From Volker: It seems that our is_delete_request handling in delay_for_exclusive_oplocks is wrong. It seems that the only difference between batch and exclusive oplocks is the time of the check: Batch oplocks are broken before the share mode check, exclusive ones after. If you run the raw.oplock.exclusive8 test against current versions of smbd, you will also find that we get a panic due to an invalid share mode entry. Attached also find the patches against v3-6-test and v3-5-test.
Created attachment 7872 [details] git-am fix for 3.6.x Volker's patch for 3.6.next. Already approved by jra.
Created attachment 7873 [details] git-am fix for 3.5.next.
Pushed to v3-5-test and v3-6-test. Closing out bug report. Thanks!