Init jitTOC on ppc64 #16014
Init jitTOC on ppc64 #16014
Conversation
|
@gacholio changes are ready for another look |
|
jenkins compile plinux jdk19 |
|
I'm thinking a better solution to this problem would be to not swap in the JIT registers on the initial continuation entry. This would allow any dedicated registers on any platform to be inherited. |
|
To be clear, on first mount, copy the live JIT registers into the continuation structure (which will now represent the carrier state) but do not copy the continuation ones back to live. All subsequent mount/unmount operations continue to work as they do today, swaping the values. |
|
Here's a completely untested suggestion: https://github.com/gacholio/openj9/tree/firstmount |
tajila
force-pushed
the
loom
branch
3 times, most recently
from
7421290
to
d3c1f82
Compare
- Enable JIT in VirtualThread tests - Disable testing on PPC due to known test failure - PPC testing will be re-enabled with eclipse-openj9#16014 Signed-off-by: Jack Lu <Jack.S.Lu@ibm.com>
Signed-off-by: Tobi Ajila <atobia@ca.ibm.com>
|
jenkins test sanity plinux |
|
jenkins test sanity plinux jdk19 |
|
Relaunching. Infra issue in the previous PR build: https://openj9-jenkins.osuosl.org/job/Build_JDK19_ppc64le_linux_Personal/40/. jenkins test sanity plinux jdk19 |
- Enable JIT in VirtualThread tests - Disable testing on PPC due to known test failure - PPC testing will be re-enabled with eclipse-openj9#16014 Signed-off-by: Jack Lu <Jack.S.Lu@ibm.com>
- Enable JIT in VirtualThread tests - Disable testing on PPC due to known test failure - PPC testing will be re-enabled with eclipse-openj9#16014 Signed-off-by: Jack Lu <Jack.S.Lu@ibm.com>
Init jitTOC on ppc64
Signed-off-by: Tobi Ajila atobia@ca.ibm.com