PR, we set LIGHT_WEIGHT_CEHCKOUT to false and use Repository URL and Branch as build parameters by default. If LIGHT_WEIGHT_CEHCKOUT is set to true, set Repository URL and Branch to its explicit value Signed-off-by: lanxia <[email protected]> * Create ParallelStreamsLoadTest (adoptium#2229) Signed-off-by: Simon Rushton <[email protected]> * Remove _5m from ParallelStreamsLoadTest targets (adoptium#2231) Signed-off-by: Simon Rushton <[email protected]> * Enable java/lang/reflect/DefaultMethodMembers/FilterNotMostSpecific.java (adoptium#2230) This was forgot after resolving openj9 issue #7623 Signed-off-by: Jason Feng <[email protected]> * Merge openj9_custom into jdk_custom (adoptium#2232) - Remove openj9_custom to merge it into jdk_custom Signed-off-by: Longyu Zhang <[email protected]> * Move JdiTest playlist entry from lambdaLoadTest to otherLoadTest (adoptium#2233) Signed-off-by: Simon Rushton <[email protected]> * Reinstate ChaCha20Poly1305ParamTest (adoptium#2236) Issue eclipse-openj9/openj9#11689 Signed-off-by: Peter Shipton <[email protected]> * Replace deprecated triggers with new syntax (adoptium#2237) See https://github.com/jenkinsci/job-dsl-plugin/blob/master/docs/Migration.md#migrating-to-177 Signed-off-by: Adam Brousseau <[email protected]> * Update userGuide to include vendor disable feature (adoptium#2239) Signed-off-by: renfeiw <[email protected]> * Use the new format of disable in playlist file (adoptium#2240) Signed-off-by: renfeiw <[email protected]> * Add a custom playlist target to run system tests (adoptium#2234) Signed-off-by: [email protected] <[email protected]> * OpenJ9 AArch64: Enable a BigInteger test back again (adoptium#2242) This commit enables the following test, which used to fail with timeouts on AArch64: - java/math/BigInteger/LargeValueExceptions.java (eclipse-openj9/openj9#9082) Signed-off-by: KONNO Kazuhiro <[email protected]> * OpenJ9 AArch64: excluding stream/CountLargeTest (adoptium#2243) The following test is known to timeout with OpenJ9 AArch64. This commit excludes it. - java/util/stream/test/org/openjdk/tests/java/util/stream/CountLargeTest.java eclipse-openj9/openj9#9040 Signed-off-by: KONNO Kazuhiro <[email protected]> * Allow test autogen user to specify parallel params (adoptium#2248) - Param PARALLEL_DEFAULT will be set as default PARALLEL option - Param NUM_MACHINES will be set if passed Signed-off-by: Adam Brousseau <[email protected]> * Exclude com/sun/crypto AEAD ByteBuffer tests (adoptium#2249) Issue eclipse-openj9/openj9#11390 Signed-off-by: Peter Shipton <[email protected]> * Fix command injection vulnerability in autoTestPR (adoptium#2244) Fixes a command injection vulnerability with the auto test PR GitHub Workflow. See https://securitylab.github.com/research/github-actions-untrusted-input Also removes the unecessary `&& github.event_name != 'pull_request'` condition from the autoTestPR job because it will always be satisfied. (This workflow only triggers upon issue_comment, so `github.event_name` will never be 'pull_request') * Update the userGuide (adoptium#2259) - add a one line to cover the special case of auto disable Signed-off-by: renfeiw <[email protected]> * Add certain directories trigger for PR/push testing (adoptium#2241) * Add certain directories trigger for PR/push testing * use run-aqa with the target specified by the changed dirs * remove unnecessary push trigger, leave only PR trigger * merge openj9 and hotspot, deal with duplication situation * Combine getChangeLocation and getBuildLists into one getBuildLists, remove unnecessary code Co-authored-by: xichen1 <[email protected]> * Add 1 hour timeouts to all copyArtifacts steps (adoptium#2261) Avoid long hanging builds when copyArtifacts hangs Signed-off-by: Adam Brousseau <[email protected]> * Set Lightweight checkout on child jobs (adoptium#2251) Pass along the LIGHT_WEIGHT_CHECKOUT parameter to the generater in order to have children with the same value. Generally this will be false for Ginders and true eveywhere else Fixes adoptium#2250 Signed-off-by: Adam Brousseau <[email protected]> * Revert "Exclude com/sun/crypto AEAD ByteBuffer tests (adoptium#2249)" (adoptium#2266) This reverts commit 2227228. * AUTO: auto exclude test dacapo-jython impl=openj9 ver=16+ (adoptium#2269) - related: adoptium#2256 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: smlambert <[email protected]> * Only pass LIGHTWEIGHT to DSL if defined (adoptium#2270) - Currently it will put NULL which will cause toBoolean() to fail Related adoptium#2251 Signed-off-by: Adam Brousseau <[email protected]> * Add jdk_vector for jdk16+ (adoptium#2267) Signed-off-by: lanxia <[email protected]> * AUTO: auto exclude test jdk_foreign_native plat=.*(ppc|arm|390).* (adoptium#2280) - related: adoptium#2265 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: adamfarley <[email protected]> * temporarily fix bug of get changed files (adoptium#2277) * Exclude jdk_net related failed tests temporarily (adoptium#2282) - exclude jdk_net failed tests temporarily - for jdk 11, 15, and 16 on related platforms Signed-off-by: Longyu Zhang <[email protected]> * add suffix to ignore native tests (adoptium#2283) * add suffix to ignore native tests * remove unnecessary matrix.bulid_list in target list * Remove maketest.sh (adoptium#2285) Signed-off-by: [email protected] <[email protected]> * Add comment-triggered PR build action (adoptium#2222) * Add comment-triggered PR build action (adoptium#2199) Add workflow `action.yml` that triggers on a PR comment starting with the word `action` then runs AQA tests based on the parameters. Add `action_argparse.py` that parses PR comments to determine the parameters for AQA tests in the `action.yml` workflow. Co-authored-by: Xichen Pan <[email protected]> Co-authored-by: patkarns <[email protected]> * Update indentation for exclusion user guide (adoptium#2288) Signed-off-by: renfeiw <[email protected]> * Fix reportStatus job running on every comment (adoptium#2290) Prevents reportStatus from running on every comment by checking for the same conditions as parseComment: - The comment must begin with `run aqa` keyword - The comment must be a PR comment * Use any node for copying TAP files in parallel test builds (adoptium#2295) Resolve: adoptium#2262 Signed-off-by: lanxia <[email protected]> * Set ulimit for sharedclasscache multithread tests (adoptium#2274) Signed-off-by: Simon Rushton <[email protected]> * Add test excluded label for auto exclusion (adoptium#2300) Signed-off-by: renfeiw <[email protected]> * Change subset to version (adoptium#2301) Signed-off-by: renfeiw <[email protected]> * Fix GitHub head ref error (adoptium#2308) * test error * test error * remove parentheses * use github.head_ref * remove test file * Update userGuide to change subset to version (adoptium#2307) Signed-off-by: renfeiw <[email protected]> * remove PushOrPRtestbuild.yml (adoptium#2309) * Add alpine-linux and x86-64_solaris platforms (adoptium#2304) * Trial external_custom target Signed-off-by: Shelley Lambert <[email protected]> * Update build.xml Signed-off-by: Shelley Lambert <[email protected]> * Add alpine-linux platform label Signed-off-by: Shelley Lambert <[email protected]> * add bits Signed-off-by: Shelley Lambert <[email protected]> * Add x86_solaris Signed-off-by: Shelley Lambert <[email protected]> * Remove waitForANodeToBecomeActive post-parallel (adoptium#2312) This step is unnecessary now that we run on any node. Also remove function as it was only used for this step. Follow up to adoptium#2295 Signed-off-by: Adam Brousseau <[email protected]> * Remove ParallelStreamsLoadTest_special_J9 long running modes 113, 614 (adoptium#2315) Modes 113 and 614 take too long to run across all platforms, and contribute to special.system timeout problems on Windows. (5) 113: -Xgcpolicy:gencon -Xjit:count=0,optlevel=warm,gcOnResolve,rtResolve -Xmn512k -Xnocompressedrefs aix 8: 23min, 11: 31min osx 8: 10min, 11: 20min plinux 8: 18min, 11: 23min xlinux 8: 25min, 11: 40min zlinux 8: 12min, 11: 14min Windows 11: 30min (20) 614: -Xcompressedrefs -Xgcpolicy:gencon -Xjit:counts="- - - - - - 1 1 1 1000 250 250 - - - 10000 100000 10000",gcOnResolve,rtResolve,sampleInterval=2,scorchingSampleThreshold=10000,quickProfile -Xmn512k -Xcheck:gc:vmthreads:all:quiet aix 8: 19min, 11: 27min osx 8: 17min, 11: 17min plinux 8: 18min, 11: 22min xlinux 8: 24min, 11: 33min zlinux 8: 10min, 11: 14min Windows 11: 27min Signed-off-by: Peter Shipton <[email protected]> * Change Renaissance 0.9.0 to 0.11.0 in build.xml (adoptium#2314) Fixes adoptium#2298 * Disable ParallelStreamsLoadTest slow OSRG modes on Windows (adoptium#2316) The OSRG modes run slow only on Windows, and contribute to Windows special.system tests timing out. The Mode Mode107-OSRG (24) can timeout the 60min limit and fail. Issue eclipse-openj9/openj9#11904 variation: (24) Mode107-OSRG JVM_OPTIONS: -Xgcpolicy:optthruput -Xdebug -Xrunjdwp:transport=dt_socket,address=8888,server=y,onthrow=no.pkg.foo,launch=echo -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 67min aix 8: 9min, 11: 4min osx 8: 3min, 11: 5min plinux 8: 2min, 11: 4min xlinux 8: 4min, 11: 10min zlinux 8: 4min, 11: 6min variation: (25) Mode110-OSRG JVM_OPTIONS: -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 32min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 4min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min variation: (26) Mode610-OSRG JVM_OPTIONS: -Xcompressedrefs -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 25min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 3min, 11: 4min xlinux 8: 5min, 11: 6min zlinux 8: 2min, 11: 2min variation: (27) Mode612-OSRG JVM_OPTIONS: -Xcompressedrefs -Xgcpolicy:gencon -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 33min aix 8: 4min, 11: 5min osx 8: 2min, 11: 4min plinux 8: 3min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min Signed-off-by: Peter Shipton <[email protected]> * AUTO: auto exclude test renaissance-dec-tree ver=16+ (adoptium#2319) - related: adoptium#2255 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: smlambert <[email protected]> * Disable ParallelStreamsLoadTest slow OSRG modes on Windows (adoptium#2325) Fixes adoptium#2316 (comment) The OSRG modes run slow only on Windows, and contribute to Windows special.system tests timing out. The Mode Mode107-OSRG (24) can timeout the 60min limit and fail. Issue eclipse-openj9/openj9#11904 variation: (24) Mode107-OSRG JVM_OPTIONS: -Xgcpolicy:optthruput -Xdebug -Xrunjdwp:transport=dt_socket,address=8888,server=y,onthrow=no.pkg.foo,launch=echo -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 67min aix 8: 9min, 11: 4min osx 8: 3min, 11: 5min plinux 8: 2min, 11: 4min xlinux 8: 4min, 11: 10min zlinux 8: 4min, 11: 6min variation: (25) Mode110-OSRG JVM_OPTIONS: -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 32min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 4min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min variation: (26) Mode610-OSRG JVM_OPTIONS: -Xcompressedrefs -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 25min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 3min, 11: 4min xlinux 8: 5min, 11: 6min zlinux 8: 2min, 11: 2min variation: (27) Mode612-OSRG JVM_OPTIONS: -Xcompressedrefs -Xgcpolicy:gencon -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 33min aix 8: 4min, 11: 5min osx 8: 2min, 11: 4min plinux 8: 3min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min Signed-off-by: Peter Shipton <[email protected]> * Set parallel machine limit to the total number of machines (adoptium#2330) Resolve: eclipse-openj9/openj9#12088 Signed-off-by: lanxia <[email protected]> * Convert subset to version (adoptium#2329) - using version in playlists Signed-off-by: renfeiw <[email protected]> * Handle a git command's request for a username on Windows (adoptium#2287) If we call git ls-remote on Window, when the repo doesn't exist, the command asks for a username and hangs. This change disables the request for a username and provides more useful error messages. It also handles the problem appropriately, so instead of failing the build, it correctly sets the repo location to the non-'u' version. Added some timeout-handling code as well, in case someone's running this on a version of git older than 2.3, which may ignore the env var we're using to suppress the username request. Signed-off-by: Adam Farley <[email protected]> * OpenJ9 JDK17 Exclude java/lang/Thread/UncaughtExceptionsTest.java (adoptium#2333) Signed-off-by: Jason Feng <[email protected]> * Only run foreign native test on x86_64 or aarch64 platforms (adoptium#2334) This test requires CLinker, and the upstream community has no current intention to port it to new platforms. So we need to exclude the test on platforms that aren't x86-64 or aarch64. Signed-off-by: Adam Farley <[email protected]> * Enable download testimage for jdk8 (adoptium#2320) Signed-off-by: Sophia Guo <[email protected]> * Ignore leading and trailing whitespace in parsing (adoptium#2318) Also simplifies the mapping of comment/command arguments to outputs. * os upate to latest (adoptium#2331) * use sw.os.alpine-linux label instead of sw.os.musl for Alpine (adoptium#2336) Signed-off-by: Stewart X Addison <[email protected]> Co-authored-by: LongyuZhang <[email protected]> Co-authored-by: Mesbah Alam <[email protected]> Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: smlambert <[email protected]> Co-authored-by: Stewart X Addison <[email protected]> Co-authored-by: Babneet Singh <[email protected]> Co-authored-by: Peter Shipton <[email protected]> Co-authored-by: KKES335 <[email protected]> Co-authored-by: Kentaro Kuramochi <[email protected]> Co-authored-by: Lan Xia <[email protected]> Co-authored-by: Simon Rushton <[email protected]> Co-authored-by: Jason Feng <[email protected]> Co-authored-by: AdamBrousseau <[email protected]> Co-authored-by: Renfei Wang <[email protected]> Co-authored-by: KONNO Kazuhiro <[email protected]> Co-authored-by: Deric Cheung <[email protected]> Co-authored-by: XichenPan <[email protected]> Co-authored-by: xichen1 <[email protected]> Co-authored-by: adamfarley <[email protected]> Co-authored-by: yanshan24 <[email protected]> Co-authored-by: Shelley Lambert <[email protected]> Co-authored-by: YananW <[email protected]> Co-authored-by: Adam Farley <[email protected]> Co-authored-by: sophia-guo <[email protected]>
|
This won't be resolved for 0.26. |
0xdaryl
modified the milestones:
Release 0.26 (Java 8, 11, 16) Apr refresh,
Release 0.27 (Java 8, 11, 16) July refresh
smlambert
added a commit
to adoptium/aqa-tests
that referenced
this issue
Apr 28, 2021
* Create createTestList.py * Delete createTestList.py * Sync original repo (#5) * Add add-exports to HCRLateAttachWorkload_previewEnabled (#2208) - Add add-exports to HCRLateAttachWorkload_previewEnabled Signed-off-by: Longyu Zhang <[email protected]> * Convert iteration based STF based load tests to time-based tests (#2203) Signed-off-by: [email protected] <[email protected]> * AUTO: auto exclude test jdk_foreign_native impl=openj9 ver=16+ (#2219) - related: #2218 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: smlambert <[email protected]> * get.sh: stop mandating GNU tar --strip when extracting debug images (#2184) Signed-off-by: Stewart X Addison <[email protected]> * Remove references to eclipse-openj9/openj9#11135 (#2215) eclipse-openj9/openj9#11135 refers to an uber issue, which will be closed soon. So, the references to eclipse-openj9/openj9#11135 have been replaced with stand-alone trackers for the respective issues. Also, a duplicate for SuperMethodTest.java has been removed. Signed-off-by: Babneet Singh <[email protected]> * Exclude ChaCha20/ChaCha20Poly1305ParamTest.java (#2221) Issue eclipse-openj9/openj9#11689 Signed-off-by: Peter Shipton <[email protected]> * Update OpenJ9 JDK17 exclude list (#2217) * Apply #2188 to OpenJ9 JDK17 exclude list Apply #2188 to OpenJ9 JDK17 exclude list. Signed-off-by: Babneet Singh <[email protected]> * Apply #2192 to OpenJ9 JDK17 exclude list Apply #2192 to OpenJ9 JDK17 exclude list. Signed-off-by: Babneet Singh <[email protected]> * Apply #2215 to OpenJ9 JDK17 exclude list Apply #2215 to OpenJ9 JDK17 exclude list. Signed-off-by: Babneet Singh <[email protected]> * Fix charsets,i18n,language_tag and locale_matching for 16 and Create new 4 test cases (#2214) Co-authored-by: Kentaro Kuramochi <[email protected]> * Allow users to set lightweight option (#2216) The benefit of having lightweight checkout is to save space on Jenkins master. However, Jenkins does not support using Repository URL and Branch build parameters with lightweight checkout together (see https://issues.jenkins.io/browse/JENKINS-48431). In this PR, we set LIGHT_WEIGHT_CEHCKOUT to false and use Repository URL and Branch as build parameters by default. If LIGHT_WEIGHT_CEHCKOUT is set to true, set Repository URL and Branch to its explicit value Signed-off-by: lanxia <[email protected]> * Create ParallelStreamsLoadTest (#2229) Signed-off-by: Simon Rushton <[email protected]> * Remove _5m from ParallelStreamsLoadTest targets (#2231) Signed-off-by: Simon Rushton <[email protected]> * Enable java/lang/reflect/DefaultMethodMembers/FilterNotMostSpecific.java (#2230) This was forgot after resolving openj9 issue #7623 Signed-off-by: Jason Feng <[email protected]> * Merge openj9_custom into jdk_custom (#2232) - Remove openj9_custom to merge it into jdk_custom Signed-off-by: Longyu Zhang <[email protected]> * Move JdiTest playlist entry from lambdaLoadTest to otherLoadTest (#2233) Signed-off-by: Simon Rushton <[email protected]> * Reinstate ChaCha20Poly1305ParamTest (#2236) Issue eclipse-openj9/openj9#11689 Signed-off-by: Peter Shipton <[email protected]> * Replace deprecated triggers with new syntax (#2237) See https://github.com/jenkinsci/job-dsl-plugin/blob/master/docs/Migration.md#migrating-to-177 Signed-off-by: Adam Brousseau <[email protected]> * Update userGuide to include vendor disable feature (#2239) Signed-off-by: renfeiw <[email protected]> * Use the new format of disable in playlist file (#2240) Signed-off-by: renfeiw <[email protected]> * Add a custom playlist target to run system tests (#2234) Signed-off-by: [email protected] <[email protected]> * OpenJ9 AArch64: Enable a BigInteger test back again (#2242) This commit enables the following test, which used to fail with timeouts on AArch64: - java/math/BigInteger/LargeValueExceptions.java (eclipse-openj9/openj9#9082) Signed-off-by: KONNO Kazuhiro <[email protected]> * OpenJ9 AArch64: excluding stream/CountLargeTest (#2243) The following test is known to timeout with OpenJ9 AArch64. This commit excludes it. - java/util/stream/test/org/openjdk/tests/java/util/stream/CountLargeTest.java eclipse-openj9/openj9#9040 Signed-off-by: KONNO Kazuhiro <[email protected]> * Allow test autogen user to specify parallel params (#2248) - Param PARALLEL_DEFAULT will be set as default PARALLEL option - Param NUM_MACHINES will be set if passed Signed-off-by: Adam Brousseau <[email protected]> * Exclude com/sun/crypto AEAD ByteBuffer tests (#2249) Issue eclipse-openj9/openj9#11390 Signed-off-by: Peter Shipton <[email protected]> * Fix command injection vulnerability in autoTestPR (#2244) Fixes a command injection vulnerability with the auto test PR GitHub Workflow. See https://securitylab.github.com/research/github-actions-untrusted-input Also removes the unecessary `&& github.event_name != 'pull_request'` condition from the autoTestPR job because it will always be satisfied. (This workflow only triggers upon issue_comment, so `github.event_name` will never be 'pull_request') * Update the userGuide (#2259) - add a one line to cover the special case of auto disable Signed-off-by: renfeiw <[email protected]> * Add certain directories trigger for PR/push testing (#2241) * Add certain directories trigger for PR/push testing * use run-aqa with the target specified by the changed dirs * remove unnecessary push trigger, leave only PR trigger * merge openj9 and hotspot, deal with duplication situation * Combine getChangeLocation and getBuildLists into one getBuildLists, remove unnecessary code Co-authored-by: xichen1 <[email protected]> * Add 1 hour timeouts to all copyArtifacts steps (#2261) Avoid long hanging builds when copyArtifacts hangs Signed-off-by: Adam Brousseau <[email protected]> * Set Lightweight checkout on child jobs (#2251) Pass along the LIGHT_WEIGHT_CHECKOUT parameter to the generater in order to have children with the same value. Generally this will be false for Ginders and true eveywhere else Fixes #2250 Signed-off-by: Adam Brousseau <[email protected]> * Revert "Exclude com/sun/crypto AEAD ByteBuffer tests (#2249)" (#2266) This reverts commit 2227228. * AUTO: auto exclude test dacapo-jython impl=openj9 ver=16+ (#2269) - related: #2256 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: smlambert <[email protected]> * Only pass LIGHTWEIGHT to DSL if defined (#2270) - Currently it will put NULL which will cause toBoolean() to fail Related #2251 Signed-off-by: Adam Brousseau <[email protected]> * Add jdk_vector for jdk16+ (#2267) Signed-off-by: lanxia <[email protected]> * AUTO: auto exclude test jdk_foreign_native plat=.*(ppc|arm|390).* (#2280) - related: #2265 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: adamfarley <[email protected]> * temporarily fix bug of get changed files (#2277) * Exclude jdk_net related failed tests temporarily (#2282) - exclude jdk_net failed tests temporarily - for jdk 11, 15, and 16 on related platforms Signed-off-by: Longyu Zhang <[email protected]> * add suffix to ignore native tests (#2283) * add suffix to ignore native tests * remove unnecessary matrix.bulid_list in target list * Remove maketest.sh (#2285) Signed-off-by: [email protected] <[email protected]> * Add comment-triggered PR build action (#2222) * Add comment-triggered PR build action (#2199) Add workflow `action.yml` that triggers on a PR comment starting with the word `action` then runs AQA tests based on the parameters. Add `action_argparse.py` that parses PR comments to determine the parameters for AQA tests in the `action.yml` workflow. Co-authored-by: Xichen Pan <[email protected]> Co-authored-by: patkarns <[email protected]> * Update indentation for exclusion user guide (#2288) Signed-off-by: renfeiw <[email protected]> * Fix reportStatus job running on every comment (#2290) Prevents reportStatus from running on every comment by checking for the same conditions as parseComment: - The comment must begin with `run aqa` keyword - The comment must be a PR comment * Use any node for copying TAP files in parallel test builds (#2295) Resolve: #2262 Signed-off-by: lanxia <[email protected]> * Set ulimit for sharedclasscache multithread tests (#2274) Signed-off-by: Simon Rushton <[email protected]> * Add test excluded label for auto exclusion (#2300) Signed-off-by: renfeiw <[email protected]> * Change subset to version (#2301) Signed-off-by: renfeiw <[email protected]> * Fix GitHub head ref error (#2308) * test error * test error * remove parentheses * use github.head_ref * remove test file * Update userGuide to change subset to version (#2307) Signed-off-by: renfeiw <[email protected]> * remove PushOrPRtestbuild.yml (#2309) * Add alpine-linux and x86-64_solaris platforms (#2304) * Trial external_custom target Signed-off-by: Shelley Lambert <[email protected]> * Update build.xml Signed-off-by: Shelley Lambert <[email protected]> * Add alpine-linux platform label Signed-off-by: Shelley Lambert <[email protected]> * add bits Signed-off-by: Shelley Lambert <[email protected]> * Add x86_solaris Signed-off-by: Shelley Lambert <[email protected]> * Remove waitForANodeToBecomeActive post-parallel (#2312) This step is unnecessary now that we run on any node. Also remove function as it was only used for this step. Follow up to #2295 Signed-off-by: Adam Brousseau <[email protected]> * Remove ParallelStreamsLoadTest_special_J9 long running modes 113, 614 (#2315) Modes 113 and 614 take too long to run across all platforms, and contribute to special.system timeout problems on Windows. (5) 113: -Xgcpolicy:gencon -Xjit:count=0,optlevel=warm,gcOnResolve,rtResolve -Xmn512k -Xnocompressedrefs aix 8: 23min, 11: 31min osx 8: 10min, 11: 20min plinux 8: 18min, 11: 23min xlinux 8: 25min, 11: 40min zlinux 8: 12min, 11: 14min Windows 11: 30min (20) 614: -Xcompressedrefs -Xgcpolicy:gencon -Xjit:counts="- - - - - - 1 1 1 1000 250 250 - - - 10000 100000 10000",gcOnResolve,rtResolve,sampleInterval=2,scorchingSampleThreshold=10000,quickProfile -Xmn512k -Xcheck:gc:vmthreads:all:quiet aix 8: 19min, 11: 27min osx 8: 17min, 11: 17min plinux 8: 18min, 11: 22min xlinux 8: 24min, 11: 33min zlinux 8: 10min, 11: 14min Windows 11: 27min Signed-off-by: Peter Shipton <[email protected]> * Change Renaissance 0.9.0 to 0.11.0 in build.xml (#2314) Fixes #2298 * Disable ParallelStreamsLoadTest slow OSRG modes on Windows (#2316) The OSRG modes run slow only on Windows, and contribute to Windows special.system tests timing out. The Mode Mode107-OSRG (24) can timeout the 60min limit and fail. Issue eclipse-openj9/openj9#11904 variation: (24) Mode107-OSRG JVM_OPTIONS: -Xgcpolicy:optthruput -Xdebug -Xrunjdwp:transport=dt_socket,address=8888,server=y,onthrow=no.pkg.foo,launch=echo -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 67min aix 8: 9min, 11: 4min osx 8: 3min, 11: 5min plinux 8: 2min, 11: 4min xlinux 8: 4min, 11: 10min zlinux 8: 4min, 11: 6min variation: (25) Mode110-OSRG JVM_OPTIONS: -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 32min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 4min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min variation: (26) Mode610-OSRG JVM_OPTIONS: -Xcompressedrefs -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 25min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 3min, 11: 4min xlinux 8: 5min, 11: 6min zlinux 8: 2min, 11: 2min variation: (27) Mode612-OSRG JVM_OPTIONS: -Xcompressedrefs -Xgcpolicy:gencon -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 33min aix 8: 4min, 11: 5min osx 8: 2min, 11: 4min plinux 8: 3min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min Signed-off-by: Peter Shipton <[email protected]> * AUTO: auto exclude test renaissance-dec-tree ver=16+ (#2319) - related: #2255 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: smlambert <[email protected]> * Disable ParallelStreamsLoadTest slow OSRG modes on Windows (#2325) Fixes #2316 (comment) The OSRG modes run slow only on Windows, and contribute to Windows special.system tests timing out. The Mode Mode107-OSRG (24) can timeout the 60min limit and fail. Issue eclipse-openj9/openj9#11904 variation: (24) Mode107-OSRG JVM_OPTIONS: -Xgcpolicy:optthruput -Xdebug -Xrunjdwp:transport=dt_socket,address=8888,server=y,onthrow=no.pkg.foo,launch=echo -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 67min aix 8: 9min, 11: 4min osx 8: 3min, 11: 5min plinux 8: 2min, 11: 4min xlinux 8: 4min, 11: 10min zlinux 8: 4min, 11: 6min variation: (25) Mode110-OSRG JVM_OPTIONS: -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 32min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 4min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min variation: (26) Mode610-OSRG JVM_OPTIONS: -Xcompressedrefs -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation -Xgcpolicy:gencon Windows 11: 25min aix 8: 4min, 11: 5min osx 8: 3min, 11: 2min plinux 8: 3min, 11: 4min xlinux 8: 5min, 11: 6min zlinux 8: 2min, 11: 2min variation: (27) Mode612-OSRG JVM_OPTIONS: -Xcompressedrefs -Xgcpolicy:gencon -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation Windows 11: 33min aix 8: 4min, 11: 5min osx 8: 2min, 11: 4min plinux 8: 3min, 11: 5min xlinux 8: 4min, 11: 6min zlinux 8: 1min, 11: 2min Signed-off-by: Peter Shipton <[email protected]> * Set parallel machine limit to the total number of machines (#2330) Resolve: eclipse-openj9/openj9#12088 Signed-off-by: lanxia <[email protected]> * Convert subset to version (#2329) - using version in playlists Signed-off-by: renfeiw <[email protected]> * Handle a git command's request for a username on Windows (#2287) If we call git ls-remote on Window, when the repo doesn't exist, the command asks for a username and hangs. This change disables the request for a username and provides more useful error messages. It also handles the problem appropriately, so instead of failing the build, it correctly sets the repo location to the non-'u' version. Added some timeout-handling code as well, in case someone's running this on a version of git older than 2.3, which may ignore the env var we're using to suppress the username request. Signed-off-by: Adam Farley <[email protected]> * OpenJ9 JDK17 Exclude java/lang/Thread/UncaughtExceptionsTest.java (#2333) Signed-off-by: Jason Feng <[email protected]> * Only run foreign native test on x86_64 or aarch64 platforms (#2334) This test requires CLinker, and the upstream community has no current intention to port it to new platforms. So we need to exclude the test on platforms that aren't x86-64 or aarch64. Signed-off-by: Adam Farley <[email protected]> * Enable download testimage for jdk8 (#2320) Signed-off-by: Sophia Guo <[email protected]> * Ignore leading and trailing whitespace in parsing (#2318) Also simplifies the mapping of comment/command arguments to outputs. * os upate to latest (#2331) * use sw.os.alpine-linux label instead of sw.os.musl for Alpine (#2336) Signed-off-by: Stewart X Addison <[email protected]> Co-authored-by: LongyuZhang <[email protected]> Co-authored-by: Mesbah Alam <[email protected]> Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: smlambert <[email protected]> Co-authored-by: Stewart X Addison <[email protected]> Co-authored-by: Babneet Singh <[email protected]> Co-authored-by: Peter Shipton <[email protected]> Co-authored-by: KKES335 <[email protected]> Co-authored-by: Kentaro Kuramochi <[email protected]> Co-authored-by: Lan Xia <[email protected]> Co-authored-by: Simon Rushton <[email protected]> Co-authored-by: Jason Feng <[email protected]> Co-authored-by: AdamBrousseau <[email protected]> Co-authored-by: Renfei Wang <[email protected]> Co-authored-by: KONNO Kazuhiro <[email protected]> Co-authored-by: Deric Cheung <[email protected]> Co-authored-by: XichenPan <[email protected]> Co-authored-by: xichen1 <[email protected]> Co-authored-by: adamfarley <[email protected]> Co-authored-by: yanshan24 <[email protected]> Co-authored-by: Shelley Lambert <[email protected]> Co-authored-by: YananW <[email protected]> Co-authored-by: Adam Farley <[email protected]> Co-authored-by: sophia-guo <[email protected]> * Update testJobTemplate Testing param separator * Update testJobTemplate adding styles to header * Update testJobTemplate Categorized all parameters * Update testJobTemplate updated CSS variables * Update testJobTemplate * Update testJobTemplate * Update testJobTemplate * Update testJobTemplate * Update testJobTemplate * Update testJobTemplate * Update testJobTemplate * Update testJobTemplate * Update testJobTemplate * Update testJobTemplate * Update testJobTemplate Testing parameter separator * Update testJobTemplate added CSS styles to parameter separators * Update testJobTemplate to better reflect mockup * Update testJobTemplate * update testJobTemplate, added subtitle styling * fixed typo * Update testJobTemplate * Update testJobTemplate * Update testJobTemplate * Style parameters in testJobTemplate * Add link to Grinder Wiki * edited styling for Grinder wiki link * edited 'grinder wiki' positioning * Added padding to 'Grinder Wiki' link * Updated JDK_BRANCH parameter order * added spaces back to copyright * Update some parameters' default values * Removed extra tab * Removed Grinder wiki link and separator bg color * removed extra spaces in testJobTemplate * added examples in CUSTOM_TARGET * fixed /n error * changed multiline string declaration * added triple quotations around description * moved CUSTOM_TARGET examples to the right spot Co-authored-by: LongyuZhang <[email protected]> Co-authored-by: Mesbah Alam <[email protected]> Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: smlambert <[email protected]> Co-authored-by: Stewart X Addison <[email protected]> Co-authored-by: Babneet Singh <[email protected]> Co-authored-by: Peter Shipton <[email protected]> Co-authored-by: KKES335 <[email protected]> Co-authored-by: Kentaro Kuramochi <[email protected]> Co-authored-by: Lan Xia <[email protected]> Co-authored-by: Simon Rushton <[email protected]> Co-authored-by: Jason Feng <[email protected]> Co-authored-by: AdamBrousseau <[email protected]> Co-authored-by: Renfei Wang <[email protected]> Co-authored-by: KONNO Kazuhiro <[email protected]> Co-authored-by: Deric Cheung <[email protected]> Co-authored-by: XichenPan <[email protected]> Co-authored-by: xichen1 <[email protected]> Co-authored-by: adamfarley <[email protected]> Co-authored-by: yanshan24 <[email protected]> Co-authored-by: Shelley Lambert <[email protected]> Co-authored-by: YananW <[email protected]> Co-authored-by: Adam Farley <[email protected]> Co-authored-by: sophia-guo <[email protected]>
|
No progress on this for 0.27. Moving to 0.28. |
0xdaryl
modified the milestones:
Release 0.27 (Java 8, 11, 16) July refresh,
Release 0.28 (Java 17)
tajila
modified the milestones:
Release 0.28 (Java 17),
Release 0.29 (Java 8, 11, 17) Oct refresh
pshipton
modified the milestones:
Release 0.29 (Java 8, 11, 17) Oct refresh,
Release 0.30 (Java 8, 11, 17) Jan refresh
0xdaryl
modified the milestones:
Release 0.30 (Java 8, 11, 17) Jan refresh,
Release 0.31 (Java 18)
pshipton
modified the milestones:
Release 0.31 (Java 18),
Release 0.32 (Java 8, 11, 17, 18) Apr refresh
pshipton
modified the milestones:
Release 0.32 (Java 8, 11, 17, 18) Apr refresh,
Tracking
|
JDK11 ppc64le_linux milestone 2( |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_special.system_x86-32_windows_Personal_testList_0/10
ParallelStreamsLoadTest_special_J9_24
variation: Mode107-OSRG
JVM_OPTIONS: -Xgcpolicy:optthruput -Xdebug -Xrunjdwp:transport=dt_socket,address=8888,server=y,onthrow=no.pkg.foo,launch=echo -Xjit:enableOSR,enableOSROnGuardFailure,count=1,disableAsyncCompilation
There are 3 core files
https://140-211-168-230-openstack.osuosl.org/artifactory/ci-eclipse-openj9/Test/Test_openjdk8_j9_special.system_x86-32_windows_Personal_testList_0/10/system_test_output.tar.gz
First core shows these interesting threads:
Next two cores show the following:
The text was updated successfully, but these errors were encountered: