kernel - Fix panic/crash in vm_object_page_remove()
commit51424739a9a596084144dc78197ad06b53f2f293
authorMatthew Dillon <dillon@apollo.backplane.com>
Mon, 17 Oct 2016 19:15:57 +0000 (17 12:15 -0700)
committerMatthew Dillon <dillon@apollo.backplane.com>
Mon, 17 Oct 2016 19:15:57 +0000 (17 12:15 -0700)
tree12eb8631d10020e2cc86ed097b94dc5e0a8b0da7
parent4a5093f732f5b3445e862ef3318d0193567985a3
kernel - Fix panic/crash in vm_object_page_remove()

* When vm_object_page_remove() is called with one page, the caller is allowed
  to hold a spinlock (typically on that page).  The callback is not supposed
  to call user_yield in that situation.

* Fix missing initialization of info.count which could result in the callback
  issuing a yield when it was not supposed to.

Reported-by: htse
sys/vm/vm_object.c