hugetlbfs: accept subpool min_size mount option and setup accordingly

Make 'min_size=<value>' be an option when mounting a hugetlbfs.  This
option takes the same value as the 'size' option.  min_size can be
specified without specifying size.  If both are specified, min_size must
be less that or equal to size else the mount will fail.  If min_size is
specified, then at mount time an attempt is made to reserve min_size
pages.  If the reservation fails, the mount fails.  At umount time, the
reserved pages are released.

Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Aneesh Kumar <aneesh.kumar@linux.vnet.ibm.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c
index db76cec3ce21268bbeffcb45c479bfadf7c33e73..3a8f12762821f0f82f68da70c3e4b3aa50034b42 100644 (file)
--- a/fs/hugetlbfs/inode.c
+++ b/fs/hugetlbfs/inode.c
@@ -47,9 +47,10 @@ struct hugetlbfs_config {
        kuid_t   uid;
        kgid_t   gid;
        umode_t mode;
-       long    nr_blocks;
+       long    max_hpages;
        long    nr_inodes;
        struct hstate *hstate;
+       long    min_hpages;
 };
 
 struct hugetlbfs_inode_info {
@@ -67,7 +68,7 @@ int sysctl_hugetlb_shm_group;
 enum {
        Opt_size, Opt_nr_inodes,
        Opt_mode, Opt_uid, Opt_gid,
-       Opt_pagesize,
+       Opt_pagesize, Opt_min_size,
        Opt_err,
 };
 
@@ -78,6 +79,7 @@ static const match_table_t tokens = {
        {Opt_uid,       "uid=%u"},
        {Opt_gid,       "gid=%u"},
        {Opt_pagesize,  "pagesize=%s"},
+       {Opt_min_size,  "min_size=%s"},
        {Opt_err,       NULL},
 };
 
@@ -754,14 +756,38 @@ static const struct super_operations hugetlbfs_ops = {
        .show_options   = generic_show_options,
 };
 
+enum { NO_SIZE, SIZE_STD, SIZE_PERCENT };
+
+/*
+ * Convert size option passed from command line to number of huge pages
+ * in the pool specified by hstate.  Size option could be in bytes
+ * (val_type == SIZE_STD) or percentage of the pool (val_type == SIZE_PERCENT).
+ */
+static long long
+hugetlbfs_size_to_hpages(struct hstate *h, unsigned long long size_opt,
+                                                               int val_type)
+{
+       if (val_type == NO_SIZE)
+               return -1;
+
+       if (val_type == SIZE_PERCENT) {
+               size_opt <<= huge_page_shift(h);
+               size_opt *= h->max_huge_pages;
+               do_div(size_opt, 100);
+       }
+
+       size_opt >>= huge_page_shift(h);
+       return size_opt;
+}
+
 static int
 hugetlbfs_parse_options(char *options, struct hugetlbfs_config *pconfig)
 {
        char *p, *rest;
        substring_t args[MAX_OPT_ARGS];
        int option;
-       unsigned long long size = 0;
-       enum { NO_SIZE, SIZE_STD, SIZE_PERCENT } setsize = NO_SIZE;
+       unsigned long long max_size_opt = 0, min_size_opt = 0;
+       int max_val_type = NO_SIZE, min_val_type = NO_SIZE;
 
        if (!options)
                return 0;
@@ -799,10 +825,10 @@ hugetlbfs_parse_options(char *options, struct hugetlbfs_config *pconfig)
                        /* memparse() will accept a K/M/G without a digit */
                        if (!isdigit(*args[0].from))
                                goto bad_val;
-                       size = memparse(args[0].from, &rest);
-                       setsize = SIZE_STD;
+                       max_size_opt = memparse(args[0].from, &rest);
+                       max_val_type = SIZE_STD;
                        if (*rest == '%')
-                               setsize = SIZE_PERCENT;
+                               max_val_type = SIZE_PERCENT;
                        break;
                }
 
@@ -825,6 +851,17 @@ hugetlbfs_parse_options(char *options, struct hugetlbfs_config *pconfig)
                        break;
                }
 
+               case Opt_min_size: {
+                       /* memparse() will accept a K/M/G without a digit */
+                       if (!isdigit(*args[0].from))
+                               goto bad_val;
+                       min_size_opt = memparse(args[0].from, &rest);
+                       min_val_type = SIZE_STD;
+                       if (*rest == '%')
+                               min_val_type = SIZE_PERCENT;
+                       break;
+               }
+
                default:
                        pr_err("Bad mount option: \"%s\"\n", p);
                        return -EINVAL;
@@ -832,15 +869,22 @@ hugetlbfs_parse_options(char *options, struct hugetlbfs_config *pconfig)
                }
        }
 
-       /* Do size after hstate is set up */
-       if (setsize > NO_SIZE) {
-               struct hstate *h = pconfig->hstate;
-               if (setsize == SIZE_PERCENT) {
-                       size <<= huge_page_shift(h);
-                       size *= h->max_huge_pages;
-                       do_div(size, 100);
-               }
-               pconfig->nr_blocks = (size >> huge_page_shift(h));
+       /*
+        * Use huge page pool size (in hstate) to convert the size
+        * options to number of huge pages.  If NO_SIZE, -1 is returned.
+        */
+       pconfig->max_hpages = hugetlbfs_size_to_hpages(pconfig->hstate,
+                                               max_size_opt, max_val_type);
+       pconfig->min_hpages = hugetlbfs_size_to_hpages(pconfig->hstate,
+                                               min_size_opt, min_val_type);
+
+       /*
+        * If max_size was specified, then min_size must be smaller
+        */
+       if (max_val_type > NO_SIZE &&
+           pconfig->min_hpages > pconfig->max_hpages) {
+               pr_err("minimum size can not be greater than maximum size\n");
+               return -EINVAL;
        }
 
        return 0;
@@ -859,12 +903,13 @@ hugetlbfs_fill_super(struct super_block *sb, void *data, int silent)
 
        save_mount_options(sb, data);
 
-       config.nr_blocks = -1; /* No limit on size by default */
+       config.max_hpages = -1; /* No limit on size by default */
        config.nr_inodes = -1; /* No limit on number of inodes by default */
        config.uid = current_fsuid();
        config.gid = current_fsgid();
        config.mode = 0755;
        config.hstate = &default_hstate;
+       config.min_hpages = -1; /* No default minimum size */
        ret = hugetlbfs_parse_options(data, &config);
        if (ret)
                return ret;
@@ -878,8 +923,15 @@ hugetlbfs_fill_super(struct super_block *sb, void *data, int silent)
        sbinfo->max_inodes = config.nr_inodes;
        sbinfo->free_inodes = config.nr_inodes;
        sbinfo->spool = NULL;
-       if (config.nr_blocks != -1) {
-               sbinfo->spool = hugepage_new_subpool(config.nr_blocks);
+       /*
+        * Allocate and initialize subpool if maximum or minimum size is
+        * specified.  Any needed reservations (for minimim size) are taken
+        * taken when the subpool is created.
+        */
+       if (config.max_hpages != -1 || config.min_hpages != -1) {
+               sbinfo->spool = hugepage_new_subpool(config.hstate,
+                                                       config.max_hpages,
+                                                       config.min_hpages);
                if (!sbinfo->spool)
                        goto out_free;
        }

diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index d1a77b87408dc689936097c2231188ee45879722..5713c49a4a5c4e1544ee63e7f9dc7a54364cbf84 100644 (file)
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -44,7 +44,8 @@ extern int hugetlb_max_hstate __read_mostly;
 #define for_each_hstate(h) \
        for ((h) = hstates; (h) < &hstates[hugetlb_max_hstate]; (h)++)
 
-struct hugepage_subpool *hugepage_new_subpool(long nr_blocks);
+struct hugepage_subpool *hugepage_new_subpool(struct hstate *h, long max_hpages,
+                                               long min_hpages);
 void hugepage_put_subpool(struct hugepage_subpool *spool);
 
 void reset_vma_resv_huge_pages(struct vm_area_struct *vma);

diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 499cb72c74b1d4c70dfb05610c0fe1248a15f166..995c8d65a95c019d9b86719cd5e0935144b5cc05 100644 (file)
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -61,6 +61,9 @@ DEFINE_SPINLOCK(hugetlb_lock);
 static int num_fault_mutexes;
 static struct mutex *htlb_fault_mutex_table ____cacheline_aligned_in_smp;
 
+/* Forward declaration */
+static int hugetlb_acct_memory(struct hstate *h, long delta);
+
 static inline void unlock_or_release_subpool(struct hugepage_subpool *spool)
 {
        bool free = (spool->count == 0) && (spool->used_hpages == 0);
@@ -68,12 +71,18 @@ static inline void unlock_or_release_subpool(struct hugepage_subpool *spool)
        spin_unlock(&spool->lock);
 
        /* If no pages are used, and no other handles to the subpool
-        * remain, free the subpool the subpool remain */
-       if (free)
+        * remain, give up any reservations mased on minimum size and
+        * free the subpool */
+       if (free) {
+               if (spool->min_hpages != -1)
+                       hugetlb_acct_memory(spool->hstate,
+                                               -spool->min_hpages);
                kfree(spool);
+       }
 }
 
-struct hugepage_subpool *hugepage_new_subpool(long nr_blocks)
+struct hugepage_subpool *hugepage_new_subpool(struct hstate *h, long max_hpages,
+                                               long min_hpages)
 {
        struct hugepage_subpool *spool;
 
@@ -83,7 +92,15 @@ struct hugepage_subpool *hugepage_new_subpool(long nr_blocks)
 
        spin_lock_init(&spool->lock);
        spool->count = 1;
-       spool->max_hpages = nr_blocks;
+       spool->max_hpages = max_hpages;
+       spool->hstate = h;
+       spool->min_hpages = min_hpages;
+
+       if (min_hpages != -1 && hugetlb_acct_memory(h, min_hpages)) {
+               kfree(spool);
+               return NULL;
+       }
+       spool->rsv_hpages = min_hpages;
 
        return spool;
 }