[PATCH] hugepage: Strict page reservation for hugepage inodes

These days, hugepages are demand-allocated at first fault time.  There's a
somewhat dubious (and racy) heuristic when making a new mmap() to check if
there are enough available hugepages to fully satisfy that mapping.

A particularly obvious case where the heuristic breaks down is where a
process maps its hugepages not as a single chunk, but as a bunch of
individually mmap()ed (or shmat()ed) blocks without touching and
instantiating the pages in between allocations.  In this case the size of
each block is compared against the total number of available hugepages.
It's thus easy for the process to become overcommitted, because each block
mapping will succeed, although the total number of hugepages required by
all blocks exceeds the number available.  In particular, this defeats such
a program which will detect a mapping failure and adjust its hugepage usage
downward accordingly.

The patch below addresses this problem, by strictly reserving a number of
physical hugepages for hugepage inodes which have been mapped, but not
instatiated.  MAP_SHARED mappings are thus "safe" - they will fail on
mmap(), not later with an OOM SIGKILL.  MAP_PRIVATE mappings can still
trigger an OOM.  (Actually SHARED mappings can technically still OOM, but
only if the sysadmin explicitly reduces the hugepage pool between mapping
and instantiation)

This patch appears to address the problem at hand - it allows DB2 to start
correctly, for instance, which previously suffered the failure described
above.

This patch causes no regressions on the libhugetblfs testsuite, and makes a
test (designed to catch this problem) pass which previously failed (ppc64,
POWER5).

Signed-off-by: David Gibson <dwg@au1.ibm.com>
Cc: William Lee Irwin III <wli@holomorphy.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

1 files changed, 22 insertions, 52 deletions

diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c
index b35195289945..1a1c2fcb7823 100644
--- a/fs/hugetlbfs/inode.c
+++ b/fs/hugetlbfs/inode.c
@@ -56,48 +56,10 @@ static void huge_pagevec_release(struct pagevec *pvec)
         pagevec_reinit(pvec);
 }
 
-/*
- * huge_pages_needed tries to determine the number of new huge pages that
- * will be required to fully populate this VMA.  This will be equal to
- * the size of the VMA in huge pages minus the number of huge pages
- * (covered by this VMA) that are found in the page cache.
- *
- * Result is in bytes to be compatible with is_hugepage_mem_enough()
- */
-static unsigned long
-huge_pages_needed(struct address_space *mapping, struct vm_area_struct *vma)
-{
-        int i;
-        struct pagevec pvec;
-        unsigned long start = vma->vm_start;
-        unsigned long end = vma->vm_end;
-        unsigned long hugepages = (end - start) >> HPAGE_SHIFT;
-        pgoff_t next = vma->vm_pgoff >> (HPAGE_SHIFT - PAGE_SHIFT);
-        pgoff_t endpg = next + hugepages;
-
-        pagevec_init(&pvec, 0);
-        while (next < endpg) {
-                if (!pagevec_lookup(&pvec, mapping, next, PAGEVEC_SIZE))
-                        break;
-                for (i = 0; i < pagevec_count(&pvec); i++) {
-                        struct page *page = pvec.pages[i];
-                        if (page->index > next)
-                                next = page->index;
-                        if (page->index >= endpg)
-                                break;
-                        next++;
-                        hugepages--;
-                }
-                huge_pagevec_release(&pvec);
-        }
-        return hugepages << HPAGE_SHIFT;
-}
-
 static int hugetlbfs_file_mmap(struct file *file, struct vm_area_struct *vma)
 {
         struct inode *inode = file->f_dentry->d_inode;
-        struct address_space *mapping = inode->i_mapping;
-        unsigned long bytes;
+        struct hugetlbfs_inode_info *info = HUGETLBFS_I(inode);
         loff_t len, vma_len;
         int ret;
 
@@ -113,10 +75,6 @@ static int hugetlbfs_file_mmap(struct file *file, struct vm_area_struct *vma)
         if (vma->vm_end - vma->vm_start < HPAGE_SIZE)
                 return -EINVAL;
 
-        bytes = huge_pages_needed(mapping, vma);
-        if (!is_hugepage_mem_enough(bytes))
-                return -ENOMEM;
-
         vma_len = (loff_t)(vma->vm_end - vma->vm_start);
 
         mutex_lock(&inode->i_mutex);
@@ -129,6 +87,10 @@ static int hugetlbfs_file_mmap(struct file *file, struct vm_area_struct *vma)
         if (!(vma->vm_flags & VM_WRITE) && len > inode->i_size)
                 goto out;
 
+        if (vma->vm_flags & VM_MAYSHARE)
+                if (hugetlb_extend_reservation(info, len >> HPAGE_SHIFT) != 0)
+                        goto out;
+
         ret = 0;
         hugetlb_prefault_arch_hook(vma->vm_mm);
         if (inode->i_size < len)
@@ -227,13 +189,18 @@ static void truncate_huge_page(struct page *page)
         put_page(page);
 }
 
-static void truncate_hugepages(struct address_space *mapping, loff_t lstart)
+static void truncate_hugepages(struct inode *inode, loff_t lstart)
 {
+        struct address_space *mapping = &inode->i_data;
         const pgoff_t start = lstart >> HPAGE_SHIFT;
         struct pagevec pvec;
         pgoff_t next;
         int i;
 
+        hugetlb_truncate_reservation(HUGETLBFS_I(inode),
+                                     lstart >> HPAGE_SHIFT);
+        if (!mapping->nrpages)
+                return;
         pagevec_init(&pvec, 0);
         next = start;
         while (1) {
@@ -262,8 +229,7 @@ static void truncate_hugepages(struct address_space *mapping, loff_t lstart)
 
 static void hugetlbfs_delete_inode(struct inode *inode)
 {
-        if (inode->i_data.nrpages)
-                truncate_hugepages(&inode->i_data, 0);
+        truncate_hugepages(inode, 0);
         clear_inode(inode);
 }
 
@@ -296,8 +262,7 @@ static void hugetlbfs_forget_inode(struct inode *inode)
         inode->i_state |= I_FREEING;
         inodes_stat.nr_inodes--;
         spin_unlock(&inode_lock);
-        if (inode->i_data.nrpages)
-                truncate_hugepages(&inode->i_data, 0);
+        truncate_hugepages(inode, 0);
         clear_inode(inode);
         destroy_inode(inode);
 }
@@ -356,7 +321,7 @@ static int hugetlb_vmtruncate(struct inode *inode, loff_t offset)
         if (!prio_tree_empty(&mapping->i_mmap))
                 hugetlb_vmtruncate_list(&mapping->i_mmap, pgoff);
         spin_unlock(&mapping->i_mmap_lock);
-        truncate_hugepages(mapping, offset);
+        truncate_hugepages(inode, offset);
         return 0;
 }
 
@@ -573,6 +538,7 @@ static struct inode *hugetlbfs_alloc_inode(struct super_block *sb)
                 hugetlbfs_inc_free_inodes(sbinfo);
                 return NULL;
         }
+        p->prereserved_hpages = 0;
         return &p->vfs_inode;
 }
 
@@ -805,9 +771,6 @@ struct file *hugetlb_zero_setup(size_t size)
         if (!can_do_hugetlb_shm())
                 return ERR_PTR(-EPERM);
 
-        if (!is_hugepage_mem_enough(size))
-                return ERR_PTR(-ENOMEM);
-
         if (!user_shm_lock(size, current->user))
                 return ERR_PTR(-ENOMEM);
 
@@ -831,6 +794,11 @@ struct file *hugetlb_zero_setup(size_t size)
         if (!inode)
                 goto out_file;
 
+        error = -ENOMEM;
+        if (hugetlb_extend_reservation(HUGETLBFS_I(inode),
+                                       size >> HPAGE_SHIFT) != 0)
+                goto out_inode;
+
         d_instantiate(dentry, inode);
         inode->i_size = size;
         inode->i_nlink = 0;
@@ -841,6 +809,8 @@ struct file *hugetlb_zero_setup(size_t size)
         file->f_mode = FMODE_WRITE | FMODE_READ;
         return file;
 
+out_inode:
+        iput(inode);
 out_file:
         put_filp(file);
 out_dentry: