Split many utils

Author: cirosantilli

README.md

@@ -39,6 +39,7 @@ Includes Linux concepts and utilities that work on Linux, not necessarily in the
     1.  [File permissions](file-permissions.md)
     1.  [Hardlink](hardlink.md)
     1.  [badblocks](badblocks.md)
+    1.  [bindfs](bindfs.md)
     1.  [blkid](blkid.md)
     1.  [chown](chown.sh)
     1.  [chmod](chmod.sh)
@@ -49,9 +50,11 @@ Includes Linux concepts and utilities that work on Linux, not necessarily in the
         1. [e2fsprogs](e2progs.md)
         1. [mke2fs](mke2fs.md)
     1.  [fdisk](fdisk.md)
+    1.  [FUSE](FUSE.md)
     1.  [fuser](fuser.md)
     1.  [lsblk](lsblk.md)
     1.  [mktemp](mktemp.md)
+    1.  [mountall](mountall.md)
     1.  [pathchk](pathchk.md)
     1.  [ramfs](ramfs.md)
     1.  [swap partition](swap-partition.md)
@@ -115,6 +118,7 @@ Includes Linux concepts and utilities that work on Linux, not necessarily in the
 1.  Programming
     1. [ack](ack.sh)
     1. [ctags](ctags/)
+    1. [cscope](cscope.md)
 1.  Processes
     1. [chroot](chroot.sh)
     1. [env](env.sh)
@@ -130,8 +134,10 @@ Includes Linux concepts and utilities that work on Linux, not necessarily in the
     1. [pstree](pstree.sh)
     1. [pwd](pwd.sh)
     1. [sleep](sleep.sh)
+    1. [strace](strace.md)
     1. [timeout](timeout.md)
     1. [top](top.md)
+    1. [umask](umask.md)
     1. [ulimit](ulimit.md)
     1. [wait](wait.sh)
 1.  Date and time
@@ -161,6 +167,7 @@ Includes Linux concepts and utilities that work on Linux, not necessarily in the
     1. [gzip](gzip.md)
     1. [tar](tar.md)
     1. [zip](zip.md)
+    1. [ar](ar.md)
 1.  User operations
     1. [id](id.md)
     1. [usermod](usermod.md)
@@ -187,6 +194,7 @@ Includes Linux concepts and utilities that work on Linux, not necessarily in the
     1. [RST](rst/)
 1.  [Virtual machine](virtual-machine/)
     1. [Docker](docker/)
+    1. [QEMU](qemu.md)
     1. [User mode Linux](user-mode-linux.md)
     1. [Vagrant](vagrang/)
 1.  [Configuration automation](config-automation.md)
@@ -207,9 +215,12 @@ Includes Linux concepts and utilities that work on Linux, not necessarily in the
     1.  Sorting
         1. [sort](sort.md)
         1. [tsort](tsort.md)
+1.  Packages
+    1. [util-linux](util-linux.md)
 1.  Misc
     1. [cron](cron.sh)
     1. [logrotate](logrotate.md)
+    1. [BusyBox](busybox.md)
     1. [xargs](xargs.md)
 1.  [Bibliography](bibliography.md)
 

ar.md

@@ -0,0 +1,13 @@
+# ar
+
+<https://en.wikipedia.org/wiki/Ar_%28Unix%29>
+
+Yet another archive format and CLI tool.
+
+Used by `.deb` files.
+
+Similar command line options to tar.
+
+Extract:
+
+    ar xf a.ar

bindfs.md

@@ -0,0 +1,21 @@
+# bindfs
+
+<https://github.com/mpartel/bindfs>
+
+FUSER filesystem similar to `mount --bind`, but allows you to mess with ownership and permissions.
+
+Useful command:
+
+    sudo bindfs -u a -g a --create-for-user=b --create-for-group=b from to
+
+Mounts `from/` on `to/`.
+
+`b` must exist.
+
+`-u a -g a`: everything seen on `to` is owned by `a:a`
+
+Everything created / saved on `b`, is created on `a` with owner `b:b`.
+
+Unmount:
+
+    fusermount -u /home/johnc/ISO-images

blkid.md

@@ -2,7 +2,9 @@
 
 `util-linux` package.
 
-Get UUID, label and filesystem type for all partitions
+There is also a `liblkid` package in Ubuntu which offers functions like disk detection.
+
+Get UUID, label and filesystem type for all partitions:
 
     sudo blkid
 

busybox.md

@@ -0,0 +1,30 @@
+# BusyBox
+
+Single executable that includes all POSIX command line utilities, including `sh`, and no libc dependency.
+
+Popular for minimal distributions, e.g. <https://github.com/ivandavidov/minimal>
+
+Source:
+
+    git clone git://git.busybox.net/busybox
+
+Once compiled, this generates a single `busybox` 2Mb executable, possibly with not external dependencies, often built with uClibc. E.g., if we do:
+
+    ldd busybox
+
+we get:
+
+
+
+It is also possible to configure exactly which utilities will be present on the output.
+
+You can use it either with subcommands:
+
+    ./busybox echo a
+
+or with symlinks, which is the standard approach when deploying it:
+
+    ln -s busybox echo
+    ./echo a
+
+`busybox` uses `argv[0]` to decide the executable in this case.

cpio.md

@@ -0,0 +1,18 @@
+# cpio
+
+GNU: <http://www.gnu.org/software/cpio/manual/cpio.html>. Ubuntu `cpio` package.
+
+Was POSIX.1-1988, but it was omitted from POSIX.1-2001. The POSIX 2008 `pax` utility [defines](http://pubs.opengroup.org/onlinepubs/9699919799/utilities/pax.html#tag_20_92_13_07) and implements the cpio format however. The file format is very simple.
+
+Copy files between archives. Also the name of a specialized archive format of the program.
+
+Generate a `.cpio` archive containing files `a` and `b`:
+
+    printf '0' > /tmp/a
+    printf '1' > /tmp/b
+    printf '/tmp/a\n/tmp/b\n' | cpio -o > /tmp/ab.cpio
+    hd /tmp/ab.cpio
+
+Find selected files to destination, building and keeping their relative directory structure:
+
+    find . ! -iname '* - *' -type f -print | cpio -pvdumB './no author'

cscope.md

@@ -0,0 +1,94 @@
+# Cscope
+
+Cscope <http://cscope.sourceforge.net/>.
+
+C-only, but more powerful, e.g. can list references.
+
+## Basic usage
+
+    cd project/root
+    cscope -R
+
+This generates a `cscope.out` file.
+
+For the Linux kernel 4.0, it took 2 minutes on a 2013 machine, and generated a 350M file, and leaves you in the ncurses interface.
+
+The `cscope.out` file is binary and not directly readable on text editors.
+
+Next time you want to open the ncurses interface, you can use either:
+
+    cscope -R
+
+which will update the database where needed and reuse existing `cscope.out`, or:
+
+    cscope -d
+
+which will skip the database update.
+
+## -b
+
+Only build the database, don't enter the ncurses.
+
+## Ignore struct declarations and see only definitions
+
+Impossible: <http://stackoverflow.com/questions/1175610>
+
+## Find struct assignments while ignoring local variables
+
+Impossible: <http://stackoverflow.com/questions/6190955/how-to-find-struct-member-uses-with-cscope>
+
+## ncurses interface
+
+You are now left on a ncurses interface with the following options:
+
+    Find this C symbol: <cursor>
+    Find this global definition:
+    Find functions called by this function:
+    Find functions calling this function:
+    Find this text string:
+    Change this text string:
+    Find this egrep pattern:
+    Find this file:
+    Find files #including this file:
+    Find assignments to this symbol:
+
+Use arrows to navigate up and down.
+
+### Panes
+
+There are two panes:
+
+- input pane, at the bottom
+- output pane at the top
+
+### ?
+
+Help
+
+### Ctrl+D
+
+Quit.
+
+### tab
+
+Switch between input and output panes.
+
+Suppose we want to find a definition. Move to `Find this global definition` and type `task_struct`.
+
+### Enter
+
+On input pane, make a query.
+
+On the output pane, open the given file on your selected terminal editor.
+
+### Space
+
+On the output pane, go one page down.
+
+### ^
+
+Filter output through shell command. E.g.:
+
+    ^grep sched
+
+Also 

ctags/.gitignore

@@ -0,0 +1,2 @@
+cscope.out
+tags

ctags/Makefile

@@ -1,5 +1,25 @@
-.PHONY: all
+.POSIX:
 
-all:
-	ctags -R -o -
-	#ctags --c-kinds=+l -R -o -
+IN_EXT ?= .c
+OUT_EXT ?= .out
+RUN ?= main
+
+INS := $(wildcard *$(IN_EXT))
+OUTS_NOEXT := $(basename $(INS))
+OUTS := $(addsuffix $(OUT_EXT), $(OUTS_NOEXT))
+
+.PHONY: all clean run
+
+all: $(OUTS)
+	-ctags -R
+	#ctags --c-kinds=+l -R
+	-cscope -Rb
+
+%$(OUT_EXT): %$(IN_EXT)
+	gcc -o '$@' -pedantic-errors -std=c89 -Wall '$<'
+
+clean:
+	rm -f *$(OUT_EXT) cscope.out tags
+
+run: all
+	./'$(RUN)$(OUT_EXT)'

ctags/README.md

@@ -2,29 +2,21 @@
 
 POSIX 7, with many GNU extensions.
 
-Reads source files and generates a list of objects
-that can be later searched.
+Reads source files and generates a list of objects that can be later searched.
 
-Objects are things like C symbols: variable names,
-function names, structure names, macros.
+Objects are things like C symbols: variable names, function names, structure names, macros.
 
-Therefore the output of this program is useful if you want to find
-where something is defined in a large source code, such as the Linux kernel for example.
+Therefore the output of this program is useful if you want to find where something is defined in a large source code, such as the Linux kernel for example.
 
 The output is both human and computer readable, therefore to use it you can either:
 
 -   look at the file yourself and search.
 
-    Useful when there are many possible definition places,
-    and you might be able to guess the correct one by looking
-    at the extra fields of the file.
+    Useful when there are many possible definition places, and you might be able to guess the correct one by looking at the extra fields of the file.
 
--   use an editor such as VIM which has built-in functionalities
-    that allows you to move cursor on top of the symbol, click something (`<C-]>`),
-    and magically jump to the definition
+-   use an editor such as VIM which has built-in functionalities that allows you to move cursor on top of the symbol, click something (`<C-]>`), and magically jump to the definition
 
-    There could however be many candidate objects with the same name
-    so you might have to loop a large list of tags.
+    There could however be many candidate objects with the same name so you might have to loop a large list of tags.
 
     If that is the case, it might be easier to read the `tags` file directly.
 
@@ -46,18 +38,15 @@ Makes a single `tags` in current directory recursing into all child directories.
 
 ## Supported languages
 
-POSIX requires only a that small set of C and FORTRAN symbols be processed.
-GNU extends it considerably, and adds support for many languages.
-The language list on the GNU implementation can be found with:
+POSIX requires only a that small set of C and FORTRAN symbols be processed. GNU extends it considerably, and adds support for many languages. The language list on the GNU implementation can be found with:
 
     ctags --list-languages
 
 For me, this includes languages such as `C++`, `python`, `java` and many more.
 
 ## Objects of each language
 
-To view what kinds of objects can be processed, which are processed by default,
-and the one letter codes for each object type use:
+To view what kinds of objects can be processed, which are processed by default, and the one letter codes for each object type use:
 
     ctags --list-kinds=<lang>
 
@@ -98,12 +87,9 @@ Take defaults, add `c`, `d` and `f`, but remove `e`:
 
     ctags -R --c-kinds=+cd-e+f
 
-One field which is tempting to remove for C is the `m`, which gives lots of id dupes
-and is not very useful.
+One field which is tempting to remove for C is the `m`, which gives lots of id dupes and is not very useful.
 
-For example, you might want to find the definition of `x`,
-and there are twenty structs which contain a member
-named `x` besides the definition which you want.
+For example, you might want to find the definition of `x`, and there are twenty structs which contain a member named `x` besides the definition which you want.
 
 This happens often on the Linux kernel code for example.
 
@@ -129,8 +115,7 @@ Can be modified with `-h`.
 
     May be either a regex (`-N`, default) or a line number (`-n`).
 
-    Advantage of regex: changes elsewhere in the file are unlikely to change the regex,
-    but can easily change the exact line number.
+    Advantage of regex: changes elsewhere in the file are unlikely to change the regex, but can easily change the exact line number.
 
     Advantage of line numbers: file is smaller.
 
@@ -142,9 +127,17 @@ Can be modified with `-h`.
 
     Can be controlled by the `--fields` option.
 
+## Find references
+
+Not possible only definitions: <http://ctags.sourceforge.net/faq.html#12>
+
+As any Eclipse user will know, this capability is fundamentally important to understand code.
+
+One of the alternative tools must be used instead.
+
 ## Alternatives
 
-- Cscope
+- ID Utils <http://www.gnu.org/software/idutils/idutils.html>
 
 ### exuberant-ctags
 

ctags/a.h

@@ -1,2 +1,4 @@
-//headers are included by default
+/* Headers are included by default. */
 int a1;
+struct s;
+void f();