This prevents the shell user from injecting input as if it were coming
from an existing input device, which in turn makes it easier for malware
/ abuse analysis to detect when someone is injecting input via the
command line vs a real user using a device.
(cherrypick of 95637eb2a332b9a09914b1bf32c10fccfba4a175.)
Bug: 30861057
Test: ran getevent, saw correct output, played with device
Change-Id: Ib53eea1b7767f25510b5d36fe6109101a9fad8e0
/dev/mem has been turned off on all Android kernels (b/19549480). This
utility cannot be used.
Test: android compiles.
Change-Id: I500021f224febb15dd8bfe6e1f6e5c398fe6f552
Some of the less-used stuff stayed behind in input.h, so we need to
parse both files.
Change-Id: Iff2b6e3fc4f5a6584169d12f16427b4f235da21a
Test: manual inspection of the generated file.
Let's remove this convenient tool for attacking buggy kernels and see if
anyone with a legitimate use notices. I suspect most potential legitimate
users write a short C program instead anyway.
Bug: http://b/29128170
Change-Id: I14e8b8594902951fe0b94c9ce13baa2c4d7b9e6e
LOCAL_LDLIBS should only be used with -l flags for NDK and some host
libraries. Other uses should use LOCAL_SHARED_LIBRARIES.
LOCAL_LDFLAGS should not contain -l entries, they should be in
LOCAL_LDLIBS instead.
LOCAL_CFLAGS should not contain -I entries, those should use
LOCAL_C_INCLUDES instead.
Change-Id: Ic901aa14651572ac682166becd4f0cd86baecc34
Hopefully this will let us remove the empty file from bionic. It shouldn't
be there because -- unlike BSD -- we don't actually have a libutil.
Change-Id: I7e3042ca120aafa877709797866b52ac62675870
Insert a space between 'if'/'while' and their clause.
Also insert a space after ',' for arguments.
This was done with clang-format.
Bug: None
Change-Id: I5315eba3d5d8ac68996a373dc7c8e1bc45165824
Currently, 'ps' will segfault when called with '--ppid' and no argument.
Make it return an error instead.
Bug: 28383024
Change-Id: I71f2d1ba9502996ae75b17605234eb99ebb589f2
It's nothing like the real iftop, and there's no evidence anyone uses it.
In particular, it hasn't been touched since it was first checked in, back
in 2008. So it's either perfect or unused. Let's find out...
Change-Id: I941adbb67f0166b9b3bd796291060bd67eafa867
Remove the weird @command syntax, and make toolbox itself list the available
commands, like toybox does. Also stop special-casing toolbox itself.
Change-Id: I77dfaaf0cf42e375d866b77362dd8afa717fb2e1
Also cleans up two instances of open() with useless mode params, and
changes a few uses of snprintf to use sizeof(buffer) instead of
hardcoded buffer sizes.
Change-Id: If11591003d910c995e72ad8f75afd072c255a3c5
The use of PR to mean PSR rather than PR was confusing, so let's just
use PR and NI the same way everyone else does.
Also remove bogus blank lines and unnecessary delays (both before first
showing anything and after having shown the final iteration).
Bug: http://b/27447491
Change-Id: I676131956facf2167abb5d0852c55b0670b5eaa6
Allow the scripts to run with both python 2.x and 3.x
Change-Id: I911118bcf370d09bdb2fb46afa21af64257f1ffb
Signed-off-by: Bernhard Rosenkränzer <Bernhard.Rosenkranzer@linaro.org>
The recent uapi header update moved all the interesting constants
out into another file. Track that, and add the missing makefile
dependency.
Bug: http://b/27073568
Change-Id: I8ac1dd1122865479f69a76cbf427d4aacdf7b790
There is at least one app that assumes that you will always get at least
one line of output in stdout from ps.
To fix this, move error output to stdout, and move the check of whether
/proc can be opened until after the ps header is printed.
Bug: 26554285
Change-Id: I6d9342aafd5c6f728735507cdd87a48a8e0373ac
The toolbox -t option is really the standard -H option. Add -H so
I can switch dumpstate over.
We should also flush after each run so we don't arbitrarily truncate
at the stdio buffer size. (I'm pretty sure no one uses toolbox top
interactively, but it annoyed me while testing -H.)
Change-Id: Ib87cb541d1b21c48468614b2ef60b0cc62694c95
This is another step towards being able to swap toolbox ps out for
a more complete ps. This should also let us fix dumpstate to not
need a separate "ps -Z" run --- we can just add -Z to the previous
run of ps instead.
Change-Id: I0b96dff51988edd2b33879f8c31faccebeeffddd
Regular ps doesn't have this, and at least in our tree everyone
seems to actually pipe to grep instead of using this anyway. But
since the switch to a more full-featured ps is going to be a nightmare
anyway, let's dip a toe in the water by removing something we don't plan
on supporting.
Change-Id: Id220ae0cfdad8368341730f643a2bfbaad3c9716
Since watchprops would have to be rewritten anyway in the upcoming properties
rewrite, there seems little point in rewriting it for toybox first. Let's
see how many people actually use watchprops and decide whether it's worth
the effort...
Change-Id: I996e7209bc81633ee66602f09f130226336d9987
Michael Wright