<joaocorreia[m]>
headius: Good Morning and thank you for your response. Answering your question, yes I have to use tomcat as it's where my client can have support.
<joaocorreia[m]>
I am using warble to create the war for tomcat, I already made a few tests with multiple config options for the min and max jruby runtimes but it doesn't seem to change the final results.
<headius[m]>
I hope this is a temporary workaround, it's pretty silly
<headius[m]>
I also see output for jitted "foo" that inlines parts of the interpreter for "bar"... that's interesting
<headius[m]>
enebo: so Vladimir has an explanation for why we need this even though we can call the method beforehand
<headius[m]>
apparently for this case, since there's no super, we never use the passed-in method name String
<headius[m]>
so String never is force to resolve in that classloader, and as a result the method signature doesn't look fully-resolved when it goes to inline it
<enebo[m]>
hmm
<headius[m]>
enebo: did you get 9.2 merged?
<headius[m]>
I just tried to merge the inlining workaround fix from 9.2 and all the same classes appear to still conflict
<enebo[m]>
headius: yeah it is merged
<enebo[m]>
I had some really weird behavior during the merge and seemingly redid the conflicts twice but it is correct now
<headius[m]>
Hmmm ok maybe I am behind
<headius[m]>
enebo: yeah it does not merge
<enebo[m]>
well it was a weird experience I will give you that
<enebo[m]>
but it for sure is all on master right now
<headius[m]>
I think you may gotten equivalent changes over but the branch did not merge
<enebo[m]>
yeah I did a merge on what I thought was head and then when I went to push it did not work so then I did a rebase but after the merge commit was already on my copy of master
<enebo[m]>
I have no idea what went wrong
<headius[m]>
ah
<headius[m]>
that would not actually merge the branches then
<headius[m]>
I can try to fix it by just keeping what's on master
<enebo[m]>
the rebase definitely is the issue but I did finish the merge
sagax has quit [Ping timeout: 258 seconds]
<enebo[m]>
before the rebase
<headius[m]>
yeah rebase breaks merge
<headius[m]>
because it's now new commits based on this branch
<headius[m]>
not commits from the other branch
<enebo[m]>
but the rebase probably overwrote the commit history fo the merge
<headius[m]>
right
<headius[m]>
if you are satisfied it's all there then I'll just merge and make no changes
<enebo[m]>
yeah I did it twice :)
<headius[m]>
there weren't any other changes on branch since then it seems other than my inlining hack which is already on master
<headius[m]>
ok
<enebo[m]>
and the only question at the end was some changes to tests which had nothing to do with 9.2
<headius[m]>
ok
<enebo[m]>
but I think I added them vs deleting them as I was confused on what was happening there so if that is a mistake we will just delete them again
<headius[m]>
I'm basically just committing a merge commit with no changes
<headius[m]>
so there's a new base for future merges
<headius[m]>
git is so intuitive
<enebo[m]>
sure it is possible 9.2.14 could happen and that would be easier (although honestly we could just cp those few commits at this point)
<headius[m]>
yeah I just wanted to make sure that inlining fix landed on 9.2 in case we do .14
<enebo[m]>
merging is a better place to leave things though
<headius[m]>
should be good now
<enebo[m]>
ok
schmendrick has joined #jruby
<headius[m]>
enebo: let's merge byteit101 PR and do any additional refinement in place
<headius[m]>
oh hmm, JI tests failed after rebase
<headius[m]>
byteit101: have you looked at failures? If we can get this green I'm fine merging it
<headius[m]>
ahorek: wow you've been busy
<enebo[m]>
headius: sure once it is green
<byteit101[m]>
headius: oh, hmm. the annotations are no longer on test classpath?
<headius[m]>
hmm looking
<byteit101[m]>
pec/java_integration/fixtures/EveryTypeAnnotations.java:52: error: package javax.annotation does not exist
<byteit101[m]>
import javax.annotation.Resource;
<headius[m]>
yeah these are all on java 11 and 14
<headius[m]>
that CI is new since you started this I think
<byteit101[m]>
yes, it was gereen back in the spring
<headius[m]>
yeah your princess is in another castle
<headius[m]>
annotations are an optional module now
<headius[m]>
I have to look this up every time
<byteit101[m]>
eating now, but i think i used that because it was a built in with some type or another . will page in after dinner
<headius[m]>
yeah need to get the annotations library on classpath for these builds
<headius[m]>
I'll try to find how
<byteit101[m]>
thanks!
schmendrick has quit [Remote host closed the connection]
<byteit101[m]>
looking through my zsh_history, I was looking for rgrep 'RetentionPolicy.RUNTIME' -C 3 | grep '@Target.*FIELD' in the jdk sources, and only Resources and Deprecated fit that bill, and only Resources was non-trivial
<headius[m]>
I think I have a build patch that will work