1*4882a593SmuzhiyunKernel Support for miscellaneous Binary Formats (binfmt_misc) 2*4882a593Smuzhiyun============================================================= 3*4882a593Smuzhiyun 4*4882a593SmuzhiyunThis Kernel feature allows you to invoke almost (for restrictions see below) 5*4882a593Smuzhiyunevery program by simply typing its name in the shell. 6*4882a593SmuzhiyunThis includes for example compiled Java(TM), Python or Emacs programs. 7*4882a593Smuzhiyun 8*4882a593SmuzhiyunTo achieve this you must tell binfmt_misc which interpreter has to be invoked 9*4882a593Smuzhiyunwith which binary. Binfmt_misc recognises the binary-type by matching some bytes 10*4882a593Smuzhiyunat the beginning of the file with a magic byte sequence (masking out specified 11*4882a593Smuzhiyunbits) you have supplied. Binfmt_misc can also recognise a filename extension 12*4882a593Smuzhiyunaka ``.com`` or ``.exe``. 13*4882a593Smuzhiyun 14*4882a593SmuzhiyunFirst you must mount binfmt_misc:: 15*4882a593Smuzhiyun 16*4882a593Smuzhiyun mount binfmt_misc -t binfmt_misc /proc/sys/fs/binfmt_misc 17*4882a593Smuzhiyun 18*4882a593SmuzhiyunTo actually register a new binary type, you have to set up a string looking like 19*4882a593Smuzhiyun``:name:type:offset:magic:mask:interpreter:flags`` (where you can choose the 20*4882a593Smuzhiyun``:`` upon your needs) and echo it to ``/proc/sys/fs/binfmt_misc/register``. 21*4882a593Smuzhiyun 22*4882a593SmuzhiyunHere is what the fields mean: 23*4882a593Smuzhiyun 24*4882a593Smuzhiyun- ``name`` 25*4882a593Smuzhiyun is an identifier string. A new /proc file will be created with this 26*4882a593Smuzhiyun ``name below /proc/sys/fs/binfmt_misc``; cannot contain slashes ``/`` for 27*4882a593Smuzhiyun obvious reasons. 28*4882a593Smuzhiyun- ``type`` 29*4882a593Smuzhiyun is the type of recognition. Give ``M`` for magic and ``E`` for extension. 30*4882a593Smuzhiyun- ``offset`` 31*4882a593Smuzhiyun is the offset of the magic/mask in the file, counted in bytes. This 32*4882a593Smuzhiyun defaults to 0 if you omit it (i.e. you write ``:name:type::magic...``). 33*4882a593Smuzhiyun Ignored when using filename extension matching. 34*4882a593Smuzhiyun- ``magic`` 35*4882a593Smuzhiyun is the byte sequence binfmt_misc is matching for. The magic string 36*4882a593Smuzhiyun may contain hex-encoded characters like ``\x0a`` or ``\xA4``. Note that you 37*4882a593Smuzhiyun must escape any NUL bytes; parsing halts at the first one. In a shell 38*4882a593Smuzhiyun environment you might have to write ``\\x0a`` to prevent the shell from 39*4882a593Smuzhiyun eating your ``\``. 40*4882a593Smuzhiyun If you chose filename extension matching, this is the extension to be 41*4882a593Smuzhiyun recognised (without the ``.``, the ``\x0a`` specials are not allowed). 42*4882a593Smuzhiyun Extension matching is case sensitive, and slashes ``/`` are not allowed! 43*4882a593Smuzhiyun- ``mask`` 44*4882a593Smuzhiyun is an (optional, defaults to all 0xff) mask. You can mask out some 45*4882a593Smuzhiyun bits from matching by supplying a string like magic and as long as magic. 46*4882a593Smuzhiyun The mask is anded with the byte sequence of the file. Note that you must 47*4882a593Smuzhiyun escape any NUL bytes; parsing halts at the first one. Ignored when using 48*4882a593Smuzhiyun filename extension matching. 49*4882a593Smuzhiyun- ``interpreter`` 50*4882a593Smuzhiyun is the program that should be invoked with the binary as first 51*4882a593Smuzhiyun argument (specify the full path) 52*4882a593Smuzhiyun- ``flags`` 53*4882a593Smuzhiyun is an optional field that controls several aspects of the invocation 54*4882a593Smuzhiyun of the interpreter. It is a string of capital letters, each controls a 55*4882a593Smuzhiyun certain aspect. The following flags are supported: 56*4882a593Smuzhiyun 57*4882a593Smuzhiyun ``P`` - preserve-argv[0] 58*4882a593Smuzhiyun Legacy behavior of binfmt_misc is to overwrite 59*4882a593Smuzhiyun the original argv[0] with the full path to the binary. When this 60*4882a593Smuzhiyun flag is included, binfmt_misc will add an argument to the argument 61*4882a593Smuzhiyun vector for this purpose, thus preserving the original ``argv[0]``. 62*4882a593Smuzhiyun e.g. If your interp is set to ``/bin/foo`` and you run ``blah`` 63*4882a593Smuzhiyun (which is in ``/usr/local/bin``), then the kernel will execute 64*4882a593Smuzhiyun ``/bin/foo`` with ``argv[]`` set to ``["/bin/foo", "/usr/local/bin/blah", "blah"]``. The interp has to be aware of this so it can 65*4882a593Smuzhiyun execute ``/usr/local/bin/blah`` 66*4882a593Smuzhiyun with ``argv[]`` set to ``["blah"]``. 67*4882a593Smuzhiyun ``O`` - open-binary 68*4882a593Smuzhiyun Legacy behavior of binfmt_misc is to pass the full path 69*4882a593Smuzhiyun of the binary to the interpreter as an argument. When this flag is 70*4882a593Smuzhiyun included, binfmt_misc will open the file for reading and pass its 71*4882a593Smuzhiyun descriptor as an argument, instead of the full path, thus allowing 72*4882a593Smuzhiyun the interpreter to execute non-readable binaries. This feature 73*4882a593Smuzhiyun should be used with care - the interpreter has to be trusted not to 74*4882a593Smuzhiyun emit the contents of the non-readable binary. 75*4882a593Smuzhiyun ``C`` - credentials 76*4882a593Smuzhiyun Currently, the behavior of binfmt_misc is to calculate 77*4882a593Smuzhiyun the credentials and security token of the new process according to 78*4882a593Smuzhiyun the interpreter. When this flag is included, these attributes are 79*4882a593Smuzhiyun calculated according to the binary. It also implies the ``O`` flag. 80*4882a593Smuzhiyun This feature should be used with care as the interpreter 81*4882a593Smuzhiyun will run with root permissions when a setuid binary owned by root 82*4882a593Smuzhiyun is run with binfmt_misc. 83*4882a593Smuzhiyun ``F`` - fix binary 84*4882a593Smuzhiyun The usual behaviour of binfmt_misc is to spawn the 85*4882a593Smuzhiyun binary lazily when the misc format file is invoked. However, 86*4882a593Smuzhiyun this doesn``t work very well in the face of mount namespaces and 87*4882a593Smuzhiyun changeroots, so the ``F`` mode opens the binary as soon as the 88*4882a593Smuzhiyun emulation is installed and uses the opened image to spawn the 89*4882a593Smuzhiyun emulator, meaning it is always available once installed, 90*4882a593Smuzhiyun regardless of how the environment changes. 91*4882a593Smuzhiyun 92*4882a593Smuzhiyun 93*4882a593SmuzhiyunThere are some restrictions: 94*4882a593Smuzhiyun 95*4882a593Smuzhiyun - the whole register string may not exceed 1920 characters 96*4882a593Smuzhiyun - the magic must reside in the first 128 bytes of the file, i.e. 97*4882a593Smuzhiyun offset+size(magic) has to be less than 128 98*4882a593Smuzhiyun - the interpreter string may not exceed 127 characters 99*4882a593Smuzhiyun 100*4882a593SmuzhiyunTo use binfmt_misc you have to mount it first. You can mount it with 101*4882a593Smuzhiyun``mount -t binfmt_misc none /proc/sys/fs/binfmt_misc`` command, or you can add 102*4882a593Smuzhiyuna line ``none /proc/sys/fs/binfmt_misc binfmt_misc defaults 0 0`` to your 103*4882a593Smuzhiyun``/etc/fstab`` so it auto mounts on boot. 104*4882a593Smuzhiyun 105*4882a593SmuzhiyunYou may want to add the binary formats in one of your ``/etc/rc`` scripts during 106*4882a593Smuzhiyunboot-up. Read the manual of your init program to figure out how to do this 107*4882a593Smuzhiyunright. 108*4882a593Smuzhiyun 109*4882a593SmuzhiyunThink about the order of adding entries! Later added entries are matched first! 110*4882a593Smuzhiyun 111*4882a593Smuzhiyun 112*4882a593SmuzhiyunA few examples (assumed you are in ``/proc/sys/fs/binfmt_misc``): 113*4882a593Smuzhiyun 114*4882a593Smuzhiyun- enable support for em86 (like binfmt_em86, for Alpha AXP only):: 115*4882a593Smuzhiyun 116*4882a593Smuzhiyun echo ':i386:M::\x7fELF\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x02\x00\x03:\xff\xff\xff\xff\xff\xfe\xfe\xff\xff\xff\xff\xff\xff\xff\xff\xff\xfb\xff\xff:/bin/em86:' > register 117*4882a593Smuzhiyun echo ':i486:M::\x7fELF\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x02\x00\x06:\xff\xff\xff\xff\xff\xfe\xfe\xff\xff\xff\xff\xff\xff\xff\xff\xff\xfb\xff\xff:/bin/em86:' > register 118*4882a593Smuzhiyun 119*4882a593Smuzhiyun- enable support for packed DOS applications (pre-configured dosemu hdimages):: 120*4882a593Smuzhiyun 121*4882a593Smuzhiyun echo ':DEXE:M::\x0eDEX::/usr/bin/dosexec:' > register 122*4882a593Smuzhiyun 123*4882a593Smuzhiyun- enable support for Windows executables using wine:: 124*4882a593Smuzhiyun 125*4882a593Smuzhiyun echo ':DOSWin:M::MZ::/usr/local/bin/wine:' > register 126*4882a593Smuzhiyun 127*4882a593SmuzhiyunFor java support see Documentation/admin-guide/java.rst 128*4882a593Smuzhiyun 129*4882a593Smuzhiyun 130*4882a593SmuzhiyunYou can enable/disable binfmt_misc or one binary type by echoing 0 (to disable) 131*4882a593Smuzhiyunor 1 (to enable) to ``/proc/sys/fs/binfmt_misc/status`` or 132*4882a593Smuzhiyun``/proc/.../the_name``. 133*4882a593SmuzhiyunCatting the file tells you the current status of ``binfmt_misc/the_entry``. 134*4882a593Smuzhiyun 135*4882a593SmuzhiyunYou can remove one entry or all entries by echoing -1 to ``/proc/.../the_name`` 136*4882a593Smuzhiyunor ``/proc/sys/fs/binfmt_misc/status``. 137*4882a593Smuzhiyun 138*4882a593Smuzhiyun 139*4882a593SmuzhiyunHints 140*4882a593Smuzhiyun----- 141*4882a593Smuzhiyun 142*4882a593SmuzhiyunIf you want to pass special arguments to your interpreter, you can 143*4882a593Smuzhiyunwrite a wrapper script for it. 144*4882a593SmuzhiyunSee :doc:`Documentation/admin-guide/java.rst <./java>` for an example. 145*4882a593Smuzhiyun 146*4882a593SmuzhiyunYour interpreter should NOT look in the PATH for the filename; the kernel 147*4882a593Smuzhiyunpasses it the full filename (or the file descriptor) to use. Using ``$PATH`` can 148*4882a593Smuzhiyuncause unexpected behaviour and can be a security hazard. 149*4882a593Smuzhiyun 150*4882a593Smuzhiyun 151*4882a593SmuzhiyunRichard Günther <rguenth@tat.physik.uni-tuebingen.de> 152