Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 26 Aug 2019 20:20:10 +0000 (UTC)
From:      Mark Johnston <markj@FreeBSD.org>
To:        src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org
Subject:   svn commit: r351518 - head/sys/vm
Message-ID:  <201908262020.x7QKKAtS021698@repo.freebsd.org>

next in thread | raw e-mail | index | archive | help
Author: markj
Date: Mon Aug 26 20:20:10 2019
New Revision: 351518
URL: https://svnweb.freebsd.org/changeset/base/351518

Log:
  Fix a few nits in vm_pqbatch_process_page().
  
  - Don't bother masking off non-queue state flags when loading the
    page's atomic state, since it is only required for one of the
    function's assertions.  Update the assertion instead.
  - Remove an incorrect comment regarding synchronization with the
    page daemon.  The page daemon only ever checks for PGA_ENQUEUED
    with the page queue lock held.
  - When clearing requeue flags, only clear the flags that have been
    acted upon.
  
  Reviewed by:	kib (previous version)
  Discussed with:	alc
  Tested by:	pho (part of a larger patch)
  MFC after:	1 week
  Sponsored by:	Netflix
  Differential Revision:	https://reviews.freebsd.org/D21368

Modified:
  head/sys/vm/vm_page.c

Modified: head/sys/vm/vm_page.c
==============================================================================
--- head/sys/vm/vm_page.c	Mon Aug 26 18:57:29 2019	(r351517)
+++ head/sys/vm/vm_page.c	Mon Aug 26 20:20:10 2019	(r351518)
@@ -3081,9 +3081,10 @@ vm_pqbatch_process_page(struct vm_pagequeue *pq, vm_pa
 	 * the page queue lock held.  In this case it is about to free the page,
 	 * which must not have any queue state.
 	 */
-	qflags = atomic_load_8(&m->aflags) & PGA_QUEUE_STATE_MASK;
-	KASSERT(pq == vm_page_pagequeue(m) || qflags == 0,
-	    ("page %p doesn't belong to queue %p but has queue state %#x",
+	qflags = atomic_load_8(&m->aflags);
+	KASSERT(pq == vm_page_pagequeue(m) ||
+	    (qflags & PGA_QUEUE_STATE_MASK) == 0,
+	    ("page %p doesn't belong to queue %p but has aflags %#x",
 	    m, pq, qflags));
 
 	if ((qflags & PGA_DEQUEUE) != 0) {
@@ -3097,6 +3098,13 @@ vm_pqbatch_process_page(struct vm_pagequeue *pq, vm_pa
 			vm_pagequeue_cnt_inc(pq);
 			vm_page_aflag_set(m, PGA_ENQUEUED);
 		}
+
+		/*
+		 * Give PGA_REQUEUE_HEAD precedence over PGA_REQUEUE.
+		 * In particular, if both flags are set in close succession,
+		 * only PGA_REQUEUE_HEAD will be applied, even if it was set
+		 * first.
+		 */
 		if ((qflags & PGA_REQUEUE_HEAD) != 0) {
 			KASSERT(m->queue == PQ_INACTIVE,
 			    ("head enqueue not supported for page %p", m));
@@ -3105,12 +3113,8 @@ vm_pqbatch_process_page(struct vm_pagequeue *pq, vm_pa
 		} else
 			TAILQ_INSERT_TAIL(&pq->pq_pl, m, plinks.q);
 
-		/*
-		 * PGA_REQUEUE and PGA_REQUEUE_HEAD must be cleared after
-		 * setting PGA_ENQUEUED in order to synchronize with the
-		 * page daemon.
-		 */
-		vm_page_aflag_clear(m, PGA_REQUEUE | PGA_REQUEUE_HEAD);
+		vm_page_aflag_clear(m, qflags & (PGA_REQUEUE |
+		    PGA_REQUEUE_HEAD));
 	}
 }
 



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201908262020.x7QKKAtS021698>