Age | Commit message (Collapse) | Author | Files | Lines |
|
The BufferedReader's .read() doesn't behave as expected. It reads
EXACTLY size bytes and will block until there are enough available to
read.
os.read() does what we expect. It will read UP TO size bytes and only
block if there is nothing available to read.
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
If the user presses Ctrl+C while waiting on a connection, we want to
gracefully exit.
If the user presses Ctrl+C during the script, we want to stop executing
the script and restart the loop.
If any other exception happens during the script, we want to print out
the stacktrace as normal, but continue the loop.
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
With the "read rest of output" code in the Comm destructor, it would
continue to read output even in situations where some error happened and
we expect sploit to die or when the user presses Ctrl+C to end sploit.
By moving it to the end of the script running code in main, it behaves
more intuitively.
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
Handle all of the edge cases when shutting down in Pipes mode.
e.g.
If the pipes are broken (tried to write after the program died)
If the fifos don't exist anymore (sometimes tempfile cleans them up
before the destructor finishes when certain errors happen)
If the object attributes for the streams and fifo paths aren't set (this
can happen if the constructor didn't finish. e.g. the user cancels while
waiting on a connection)
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
If we need to wait on the target program to die, we don't want to just
wait forever with no indication to the user. Instead, only call wait if
the program is still alive, inform the user that we are doing this, and
give them the ability to forcefully kill the target program with Ctrl+C.
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
Previously, you could specify a directory which must exist under /tmp.
Now, you can give the full path to a directory to be used by Pipes.
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
comm.interact() will drop the user into an "interactive" mode where they
can directly control what is sent. A SIGINT (Ctrl+C) will drop the
script out of interactive mode and continue executing the rest of the
script. If the output of the program (input into our script) goes into
a broken state (such as when the target program exits), interactive mode
will automatically exit.
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
Both new functions check the input for a predicate and keep reading
until the predicate is true.
readuntil() will consume input byte by byte and use the entire string
read to check the predicate. It will then return that entire string.
readlineuntil() consumes input line by line and only uses the last line
to check the predicate. The line that satisfies the predicate is all
that is returned.
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
First part of the MVP for the larger Sploit rework effort.
Add project structure, python packaging, basic comms, and "log" hook.
From in or out of the sploit directory, you can run the "sploit.py"
script, run python -m sploit, or import the sploit modules from the
python3 shell.
You can also pip install Sploit and from anywhere you can run the sploit
command, run python -m sploit, or import the sploit modules from the
python3 shell.
Running as a standalone application, Sploit can run in a "target" mode,
a "pipe" mode, and a "pipe daemon" mode. In "target" mode, Sploit will
launch a target program as a subprocess and run an exploit script
against its I/O. In "pipe" mode, Sploit will create named fifos and
wait for a program to connect to them to run an exploit script against
them. In "pipe daemon" mode, Sploit will run similar to the "pipe" mode,
but automatically recreate the fifos with the same name after each
execution.
Basic comm operations of read, readline, write, and writeline are
available to the exploit script.
A "log" hook is executed whenever data is read in from the target
program. This will just print the data out, but it can be configured to
decode it with a specific encoding or you could replace the function for
different behavior.
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|
|
Signed-off-by: dusoleil <howcansocksbereal@gmail.com>
|