Print Memory Stats #2101
Print Memory Stats #2101
Conversation
|
@mpirvu for your review. |
| @@ -2711,6 +2711,9 @@ void TR::CompilationInfo::stopCompilationThreads() | |||
|
|
|||
| if (getJProfilerThread()) | |||
| fprintf(stderr, "Allocated memory for profile info = %d KB\n", getJProfilerThread()->getProfileInfoFootprint()/1024); | |||
|
|
|||
| if (trPersistentMemory) | |||
There was a problem hiding this comment.
These statistics are probably useful only when tracking a memory leak. Also, the list of memory objects is very long.
For these two reasons I would like the printout to be enabled by another environment variable.
|
Changes made. |
|
@dsouzai a rebase forward should clear up the travis failure (I hope) so we can get this tested and merged. |
|
Because this depends on eclipse-omr/omr#2542 until that code is merged in openj9-omr, I think travis will still be broken EDIT: It's been added to |
|
Jenkins test sanity depends eclipse-omr/omr#2542 |
|
Jenkins test sanity depends eclipse/omr#master |
|
@vijaysun-omr This PR should be ok to merge now. I can rebase the changes if you'd like the travis build to run again, but given that sanity tests passed, it doesn't seem necessary (given how small this change is). The windows failure is unrelated and caused by something else (other PR builds are also failing on windows). |
|
Windows issue caused by #2129 |
andrewcraik
added
the
depends:omr
|
Rebased onto latest master for clean travis build. |
So much for that... failed because eclipse-omr/omr#2490 wasn't in openj9-omr . It looks like that change is now in openj9-omr, so let's try again.. |
Signed-off-by: Irwin D'Souza <dsouzai@ca.ibm.com>
|
@vijaysun-omr This should be good to merge now |
Added the printing of JIT memory stats at shutdown if the
TR_PrintPersistentMemenv var is set, and every 5 minutes in the vlog ifverbose={jitMemory}is set.Depends on eclipse-omr/omr#2542