[src/trunk]: src/sys/uvm uvm_pagealloc_pgb(): don't fill cache if we're into ...

[ad <ad%NetBSD.org@localhost>]

details:   https://anonhg.NetBSD.org/src/rev/e847580d4350
branches:  trunk
changeset: 847614:e847580d4350
user:      ad <ad%NetBSD.org@localhost>
date:      Mon Dec 30 17:45:53 2019 +0000

description:
uvm_pagealloc_pgb(): don't fill cache if we're into the reserves.

uvm_pagereplace(): use radix_tree_replace_node() to avoid alloc/free.

diffstat:

 sys/uvm/uvm_page.c |  16 +++++++++++-----
 1 files changed, 11 insertions(+), 5 deletions(-)

diffs (65 lines):

diff -r 1db97a448d3a -r e847580d4350 sys/uvm/uvm_page.c
--- a/sys/uvm/uvm_page.c        Mon Dec 30 16:41:38 2019 +0000
+++ b/sys/uvm/uvm_page.c        Mon Dec 30 17:45:53 2019 +0000
@@ -1,4 +1,4 @@
-/*     $NetBSD: uvm_page.c,v 1.216 2019/12/28 16:07:41 ad Exp $        */
+/*     $NetBSD: uvm_page.c,v 1.217 2019/12/30 17:45:53 ad Exp $        */
 
 /*-
  * Copyright (c) 2019 The NetBSD Foundation, Inc.
@@ -95,7 +95,7 @@
  */
 
 #include <sys/cdefs.h>
-__KERNEL_RCSID(0, "$NetBSD: uvm_page.c,v 1.216 2019/12/28 16:07:41 ad Exp $");
+__KERNEL_RCSID(0, "$NetBSD: uvm_page.c,v 1.217 2019/12/30 17:45:53 ad Exp $");
 
 #include "opt_ddb.h"
 #include "opt_uvm.h"
@@ -1028,6 +1028,7 @@
        struct pgflbucket *pgb;
        struct vm_page *pg;
        kmutex_t *lock;
+       bool fill;
 
        /*
         * Skip the bucket if empty, no lock needed.  There could be many
@@ -1048,6 +1049,9 @@
                        mutex_spin_exit(lock);
                        return NULL;
                }
+               fill = false;
+       } else {
+               fill = true;
        }
 
        /* Try all page colors as needed. */
@@ -1073,7 +1077,7 @@
                         * so if we found pages in this CPU's preferred
                         * bucket.
                         */
-                       if (__predict_true(b == ucpu->pgflbucket)) {
+                       if (__predict_true(b == ucpu->pgflbucket && fill)) {
                                uvm_pgflcache_fill(ucpu, f, b, c);
                        }
                        mutex_spin_exit(lock);
@@ -1347,6 +1351,7 @@
 uvm_pagereplace(struct vm_page *oldpg, struct vm_page *newpg)
 {
        struct uvm_object *uobj = oldpg->uobject;
+       struct vm_page *pg __diagused;
 
        KASSERT((oldpg->flags & PG_TABLED) != 0);
        KASSERT(uobj != NULL);
@@ -1355,8 +1360,9 @@
        KASSERT(mutex_owned(uobj->vmobjlock));
 
        newpg->offset = oldpg->offset;
-       uvm_pageremove_tree(uobj, oldpg);
-       uvm_pageinsert_tree(uobj, newpg);
+       pg = radix_tree_replace_node(&uobj->uo_pages,
+           newpg->offset >> PAGE_SHIFT, newpg);
+       KASSERT(pg == oldpg);
 
        /* take page interlocks during rename */
        if (oldpg < newpg) {