-
-
Notifications
You must be signed in to change notification settings - Fork 314
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
General Triage of extended.openjdk tests #2137
Comments
Fresh run of extended.openjdk on aarch64 (https://ci.adoptopenjdk.net/computer/test-packet-ubuntu1604-armv8-1/), has 364 test failures, most are java/awt test cases with 'not a focus owner' exceptions ( |
FYI, @llxia @renfeiw
Assume related to adoptium/TKG#131 and #2146 |
Mode150/Mode650 was set the same as Mode110/Mode610 in ottawa.csv. It appears that the behavior of Mode150 and Mode110 are different on win32. I suspect Mode150 cannot run on all 32-bit. I will update ottawa.csv. Mode650 (-XX:-UseCompressedOops) passed on win32 OpenJ9 jvm: |
@llxia - all extended.openjdk targets are also failing against corretto trying to run Mode150 (see: https://trss.adoptopenjdk.net/allTestsInfo?buildId=5ffa04462bad7850660759cd) |
All openjdk tests with hotspot on 32 platform failed.
-XX:-UseCompressedOops works only for 64-bit JVMs. |
#2146 why Openjdk test targets limited to impl = hotspot need to add both mode150 and Mode650 with mode150 skipped? By default +CheckCompressedOops option is enabled, and compressed pointers are used. |
Re #2137 (comment), For hotspot builds, Mode150 is skipped due to the platform |
Missing utilities on aarch64 machines, some tests seem to expect certain utility commands ("which" and "hostname" expected). Example: https://ci.adoptopenjdk.net/job/Test_openjdk11_j9_extended.openjdk_aarch64_linux_xl/2/testReport/java_nio_file_attribute_BasicFileAttributeView_UnixSocketFile/java/UnixSocketFile/ - requires "which" to be present https://ci.adoptopenjdk.net/job/Test_openjdk11_j9_extended.openjdk_aarch64_linux_xl/2/testReport/sun_net_www_protocol_https_HttpsURLConnection_PostThruProxyWithAuth/java/PostThruProxyWithAuth/ and PostThruProxy.java- expects hostname to be present |
From jdk_client_sanity target (all 25 tests fail), output from xlinux jtr file for ColorChooserDemoTest:
|
auto exclude test jdk_client_sanity |
- related: #2137 (comment) Signed-off-by: GitHub <[email protected]>
- related: #2137 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: smlambert <[email protected]>
jdk_security3 failures tracked and excluded under #2123 |
jdk_2d suffers from the same type of failure seen with the jdk_client_sanity target, example of output from that group of tests, from an xlinux jdk11 hs run:
|
auto exclude test jdk_2d |
- related: #2137 (comment) Signed-off-by: GitHub <[email protected]>
- related: #2137 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: smlambert <[email protected]>
auto exclude test jdk_awt |
- related: #2137 (comment) Signed-off-by: GitHub <[email protected]>
- related: #2137 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: smlambert <[email protected]>
jdk_tools target is missing $(JDK_NATIVE_OPTIONS) which causes several tests to fail in that target. |
This is happening on the new Fedora installs in docker containers. Ubuntu seems to have them by default. Fixing now. |
auto exclude test jdk_imageio |
- related: #2137 (comment) Signed-off-by: GitHub <[email protected]>
auto exclude test jdk_imageio plat=.mac. |
- related: #2137 (comment) Signed-off-by: GitHub <[email protected]>
- related: #2137 (comment) Signed-off-by: GitHub <[email protected]> Co-authored-by: smlambert <[email protected]>
As we wish to enable extended.openjdk tests in weekly builds (and release builds), we must triage the various failures within them. There are any where between 70 to nearly 300 test cases failures depending on the platform, but the number of core issues that cause this number of failures seems to be small. This issue lists the set of core issues that account for most of the failures (and will allow us to temporarily disable tests while the issues are investigated and fixed):
Individually raised issues for individually failing test classes (rather than by test target):
The text was updated successfully, but these errors were encountered: