Change Details

Global ordering introduced in D22997 makes it possible. Since the current version of the routine discussed issues with it, I did the same with the new routine. I consider it to be enough of an improvement to go in as it is. Sample result after setting kern.maxvnodes=5000 and doing buildkernel: stock: vfs.numvnodes: 4742 vfs.recycles: 199491 vfs.recycles_free: 5730 patched: vfs.numvnodes: 4439 vfs.recycles: 64244 vfs.recycles_free: 6901

Global ordering introduced in D22997 makes it possible. Since the current version of the routine discussed issues with it, I did the same with the new routine. I consider it to be enough of an improvement to go in as it is. Tested by setting kern.maxvnodes=5000 and doing buildkernel. Results were similar to the stock kernel since vnode both vnode pressure is too much AND vlrureclaim only has one mode to go over everything (instead of only part of the list). This will be fixed later.

Global ordering introduced in D22997 makes it possible. Since the current version of the routine discussed issues with it, I did the same with the new routine. I consider it to be enough of an improvement to go in as it is. Sample result afterTested by setting kern.maxvnodes=5000 and doing buildkernel: stock: vfs.numvnodes: 4742 vfs.recycles: 199491 vfs.recycles_free: 5730 patched: vfs.numvnodes: 4439 vfs.recycles: 64244 vfs.recycles_free: 6901. Results were similar to the stock kernel since vnode both vnode pressure is too much AND vlrureclaim only has one mode to go over everything (instead of only part of the list). This will be fixed later.