| Issue | Fix Version | Assignee | Priority | Status | Resolution | Resolved In Build |
|---|---|---|---|---|---|---|
| JDK-8204014 | 11.0.1 | Tobias Hartmann | P2 | Resolved | Fixed | team |
| JDK-8263222 | openjdk8u302 | Fei Yang | P2 | Resolved | Fixed | b01 |
We see two test failures with -Xcomp -Xbatch: vm/instr/tableswitch/tableswitch006/tableswitch00610m1/tableswitch00610m1.html and vm/instr/tableswitch/tableswitch006/tableswitch00610m1t/tableswitch00610m1t.html.
This does not fail if we test with C2 only, nor without -Xcomp, so I assume it's related to C1.
Find the .jtr file illustrating the failure attached.
It fails only on linux (ppc, ppcle, s390, x86_64). I see the failures for a while now in our nightly tests of jdk/jdk. We update our jck test suite regularly.
I could imagine this is related to the gcc version (the only in common to these platforms). We use SUSE gcc48-4.8.5-24.1 on SLES 11.3.
This does not fail if we test with C2 only, nor without -Xcomp, so I assume it's related to C1.
Find the .jtr file illustrating the failure attached.
It fails only on linux (ppc, ppcle, s390, x86_64). I see the failures for a while now in our nightly tests of jdk/jdk. We update our jck test suite regularly.
I could imagine this is related to the gcc version (the only in common to these platforms). We use SUSE gcc48-4.8.5-24.1 on SLES 11.3.
- backported by
-
JDK-8204014 C1 emits incorrect code due to integer overflow in _tableswitch keys
-
- Resolved
-
-
-
- Resolved
-
- relates to
-
JDK-8200303 C2 should leverage profiling for lookupswitch/tableswitch
-
- Resolved
-