aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorDavid Oberhollenzer <david.oberhollenzer@sigma-star.at>2023-09-03 14:41:26 +0200
committerDavid Oberhollenzer <david.oberhollenzer@sigma-star.at>2023-09-03 14:41:26 +0200
commitfb7b905de6b9154b3cc46a4ae39a40f9385749f2 (patch)
tree99038ae2046265c351732b12d50349de2c3e3d16
parentb7a102881f963baa2e18381e37a2769ccdbf4cf2 (diff)
Add a README file
Signed-off-by: David Oberhollenzer <david.oberhollenzer@sigma-star.at>
-rw-r--r--Makefile.am2
-rw-r--r--README.txt143
2 files changed, 144 insertions, 1 deletions
diff --git a/Makefile.am b/Makefile.am
index 43bac49..dd14d96 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -41,7 +41,7 @@ test_PROGRAMS =
test_SCRIPTS =
TESTS =
-EXTRA_DIST = COPYING CHANGELOG.md
+EXTRA_DIST = COPYING CHANGELOG.md README.txt
include lib/Makemodule.am
include ubi-utils/Makemodule.am
diff --git a/README.txt b/README.txt
new file mode 100644
index 0000000..6ee0136
--- /dev/null
+++ b/README.txt
@@ -0,0 +1,143 @@
+
+ mtd-utils
+ =========
+
+ This package provides userspace utilities for working with Linux MTD and UBI
+ infrastructure, as well as related filesystems (JFFS2, UBIFS).
+
+ MTD stands for "Memory Technology Devices" and means flash, RAM and similar
+ chips, used for solid state storage on embedded devices. Think of e.g. a raw
+ nand-flash chip attached directly to an SoC via parallel or SPI bus, but
+ without _any_ special management like an eMMC has. The MTD subsystem exposes
+ those devices to userspace as character devices. It provides a unified
+ interface, but does not attempt to hide the page vs erase-block device
+ geometry or issues involved in dealing with the underlying storage, like wear
+ or bad blocks.
+
+ For this purpose, UBI provides a layer on top of MTD that hides away some
+ of those complexities. UBI also implements LVM-like logical volume partitioning
+ and some other useful capabilities.
+
+ UBI does not emulate block device like storage. UBI volumes are exposed as
+ character devices with page vs erase-block access geometry, much like MTD
+ partitions. But they behave more like a flash chip with "idealized" properties
+ and transparent wear-leveling.
+
+ JFFS2 is a flash friendly filesystem. It predates UBI and is designed to run
+ on top of raw MTD, implementing its own wear leveling.
+
+ The more recent UBIFS (formerly JFFS3) can be used on top of UBI volumes and
+ offers a number of improvements over JFFS2, offers newer compression schemes
+ and advanced features like file-level encryption and authentication.
+
+
+ Installing
+ ----------
+
+ mtd-utils are packaged by a number of Linux distributions[0] and can be
+ installed via the package management system. A Yocto recipe is available
+ as well[1] as part of the openembedded-core layer.
+
+ Should the tools be too big for on-device use, the BusyBox project provides
+ compatible clone versions.
+
+ Before making this choice however, please note that some "heavy weight" tools
+ like mkfs.ubifs are typically not needed on the device itself. The kernel
+ module can format a volume on first mount. Other, heavy weight tools like lsmtd
+ can be disabled via a configure flag.
+
+ [0] https://repology.org/project/mtd-utils/versions
+ [1] https://layers.openembedded.org/layerindex/recipe/262/
+
+
+ How to get the source
+ ---------------------
+
+ The official git repository and release tarballs are hosted on infradead.org:
+
+ Relase tarballs: ftp://ftp.infradead.org/pub/mtd-utils
+ git repository: git://git.infradead.org/mtd-utils.git
+
+ This is also where the website and documentation are hosted:
+
+ http://www.linux-mtd.infradead.org/
+
+ A http mirror is also provided on infraroot.at:
+
+ Relase tarballs: https://infraroot.at/pub/mtd
+ git repository: https://git.infraroot.at/mtd-utils.git
+
+
+ There are a number of mirrored repositories on GitHub, but this is not where
+ official development takes place. If you want to contribute upstream, please
+ do not open any issues or pull requests there. Upstream development is done
+ via mailing list (see below).
+
+
+ How to build
+ ------------
+
+ mtd-utils uses an autotools based build system. If you are building the git
+ tree, you first need to run `./autogen.sh`. This sets up the build system
+ and requires autoconf, automake and possibly other autotools to be installed.
+
+ After unpacking a tarball (or running autogen.sh) simply run:
+
+ $ ./configure
+ $ make
+
+ to build the package. This does not require autotools to be installed, but
+ does need pkg-config, make, etc...
+
+ There are a number of configure switches to tune the build.
+ Run `./configure --help` to get an overview.
+
+
+ For compiling mtd-utils, you need development packages for the following
+ dependency libraries:
+
+ zlib (required by mkfs.ubifs, mkfs.jffs2
+ lzo2 (optional dependency for mkfs.ubifs, mkfs.jffs2)
+ zstd (optional dependency for mkfs.ubifs)
+ libuuid* (required by mkfs.ubifs)
+ libselinux (optional dependency for mkfs.ubifs)
+ openssl (optional dependency for mkfs.ubifs)
+
+ * this library is part of util-linux, aka util-linux-ng and some distributions
+ package it under that name.
+
+ If you don't explicitly disable xattr support, mkfs.ubifs and mkfs.jffs2
+ require the "sys/xattr.h" and "sys/acl.h" header files, depending on your
+ Distributions, those may be packaged as part of libattr and libacl
+ respectively.
+
+
+ Please note that the mkfs tools are optional and can be disabled via a
+ configure flag. This should also remove the need for the dependency
+ libraries.
+
+
+ How to contribute
+ -----------------
+
+ Development of mtd-utils takes place on the linux-mtd mailing list:
+
+ mailto: linux-mtd@lists.infradead.org
+ archive: https://lists.infradead.org/pipermail/linux-mtd/
+
+ You can subscribe here: http://lists.infradead.org/mailman/listinfo/linux-mtd/
+ or simply send "subscribe" to linux-mtd-request@lists.infradead.org
+
+
+ Contributions are submitted in the form of plain text patches,
+ using `git send-email` or similar.
+
+ Please prefix your subject with "mtd-utils: " to make them easier to spot.
+
+ It may take a few days for a patch to be picked up. This is in part done
+ intentionally to allow other people to comment on it. If it's been more
+ than a week and you feel your patch might have been overlooked, please send
+ a friendly ping, or re-submit the patch series with "[RESEND]" in the subject.
+
+
+ There is also a #mtd IRC channel on irc.oftc.net