Currently --userns is only allowed when pasta is attaching to an existing
netns or PID, and is prohibited when creating a new netns by spawning a
command or shell.
With the new handling of userns, this check isn't neccessary. I'm not sure
if there's any use case for --userns with a spawned command, but it's
strictly more flexible and requires zero extra code, so we might as well.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
passt/pasta can interact with user namespaces in a number of ways:
1) With --netns-only we'll remain in our original user namespace
2) With --userns or a PID option to pasta we'll join either the given
user namespace or that of the PID
3) When pasta spawns a shell or command we'll start a new user namespace
for the command and then join it
4) With passt we'll create a new user namespace when we sandbox()
ourself
However (3) and (4) turn out to have essentially the same effect. In both
cases we create one new user namespace. The spawned command starts there,
and passt/pasta itself will live there from sandbox() onwards.
Because of this, we can simplify user namespace handling by moving the
userns handling earlier, to the same point we drop root in the original
namespace. Extend the drop_user() function to isolate_user() which does
both.
After switching UID and GID in the original userns, isolate_user() will
either join or create the userns we require. When we spawn a command with
pasta_start_ns()/pasta_setup_ns() we no longer need to create a userns,
because we're already made one. sandbox() likewise no longer needs to
create (or join) an userns because we're already in the one we need.
We no longer need c->pasta_userns_fd, since the fd is only used locally
in isolate_user(). Likewise we can replace c->netns_only with a local
in conf(), since it's not used outside there.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
conf_ns_open() opens file descriptors for the namespaces pasta needs, but
it doesnt really have anything to do with configuration any more. For
better clarity, move it to pasta.c and rename it pasta_open_ns(). This
makes the symmetry between it and pasta_start_ns() more clear, since these
represent the two basic ways that pasta can operate, either attaching to
an existing namespace/process or spawning a new one.
Since its no longer validating options, the errors it could return
shouldn't cause a usage message. Just exit directly with an error instead.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
There are a number of different ways to specify namespaces for pasta to
use. Some combinations are valid and some are not. Currently validation
for these is spread across several places: conf_ns_pid() validates PID
options specifically. Near its callsite in conf() several other checks
are made. Some additional checks are made in conf_ns_open() and finally
theres a check just before the call to pasta_start_ns().
This is quite hard to follow. Make it easier by putting all the validation
logic together in a new conf_pasta_ns() function, which subsumes
conf_ns_pid(). This reveals that some of the checks were redundant with
each other, so remove those.
For good measure, rename conf_netns() to conf_netns_opt() to make it
clearer its handling just the --netns option specifically, not overall
configuration of the netns.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
passt/pasta contains a number of routines designed to isolate passt from
the rest of the system for security. These are spread through util.c and
passt.c. Move them together into a new isolation.c file.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
passt is allowed to run as "root" (UID 0) in a user namespace, but notas
real root in the init namespace. We read /proc/self/uid_map to determine
if we're in the init namespace or not.
If we're unable to open /proc/self/uid_map we assume we're ok and
continue running as UID 0. This seems unwise. The only instances I
can think of where uid_map won't be available are if the host kernel
doesn't support namespaces, or /proc is not mounted. In neither case
is it safe to assume we're "not really" root and continue (although in
practice we'd likely fail for other reasons pretty soon anyway).
Therefore, fail with an error in this case, instead of carrying on.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Currently the logic to work out what UID and GID we will run as is spread
across conf(). If --runas is specified it's handled in conf_runas(),
otherwise it's handled by check_root(), which depends on initialization of
the uid and gid variables by either conf() itself or conf_runas().
Make this clearer by putting all the UID and GID logic into a single
conf_ugid() function.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
check_root() both checks to see if we are root (in the init namespace),
and if we are drops to an unprivileged user. To make future cleanups
simpler, split the checking for root (now in check_root()) from the actual
dropping of privilege (now in drop_root()).
Note that this does slightly alter semantics. Previously we would only
setuid() if we were originally root (in the init namespace). Now we will
always setuid() and setgid(), though it won't actually change anything if
we weren't privileged to begin with. This also means that we will now
always attempt to switch to the user specified with --runas, even if we
aren't (init namespace) root to begin with. Obviously this will fail with
an error if we weren't privileged to start with. --help and the man page
are updated accordingly.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
c->uid and c->gid are first set in conf(), and last used in check_root()
itself called from conf(). Therefore these don't need to be fields in the
long lived context structure and can instead be locals in conf().
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Declaring them as required_argument in the longopts array specifies
validation, but doesn't affect how optind is increased after parsing
their values.
Currently, passing one of these options as last option causes pasta
to handle their own values as path to a binary to execute.
Fixes: aae2a9bbf7 ("conf: Use "-D none" and "-S none" instead of missing empty option arguments")
Fixes: bf95322fc1 ("conf: Make the argument to --pcap option mandatory")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
When not given an existing PID or network namspace to attach to, pasta
spawns a shell. Most commands which can spawn a shell in an altered
environment can also run other commands in that same environment, which can
be useful in automation.
Allow pasta to do the same thing; it can be given an arbitrary command to
run in the network and user namespace which pasta creates. If neither a
command nor an existing PID or netns to attach to is given, continue to
spawn a default shell, as before.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
When attaching to an existing namespace, pasta can take a PID or the name
or path of a network namespace as a non-option parameter. We disambiguate
based on what the parameter looks like. Make this more explicit by using
a --netns option for explicitly giving the path or name, and treating a
non-option argument always as a PID.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
[sbrivio: Fix typo in man page]
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
pasta takes as its only non-option argument either a PID to attach to the
namespaces of, a PATH to a network namespace or a NAME of a network
namespace (relative to /run/netns). Currently to determine which it is
we try all 3 in that order, and if anything goes wrong we move onto the
next.
This has the potential to cause very confusing failure modes. e.g. if the
argument is intended to be a network namespace name, but a (non-namespace)
file of the same name exists in the current directory.
Make behaviour more predictable by choosing how to treat the argument based
only on the argument's contents, not anything else on the system:
- If it's a decimal integer treat it as a PID
- Otherwise, if it has no '/' characters, treat it as a netns name
(ip-netns doesn't allow '/' in netns names)
- Otherwise, treat it as a netns path
If you want to open a persistent netns in the current directory, you can
use './netns'.
This also allows us to split the parsing of the PID|PATH|NAME option from
the actual opening of the namespaces. In turn that allows us to put the
opening of existing namespaces next to the opening of new namespaces in
pasta_start_ns. That makes the logical flow easier to follow and will
enable later cleanups.
Caveats:
- The separation of functions mean we will always generate the basename
and dirname for the netns_quit system, even when using PID namespaces.
This is pointless, since the netns_quit system doesn't work for non
persistent namespaces, but is harmless.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
After calling conf_ns_opt() we check for -ENOENT and print an error
message, but conf_ns_opt() prints messages for other errors itself. For
consistency move the ENOENT message into conf_ns_opt() as well.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
pasta can identify a netns as a "name", which is to say a path relative to
(usually) /run/netns, which is the place that ip(8) creates persistent
network namespaces. Alternatively a full path to a netns can be given.
The --nsrun-dir option allows the user to change the standard path where
netns names are resolved. However, there's no real point to this, if the
user wants to override the location of the netns, they can just as easily
use the full path to specify the netns.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Both the -D (--dns) and -S (--search) options take an optional argument.
If the argument is omitted the option is disabled entirely. However,
handling the optional argument requires some ugly special case handling if
it's the last option on the command line, and has potential ambiguity with
non-option arguments used with pasta. It can also make it more confusing
to read command lines.
Simplify the logic here by replacing the non-argument versions with an
explicit "-D none" or "-S none".
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
[sbrivio: Reworked logic to exclude redundant/conflicting options]
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
The --pcap or -p option can be used with or without an argument. If given,
the argument gives the name of the file to save a packet trace to. If
omitted, we generate a default name in /tmp.
Generating the default name isn't particularly useful though, since making
a suitable name can easily be done by the caller. Remove this feature.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
There are several places in the passt code where we have lint overrides
because we're not adding CLOEXEC flags to open or other operations.
Comments suggest this is because it's before we fork() into the background
but we'll need those file descriptors after we're in the background.
However, as the name suggests CLOEXEC closes on exec(), not on fork(). The
only place we exec() is either super early invoke the avx2 version of the
binary, or when we start a shell in pasta mode, which certainly *doesn't*
require the fds in question.
Add the CLOEXEC flag in those places, and remove the lint overrides.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Looks like a copy-paste error where we're checking against the size of the
pcap field, rather than the sock_path field.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
The context structure contains a batch of fields specific to IPv4 and to
IPv6 connectivity. Split those out into a sub-structure.
This allows the conf_ip4() and conf_ip6() functions, which take the
entire context but touch very little of it, to be given more specific
parameters, making it clearer what it affects without stepping through the
code.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
After recent changes, conf_ip() now has essentially entirely disjoint paths
for IPv4 and IPv6 configuration. So, it's cleaner to split them out into
different functions conf_ip4() and conf_ip6().
Splitting these out also lets us make the interface a bit nicer, having
them return success or failure directly, rather than manipulating c->v4
and c->v6 to indicate success/failure of the two versions.
Since these functions may also initialize the interface index for each
protocol, it turns out we can then drop c->v4 and c->v6 entirely, replacing
tests on those with tests on whether c->ifi4 or c->ifi6 is non-zero (since
a 0 interface index is never valid).
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
[sbrivio: Whitespace fixes]
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
The v4 and v6 fields of the context structure can be confusing, because
they change meaning part way through the code: Before conf_ip(), they are
booleans which indicate whether the -4 or -6 options have been given.
After conf_ip() they are DISABLED|ENABLED|PROBE enums which indicate
whether the IP version is available (which means both that it was allowed
on the command line and we were able to configure it). The PROBE variant
of the enum is only used locally within conf_ip() and since recent changes
there it no longer has a real purpose different from ENABLED.
Simplify this all by making the context fields always just a boolean
indicating the availability of the IP version. They both default to 1, but
can be set to 0 by either command line options or configuration failures.
We use some local variables in conf() for tracking the state of the command
line options on their own.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
[sbrivio: Minor coding style fix in conf.c]
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
In pasta mode, the guest's MAC address is set up in pasta_ns_cobf() called
from tap_sock_tun_init(). If we have a guest MAC configured with
--ns-mac-addr, this will set the given MAC on the kernel tuntap device, or
if we haven't configured one it will update our record of the guest MAC to
the kernel assigned one from the device.
For passt, we don't initially know the guest's MAC until we receive packets
from it, so we have to initially use a broadcast address. This is - oddly
- set up in an entirely different place, in conf_ip() conditional on the
mode.
Move it to the logically matching place for passt - tap_sock_unix_init().
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
When sending packets to the guest we need a source MAC address, which we
currently take from the host side interface we're using (though it's
basically arbitrary). However if not given on the command line this MAC
is initialized in an IPv4 specific path, and will end up as
00:00:00:00:00:00 when running "passt 6". The MAC address is also used
for IPv6 packets, though.
Interestingly, we largely seem to get away with using an all-zero MAC, but
it's probably not a good idea. Make the IPv6 path pick the MAC address
from its interface if the IPv4 path hasn't already done so.
While we're there, use the existing MAC_IS_ZERO macro to make the code a
little clearer.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Now that the back end allows passt/pasta to use different external
interfaces for IPv4 and IPv6, use that to do the right thing in the case
that the host has IPv4 and IPv6 connectivity via different interfaces.
If the user hasn't explicitly chosen an interface, separately search for
a suitable external interface for each protocol.
As a bonus, this substantially simplifies the external interface probe. It
also eliminates a subtle confusing case where in some circumstances we
would pick the first interface in interface index order, and sometimes in
order of routes returned from netlink. On some network configurations that
could cause tests to fail, because the logic in the tests was subtly
different (it always used route order).
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
It's quite plausible for a host to have both IPv4 and IPv6 connectivity,
but only via different interfaces. For example, this will happen in the
case that IPv6 connectivity is via a tunnel (e.g. 6in4 or 6rd). It would
also happen in the case that IPv4 access is via a tunnel on an otherwise
IPv6 only local network, which is a setup that might become more common in
the post IPv4 address exhaustion world.
In turns out there's no real need for passt/pasta to get its IPv4 and IPv6
connectivity via the same interface, so we can handle this situation fairly
easily. Change the core to allow eparate external interfaces for IPv4 and
IPv6. We don't actually set these separately for now.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
I forgot to reset the range endpoints after parsing an item of the
comma-separated list in commit 220759efb8 ("conf: Allow to specify
ranges and ports excluded from given ranges") -- fix that.
Fixes: 220759efb8 ("conf: Allow to specify ranges and ports excluded from given ranges")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This is useful in environments where we want to forward a large
number of ports, or all non-ephemeral ones, and some other service
running on the host needs a few selected ports.
I'm using ~ as prefix for the specification of excluded ranges and
ports to avoid the need for explicit command line quoting.
Ranges and ports can be excluded from given ranges by adding them
in the comma-separated list, prefixed by ~. Some quick examples:
-t 5000-6000,~5555: forward ports 5000 to 6000, but not 5555
-t ~20000-20010: forward all non-ephemeral, allowed ports, except
for ports 20000 to 20010
...more details in usage message and man page.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
In commit 675174d4ba ("conf, tap: Split netlink and pasta
functions, allow interface configuration"), I broke the initial
setting of the observed IPv6 addresses in two ways:
- the size copied from the configured addresses corresponds to an
IPv4 address, not to an IPv6 address
- the observed link-local address is initialised to the configured
unicast address, not the link-local one
If we haven't seen the guest using some type of addresses yet, we
should default to the configured values, hence these initial
settings: fix both.
This resulted in UDP flows to the guest from a unique local address
on the network not working before the guest shows passt a valid
address itself, as reported by Alona.
Reported-by: Alona Paz <alkaplan@redhat.com>
Link: https://bugs.passt.top/show_bug.cgi?id=16
Fixes: 675174d4ba ("conf, tap: Split netlink and pasta functions, allow interface configuration")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
By default, passt detects the nameserver used by the host system by reading
/etc/resolv.conf, and advertises that to the guest via DHCP. However this
breaks down if the host's nameserver is local (on 127.0.0.1 or ::1);
connecting to localhost on the guest won't reach the host's nameserver.
Using a local nameserver is a reasonably common case when using dnsmasq
or similar to merge name resolution on a home network with name resolution
from an organization-private VPN.
We already have the gateway mapping support to allow reaching host-local
services from the guest via the address of the default gateway. Add code
to detect the case of a local DNS server and use the gateway mapping to
advertise it usefully to the guest.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Switch the resolv.conf parsing in conf.c to use the new lineread
implementation. This means that it can now handle a resolv.conf file which
contains blank lines.
There are quite a few other fragilities with the resolv.conf parsing, but
that's out of scope for this patch.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
In conf_runas(), Coverity reports that we might dereference uid and
gid despite possibly being NULL (CWE-476) because of the check after
the first sscanf(). They can't be NULL, but I actually wanted to
check that UID and GID are non-zero (the user could otherwise pass
--runas root:root and defy the whole mechanism).
Later on, we have the same type of warning for 'gr': it's compared
against NULL, so it might be NULL, which is actually the case: but
in that case, we don't dereference it, because we'll return -ENOENT
right away. Rewrite the clause to silence the warning.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
On some systems, user and group "nobody" might not be available. The
new --runas option allows to override the default "nobody" choice if
started as root.
Now that we allow this, drop the initgroups() call that was used to
add any additional groups for the given user, as that might now
grant unnecessarily broad permissions. For instance, several
distributions have a "kvm" group to allow regular user access to
/dev/kvm, and we don't need that in passt or pasta.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This feature is available in slirp4netns but was missing in passt and
pasta.
Given that we don't do dynamic memory allocation, we need to bind
sockets while parsing port configuration. This means we need to
process all other options first, as they might affect addressing and
IP version support. It also implies a minor rework of how TCP and UDP
implementations bind sockets.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This really just needs to be an assignment before line_read() --
turn it into a for loop. Reported by Coverity.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
--debug can be a bit too noisy, especially as single packets or
socket messages are logged: implement a new option, --trace,
implying --debug, that enables all debug messages.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
If we statically configure a default route, and also advertise it for
SLAAC, the kernel will try moments later to add the same route:
ICMPv6: RA: ndisc_router_discovery failed to add default route
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This should be convenient for users managing filesystem-bound network
namespaces: monitor the base directory of the namespace and exit if
the namespace given as PATH or NAME target is deleted. We can't add
an inotify watch directly on the namespace directory, that won't work
with nsfs.
Add an option to disable this behaviour, --no-netns-quit.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
For compatibility with libslirp/slirp4netns users: introduce a
mechanism to map, in the UDP routines, an address facing guest or
namespace to the first IPv4 or IPv6 address resulting from
configuration as resolver. This can be enabled with the new
--dns-forward option.
This implies that sourcing and using DNS addresses and search lists,
passed via command line or read from /etc/resolv.conf, is not bound
anymore to DHCP/DHCPv6/NDP usage: for example, pasta users might just
want to use addresses from /etc/resolv.conf as mapping target, while
not passing DNS options via DHCP.
Reflect this in all the involved code paths by differentiating
DHCP/DHCPv6/NDP usage from DNS configuration per se, and in the new
options --dhcp-dns, --dhcp-search for pasta, and --no-dhcp-dns,
--no-dhcp-search for passt.
This should be the last bit to enable substantial compatibility
between slirp4netns.sh and slirp4netns(1): pass the --dns-forward
option from the script too.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Provide a sane default, instead of /0, if an address is given, and it
doesn't correspond to any host address we could find via netlink.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Nobody currently calls this as passt4netns, that was the name I used
before 'pasta', drop any reference before it's too late.
While at it, explicitly check that argc is bigger than or equal to
one, just as a defensive measure: argv[0] being NULL is not an issue
anyway.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
To reach (at least) a conceptually equivalent security level as
implemented by --enable-sandbox in slirp4netns, we need to create a
new mount namespace and pivot_root() into a new (empty) mountpoint, so
that passt and pasta can't access any filesystem resource after
initialisation.
While at it, also detach IPC, PID (only for passt, to prevent
vulnerabilities based on the knowledge of a target PID), and UTS
namespaces.
With this approach, if we apply the seccomp filters right after the
configuration step, the number of allowed syscalls grows further. To
prevent this, defer the application of seccomp policies after the
initialisation phase, before the main loop, that's where we expect bad
things to happen, potentially. This way, we get back to 22 allowed
syscalls for passt and 34 for pasta, on x86_64.
While at it, move #syscalls notes to specific code paths wherever it
conceptually makes sense.
We have to open all the file handles we'll ever need before
sandboxing:
- the packet capture file can only be opened once, drop instance
numbers from the default path and use the (pre-sandbox) PID instead
- /proc/net/tcp{,v6} and /proc/net/udp{,v6}, for automatic detection
of bound ports in pasta mode, are now opened only once, before
sandboxing, and their handles are stored in the execution context
- the UNIX domain socket for passt is also bound only once, before
sandboxing: to reject clients after the first one, instead of
closing the listening socket, keep it open, accept and immediately
discard new connection if we already have a valid one
Clarify the (unchanged) behaviour for --netns-only in the man page.
To actually make passt and pasta processes run in a separate PID
namespace, we need to unshare(CLONE_NEWPID) before forking to
background (if configured to do so). Introduce a small daemon()
implementation, __daemon(), that additionally saves the PID file
before forking. While running in foreground, the process itself can't
move to a new PID namespace (a process can't change the notion of its
own PID): mention that in the man page.
For some reason, fork() in a detached PID namespace causes SIGTERM
and SIGQUIT to be ignored, even if the handler is still reported as
SIG_DFL: add a signal handler that just exits.
We can now drop most of the pasta_child_handler() implementation,
that took care of terminating all processes running in the same
namespace, if pasta started a shell: the shell itself is now the
init process in that namespace, and all children will terminate
once the init process exits.
Issuing 'echo $$' in a detached PID namespace won't return the
actual namespace PID as seen from the init namespace: adapt
demo and test setup scripts to reflect that.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
clang-tidy from LLVM 13.0.1 reports some new warnings from these
checkers:
- altera-unroll-loops, altera-id-dependent-backward-branch: ignore
for the moment being, add a TODO item
- bugprone-easily-swappable-parameters: ignore, nothing to do about
those
- readability-function-cognitive-complexity: ignore for the moment
being, add a TODO item
- altera-struct-pack-align: ignore, alignment is forced in protocol
headers
- concurrency-mt-unsafe: ignore for the moment being, add a TODO
item
Fix bugprone-implicit-widening-of-multiplication-result warnings,
though, that's doable and they seem to make sense.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
