aboutsummaryrefslogtreecommitdiffstats
path: root/mm
diff options
context:
space:
mode:
authorDavid Rientjes <rientjes@google.com>2008-04-28 02:12:27 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2008-04-28 08:58:19 -0700
commitf5b087b52f1710eb0bf15a2d2b030c51a6a1ca9e (patch)
tree66336235822f59215707dfa501e1d2b66b38a015 /mm
parent028fec414d803117eb4b2ed12acb4dd5da65b32d (diff)
downloadkernel_samsung_aries-f5b087b52f1710eb0bf15a2d2b030c51a6a1ca9e.zip
kernel_samsung_aries-f5b087b52f1710eb0bf15a2d2b030c51a6a1ca9e.tar.gz
kernel_samsung_aries-f5b087b52f1710eb0bf15a2d2b030c51a6a1ca9e.tar.bz2
mempolicy: add MPOL_F_STATIC_NODES flag
Add an optional mempolicy mode flag, MPOL_F_STATIC_NODES, that suppresses the node remap when the policy is rebound. Adds another member to struct mempolicy, nodemask_t user_nodemask, as part of a union with cpuset_mems_allowed: struct mempolicy { ... union { nodemask_t cpuset_mems_allowed; nodemask_t user_nodemask; } w; } that stores the the nodemask that the user passed when he or she created the mempolicy via set_mempolicy() or mbind(). When using MPOL_F_STATIC_NODES, which is passed with any mempolicy mode, the user's passed nodemask intersected with the VMA or task's allowed nodes is always used when determining the preferred node, setting the MPOL_BIND zonelist, or creating the interleave nodemask. This happens whenever the policy is rebound, including when a task's cpuset assignment changes or the cpuset's mems are changed. This creates an interesting side-effect in that it allows the mempolicy "intent" to lie dormant and uneffected until it has access to the node(s) that it desires. For example, if you currently ask for an interleaved policy over a set of nodes that you do not have access to, the mempolicy is not created and the task continues to use the previous policy. With this change, however, it is possible to create the same mempolicy; it is only effected when access to nodes in the nodemask is acquired. It is also possible to mount tmpfs with the static nodemask behavior when specifying a node or nodemask. To do this, simply add "=static" immediately following the mempolicy mode at mount time: mount -o remount mpol=interleave=static:1-3 Also removes mpol_check_policy() and folds its logic into mpol_new() since it is now obsoleted. The unused vma_mpol_equal() is also removed. 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> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'mm')
-rw-r--r--mm/mempolicy.c172
-rw-r--r--mm/shmem.c2
2 files changed, 90 insertions, 84 deletions
diff --git a/mm/mempolicy.c b/mm/mempolicy.c
index 1f6ff9c..d59b1e7 100644
--- a/mm/mempolicy.c
+++ b/mm/mempolicy.c
@@ -113,58 +113,6 @@ struct mempolicy default_policy = {
static void mpol_rebind_policy(struct mempolicy *pol,
const nodemask_t *newmask);
-/* Do sanity checking on a policy */
-static int mpol_check_policy(unsigned short mode, nodemask_t *nodes)
-{
- int was_empty, is_empty;
-
- if (!nodes)
- return 0;
-
- /*
- * "Contextualize" the in-coming nodemast for cpusets:
- * Remember whether in-coming nodemask was empty, If not,
- * restrict the nodes to the allowed nodes in the cpuset.
- * This is guaranteed to be a subset of nodes with memory.
- */
- cpuset_update_task_memory_state();
- is_empty = was_empty = nodes_empty(*nodes);
- if (!was_empty) {
- nodes_and(*nodes, *nodes, cpuset_current_mems_allowed);
- is_empty = nodes_empty(*nodes); /* after "contextualization" */
- }
-
- switch (mode) {
- case MPOL_DEFAULT:
- /*
- * require caller to specify an empty nodemask
- * before "contextualization"
- */
- if (!was_empty)
- return -EINVAL;
- break;
- case MPOL_BIND:
- case MPOL_INTERLEAVE:
- /*
- * require at least 1 valid node after "contextualization"
- */
- if (is_empty)
- return -EINVAL;
- break;
- case MPOL_PREFERRED:
- /*
- * Did caller specify invalid nodes?
- * Don't silently accept this as "local allocation".
- */
- if (!was_empty && is_empty)
- return -EINVAL;
- break;
- default:
- BUG();
- }
- return 0;
-}
-
/* Check that the nodemask contains at least one populated zone */
static int is_valid_nodemask(nodemask_t *nodemask)
{
@@ -186,48 +134,60 @@ static int is_valid_nodemask(nodemask_t *nodemask)
return 0;
}
+static inline int mpol_store_user_nodemask(const struct mempolicy *pol)
+{
+ return pol->flags & MPOL_F_STATIC_NODES;
+}
+
/* Create a new policy */
static struct mempolicy *mpol_new(unsigned short mode, unsigned short flags,
nodemask_t *nodes)
{
struct mempolicy *policy;
+ nodemask_t cpuset_context_nmask;
pr_debug("setting mode %d flags %d nodes[0] %lx\n",
mode, flags, nodes ? nodes_addr(*nodes)[0] : -1);
if (mode == MPOL_DEFAULT)
- return NULL;
+ return (nodes && nodes_weight(*nodes)) ? ERR_PTR(-EINVAL) :
+ NULL;
policy = kmem_cache_alloc(policy_cache, GFP_KERNEL);
if (!policy)
return ERR_PTR(-ENOMEM);
atomic_set(&policy->refcnt, 1);
+ cpuset_update_task_memory_state();
+ nodes_and(cpuset_context_nmask, *nodes, cpuset_current_mems_allowed);
switch (mode) {
case MPOL_INTERLEAVE:
- policy->v.nodes = *nodes;
- if (nodes_weight(policy->v.nodes) == 0) {
- kmem_cache_free(policy_cache, policy);
- return ERR_PTR(-EINVAL);
- }
+ if (nodes_empty(*nodes) || nodes_empty(cpuset_context_nmask))
+ goto free;
+ policy->v.nodes = cpuset_context_nmask;
break;
case MPOL_PREFERRED:
- policy->v.preferred_node = first_node(*nodes);
+ policy->v.preferred_node = first_node(cpuset_context_nmask);
if (policy->v.preferred_node >= MAX_NUMNODES)
- policy->v.preferred_node = -1;
+ goto free;
break;
case MPOL_BIND:
- if (!is_valid_nodemask(nodes)) {
- kmem_cache_free(policy_cache, policy);
- return ERR_PTR(-EINVAL);
- }
- policy->v.nodes = *nodes;
+ if (!is_valid_nodemask(&cpuset_context_nmask))
+ goto free;
+ policy->v.nodes = cpuset_context_nmask;
break;
default:
BUG();
}
policy->policy = mode;
policy->flags = flags;
- policy->cpuset_mems_allowed = cpuset_mems_allowed(current);
+ if (mpol_store_user_nodemask(policy))
+ policy->w.user_nodemask = *nodes;
+ else
+ policy->w.cpuset_mems_allowed = cpuset_mems_allowed(current);
return policy;
+
+free:
+ kmem_cache_free(policy_cache, policy);
+ return ERR_PTR(-EINVAL);
}
static void gather_stats(struct page *, void *, int pte_dirty);
@@ -473,15 +433,14 @@ static long do_set_mempolicy(unsigned short mode, unsigned short flags,
{
struct mempolicy *new;
- if (mpol_check_policy(mode, nodes))
- return -EINVAL;
new = mpol_new(mode, flags, nodes);
if (IS_ERR(new))
return PTR_ERR(new);
mpol_free(current->mempolicy);
current->mempolicy = new;
mpol_set_task_struct_flag();
- if (new && new->policy == MPOL_INTERLEAVE)
+ if (new && new->policy == MPOL_INTERLEAVE &&
+ nodes_weight(new->v.nodes))
current->il_next = first_node(new->v.nodes);
return 0;
}
@@ -796,9 +755,6 @@ static long do_mbind(unsigned long start, unsigned long len,
if (end == start)
return 0;
- if (mpol_check_policy(mode, nmask))
- return -EINVAL;
-
new = mpol_new(mode, mode_flags, nmask);
if (IS_ERR(new))
return PTR_ERR(new);
@@ -1206,7 +1162,8 @@ static unsigned interleave_nodes(struct mempolicy *policy)
next = next_node(nid, policy->v.nodes);
if (next >= MAX_NUMNODES)
next = first_node(policy->v.nodes);
- me->il_next = next;
+ if (next < MAX_NUMNODES)
+ me->il_next = next;
return nid;
}
@@ -1252,10 +1209,13 @@ static unsigned offset_il_node(struct mempolicy *pol,
struct vm_area_struct *vma, unsigned long off)
{
unsigned nnodes = nodes_weight(pol->v.nodes);
- unsigned target = (unsigned)off % nnodes;
+ unsigned target;
int c;
int nid = -1;
+ if (!nnodes)
+ return numa_node_id();
+ target = (unsigned int)off % nnodes;
c = 0;
do {
nid = next_node(nid, pol->v.nodes);
@@ -1465,6 +1425,16 @@ struct mempolicy *__mpol_copy(struct mempolicy *old)
return new;
}
+static int mpol_match_intent(const struct mempolicy *a,
+ const struct mempolicy *b)
+{
+ if (a->flags != b->flags)
+ return 0;
+ if (!mpol_store_user_nodemask(a))
+ return 1;
+ return nodes_equal(a->w.user_nodemask, b->w.user_nodemask);
+}
+
/* Slow path of a mempolicy comparison */
int __mpol_equal(struct mempolicy *a, struct mempolicy *b)
{
@@ -1472,6 +1442,8 @@ int __mpol_equal(struct mempolicy *a, struct mempolicy *b)
return 0;
if (a->policy != b->policy)
return 0;
+ if (a->policy != MPOL_DEFAULT && !mpol_match_intent(a, b))
+ return 0;
switch (a->policy) {
case MPOL_DEFAULT:
return 1;
@@ -1771,13 +1743,14 @@ void numa_default_policy(void)
static void mpol_rebind_policy(struct mempolicy *pol,
const nodemask_t *newmask)
{
- nodemask_t *mpolmask;
nodemask_t tmp;
+ int static_nodes;
if (!pol)
return;
- mpolmask = &pol->cpuset_mems_allowed;
- if (nodes_equal(*mpolmask, *newmask))
+ static_nodes = pol->flags & MPOL_F_STATIC_NODES;
+ if (!mpol_store_user_nodemask(pol) &&
+ nodes_equal(pol->w.cpuset_mems_allowed, *newmask))
return;
switch (pol->policy) {
@@ -1786,16 +1759,35 @@ static void mpol_rebind_policy(struct mempolicy *pol,
case MPOL_BIND:
/* Fall through */
case MPOL_INTERLEAVE:
- nodes_remap(tmp, pol->v.nodes, *mpolmask, *newmask);
+ if (static_nodes)
+ nodes_and(tmp, pol->w.user_nodemask, *newmask);
+ else {
+ nodes_remap(tmp, pol->v.nodes,
+ pol->w.cpuset_mems_allowed, *newmask);
+ pol->w.cpuset_mems_allowed = *newmask;
+ }
pol->v.nodes = tmp;
- *mpolmask = *newmask;
- current->il_next = node_remap(current->il_next,
- *mpolmask, *newmask);
+ if (!node_isset(current->il_next, tmp)) {
+ current->il_next = next_node(current->il_next, tmp);
+ if (current->il_next >= MAX_NUMNODES)
+ current->il_next = first_node(tmp);
+ if (current->il_next >= MAX_NUMNODES)
+ current->il_next = numa_node_id();
+ }
break;
case MPOL_PREFERRED:
- pol->v.preferred_node = node_remap(pol->v.preferred_node,
- *mpolmask, *newmask);
- *mpolmask = *newmask;
+ if (static_nodes) {
+ int node = first_node(pol->w.user_nodemask);
+
+ if (node_isset(node, *newmask))
+ pol->v.preferred_node = node;
+ else
+ pol->v.preferred_node = -1;
+ } else {
+ pol->v.preferred_node = node_remap(pol->v.preferred_node,
+ pol->w.cpuset_mems_allowed, *newmask);
+ pol->w.cpuset_mems_allowed = *newmask;
+ }
break;
default:
BUG();
@@ -1847,6 +1839,7 @@ static inline int mpol_to_str(char *buffer, int maxlen, struct mempolicy *pol)
int l;
nodemask_t nodes;
unsigned short mode = pol ? pol->policy : MPOL_DEFAULT;
+ unsigned short flags = pol ? pol->flags : 0;
switch (mode) {
case MPOL_DEFAULT:
@@ -1876,6 +1869,17 @@ static inline int mpol_to_str(char *buffer, int maxlen, struct mempolicy *pol)
strcpy(p, policy_types[mode]);
p += l;
+ if (flags) {
+ int need_bar = 0;
+
+ if (buffer + maxlen < p + 2)
+ return -ENOSPC;
+ *p++ = '=';
+
+ if (flags & MPOL_F_STATIC_NODES)
+ p += sprintf(p, "%sstatic", need_bar++ ? "|" : "");
+ }
+
if (!nodes_empty(nodes)) {
if (buffer + maxlen < p + 2)
return -ENOSPC;
diff --git a/mm/shmem.c b/mm/shmem.c
index 1ccf794..3e9fda0 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -1126,6 +1126,8 @@ static int shmem_parse_mpol(char *value, unsigned short *policy,
err = 0;
}
if (flags) {
+ if (!strcmp(flags, "static"))
+ *mode_flags |= MPOL_F_STATIC_NODES;
}
out:
/* Restore string for error message */