^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) .. SPDX-License-Identifier: GPL-2.0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) =====
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) Tmpfs
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) =====
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) Tmpfs is a file system which keeps all files in virtual memory.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) Everything in tmpfs is temporary in the sense that no files will be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) created on your hard drive. If you unmount a tmpfs instance,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) everything stored therein is lost.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) tmpfs puts everything into the kernel internal caches and grows and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) shrinks to accommodate the files it contains and is able to swap
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) unneeded pages out to swap space. It has maximum size limits which can
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) be adjusted on the fly via 'mount -o remount ...'
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) If you compare it to ramfs (which was the template to create tmpfs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) you gain swapping and limit checking. Another similar thing is the RAM
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) disk (/dev/ram*), which simulates a fixed size hard disk in physical
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) RAM, where you have to create an ordinary filesystem on top. Ramdisks
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) cannot swap and you do not have the possibility to resize them.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) Since tmpfs lives completely in the page cache and on swap, all tmpfs
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) pages will be shown as "Shmem" in /proc/meminfo and "Shared" in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) free(1). Notice that these counters also include shared memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) (shmem, see ipcs(1)). The most reliable way to get the count is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) using df(1) and du(1).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) tmpfs has the following uses:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) 1) There is always a kernel internal mount which you will not see at
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) all. This is used for shared anonymous mappings and SYSV shared
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) memory.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) This mount does not depend on CONFIG_TMPFS. If CONFIG_TMPFS is not
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) set, the user visible part of tmpfs is not build. But the internal
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) mechanisms are always present.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) 2) glibc 2.2 and above expects tmpfs to be mounted at /dev/shm for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) POSIX shared memory (shm_open, shm_unlink). Adding the following
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) line to /etc/fstab should take care of this::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) tmpfs /dev/shm tmpfs defaults 0 0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) Remember to create the directory that you intend to mount tmpfs on
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) if necessary.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) This mount is _not_ needed for SYSV shared memory. The internal
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) mount is used for that. (In the 2.3 kernel versions it was
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) necessary to mount the predecessor of tmpfs (shm fs) to use SYSV
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) shared memory)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) 3) Some people (including me) find it very convenient to mount it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) e.g. on /tmp and /var/tmp and have a big swap partition. And now
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) loop mounts of tmpfs files do work, so mkinitrd shipped by most
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) distributions should succeed with a tmpfs /tmp.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) 4) And probably a lot more I do not know about :-)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) tmpfs has three mount options for sizing:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) ========= ============================================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) size The limit of allocated bytes for this tmpfs instance. The
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) default is half of your physical RAM without swap. If you
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) oversize your tmpfs instances the machine will deadlock
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) since the OOM handler will not be able to free that memory.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) nr_blocks The same as size, but in blocks of PAGE_SIZE.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) nr_inodes The maximum number of inodes for this instance. The default
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) is half of the number of your physical RAM pages, or (on a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) machine with highmem) the number of lowmem RAM pages,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) whichever is the lower.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) ========= ============================================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) These parameters accept a suffix k, m or g for kilo, mega and giga and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) can be changed on remount. The size parameter also accepts a suffix %
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) to limit this tmpfs instance to that percentage of your physical RAM:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) the default, when neither size nor nr_blocks is specified, is size=50%
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) If nr_blocks=0 (or size=0), blocks will not be limited in that instance;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) if nr_inodes=0, inodes will not be limited. It is generally unwise to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) mount with such options, since it allows any user with write access to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) use up all the memory on the machine; but enhances the scalability of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) that instance in a system with many cpus making intensive use of it.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) tmpfs has a mount option to set the NUMA memory allocation policy for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) all files in that instance (if CONFIG_NUMA is enabled) - which can be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) adjusted on the fly via 'mount -o remount ...'
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) ======================== ==============================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) mpol=default use the process allocation policy
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) (see set_mempolicy(2))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) mpol=prefer:Node prefers to allocate memory from the given Node
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) mpol=bind:NodeList allocates memory only from nodes in NodeList
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) mpol=interleave prefers to allocate from each node in turn
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) mpol=interleave:NodeList allocates from each node of NodeList in turn
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) mpol=local prefers to allocate memory from the local node
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) ======================== ==============================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) NodeList format is a comma-separated list of decimal numbers and ranges,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) a range being two hyphen-separated decimal numbers, the smallest and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) largest node numbers in the range. For example, mpol=bind:0-3,5,7,9-15
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) A memory policy with a valid NodeList will be saved, as specified, for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) use at file creation time. When a task allocates a file in the file
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) system, the mount option memory policy will be applied with a NodeList,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) if any, modified by the calling task's cpuset constraints
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) [See Documentation/admin-guide/cgroup-v1/cpusets.rst] and any optional flags,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) listed below. If the resulting NodeLists is the empty set, the effective
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) memory policy for the file will revert to "default" policy.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) NUMA memory allocation policies have optional flags that can be used in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) conjunction with their modes. These optional flags can be specified
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) when tmpfs is mounted by appending them to the mode before the NodeList.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) See Documentation/admin-guide/mm/numa_memory_policy.rst for a list of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) all available memory allocation policy mode flags and their effect on
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) memory policy.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) ::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) =static is equivalent to MPOL_F_STATIC_NODES
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) =relative is equivalent to MPOL_F_RELATIVE_NODES
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) For example, mpol=bind=static:NodeList, is the equivalent of an
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) allocation policy of MPOL_BIND | MPOL_F_STATIC_NODES.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) Note that trying to mount a tmpfs with an mpol option will fail if the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) running kernel does not support NUMA; and will fail if its nodelist
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) specifies a node which is not online. If your system relies on that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) tmpfs being mounted, but from time to time runs a kernel built without
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) NUMA capability (perhaps a safe recovery kernel), or with fewer nodes
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) online, then it is advisable to omit the mpol option from automatic
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) mount options. It can be added later, when the tmpfs is already mounted
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) on MountPoint, by 'mount -o remount,mpol=Policy:NodeList MountPoint'.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) To specify the initial root directory you can use the following mount
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) options:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) ==== ==================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) mode The permissions as an octal number
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) uid The user id
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) gid The group id
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) ==== ==================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) These options do not have any effect on remount. You can change these
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) parameters with chmod(1), chown(1) and chgrp(1) on a mounted filesystem.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) tmpfs has a mount option to select whether it will wrap at 32- or 64-bit inode
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) numbers:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) ======= ========================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) inode64 Use 64-bit inode numbers
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) inode32 Use 32-bit inode numbers
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) ======= ========================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) On a 32-bit kernel, inode32 is implicit, and inode64 is refused at mount time.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) On a 64-bit kernel, CONFIG_TMPFS_INODE64 sets the default. inode64 avoids the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) possibility of multiple files with the same inode number on a single device;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) but risks glibc failing with EOVERFLOW once 33-bit inode numbers are reached -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) if a long-lived tmpfs is accessed by 32-bit applications so ancient that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) opening a file larger than 2GiB fails with EINVAL.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) So 'mount -t tmpfs -o size=10G,nr_inodes=10k,mode=700 tmpfs /mytmpfs'
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) will give you tmpfs instance on /mytmpfs which can allocate 10GB
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) RAM/SWAP in 10240 inodes and it is only accessible by root.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) :Author:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) Christoph Rohland <cr@sap.com>, 1.12.01
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) :Updated:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) Hugh Dickins, 4 June 2007
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) :Updated:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) KOSAKI Motohiro, 16 Mar 2010
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) :Updated:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) Chris Down, 13 July 2020
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags