[23.0] Synchronization with upstream 23.0 branch for July CPU release #541
Conversation
(cherry picked from commit 7156ae5)
(cherry picked from commit eb7b761)
(cherry picked from commit 6e841cd)
(cherry picked from commit ddb181a)
(cherry picked from commit 1079748)
(cherry picked from commit d38616e)
(cherry picked from commit 1c1b733)
(cherry picked from commit 5bad0de)
(cherry picked from commit 02ac1ff)
(cherry picked from commit 2db539b)
(cherry picked from commit b7b8d2b)
…or OHC (cherry picked from commit 9a30eac)
(cherry picked from commit aa7eeda)
(cherry picked from commit f4adc27)
(cherry picked from commit 8478188)
(cherry picked from commit 011acc3)
(cherry picked from commit f6980f0)
…languages repo (cherry picked from commit 463ae6f)
(cherry picked from commit da4f324)
(cherry picked from commit 188268f)
(cherry picked from commit 702ed41)
(cherry picked from commit 3c79ab0)
(cherry picked from commit 337f75f)
…on the GraalVM website. PullRequest: graal/14797
… Started guides. PullRequest: graal/14812
PullRequest: graal/14830
…ecific files which are compiled with JDK 20.
(cherry picked from commit 7c6be37)
PullRequest: graal/14842
This reverts commit bb8edf2.
(cherry picked from commit 0407384)
(cherry picked from commit a54ce8e)
(cherry picked from commit d1406b7)
(cherry picked from commit 502bf25)
(cherry picked from commit d6df93a)
(cherry picked from commit 51032a2)
(cherry picked from commit cdd5f9f)
(cherry picked from commit edcd236)
(cherry picked from commit 8b7e68f)
(cherry picked from commit 1fdd499)
…VM CE Container Images guide. Added Native Image getting started steps to GraalVM Windows installation guide./ Update required GraalVM support version in Native Image JFR doc. PullRequest: graal/14969
oracle-contributor-agreement
bot
added
the
OCA Verified
|
The mandrel integration tests are Karm/mandrel-integration-tests#160 The interesting part though is that in the 23.0 branch the single compile unit patch is not present, so we can't backport just oracle#6937. We will probably need a number of backports to get this fixes in 23.0. |
Right. Yes, this will need to be considered. Lets wait and see how the fix develops in head first, though. |
Ah, so if the single compile unit patch (oracle#6419) is not present in 23.0 then why do we want to backport oracle#6937? As far as I am aware, the debug info generated prior to oracle#6419 was not suffering from startup delays. Are there features provided by oracle#6419 or dependent follow-ups that we need? |
We are seeing Karm/mandrel-integration-tests#160 test failures in CI. See below.
See https://github.com/graalvm/mandrel/actions/runs/5620427175/job/15230467880#step:11:18570 it suggests Karm/mandrel-integration-tests#160 being present in 23.0. I don't know why. |
|
Hmm, I was wrong the single unit PR is in 23.0 oracle@53072937bd22. False alarm! |
Ok, that was what I thought. So, the difficulty with backporting oracle#6937 (revert to CU per class) appears to be that 23.0 does not contain oracle#6625 (debug info for foreign types). That is why it doesn't backport cleanly. I don't know if backporting both PRs in sequence will apply but I think it is likely(or, at least, will only require minor tweaks). If so then do we want to backport both? Or do we instead backport a modified variant of the CU per class PR? |
|
I would say both. I will try to see how hard it is. |
No description provided.