In kmem_unback(), take one small bit of code out of the region where the write lock is held.
Details
Details
Diff Detail
Diff Detail
- Repository
- rG FreeBSD src repository
- Lint
Lint Not Applicable - Unit
Tests Not Applicable
Event Timeline
Comment Actions
Something (un)related: Can we somehow automate checking for the issue where iterators are invalidated due to the object unlock? For instance, we can add the object' lock generation for INVARIANTS kernels, remember the current generation in the iterator on init, and assert that the lock generation does not change on next access to the iterator.