[
https://issues.apache.org/jira/browse/DERBY-4528?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Kathey Marsden resolved DERBY-4528.
-----------------------------------
Resolution: Invalid
Closing Invalid because this appears to be a IBM JVM JIT issue.
> Intermittent ij syntax errors with IBM 1.6 jvm (again)
> ------------------------------------------------------
>
> Key: DERBY-4528
> URL: https://issues.apache.org/jira/browse/DERBY-4528
> Project: Derby
> Issue Type: Bug
> Components: Tools
> Affects Versions: 10.5.3.0
> Environment: java.vm.info ;value: JRE 1.6.0 IBM J9 2.4 Linux
> x86-32
> jvmxi3260sr7-20091214_49398 (JIT enabled, AOT enabled)
> J9VM -
> 20091214_049398
> JIT -
> r9_20091123_13891
> GC - 20091111_AA
> java.version ;value: 1.6.0
> Reporter: Kathey Marsden
>
> This is likely an IBM jvm JIT issue, but I am filing and will close it as
> invalid so that it is part of the historical record. The user encountering
> the problem has decided to work around it by turning JIT off for the ij
> invocations and has not been able to provide a reproduction we can work on.
> The symptom is just the same as DERBY-1220 which was only seen on z/os with
> IBM 1.5. ij seems to ignore the semicolon as a separator and send it as part
> of the sql command, printing an error like
> ERROR 42X01: Syntax error:
> Encountered ";" at line 10, column 2.
> DERBY-1220 was not worked on directly by the IBM jvm team but seemed to go
> away with a later build.
> This problem came back again with a prerelease IBM 1.6 jvm on Windows but
> was much worse, instead of a few tests failing, almost every test with ij
> failed. This was version:
> java version "1.6.0-internal"
> Java(TM) SE Runtime Environment (build pwi3260-20070819_01)
> IBM J9 VM (build 2.4, J2RE 1.6.0 IBM J9 2.4 Windows XP x86-32
> jvmwi3260-20070817_13537 (JIT enabled)
> J9VM - 20070817_013537_lHdSMR
> JIT - dev_20070817_1300
> GC - 20070815_AA)
> At that time we identified it as a JIT issue, but again before the jvm team
> worked on it directly we got a new build that seemed to fix it.
> Now a user is seeing it again intermittently on Linux with a recent IBM 1.6
> build (see environment) as part of a very large and complex build process.
> Specifying -Xint on their ij invocations allows them to work around the
> issue. Performing the normal JIT diagnostics would require narrowing down
> their reproduction significantly and many, many runs to get conclusive
> results, so they have chosen to just work around it for now.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
