1# OpenBMC Flash Layout and Filesystem Documentation
2This file is focused on providing information about the flash setup that the
3code update application requires to be supported out-of-the-box, which includes
4how the Linux filesystem is setup, filesystem layouts, overlays, boot options,
5and how code update updates the flash modules and boots the new image. See
6[code-update.md](code-update.md) for details about code update interfaces.
7
8## Design considerations
9### Boot loading and init
10For system initialization and bootstrap, [Das U-Boot][] was selected as the
11bootloader.
12
13After basic initialization of the system, the bootloader may present a prompt
14and/or start automatic boot. The commands and/or data to select the boot image
15are stored in the bootloader environment. The bootloader copies the compressed
16kernel, initrd image, and device tree into memory, and then transfers control to
17the kernel. The kernel initializes itself and the system using the information
18passed in the device tree, including the flash partitions and the kernel command
19line embedded in the tree.
20
21### Runtime management
22For runtime management, the [systemd][] system and service manager was chosen
23for its configuration, dependency, and triggered action support, as well as its
24robust recovery.
25
26Before starting execution, systemd requires the root filesystem and all binaries
27to be mounted. The filesystems for /dev, /sys, and /proc may be mounted before
28starting systemd. Reference the [systemd File Hierarchy Requirements][].
29
30### Root filesystem
31For storage of the root filesystem, a read-only volume was selected. This allows
32the majority of the filesystem content, including all executables and static
33data files, to be stored in a read-only filesystem image. Replacing read-only
34filesystem images allows the space used by the content to be confirmed at build
35time and allows the selection of compressed filesystems that do not support
36mutations.
37
38An effort has been made to adhere to the Filesystem Hierarchy Standard [FHS][].
39Specifically data ephemeral to the current boot is stored in /run and most
40application data is stored under /var. Some information continues to be stored
41in the system configuration data directory /etc; this is mostly traditionally
42configuration such as network addresses, user identification, and ssh host keys.
43
44To conserve flash space, squashfs with xz compression was selected to store the
45read-only filesystem content. This applies to systems with limited attached
46flash storage (see the JFFS2 and UBI options below), not eMMC.
47
48To load the root filesystem, the initramfs locates and mounts the squashfs and
49writable filesystems, merges them with overlayfs, performs a chroot into the
50result and starts systemd. Alternatively, information to find the active image
51for the BMC can be stored in the U-Boot environment, and an init script can
52mount the images then start systemd. This choice depends on the platform
53implementation, and details are located in the Supported Filesystem Choices
54section below.
55
56## Supported Filesystem Choices
57OpenBMC supports code update for the following types of filesystems. Additional
58information is given for each filesystem such as how the filesystem is stored on
59flash, how the filesystem is instantiated during system boot/init, and how code
60update handles the filesystem.
61
62### Writable Filesystem Options
63#### JFFS2 on MTD partition
64
65The majority of the filesystem is stored in a read-only squashfs in an MTD
66partition using the block emulation driver (mtdblock). A second MTD partition is
67mounted read-write using the JFFS2 filesystem. This read-write filesystem is
68mounted over the entire filesystem space allowing all files and directories to
69be written.
70
71This filesystem stack requires the mounts to be performed from an initramfs. The
72initramfs is composed of a basic system based on busybox and three custom
73scripts (init, shutdown, and update) that locate the MTD partitions by name.
74These scripts are installed by [obmc-phosphor-initfs][].
75
76In code update mode, the squashfs image and white-listed files from the
77read-write filesystem are copied into RAM by the initramfs and used to assemble
78the root overlayfs instance, leaving the flash free to be modified by installing
79images at runtime. An orderly shutdown writes remaining images to like-named raw
80MTD partitions and white listed files to the writable overlay filesystem.
81Alternatively, if code update mode was not selected, the image updates must be
82delayed until the partitions are unmounted during an orderly shutdown.
83
84This is the default filesystem in OpenBMC. It is used in several BMC systems
85based around the AST2400 and AST2500 system-on-chip controllers from Aspeed
86Technology. These SOCs support 1 and 2 GB of DDR RAM, while the attached flash
87storage is typically in the 10s of MB, so staging the filesystem to RAM is not
88an issue.
89
90#### UBI on MTD partition
91
92The majority of the filesystem is stored in a read-only squashfs in a static UBI
93volume using the UBI block emulation driver (ubiblock). To store updates to
94files, a UBIFS volume is used for /var and mounted over the /etc and /home
95directories using overlayfs. These mounts are performed by the `init` script
96installed by the [preinit-mounts][] package before `systemd` is started.
97Selecting UBI allows the writes to the read-write overlay to be distributed over
98the full UBI area instead of just the read-write MTD partition.
99
100The environment for Das U-boot continues to be stored at fixed sectors in the
101flash. The Das U-boot environment contains enough MTD partition definition to
102read UBI volumes in a UBI device in the same flash. The bootcmd script loads a
103kernel from a FIT image and pass it to bootargs to locate and mount the squashfs
104in the paired UBI volume.
105
106This option is enabled via the `obmc-ubi-fs` OpenBMC distro feature. Used in the
107same BMC subsystems as the JFFS2 ones, but targeted for configurations that have
108enough flash storage to store at least 2 copies of the filesystem. This can be
109accomplished with dual flash storage. Some controllers, such as those in the
110AST2500, allow booting from an alternate flash on failure and this UBI option
111supports this feature. For this support, a copy of each kernel is stored on each
112flash and the U-Boot environment selects which kernel to use.
113
114#### ext4 on eMMC
115
116This is a work in progress. See the [eMMC Design Document][].
117
118### Auxiliary Filesystems
119A tmpfs is used for /tmp, /run, and similar, while /dev, /proc, and
120/sys are supported by their normal kernel special filesystems, as specified by
121the FHS.
122
123## Other
124Additional Bitbake layer configurations exist for Raspberry Pi and x86 QEMU
125machines, but are provided primarily for code development and exploration. Code
126update for these environments is not supported.
127
128[Das U-Boot]: https://www.denx.de/wiki/U-Boot
129[systemd]: https://github.com/openbmc/docs/blob/master/architecture/openbmc-systemd.md
130[systemd File Hierarchy Requirements]: https://www.freedesktop.org/wiki/Software/systemd/FileHierarchy/
131[FHS]: https://refspecs.linuxfoundation.org/fhs.shtml
132[obmc-phosphor-initfs]: https://github.com/openbmc/openbmc/blob/master/meta-phosphor/recipes-phosphor/initrdscripts/obmc-phosphor-initfs.bb
133[preinit-mounts]: https://github.com/openbmc/openbmc/tree/master/meta-phosphor/recipes-phosphor/preinit-mounts
134[eMMC Design Document]: https://github.com/openbmc/docs/blob/master/architecture/code-update/emmc-storage-design.md
135