alloc_hugepages, free_hugepages — allocate or free huge pages
void *alloc_hugepages( |
int key, |
| void *addr, | |
| size_t len, | |
| int prot, | |
int flag); |
int free_hugepages( |
void *addr); |
The system calls alloc_hugepages() and free_hugepages() were introduced in Linux
2.5.36 and removed again in 2.5.54. They existed only on i386
and ia64 (when built with CONFIG_HUGETLB_PAGE). In Linux 2.4.20 the
syscall numbers exist, but the calls fail with the error
ENOSYS.
On i386 the memory management hardware knows about ordinary pages (4 KiB) and huge pages (2 or 4 MiB). Similarly ia64 knows about huge pages of several sizes. These system calls serve to map huge pages into the process's memory or to free them again. Huge pages are locked into memory, and are not swapped.
The key argument
is an identifier. When zero the pages are private, and not
inherited by children. When positive the pages are shared
with other applications using the same key, and inherited by child
processes.
The addr argument
of free_hugepages() tells which
page is being freed: it was the return value of a call to
alloc_hugepages(). (The memory
is first actually freed when all users have released it.) The
addr argument of
alloc_hugepages() is a hint,
that the kernel may or may not follow. Addresses must be
properly aligned.
The len argument
is the length of the required segment. It must be a multiple
of the huge page size.
The prot argument
specifies the memory protection of the segment. It is one of
PROT_READ, PROT_WRITE, PROT_EXEC.
The flag argument
is ignored, unless key is positive. In that case,
if flag is
IPC_CREAT, then a new huge page
segment is created when none with the given key existed. If
this flag is not set, then ENOENT is returned when no segment with
the given key exists.
On success, alloc_hugepages() returns the allocated
virtual address, and free_hugepages() returns zero. On error,
−1 is returned, and errno
is set appropriately.
/proc/sys/vm/nr_hugepages
Number of configured hugetlb pages. This can be read and
written.
/proc/meminfo Gives info on
the number of configured hugetlb pages and on their size in
the three variables HugePages_Total, HugePages_Free,
Hugepagesize.
These calls are specific to Linux on Intel processors, and should not be used in programs intended to be portable.
These system calls are gone; they existed only in Linux 2.5.36 through to 2.5.54. Now the hugetlbfs file system can be used instead. Memory backed by huge pages (if the CPU supports them) is obtained by using mmap(2) to map files in this virtual file system.
The maximal number of huge pages can be specified using
the hugepages= boot
parameter.
This page is part of release 3.24 of the Linux man-pages project. A
description of the project, and information about reporting
bugs, can be found at
http://www.kernel.org/doc/man-pages/.
|
Copyright 2003 Andries E. Brouwer (aebcwi.nl) Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Since the Linux kernel and libraries are constantly changing, this manual page may be incorrect or out-of-date. The author(s) assume no responsibility for errors or omissions, or for damages resulting from the use of the information contained herein. The author(s) may not have taken the same level of care in the production of this manual, which is licensed free of charge, as they might when working professionally. Formatted or processed versions of this manual, if unaccompanied by the source, must acknowledge the copyright and authors of this work. |