[PATCH 0/3] Add pseudo-anonymous huge page mappings

[PATCH 0/3] Add pseudo-anonymous huge page mappings

[Eric B Munson]

This patch set adds a flag to mmap that allows the user to request
a mapping to be backed with huge pages.  This mapping will borrow
functionality from the huge page shm code to create a file on the
kernel internal mount and uses it to approximate an anonymous
mapping.  The MAP_LARGEPAGE flag is a modifier to MAP_ANONYMOUS
and will not work without both flags being preset.

A new flag is necessary because there is no other way to hook into
huge pages without creating a file on a hugetlbfs mount which
wouldn't be MAP_ANONYMOUS.

To userspace, this mapping will behave just like an anonymous mapping
because the file is not accessible outside of the kernel.

Eric B Munson (3):
  hugetlbfs: Allow the creation of files suitable for MAP_PRIVATE on
    the vfs internal mount
  Add MAP_LARGEPAGE for mmaping pseudo-anonymous huge page regions
  Add MAP_LARGEPAGE example to vm/hugetlbpage.txt

 Documentation/vm/hugetlbpage.txt  |   80 +++++++++++++++++++++++++++++++++++++
 fs/hugetlbfs/inode.c              |   22 ++++++++--
 include/asm-generic/mman-common.h |    1 +
 include/linux/hugetlb.h           |   17 +++++++-
 ipc/shm.c                         |    3 +-
 mm/mmap.c                         |   16 +++++++
 6 files changed, 133 insertions(+), 6 deletions(-)

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH 1/3] hugetlbfs: Allow the creation of files suitable for MAP_PRIVATE on the vfs internal mount

[Eric B Munson]

There are two means of creating mappings backed by huge pages:

        1. mmap() a file created on hugetlbfs
        2. Use shm which creates a file on an internal mount which essentially
           maps it MAP_SHARED

The internal mount is only used for shared mappings but there is very
little that stops it being used for private mappings. This patch extends
hugetlbfs_file_setup() to deal with the creation of files that will be
mapped MAP_PRIVATE on the internal hugetlbfs mount. This extended API is
used in a subsequent patch to implement the MAP_LARGEPAGE mmap() flag.

Signed-off-by: Eric Munson <ebmunson@us.ibm.com>
---
 fs/hugetlbfs/inode.c    |   22 ++++++++++++++++++----
 include/linux/hugetlb.h |   10 +++++++++-
 ipc/shm.c               |    3 ++-
 3 files changed, 29 insertions(+), 6 deletions(-)

diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c
index 941c842..361f536 100644
--- a/fs/hugetlbfs/inode.c
+++ b/fs/hugetlbfs/inode.c
@@ -506,6 +506,13 @@ static struct inode *hugetlbfs_get_inode(struct super_block *sb, uid_t uid,
 		inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME;
 		INIT_LIST_HEAD(&inode->i_mapping->private_list);
 		info = HUGETLBFS_I(inode);
+		/*
+		 * The policy is initialized here even if we are creating a
+		 * private inode because initialization simply creates an
+		 * an empty rb tree and calls spin_lock_init(), later when we
+		 * call mpol_free_shared_policy() it will just return because
+		 * the rb tree will still be empty.
+		 */
 		mpol_shared_policy_init(&info->policy, NULL);
 		switch (mode & S_IFMT) {
 		default:
@@ -930,12 +937,19 @@ static struct file_system_type hugetlbfs_fs_type = {
 
 static struct vfsmount *hugetlbfs_vfsmount;
 
-static int can_do_hugetlb_shm(void)
+static int can_do_hugetlb_shm(int creat_flags)
 {
-	return capable(CAP_IPC_LOCK) || in_group_p(sysctl_hugetlb_shm_group);
+	if (!(creat_flags & HUGETLB_SHMFS_INODE))
+		return 0;
+	if (capable(CAP_IPC_LOCK))
+		return 1;
+	if (in_group_p(sysctl_hugetlb_shm_group))
+		return 1;
+	return 0;
 }
 
-struct file *hugetlb_file_setup(const char *name, size_t size, int acctflag)
+struct file *hugetlb_file_setup(const char *name, size_t size, int acctflag,
+				int creat_flags)
 {
 	int error = -ENOMEM;
 	int unlock_shm = 0;
@@ -948,7 +962,7 @@ struct file *hugetlb_file_setup(const char *name, size_t size, int acctflag)
 	if (!hugetlbfs_vfsmount)
 		return ERR_PTR(-ENOENT);
 
-	if (!can_do_hugetlb_shm()) {
+	if (!can_do_hugetlb_shm(creat_flags)) {
 		if (user_shm_lock(size, user)) {
 			unlock_shm = 1;
 			WARN_ONCE(1,
diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index 2723513..78b6ddf 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -109,6 +109,14 @@ static inline void hugetlb_report_meminfo(struct seq_file *m)
 
 #endif /* !CONFIG_HUGETLB_PAGE */
 
+enum {
+	/*
+	 * The file will be used as an shm file so shmfs accounting rules
+	 * apply
+	 */
+	HUGETLB_SHMFS_INODE     = 0x01,
+};
+
 #ifdef CONFIG_HUGETLBFS
 struct hugetlbfs_config {
 	uid_t   uid;
@@ -146,7 +154,7 @@ static inline struct hugetlbfs_sb_info *HUGETLBFS_SB(struct super_block *sb)
 
 extern const struct file_operations hugetlbfs_file_operations;
 extern struct vm_operations_struct hugetlb_vm_ops;
-struct file *hugetlb_file_setup(const char *name, size_t, int);
+struct file *hugetlb_file_setup(const char *name, size_t, int, int);
 int hugetlb_get_quota(struct address_space *mapping, long delta);
 void hugetlb_put_quota(struct address_space *mapping, long delta);
 
diff --git a/ipc/shm.c b/ipc/shm.c
index 15dd238..801c68a 100644
--- a/ipc/shm.c
+++ b/ipc/shm.c
@@ -369,7 +369,8 @@ static int newseg(struct ipc_namespace *ns, struct ipc_params *params)
 		/* hugetlb_file_setup applies strict accounting */
 		if (shmflg & SHM_NORESERVE)
 			acctflag = VM_NORESERVE;
-		file = hugetlb_file_setup(name, size, acctflag);
+		file = hugetlb_file_setup(name, size, acctflag,
+					HUGETLB_SHMFS_INODE);
 		shp->mlock_user = current_user();
 	} else {
 		/*
-- 
1.6.3.2

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH 2/3] Add MAP_LARGEPAGE for mmaping pseudo-anonymous huge page regions

[Eric B Munson]

This patch adds a flag for mmap that will be used to request a huge
page region that will look like anonymous memory to user space.  This
is accomplished by using a file on the internal vfsmount.  MAP_LARGEPAGE
is a modifier of MAP_ANONYMOUS and so must be specified with it.  The
region will behave the same as a MAP_ANONYMOUS region using small pages.

Signed-off-by: Eric B Munson <ebmunson@us.ibm.com>
---
 include/asm-generic/mman-common.h |    1 +
 include/linux/hugetlb.h           |    7 +++++++
 mm/mmap.c                         |   16 ++++++++++++++++
 3 files changed, 24 insertions(+), 0 deletions(-)

diff --git a/include/asm-generic/mman-common.h b/include/asm-generic/mman-common.h
index 3b69ad3..60b6be7 100644
--- a/include/asm-generic/mman-common.h
+++ b/include/asm-generic/mman-common.h
@@ -19,6 +19,7 @@
 #define MAP_TYPE	0x0f		/* Mask for type of mapping */
 #define MAP_FIXED	0x10		/* Interpret addr exactly */
 #define MAP_ANONYMOUS	0x20		/* don't use a file */
+#define MAP_LARGEPAGE	0x40		/* create a large page mapping */
 
 #define MS_ASYNC	1		/* sync memory asynchronously */
 #define MS_INVALIDATE	2		/* invalidate the caches */
diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index 78b6ddf..b84361c 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -109,12 +109,19 @@ static inline void hugetlb_report_meminfo(struct seq_file *m)
 
 #endif /* !CONFIG_HUGETLB_PAGE */
 
+#define HUGETLB_ANON_FILE "anon_hugepage"
+
 enum {
 	/*
 	 * The file will be used as an shm file so shmfs accounting rules
 	 * apply
 	 */
 	HUGETLB_SHMFS_INODE     = 0x01,
+	/*
+	 * The file is being created on the internal vfs mount and shmfs
+	 * accounting rules do not apply
+	 */
+	HUGETLB_ANONHUGE_INODE  = 0x02,
 };
 
 #ifdef CONFIG_HUGETLBFS
diff --git a/mm/mmap.c b/mm/mmap.c
index 34579b2..c2c729a 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -29,6 +29,7 @@
 #include <linux/rmap.h>
 #include <linux/mmu_notifier.h>
 #include <linux/perf_counter.h>
+#include <linux/hugetlb.h>
 
 #include <asm/uaccess.h>
 #include <asm/cacheflush.h>
@@ -954,6 +955,21 @@ unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
 	if (mm->map_count > sysctl_max_map_count)
 		return -ENOMEM;
 
+	if (flags & MAP_LARGEPAGE) {
+		if (file)
+			return -EINVAL;
+
+		/*
+		 * VM_NORESERVE is used because the reservations will be
+		 * taken when vm_ops->mmap() is called
+		 */
+		len = ALIGN(len, huge_page_size(&default_hstate));
+		file = hugetlb_file_setup(HUGETLB_ANON_FILE, len, VM_NORESERVE,
+						HUGETLB_ANONHUGE_INODE);
+		if (IS_ERR(file))
+			return -ENOMEM;
+	}
+
 	/* Obtain the address to map to. we verify (or select) it and ensure
 	 * that it represents a valid section of the address space.
 	 */
-- 
1.6.3.2

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH 3/3] Add MAP_LARGEPAGE example to vm/hugetlbpage.txt

[Eric B Munson]

This patch adds an example of how to use the MAP_LARGEPAGE flag to
the vm documentation.

Signed-off-by: Eric B Munson <ebmunson@us.ibm.com>
---
 Documentation/vm/hugetlbpage.txt |   80 ++++++++++++++++++++++++++++++++++++++
 1 files changed, 80 insertions(+), 0 deletions(-)

diff --git a/Documentation/vm/hugetlbpage.txt b/Documentation/vm/hugetlbpage.txt
index ea8714f..fec7fc1 100644
--- a/Documentation/vm/hugetlbpage.txt
+++ b/Documentation/vm/hugetlbpage.txt
@@ -337,3 +337,83 @@ int main(void)
 
 	return 0;
 }
+
+*******************************************************************
+
+/*
+ * Example of using hugepage memory in a user application using the mmap
+ * system call with MAP_LARGEPAGE flag.  Before running this program make
+ * sure the administrator has allocated enough default sized huge pages
+ * to cover the 256 MB allocation.
+ *
+ * For ia64 architecture, Linux kernel reserves Region number 4 for hugepages.
+ * That means the addresses starting with 0x800000... will need to be
+ * specified.  Specifying a fixed address is not required on ppc64, i386
+ * or x86_64.
+ */
+#include <stdlib.h>
+#include <stdio.h>
+#include <unistd.h>
+#include <sys/mman.h>
+#include <fcntl.h>
+
+#define LENGTH (256UL*1024*1024)
+#define PROTECTION (PROT_READ | PROT_WRITE)
+
+#ifndef MAP_LARGEPAGE
+#define MAP_LARGEPAGE 0x40
+#endif
+
+/* Only ia64 requires this */
+#ifdef __ia64__
+#define ADDR (void *)(0x8000000000000000UL)
+#define FLAGS (MAP_PRIVATE | MAP_ANONYMOUS | MAP_LARGEPAGE | MAP_FIXED)
+#else
+#define ADDR (void *)(0x0UL)
+#define FLAGS (MAP_PRIVATE | MAP_ANONYMOUS | MAP_LARGEPAGE)
+#endif
+
+void check_bytes(char *addr)
+{
+	printf("First hex is %x\n", *((unsigned int *)addr));
+}
+
+void write_bytes(char *addr)
+{
+	unsigned long i;
+
+	for (i = 0; i < LENGTH; i++)
+		*(addr + i) = (char)i;
+}
+
+void read_bytes(char *addr)
+{
+	unsigned long i;
+
+	check_bytes(addr);
+	for (i = 0; i < LENGTH; i++)
+		if (*(addr + i) != (char)i) {
+			printf("Mismatch at %lu\n", i);
+			break;
+		}
+}
+
+int main(void)
+{
+	void *addr;
+
+	addr = mmap(ADDR, LENGTH, PROTECTION, FLAGS, 0, 0);
+	if (addr == MAP_FAILED) {
+		perror("mmap");
+		exit(1);
+	}
+
+	printf("Returned address is %p\n", addr);
+	check_bytes(addr);
+	write_bytes(addr);
+	read_bytes(addr);
+
+	munmap(addr, LENGTH);
+
+	return 0;
+}
-- 
1.6.3.2

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH 2/3] Add MAP_LARGEPAGE for mmaping pseudo-anonymous huge page regions

[Michael Kerrisk]

Eric,

On Wed, Aug 12, 2009 at 12:13 AM, Eric B Munson<ebmunson@us.ibm.com> wrote:
> This patch adds a flag for mmap that will be used to request a huge
> page region that will look like anonymous memory to user space.  This
> is accomplished by using a file on the internal vfsmount.  MAP_LARGEPAGE
> is a modifier of MAP_ANONYMOUS and so must be specified with it.  The
> region will behave the same as a MAP_ANONYMOUS region using small pages.

Does this flag provide functionality analogous to shmget(SHM_HUGETLB)?
If so, would iot not make sense to name it similarly (i.e.,
MAP_HUGETLB)?

Cheers,

Michael

-- 
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
Watch my Linux system programming book progress to publication!
http://blog.man7.org/

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH 2/3] Add MAP_LARGEPAGE for mmaping pseudo-anonymous huge page regions

[Eric B Munson]

[-- Attachment #1: Type: text/plain, Size: 1058 bytes --]

On Wed, 12 Aug 2009, Michael Kerrisk wrote:

> Eric,
> 
> On Wed, Aug 12, 2009 at 12:13 AM, Eric B Munson<ebmunson@us.ibm.com> wrote:
> > This patch adds a flag for mmap that will be used to request a huge
> > page region that will look like anonymous memory to user space.  This
> > is accomplished by using a file on the internal vfsmount.  MAP_LARGEPAGE
> > is a modifier of MAP_ANONYMOUS and so must be specified with it.  The
> > region will behave the same as a MAP_ANONYMOUS region using small pages.
> 
> Does this flag provide functionality analogous to shmget(SHM_HUGETLB)?
> If so, would iot not make sense to name it similarly (i.e.,
> MAP_HUGETLB)?
> 
> Cheers,
> 
> Michael
> 
> -- 
> Michael Kerrisk
> Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
> Watch my Linux system programming book progress to publication!
> http://blog.man7.org/
> 

I have no particular attachment to MAP_LARGEPAGE, I will make this chage for V2.

-- 
Eric B Munson
IBM Linux Technology Center
ebmunson@us.ibm.com


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

Re: [PATCH 2/3] Add MAP_LARGEPAGE for mmaping pseudo-anonymous huge page regions

[Pekka Enberg]

Hi Eric,

On Wed, Aug 12, 2009 at 1:13 AM, Eric B Munson<ebmunson@us.ibm.com> wrote:
> This patch adds a flag for mmap that will be used to request a huge
> page region that will look like anonymous memory to user space.  This
> is accomplished by using a file on the internal vfsmount.  MAP_LARGEPAGE
> is a modifier of MAP_ANONYMOUS and so must be specified with it.  The
> region will behave the same as a MAP_ANONYMOUS region using small pages.
>
> Signed-off-by: Eric B Munson <ebmunson@us.ibm.com>

I would love to see something like this in the kernel. Huge pages are
useful for garbage collection and JIT text in the userspace but
unfortunately obtaining them is a real PITA at the moment.

Is there any way to drop the
CAP_IPC_LOCK/in_group_p(hugetlbfs_shm_group) requirement, btw? That
would make huge pages even more accessible to user-space virtual
machines.

                        Pekka

> ---
>  include/asm-generic/mman-common.h |    1 +
>  include/linux/hugetlb.h           |    7 +++++++
>  mm/mmap.c                         |   16 ++++++++++++++++
>  3 files changed, 24 insertions(+), 0 deletions(-)
>
> diff --git a/include/asm-generic/mman-common.h b/include/asm-generic/mman-common.h
> index 3b69ad3..60b6be7 100644
> --- a/include/asm-generic/mman-common.h
> +++ b/include/asm-generic/mman-common.h
> @@ -19,6 +19,7 @@
>  #define MAP_TYPE       0x0f            /* Mask for type of mapping */
>  #define MAP_FIXED      0x10            /* Interpret addr exactly */
>  #define MAP_ANONYMOUS  0x20            /* don't use a file */
> +#define MAP_LARGEPAGE  0x40            /* create a large page mapping */
>
>  #define MS_ASYNC       1               /* sync memory asynchronously */
>  #define MS_INVALIDATE  2               /* invalidate the caches */
> diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
> index 78b6ddf..b84361c 100644
> --- a/include/linux/hugetlb.h
> +++ b/include/linux/hugetlb.h
> @@ -109,12 +109,19 @@ static inline void hugetlb_report_meminfo(struct seq_file *m)
>
>  #endif /* !CONFIG_HUGETLB_PAGE */
>
> +#define HUGETLB_ANON_FILE "anon_hugepage"
> +
>  enum {
>        /*
>         * The file will be used as an shm file so shmfs accounting rules
>         * apply
>         */
>        HUGETLB_SHMFS_INODE     = 0x01,
> +       /*
> +        * The file is being created on the internal vfs mount and shmfs
> +        * accounting rules do not apply
> +        */
> +       HUGETLB_ANONHUGE_INODE  = 0x02,
>  };
>
>  #ifdef CONFIG_HUGETLBFS
> diff --git a/mm/mmap.c b/mm/mmap.c
> index 34579b2..c2c729a 100644
> --- a/mm/mmap.c
> +++ b/mm/mmap.c
> @@ -29,6 +29,7 @@
>  #include <linux/rmap.h>
>  #include <linux/mmu_notifier.h>
>  #include <linux/perf_counter.h>
> +#include <linux/hugetlb.h>
>
>  #include <asm/uaccess.h>
>  #include <asm/cacheflush.h>
> @@ -954,6 +955,21 @@ unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
>        if (mm->map_count > sysctl_max_map_count)
>                return -ENOMEM;
>
> +       if (flags & MAP_LARGEPAGE) {
> +               if (file)
> +                       return -EINVAL;
> +
> +               /*
> +                * VM_NORESERVE is used because the reservations will be
> +                * taken when vm_ops->mmap() is called
> +                */
> +               len = ALIGN(len, huge_page_size(&default_hstate));
> +               file = hugetlb_file_setup(HUGETLB_ANON_FILE, len, VM_NORESERVE,
> +                                               HUGETLB_ANONHUGE_INODE);
> +               if (IS_ERR(file))
> +                       return -ENOMEM;
> +       }
> +
>        /* Obtain the address to map to. we verify (or select) it and ensure
>         * that it represents a valid section of the address space.
>         */
> --
> 1.6.3.2
>
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo@kvack.org.  For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>