OpenSmalltalk
/
opensmalltalk-vm
|
201911282316
|
1 hr, 29 mins, and 1 sec
|
Eliot Miranda
|
CogVM source as per VMMaker.oscog-eem.2598
Spur: VMMaker.oscog-cb.2595
Fixed a fascinating bug in Planning compactor.
Sometimes Planning decides to go for a multi-pass compaction, but in one compaction it compacts it all, so that firstFreeObject is the last object on heap, on the object after firstFreeObject is outside of the heap. Planning gets confused in this case, while everything is compacted. So I changed to abort compaction in that case, since everything is already compacted.
Avoid inlining endCompaction to prevent duplication in compact.
Make a minor refactoring to reinitializeScanFrom: to make it more readable (move assignment to firstMobileObject into scanForFirstFreeAndFirstMobileObjectFrom:).
Improve the ImageLeakChecker to bounds check objects while swizzling, and hence detect the damaged images produced by the issue 444 planning compactor bug as corrupted.
Refactor swizzleObj: to swizzleObj:in: so that ImageLeakChecker/ SpurLeakCheckingSegmentManager can produce more informative diagnostics.
Improve the leak checker to first bounds check oops against the heap extent before probing the heap map. This prevents ImageLeakChecker from crashing on issue 444 corrupted images.
|
|
|
Would you like to stay up-to-date with the upcoming Travis CI build environment updates? We set up a mailing list for you!
SIGN UP HERE
|
|
|
|
<script type="application/ld+json">
{
"@context": "http://schema.org",
"@type": "EmailMessage",
"action": {
"@type": "ViewAction",
"url": "https://travis-ci.org/OpenSmalltalk/opensmalltalk-vm/builds/621147865?utm_medium=notification&utm_source=email",
"name": "View Build"
},
"description": "View Build #1862 on Travis CI"
}
</script>