jdk_nio_0_FAILED sun/nio/cs/TestStringCoding.java RuntimeException: getBytes(csn) failed -> x-IBM29626C about openj9

[pshipton]

The jdk17 failure is repeatable when running just the single test, it fails 100%.
https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38154

Weekly history
610 - failed
605 - didn't run
596 - builds failed (no testing) due to infra
589 - passed

609 https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38158/ - failed
608 https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38160/ - failed
607 https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38161/ - failed
606 https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38162/ - crash
605 https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38159/ - crash #18979
604 https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38163/ - failed
602 https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38168/ - failed
601 https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38167/ - failed
600 https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38165/ - failed
597 (c92d583, 410ef0d23) https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38166/ - failed
596 (8dc58cc, cbf3bf87c) https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38164/ - passing

8dc58cc...c92d583
eclipse-openj9/openj9-omr@cbf3bf8...410ef0d

@hzongaro fyi, there are possibly two problems in this issue, although maybe related. The most recent is a 100% failure on Z with a short list of possibilities in the previous compares. I'll narrow it down.

eclipse/omr#7222
https://openj9-jenkins.osuosl.org/job/Pipeline-Build-Test-Personal/495/
https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38172/ - passing

eclipse/omr#7229
https://openj9-jenkins.osuosl.org/job/Pipeline-Build-Test-Personal/496/
https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38171/ - passing

#18819
https://openj9-jenkins.osuosl.org/job/Pipeline-Build-Test-Personal/497/
https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38173/ - failing

[pshipton]

@knn-k @Akira1Saitoh pls take a look. Note this is a blocker for the upcoming release.

[Akira1Saitoh]

In #18819, transformIndexOfKnownString was enabled for StringLatin1.indexOfChar. transformIndexOfKnownString seems to be incorrectly transforming the trees.

In sun/nio/cs/ext/IBM29626C$Encoder.encodeSingle(C[B)I, String.indexOf is called against the String literal "\u00A2\u00A3\u00AC". However, this call is transformed into the following tree:

   6256 [   144] O^O VALUE PROPAGATION: Replacing indexOf call node [0000FFFF88822A60] on known string receiver with equivalent iselect tree
   6257 [   145] O^O VALUE PROPAGATION: Replacing indexOf call node [0000FFFF88822A60] on known string receiver with equivalent iselect tree
   6258 [   146] O^O VALUE PROPAGATION: Replacing indexOf call node [0000FFFF88822A60] on known string receiver with equivalent iselect tree
...
   6530 n547n     treetop                                                                             [    0xffff88822ab0] bci=[4,32,213] rc=0 vc=137 vn=- li=- udi=- nc=1
   6531 n546n       PassThrough                                                                       [    0xffff88822a60] bci=[4,32,213] rc=3 vc=137 vn=- li=- udi=- nc=1
   6532 n790n         iselect ()                                                                      [    0xffff888276a0] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=3 flg=0x20
   6533 n788n           icmpeq                                                                        [    0xffff88827600] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=2
   6534 n543n             ==>iload
   6535 n787n             iconst -94 (X!=0 X<=0 )                                                     [    0xffff888275b0] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=0 flg=0x204
   6536 n789n           iconst 0 (X==0 X>=0 X<=0 )                                                    [    0xffff88827650] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=0 flg=0x302
   6537 n786n           iselect ()                                                                    [    0xffff88827560] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=3 flg=0x20
   6538 n784n             icmpeq                                                                      [    0xffff888274c0] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=2
   6539 n543n               ==>iload
   6540 n783n               iconst -93 (X!=0 X<=0 )                                                   [    0xffff88827470] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=0 flg=0x204
   6541 n785n             iconst 1 (X!=0 X>=0 )                                                       [    0xffff88827510] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=0 flg=0x104
   6542 n782n             iselect ()                                                                  [    0xffff88827420] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=3 flg=0x20
   6543 n780n               icmpeq                                                                    [    0xffff88827380] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=2
   6544 n543n                 ==>iload
   6545 n779n                 iconst -84 (X!=0 X<=0 )                                                 [    0xffff88827330] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=0 flg=0x204
   6546 n781n               iconst 2 (X!=0 X>=0 )                                                     [    0xffff888273d0] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=0 flg=0x104
   6547 n778n               iconst -1 (X!=0 X<=0 )                                                    [    0xffff888272e0] bci=[4,32,213] rc=1 vc=0 vn=- li=- udi=- nc=0 flg=0x204

The tree compares the input character with -94, -93, and -84, but those values should be 0xa2, 0xa3, and 0xac.

[pshipton]

cent8x86-svl-rt5-1, rhel8x86-svl-rt6-1 succeeded.

Failures:

rhel8x86-svl-rt7-1
1/20 sun/nio/cs/TestStringCoding.java

sles15x86-svl-rt6-1
20 failures (100%) of java/nio/file/Files/CopyMoveVariations.java
2/20 sun/nio/cs/TestStringCoding.java

ubu22x86-svl-rt4-1
3/20 sun/nio/cs/TestStringCoding.java

[JasonFengJ9]

Occurred at JDK17 s390x_linux(rhel8s390x-svl-rt1-1) as well.

[2024-02-24T17:06:18.040Z] variation: Mode150
[2024-02-24T17:06:18.040Z] JVM_OPTIONS:  -XX:+UseCompressedOops -Xverbosegclog 

[2024-02-24T17:18:13.495Z] TEST: sun/nio/cs/TestStringCoding.java

[2024-02-24T17:18:13.496Z] STDERR:
[2024-02-24T17:18:13.496Z] java.lang.RuntimeException: getBytes(csn) failed  -> x-IBM29626C
[2024-02-24T17:18:13.496Z] 	at TestStringCoding.testGetBytes(TestStringCoding.java:164)
[2024-02-24T17:18:13.496Z] 	at TestStringCoding.main(TestStringCoding.java:64)
[2024-02-24T17:18:13.496Z] 	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
[2024-02-24T17:18:13.496Z] 	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:77)
[2024-02-24T17:18:13.496Z] 	at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
[2024-02-24T17:18:13.496Z] 	at java.base/java.lang.reflect.Method.invoke(Method.java:574)
[2024-02-24T17:18:13.496Z] 	at com.sun.javatest.regtest.agent.MainWrapper$MainTask.run(MainWrapper.java:138)
[2024-02-24T17:18:13.496Z] 	at java.base/java.lang.Thread.run(Thread.java:857)
[2024-02-24T17:18:13.496Z] 
[2024-02-24T17:18:13.496Z] JavaTest Message: Test threw exception: java.lang.RuntimeException: getBytes(csn) failed  -> x-IBM29626C

[2024-02-24T17:18:21.873Z] jdk_nio_0_FAILED

JDK21 aarch64_linux(rhel9-aarch64-8)

[Akira1Saitoh]

Opened #19044.

[Akira1Saitoh]

Running sun/nio/cs/TestStringCoding.java test with the VM including the fix (#19044).
https://hyc-runtimes-jenkins.swg-devops.com/job/Grinder/38187/ - passed (25/25)

[Akira1Saitoh]

Opened #19056 for 0.44.