Move the acquisition and release of the page queues lock outside of a loop
in vm_object_split() to avoid repeated acquisition and release.
This commit is contained in:
parent
76505d5adf
commit
5ba514bc89
Notes:
svn2git
2020-12-20 02:59:44 +00:00
svn path=/head/; revision=139921
@ -1280,6 +1280,7 @@ vm_object_split(vm_map_entry_t entry)
|
||||
orig_object->backing_object_offset + entry->offset;
|
||||
new_object->backing_object = source;
|
||||
}
|
||||
vm_page_lock_queues();
|
||||
for (idx = 0; idx < size; idx++) {
|
||||
retry:
|
||||
m = vm_page_lookup(orig_object, offidxstart + idx);
|
||||
@ -1293,7 +1294,6 @@ vm_object_split(vm_map_entry_t entry)
|
||||
* We do not have to VM_PROT_NONE the page as mappings should
|
||||
* not be changed by this operation.
|
||||
*/
|
||||
vm_page_lock_queues();
|
||||
if ((m->flags & PG_BUSY) || m->busy) {
|
||||
vm_page_flag_set(m, PG_WANTED | PG_REFERENCED);
|
||||
VM_OBJECT_UNLOCK(orig_object);
|
||||
@ -1301,13 +1301,14 @@ vm_object_split(vm_map_entry_t entry)
|
||||
msleep(m, &vm_page_queue_mtx, PDROP | PVM, "spltwt", 0);
|
||||
VM_OBJECT_LOCK(new_object);
|
||||
VM_OBJECT_LOCK(orig_object);
|
||||
vm_page_lock_queues();
|
||||
goto retry;
|
||||
}
|
||||
vm_page_rename(m, new_object, idx);
|
||||
/* page automatically made dirty by rename and cache handled */
|
||||
vm_page_busy(m);
|
||||
vm_page_unlock_queues();
|
||||
}
|
||||
vm_page_unlock_queues();
|
||||
if (orig_object->type == OBJT_SWAP) {
|
||||
/*
|
||||
* swap_pager_copy() can sleep, in which case the orig_object's
|
||||
|
Loading…
Reference in New Issue
Block a user