-
Notifications
You must be signed in to change notification settings - Fork 721
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
DaaLoadTest ConcurrentScavenge segfault or assert #11616
Comments
similar failure is mentioned in #11531 , I guess one of items is duplicate of another |
#11531 (comment) is a dup of this issue. It doesn't belong in #11531. |
There are more failures in subsequent builds, not detailed here. |
@fjeremic |
Interestingly enough, I can only consistently reproduce the issue if I run the test within GDB. Outside of it, it’s intermittent. I was also able to filter the JIT methods that's causing the issue, and they are:
Used limitFile to filter out these, and if I leave any of these methods to be compiled the test fails. |
It's causing quite a few failures in the builds, but the failure is limited to an easily identified set of tests. I'm ok to leave the code in this week, but please revert it before the next weekend if it can't be fixed. |
@pshipton fyi
|
There has been some progress regarding the diagnosis of the failures. There seems to be some code related to read barrier in concurrent scavenger that thinks |
@pshipton i'm going to exclude this test (DaaLoadTest_all_ConcurrentScavenge_0), as it is causing numerous hung processes on Adopt machines, causing subsequent build/test issues. |
@pshipton @smlambert fyi, the Adopt DaaLoadTest hung Java jobs are due to the excessive DaaAllTest timeouts of 15hours, I have raised: adoptium/openj9-systemtest#121 |
@andrew-m-leonard is there a PR for excluding the test? I think it's more than one test that can fail. We could just back out the problematic change. We haven't been seeing any hangs in the OpenJ9 builds, so it didn't seem much of a problem. |
I'll keep the test excluded label until it's determined if anything was actually excluded. |
@fjeremic fyi, looks like the hang in #11616 (comment) is related to the jitdump. There is a |
Although #11507 was reverted, there was another DaaLoadTest failure. Is this a different problem, or is this Issue not caused by 11507? @fjeremic @bragaigor https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_sanity.system_ppc64_aix_Nightly/612
|
@dmitripivkine fyi as well the previous comment, since it's crashing in the GC. |
Seems this latest crash belongs in #10489, I'll add it there. |
#11507 was indeed causing the issue, maybe 11507 made the issue more obvious and to crash more often? I tested locally and eclipse-omr/omr#5751 seemed to have fixed 11507 problem. But not sure if this is last crash could be related or if Filip's fix would solve it. |
last crash is unrelated known problem #10489 |
Same commit as eclipse-openj9#11507 The issue related to concurrent scavenger (eclipse-openj9#11616) was solved by eclipse-omr/omr#5751 We can now safely re-introduce dataAddr again. Signed-off-by: Igor Braga <higorb1@gmail.com>
Same commit as eclipse-openj9#11507 The issue related to concurrent scavenger (eclipse-openj9#11616) was solved by eclipse-omr/omr#5751 We can now safely re-introduce dataAddr again. Conflicts: runtime/compiler/x/codegen/J9TreeEvaluator.cpp Change-Id: I827e6d63c33b982b828820f486a0e16a857cec3a Signed-off-by: Igor Braga <higorb1@gmail.com> Signed-off-by: Lin Hu <linhu@ca.ibm.com>
Same commit as eclipse-openj9#11507 The issue related to concurrent scavenger (eclipse-openj9#11616) was solved by eclipse-omr/omr#5751 We can now safely re-introduce dataAddr again. Signed-off-by: Igor Braga <higorb1@gmail.com>
Changes from previous build
eclipse-openj9/openj9-omr@d2a0060...bded46c
56d0a4a...a6756ee
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_sanity.system_ppc64le_linux_xl_Nightly_testList_0/34
DaaLoadTest_daa1_ConcurrentScavenge_0
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_sanity.system_s390x_linux_xl_Nightly_testList_0/37/
DaaLoadTest_daa1_ConcurrentScavenge_0
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_sanity.system_x86-64_linux_xl_Nightly_testList_1/34/
DaaLoadTest_daa1_ConcurrentScavenge_0
https://ci.eclipse.org/openj9/job/Test_openjdk8_j9_sanity.system_x86-64_linux_xl_Nightly_testList_0/34
DaaLoadTest_daa2_ConcurrentScavenge_0
https://ci.eclipse.org/openj9/job/Test_openjdk11_j9_sanity.system_ppc64_aix_xl_Nightly_testList_1/32/
DaaLoadTest_daa1_ConcurrentScavenge_0
https://ci.eclipse.org/openj9/job/Test_openjdk11_j9_extended.system_s390x_linux_xl_Nightly_testList_1/35/
DaaLoadTest_all_ConcurrentScavenge_0
https://ci.eclipse.org/openj9/job/Test_openjdk11_j9_sanity.system_x86-64_linux_xl_Nightly_testList_1/33/
DaaLoadTest_daa1_ConcurrentScavenge_0
The text was updated successfully, but these errors were encountered: