LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: David Rientjes <rientjes@google.com>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: Paul Jackson <pj@sgi.com>, Christoph Lameter <clameter@sgi.com>,
	Lee Schermerhorn <Lee.Schermerhorn@hp.com>,
	Andi Kleen <ak@suse.de>, Randy Dunlap <randy.dunlap@oracle.com>,
	linux-kernel@vger.kernel.org
Subject: [patch -mm 1/2] mempolicy: disallow static or relative flags for local preferred mode
Date: Fri, 7 Mar 2008 17:24:15 -0800 (PST)	[thread overview]
Message-ID: <alpine.DEB.1.00.0803071720460.4611@chino.kir.corp.google.com> (raw)

MPOL_F_STATIC_NODES and MPOL_F_RELATIVE_NODES don't mean anything for
MPOL_PREFERRED policies that were created with an empty nodemask (for
purely local allocations).  They'll never be invalidated because the
allowed mems of a task changes or need to be rebound relative to a
cpuset's placement.

Also fixes a bug identified by Lee Schermerhorn that disallowed empty
nodemasks to be passed to MPOL_PREFERRED to specify local allocations.

Cc: Paul Jackson <pj@sgi.com>
Cc: Christoph Lameter <clameter@sgi.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
Cc: Andi Kleen <ak@suse.de>
Cc: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: David Rientjes <rientjes@google.com>
---
 Documentation/vm/numa_memory_policy.txt |   16 ++++++++++++++--
 mm/mempolicy.c                          |   17 ++++++++++++-----
 2 files changed, 26 insertions(+), 7 deletions(-)

diff --git a/Documentation/vm/numa_memory_policy.txt b/Documentation/vm/numa_memory_policy.txt
--- a/Documentation/vm/numa_memory_policy.txt
+++ b/Documentation/vm/numa_memory_policy.txt
@@ -210,6 +210,12 @@ Components of Memory Policies
 	    local allocation for a specific range of addresses--i.e. for
 	    VMA policies.
 
+	    It is possible for the user to specify that local allocation is
+	    always preferred by passing an empty nodemask with this mode.
+	    If an empty nodemask is passed, the policy cannot use the
+	    MPOL_F_STATIC_NODES or MPOL_F_RELATIVE_NODES flags described
+	    below.
+
 	MPOL_INTERLEAVED:  This mode specifies that page allocations be
 	interleaved, on a page granularity, across the nodes specified in
 	the policy.  This mode also behaves slightly differently, based on
@@ -259,7 +265,10 @@ Components of Memory Policies
 	    occurs over that node.  If no nodes from the user's nodemask are
 	    now allowed, the Default behavior is used.
 
-	    MPOL_F_STATIC_NODES cannot be used with MPOL_F_RELATIVE_NODES.
+	    MPOL_F_STATIC_NODES cannot be combined with the
+	    MPOL_F_RELATIVE_NODES flag.  It also cannot be used for
+	    MPOL_PREFERRED policies that were created with an empty nodemask
+	    (local allocation).
 
 	MPOL_F_RELATIVE_NODES:  This flag specifies that the nodemask passed
 	by the user will be mapped relative to the set of the task or VMA's
@@ -306,7 +315,10 @@ Components of Memory Policies
 	    set of memory nodes allowed by the task's cpuset, as that may
 	    change over time.
 
-	    MPOL_F_RELATIVE_NODES cannot be used with MPOL_F_STATIC_NODES.
+	    MPOL_F_RELATIVE_NODES cannot be combined with the
+	    MPOL_F_STATIC_NODES flag.  It also cannot be used for
+	    MPOL_PREFERRED policies that were created with an empty nodemask
+	    (local allocation).
 
 MEMORY POLICY APIs
 
diff --git a/mm/mempolicy.c b/mm/mempolicy.c
--- a/mm/mempolicy.c
+++ b/mm/mempolicy.c
@@ -151,9 +151,7 @@ static int mpol_new_interleave(struct mempolicy *pol, const nodemask_t *nodes)
 
 static int mpol_new_preferred(struct mempolicy *pol, const nodemask_t *nodes)
 {
-	if (nodes_empty(*nodes))
-		return -EINVAL;
-	pol->v.preferred_node = first_node(*nodes);
+	pol->v.preferred_node = !nodes_empty(*nodes) ? first_node(*nodes) : -1;
 	return 0;
 }
 
@@ -176,7 +174,16 @@ static struct mempolicy *mpol_new(unsigned short mode, unsigned short flags,
 	pr_debug("setting mode %d flags %d nodes[0] %lx\n",
 		 mode, flags, nodes ? nodes_addr(*nodes)[0] : -1);
 
-	if (nodes && nodes_empty(*nodes) && mode != MPOL_PREFERRED)
+	/*
+	 * MPOL_PREFERRED cannot be used with MPOL_F_STATIC_NODES or
+	 * MPOL_F_RELATIVE_NODES if the nodemask is empty (local allocation).
+	 * All other modes require a valid pointer to a non-empty nodemask.
+	 */
+	if (mode == MPOL_PREFERRED) {
+		if (nodes_empty(*nodes) && ((flags & MPOL_F_STATIC_NODES) ||
+					    (flags & MPOL_F_RELATIVE_NODES)))
+			return ERR_PTR(-EINVAL);
+	} else if (nodes && nodes_empty(*nodes))
 		return ERR_PTR(-EINVAL);
 	if (mode == MPOL_DEFAULT)
 		return NULL;
@@ -250,7 +257,7 @@ static void mpol_rebind_preferred(struct mempolicy *pol,
 	} else if (pol->flags & MPOL_F_RELATIVE_NODES) {
 		mpol_relative_nodemask(&tmp, &pol->w.user_nodemask, nodes);
 		pol->v.preferred_node = first_node(tmp);
-	} else {
+	} else if (pol->v.preferred_node != -1) {
 		pol->v.preferred_node = node_remap(pol->v.preferred_node,
 						   pol->w.cpuset_mems_allowed,
 						   *nodes);

             reply	other threads:[~2008-03-08  1:26 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-03-08  1:24 David Rientjes [this message]
2008-03-08  1:24 ` [patch -mm 2/2] mempolicy: use default_policy mode instead of MPOL_DEFAULT David Rientjes
2008-03-08  1:28   ` Paul Jackson
2008-03-08  1:33     ` David Rientjes
2008-03-08  1:35       ` Paul Jackson
2008-03-08  1:59         ` Christoph Lameter
2008-03-08  2:14           ` David Rientjes
2008-03-08 19:13       ` Lee Schermerhorn
2008-03-08 22:20         ` David Rientjes
2008-03-08 23:19           ` Andi Kleen
2008-03-10 13:48             ` Lee Schermerhorn
2008-03-10 19:09 ` [patch -mm 1/2] mempolicy: disallow static or relative flags for local preferred mode Andrew Morton

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=alpine.DEB.1.00.0803071720460.4611@chino.kir.corp.google.com \
    --to=rientjes@google.com \
    --cc=Lee.Schermerhorn@hp.com \
    --cc=ak@suse.de \
    --cc=akpm@linux-foundation.org \
    --cc=clameter@sgi.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pj@sgi.com \
    --cc=randy.dunlap@oracle.com \
    --subject='Re: [patch -mm 1/2] mempolicy: disallow static or relative flags for local preferred mode' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).