Adapted from a patch by Paul Holzinger: when pasta spawns a command,
operating without a pre-existing user and network namespace, it needs
to wait for the tap device to be configured and its handler ready,
before the command is actually executed.
Otherwise, something like:
pasta --config-net nslookup passt.top
usually fails as the nslookup command is issued before the network
interface is ready.
We can't adopt a simpler approach based on SIGSTOP and SIGCONT here:
the child runs in a separate PID namespace, so it can't send SIGSTOP
to itself as the kernel sees the child as init process and blocks
the delivery of the signal.
We could send SIGSTOP from the parent, but this wouldn't avoid the
possible condition where the child isn't ready to wait for it when
the parent sends it, also raised by Paul -- and SIGSTOP can't be
blocked, so it can never be pending.
Use SIGUSR1 instead: mask it before clone(), so that the child starts
with it blocked, and can safely wait for it. Once the parent is
ready, it sends SIGUSR1 to the child. If SIGUSR1 is sent before the
child is waiting for it, the kernel will queue it for us, because
it's blocked.
Reported-by: Paul Holzinger <pholzing@redhat.com>
Fixes: 1392bc5ca0 ("Allow pasta to take a command to execute")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This passes a fully connected stream socket to passt.
Signed-off-by: Richard W.M. Jones <rjones@redhat.com>
[sbrivio: reuse fd_tap instead of adding a new descriptor,
imply --one-off on --fd, add to optstring and usage()]
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...so that we avoid printing some lines twice because log-level is
still set to LOG_EMERG, as if logging configuration didn't happen
yet.
While at it, note that logging to stderr doesn't really depend on
whether debug mode is enabled or not.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
We recently corrected some errors handling the endianness of IPv4
addresses. These are very easy errors to make since although we mostly
store them in network endianness, we sometimes need to manipulate them in
host endianness.
To reduce the chances of making such mistakes again, change to always using
a (struct in_addr) instead of a bare in_addr_t or uint32_t to store network
endian addresses. This makes it harder to accidentally do arithmetic or
comparisons on such addresses as if they were host endian.
We introduce a number of IN4_IS_ADDR_*() helpers to make it easier to
directly work with struct in_addr values. This has the additional benefit
of making the IPv4 and IPv6 paths more visually similar.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
In pasta mode, ICMP and ICMPv6 echo sockets relay back to us any
reply we send: we're on the same host as the target, after all. We
discard them by comparing the last sequence we sent with the sequence
we receive.
However, on the first reply for a given identifier, the sequence
might be zero, depending on the implementation of ping(8): we need
another value to indicate we haven't sent any sequence number, yet.
Use -1 as initialiser in the echo identifier map.
This is visible with Busybox's ping, and was reported by Paul on the
integration at https://github.com/containers/podman/pull/16141, with:
$ podman run --net=pasta alpine ping -c 2 192.168.188.1
...where only the second reply would be routed back.
Reported-by: Paul Holzinger <pholzing@redhat.com>
Fixes: 33482d5bf2 ("passt: Add PASTA mode, major rework")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
We have a number of steps of self-isolation scattered across our code.
Improve function names and add comments to make it clearer what the self
isolation model is, what the steps do, and why they happen at the points
they happen.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
In some environments, such as KubeVirt pods, we might not have a
system logger available. We could choose to run in foreground, but
this takes away the convenient synchronisation mechanism derived from
forking to background when interfaces are ready.
Add optional logging to file with -l/--log-file and --log-size.
Unfortunately, this means we need to duplicate features that are more
appropriately implemented by a system logger, such as rotation. Keep
that reasonably simple, by using fallocate() with range collapsing
where supported (Linux kernel >= 3.15, extent-based ext4 and XFS) and
falling back to an unsophisticated block-by-block moving of entries
toward the beginning of the file once we reach the (mandatory) size
limit.
While at it, clarify the role of LOG_EMERG in passt.c.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Logging to file is going to add some further complexity that we don't
want to squeeze into util.c.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
TIMER_INTERVAL is the minimum of two separately defined intervals which
happen to have the same value at present. This results in an expression
which has the same value in both branches of a ternary operator, which
cppcheck warngs about. This is logically sound in this case, so suppress
the error (we appear to already have a similar suppression for clang-tidy).
Also add an unmatchedSuppression suppression, since only some cppcheck
versions complain about this instance.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Stop ignoring the return codes from sigaction() and signal(). Unlikely to
happen in practice, but if it ever did it could lead to really hard to
debug problems. So, take clang-tidy's advice and check for errors here.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
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>
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>
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>
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>
Otherwise, if the current PID has fewer digits than a previously
written one, the content will be wrong.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
We handle SIGQUIT and SIGTERM calling exit(), which is usually
implemented with the exit_group() system call.
If we don't allow exit_group(), we'll get a SIGSYS while handling
SIGQUIT and SIGTERM, which means a misleading non-zero exit code.
Reported-by: Wenli Quan <wquan@redhat.com>
Link: https://bugzilla.redhat.com/show_bug.cgi?id=2101990
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...instead of argv[0], which might or might not contain a valid path
to the executable itself. Instead of mangling argv[0], use the same
link to find out if we're already running the AVX2 build where
supported.
Alternatively, we could use execvpe(), but that might result in
running a different installed version, in case e.g. the set of
binaries is present in both /usr/bin and /usr/local/bin, with both
being in $PATH.
Reported-by: Wenli Quan <wquan@redhat.com>
Link: https://bugzilla.redhat.com/show_bug.cgi?id=2101310
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>
It's actually quite easy to make it fail depending on the
environment, accurately report errors here.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
We can't take for granted that the hard limit for open files is
big enough as to allow to delay closing sockets to a timer.
Store the value of RTLIMIT_NOFILE we set at start, and use it to
understand if we're approaching the limit with pending, spliced
TCP connections. If that's the case, close sockets right away as
soon as they're not needed, instead of deferring this task to a
timer.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
With a lot of concurrent connections, the bitmap scan approach is
not really sustainable.
Switch to per-connection timerfd timers, set based on events and on
two new flags, ACK_FROM_TAP_DUE and ACK_TO_TAP_DUE. Timers are added
to the common epoll list, and implement the existing timeouts.
While at it, drop the CONN_ prefix from flag names, otherwise they
get quite long, and fix the logic to decide if a connection has a
local, possibly unreachable endpoint: we shouldn't go through the
rest of tcp_conn_from_tap() if we reset the connection due to a
successful bind(2), and we'll get EACCES if the port number is low.
Suggested by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Using events and flags instead of states makes the implementation
much more straightforward: actions are mostly centered on events
that occurred on the connection rather than states.
An example is given by the ESTABLISHED_SOCK_FIN_SENT and
FIN_WAIT_1_SOCK_FIN abominations: we don't actually care about
which side started closing the connection to handle closing of
connection halves.
Split out the spliced implementation, as it has very little in
common with the "regular" TCP path.
Refactor things here and there to improve clarity. Add helpers
to trace where resets and flag settings come from.
No functional changes intended.
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>
Build-time selection of AVX2 flags and routines is not practical for
distributions, but limiting AVX2 usage to checksum routines with
specific run-time detection doesn't allow for easy performance gains
from auto-vectorisation of batched packet handling routines.
For x86_64, build non-AVX2 and AVX2 binaries, and implement a simple
wrapper replacing the current executable with the AVX2 build if it's
available, and if AVX2 is supported by the current CPU.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
It looks like glibc commonly implements clock_gettime(2) with
clock_gettime64(), and uses recv() instead of recvfrom(), send()
instead of sendto(), and sigreturn() instead of rt_sigreturn() on
armv6l and armv7l.
Adjust the list of system calls for armv6l and armv7l accordingly.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
A kernel might not be configured with CONFIG_TRANSPARENT_HUGEPAGE,
especially on embedded systems. Ignore the error, it doesn't affect
functionality.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...it doesn't actually matter as we're checking errno at the very
end, but, depending on build flags, chdir() might be declared with
warn_unused_result and the compiler issues a warning.
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>
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>
Two effects:
- ptrace() on passt and pasta can only be done by root, so that even
if somebody gains access to the same user, they won't be able to
check data passed in syscalls anyway. No core dumps allowed either
- /proc/PID files are owned by root:root, and they can't be read by
the same user as the one passt or pasta are running with
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>
The existing behaviour is not really practical: an automated agent in
charge of starting both qemu and passt would need to fork itself to
start passt, because passt won't fork to background until qemu
connects, and the agent needs to unblock to start qemu.
Instead of waiting for a connection to daemonise, do it right away as
soon as a socket is available: that can be considered an initialised
state already.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Depending on the C library, but not necessarily in all the
functions we use, statx() might be used instead of stat(),
getdents() instead of getdents64(), readlinkat() instead of
readlink(), openat() instead of open().
On aarch64, it's clone() and not fork(), and dup3() instead of
dup2() -- just allow the existing alternative instead of dealing
with per-arch selections.
Since glibc commit 9a7565403758 ("posix: Consolidate fork
implementation"), we need to allow set_robust_list() for
fork()/clone(), even in a single-threaded context.
On some architectures, epoll_pwait() is provided instead of
epoll_wait(), but never both. Same with newfstat() and
fstat(), sigreturn() and rt_sigreturn(), getdents64() and
getdents(), readlink() and readlinkat(), unlink() and
unlinkat(), whereas pipe() might not be available, but
pipe2() always is, exclusively or not.
Seen on Fedora 34: newfstatat() is used on top of fstat().
syslog() is an actual system call on some glibc/arch combinations,
instead of a connect()/send() implementation.
On ppc64 and ppc64le, _llseek(), recv(), send() and getuid()
are used. For ppc64 only: ugetrlimit() for the getrlimit()
implementation, plus sigreturn() and fcntl64().
On s390x, additionally, we need to allow socketcall() (on top
of socket()), and sigreturn() also for passt (not just for
pasta).
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
On some distributions, on ppc64, ulimit -s returns 'unlimited': add a
reasonable default, and also make sure ulimit is invoked using the
default shell, which should ensure ulimit is actually implemented.
Also note that AUDIT_ARCH doesn't follow closely the naming reported
by 'uname -m': convert for i386 and ppc as needed.
While at it, move inclusion of seccomp.h after util.h, the former is
less generic (cosmetic/clang-tidy only).
Older kernel headers might lack a definition for AUDIT_ARCH_PPC64LE:
define that explicitly if it's not available.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This is the only remaining Linux-specific include -- drop it to avoid
clang-tidy warnings and to make code more portable.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...it looks like, on a recent Fedora installation, daemon() uses it.
Reported-by: Giuseppe Scrivano <gscrivan@redhat.com>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This is actually annoying: there's no way to make it fork into
background when running from a script. However, it's always
possible to keep it in foreground with -f. Make it simpler, and
always fork into background if -f is not given.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...mostly false positives, but a number of very relevant ones too,
in tcp_get_sndbuf(), tcp_conn_from_tap(), and siphash PREAMBLE().
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Unions and structs, you all have names now.
Take the chance to enable bugprone-reserved-identifier,
cert-dcl37-c, and cert-dcl51-cpp checkers in clang-tidy.
Provide a ffsl() weak declaration using gcc built-in.
Start reordering includes, but that's not enough for the
llvm-include-order checker yet.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
A mix of unchecked return values, a missing permission mask for
open(2) with O_CREAT, and some false positives from
-Wstringop-overflow and -Wmaybe-uninitialized.
Reported-by: Martin Hauke <mardnh@gmx.de>
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
While it's not recommended to give passt any capability, drop all the
ones we might have got by mistake, except for the only sensible one,
CAP_NET_BIND_SERVICE.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Replace libc functions that might dynamically allocate memory with own
implementations or wrappers.
Drop brk(2) from list of allowed syscalls in seccomp profile.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
List of allowed syscalls comes from comments in the form:
#syscalls <list>
for syscalls needed both in passt and pasta mode, and:
#syscalls:pasta <list>
#syscalls:passt <list>
for syscalls specifically needed in pasta or passt mode only.
seccomp.sh builds a list of BPF statements from those comments,
prefixed by a binary search tree to keep lookup fast.
While at it, clean up a bit the Makefile using wildcards.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Move netlink routines to their own file, and use netlink to configure
or fetch all the information we need, except for the TUNSETIFF ioctl.
Move pasta-specific functions to their own file as well, add
parameters and calls to configure the tap interface in the namespace.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
We usually have up to one additional child exiting while we receive
a SIGCHLD, instead of complicating this with tracking PIDs, just
add a second waitid() call.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Based on a patch from Giuseppe Scrivano, this adds the ability to:
- specify paths and names of target namespaces to join, instead of
a PID, also for user namespaces, with --userns
- request to join or create a network namespace only, without
entering or creating a user namespace, with --netns-only
- specify the base directory for netns mountpoints, with --nsrun-dir
Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
[sbrivio: reworked logic to actually join the given namespaces when
they're not created, implemented --netns-only and --nsrun-dir,
updated pasta demo script and man page]
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
We'll need this for TCP ACK coalescing on tap/guest-side. For
convenience, allow _handler() functions to be undefined, courtesy
of __attribute__((weak)).
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
The initial timestamp was not initialised, so timers for protocol
handlers wouldn't run at all sometimes.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
If transparent huge pages are available, madvise() will do the trick.
While at it, decrease EPOLL_EVENTS for the main loop from 10 to 8,
for slightly better socket fairness.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
If pasta created the namespace, it's probably expected that processes
started in the same namespace are terminated once pasta exits. Scan
procfs namespace links for corresponding processes, send SIGQUIT and
SIGKILL (after one second) if found.
While at it, make the signal handler reap otherwise-zombies resulting
from clone().
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...this allows processes running as the only group available in the
namespace to create ICMP Echo sockets.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Traffic with loopback source address will be forwarded to the direct
loopback connection in the namespace, and the tap interface is used
for the rest.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Allow to bind IPv4 and IPv6 ports to tap, namespace or init separately.
Port numbers of TCP ports that are bound in a namespace are also bound
for UDP for convenience (e.g. iperf3), and IPv4 ports are always bound
if the corresponding IPv6 port is bound (socket might not have the
IPV6_V6ONLY option set). This will also be configurable later.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Packets are received directly onto pre-cooked, static buffers
for IPv4 (with partial checksum pre-calculation) and IPv6 frames,
with pre-filled Ethernet addresses and, partially, IP headers,
and sent out from the same buffers with sendmmsg(), for both
passt and pasta (non-local traffic only) modes.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
PASTA (Pack A Subtle Tap Abstraction) provides quasi-native host
connectivity to an otherwise disconnected, unprivileged network
and user namespace, similarly to slirp4netns. Given that the
implementation is largely overlapping with PASST, no separate binary
is built: 'pasta' (and 'passt4netns' for clarity) both link to
'passt', and the mode of operation is selected depending on how the
binary is invoked. Usage example:
$ unshare -rUn
# echo $$
1871759
$ ./pasta 1871759 # From another terminal
# udhcpc -i pasta0 2>/dev/null
# ping -c1 pasta.pizza
PING pasta.pizza (64.190.62.111) 56(84) bytes of data.
64 bytes from 64.190.62.111 (64.190.62.111): icmp_seq=1 ttl=255 time=34.6 ms
--- pasta.pizza ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 34.575/34.575/34.575/0.000 ms
# ping -c1 spaghetti.pizza
PING spaghetti.pizza(2606:4700:3034::6815:147a (2606:4700:3034::6815:147a)) 56 data bytes
64 bytes from 2606:4700:3034::6815:147a (2606:4700:3034::6815:147a): icmp_seq=1 ttl=255 time=29.0 ms
--- spaghetti.pizza ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 28.967/28.967/28.967/0.000 ms
This entails a major rework, especially with regard to the storage of
tracked connections and to the semantics of epoll(7) references.
Indexing TCP and UDP bindings merely by socket proved to be
inflexible and unsuitable to handle different connection flows: pasta
also provides Layer-2 to Layer-2 socket mapping between init and a
separate namespace for local connections, using a pair of splice()
system calls for TCP, and a recvmmsg()/sendmmsg() pair for UDP local
bindings. For instance, building on the previous example:
# ip link set dev lo up
# iperf3 -s
$ iperf3 -c ::1 -Z -w 32M -l 1024k -P2 | tail -n4
[SUM] 0.00-10.00 sec 52.3 GBytes 44.9 Gbits/sec 283 sender
[SUM] 0.00-10.43 sec 52.3 GBytes 43.1 Gbits/sec receiver
iperf Done.
epoll(7) references now include a generic part in order to
demultiplex data to the relevant protocol handler, using 24
bits for the socket number, and an opaque portion reserved for
usage by the single protocol handlers, in order to track sockets
back to corresponding connections and bindings.
A number of fixes pertaining to TCP state machine and congestion
window handling are also included here.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
The most common case is actually that no other instance created a
socket with that name -- and that also means there is no other
instance.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...sharing the same filesystem. Instead of a fixed path for the UNIX
domain socket, passt now uses a path with a counter, probing for
existing instances, and picking the first free one.
The demo script is updated accordingly -- it can now be started several
times to create multiple namespaces with an instance of passt each,
with addressing reflecting separate subnets, and NDP proxying between
them.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
and, given that the connection table is indexed by socket number,
we also need to increase MAX_CONNS now as the ICMP implementation
needs 2^17 sockets, that will be opened before TCP connections are
accepted.
This needs to be changed later: the connection table should be
indexed by a translated number -- we're wasting 2^17 table entries
otherwise. Move initialisation of TCP listening sockets as last
per-protocol initialisation, this will make it easier.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
The socket isn't necessarily closed, make sure we close it before
getting a new one from accept(), so that we don't mix it up with
protocol sockets numbering.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
With -DDEBUG, passt now saves guest-side traffic captures in
pcap format at /tmp/passt_<ISO8601 timestamp>.pcap. The timestamp
refers to time and date of start-up.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Add support for a variable amount of DNS servers, including zero,
from /etc/resolv.conf, in DHCP, NDP and DHCPv6 implementations.
Introduce support for domain search list for DHCP (RFC 3397),
NDP (RFC 8106), and DHCPv6 (RFC 3646), also sourced from
/etc/resolv.conf.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Once passt forks to background, it should be guaranteed that the UNIX
domain socket is available, otherwise, if qemu is started right after
it, it might fail to connect.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Multiple arrays, one for each address, were needed with a single
fprintf(). Now that it's replaced by info(), we can have just one
for each protocol version.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
With glibc, we can't reliably build a static binary with
getprotobynumber(), which is currently used with -DDEBUG.
Replace that with a small array of protocol strings.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This is in preparation for scatter-gather IO on the UDP receive path:
save a getsockname() syscall by setting a flag if we get the numbering
of all bound sockets in a strict sequence (expected, in practice) and
repurpose the tap buffer to be also a socket receive buffer, passing
it down to protocol handlers.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
As we support UDP forwarding for packets that are sent to local
ports, we actually need some kind of connection tracking for UDP.
While at it, this commit introduces a number of vaguely related fixes
for issues observed while trying this out. In detail:
- implement an explicit, albeit minimalistic, connection tracking
for UDP, to allow usage of ephemeral ports by the guest and by
the host at the same time, by binding them dynamically as needed,
and to allow mapping address changes for packets with a loopback
address as destination
- set the guest MAC address whenever we receive a packet from tap
instead of waiting for an ARP request, and set it to broadcast on
start, otherwise DHCPv6 might not work if all DHCPv6 requests time
out before the guest starts talking IPv4
- split context IPv6 address into address we assign, global or site
address seen on tap, and link-local address seen on tap, and make
sure we use the addresses we've seen as destination (link-local
choice depends on source address). Similarly, for IPv4, split into
address we assign and address we observe, and use the address we
observe as destination
- introduce a clock_gettime() syscall right after epoll_wait() wakes
up, so that we can remove all the other ones and pass the current
timestamp to tap and socket handlers -- this is additionally needed
by UDP to time out bindings to ephemeral ports and mappings between
loopback address and a local address
- rename sock_l4_add() to sock_l4(), no semantic changes intended
- include <arpa/inet.h> in passt.c before kernel headers so that we
can use <netinet/in.h> macros to check IPv6 address types, and
remove a duplicate <linux/ip.h> inclusion
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Avoid a bunch of syscalls on forwarding paths by:
- storing minimum and maximum file descriptor numbers for each
protocol, fall back to SO_PROTOCOL query only on overlaps
- allocating a larger receive buffer -- this can result in more
coalesced packets than sendmmsg() can take (UIO_MAXIOV, i.e. 1024),
so make sure we don't exceed that within a single call to protocol
tap handlers
- nesting the handling loop in tap_handler() in the receive loop,
so that we have better chances of filling our receive buffer in
fewer calls
- skipping the recvfrom() in the UDP handler on EPOLLERR -- there's
nothing to be done in that case
and while at it:
- restore the 20ms timer interval for periodic (TCP) events, I
accidentally changed that to 100ms in an earlier commit
- attempt using SO_ZEROCOPY for UDP -- if it's not available,
sendmmsg() will succeed anyway
- fix the handling of the status code from sendmmsg(), if it fails,
we'll try to discard the first message, hence return 1 from the
UDP handler
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Receive packets in batches from AF_UNIX, check if they can be sent
with a single syscall, and batch them up with sendmmsg() in case.
A bit rudimentary, currently only implemented for UDP, but it seems
to work.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
This implementation, similarly to the IPv4 DHCP one, hands out a
single address, which is the same as the upstream address for the
host.
This avoids the need for address translation as long as the client
runs a DHCPv6 client. The NDP "Managed" flag is now set in Router
Advertisements.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...and mention it in the README.
While at it, remove useless escaping in the README, and fix
indentation in the syslog message with the qemu command line
example.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
...otherwise, both IPv4 and IPv6 are considered disabled, and nothing
works anymore.
While at it, don't fork to background on debug builds, and log to
stderr too in that case.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Default value for /proc/sys/fs/nr_open is 2^20, which is more than
enough: set this hard limit as current (soft) limit on start, and
drop the 'ulimit -n' from the demo script.
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
