mm: Fix SHM_HUGETLB to work with users in hugetlb_shm_group

Allow non root users with sufficient mlock rlimits to be able to allocate hugetlb backed shm for now. Deprecate this though. This is being deprecated because the mlock based rlimit checks for SHM_HUGETLB is not consistent with mmap based huge page allocations.

Signed-off-by Ravikiran Thirumalai <[email protected]>

Cc Mel Gorman <[email protected]>

Cc Wli <[email protected]>

Index: linux-2.6-tip/fs/hugetlbfs/inode.c
===================================================================

fs/hugetlbfs/inode.c +1 -1

@@ -948,6 +948,7 @@ static int can_do_hugetlb_shm(void)
 struct file *hugetlb_file_setup(const char *name, size_t size)
 {
 	int error = -ENOMEM;
+	int unlock_shm = 0;
 	struct file *file;
 	struct inode *inode;
 	struct dentry *dentry, *root;

@@ -957,8 +958,14 @@ struct file *hugetlb_file_setup(const ch
 	if (!hugetlbfs_vfsmount)
 		return ERR_PTR(-ENOENT);
 
-	if (!can_do_hugetlb_shm())

fs/hugetlbfs/inode.c +2 -0

 	root = hugetlbfs_vfsmount->mnt_root;
 	quick_string.name = name;

@@ -997,6 +1004,8 @@ out_inode:
 out_dentry:
 	dput(dentry);
 out_shm_unlock:
+	if (unlock_shm)
+		user_shm_unlock(size, user);
 	return ERR_PTR(error);
 }
 

Index: linux-2.6-tip/Documentation/feature-removal-schedule.txt
===================================================================

Documentation/feature-removal-schedule.txt +1 -0

@@ -335,3 +335,14 @@ Why:	In 2.6.18 the Secmark concept was i
 	Secmark, it is time to deprecate the older mechanism and start the
 	process of removing the old code.
 Who:	Paul Moore <[email protected]>
+---------------------------

This is a two patch series to fix a long standing inconsistency with the mechanism to allow non root users allocate hugetlb shm. The patch changelog is self explanatory. Here's a link to the previous discussion as well:

	http://lkml.org/lkml/2009/2/10/89

---
Fix hugetlb subsystem so that non root users belonging to hugetlb_shm_group can actually allocate hugetlb backed shm.

Currently non root users cannot even map one large page using SHM_HUGETLB when they belong to the gid in /proc/sys/vm/hugetlb_shm_group. This is because allocation size is verified against RLIMIT_MEMLOCK resource limit even if the user belongs to hugetlb_shm_group.

This patch
1. Fixes hugetlb subsystem so that users with CAP_IPC_LOCK and users belonging to hugetlb_shm_group don't need to be restricted with RLIMIT_MEMLOCK resource limits
2. This patch also disables mlock based rlimit checking (which will be reinstated and marked deprecated in a subsequent patch).

Signed-off-by Ravikiran Thirumalai <[email protected]>

Cc Mel Gorman <[email protected]>

Cc Wli <[email protected]>

Index: linux-2.6-tip/fs/hugetlbfs/inode.c
===================================================================

fs/hugetlbfs/inode.c +0 -3

 struct file *hugetlb_file_setup(const char *name, size_t size)

@@ -962,9 +960,6 @@ struct file *hugetlb_file_setup(const ch
 	if (!can_do_hugetlb_shm())
 		return ERR_PTR(-EPERM);
 
-	if (!user_shm_lock(size, user))
-		return ERR_PTR(-ENOMEM);
-
 	root = hugetlbfs_vfsmount->mnt_root;
 	quick_string.name = name;
 	quick_string.len = strlen(quick_string.name);

Quoted 6 lines · Ravikiran G Thirumalai

This is a two patch series to fix a long standing inconsistency with the mechanism to allow non root users allocate hugetlb shm. The patch changelog is self explanatory. Here's a link to the previous discussion as well:

	http://lkml.org/lkml/2009/2/10/89

checkpatch complains about the () around the check being unnecessary. Not a big issue though.