[PATCH] Introduce passt-repair

[PATCH] Introduce passt-repair

[Stefano Brivio]

A privileged helper to set/clear TCP_REPAIR on sockets on behalf of
passt. Not used yet.

Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
---
Note: I'm merging this right now, but posting anyway to ease review.

 .gitignore                            |   1 +
 Makefile                              |  17 ++-
 contrib/apparmor/usr.bin.passt-repair |  29 +++++
 contrib/fedora/passt.spec             |   2 +
 contrib/selinux/passt-repair.fc       |  11 ++
 contrib/selinux/passt-repair.te       |  58 ++++++++++
 hooks/pre-push                        |   1 +
 passt-repair.1                        |  70 ++++++++++++
 passt-repair.c                        | 154 ++++++++++++++++++++++++++
 seccomp.sh                            |   6 +-
 10 files changed, 342 insertions(+), 7 deletions(-)
 create mode 100644 contrib/apparmor/usr.bin.passt-repair
 create mode 100644 contrib/selinux/passt-repair.fc
 create mode 100644 contrib/selinux/passt-repair.te
 create mode 100644 passt-repair.1
 create mode 100644 passt-repair.c

diff --git a/.gitignore b/.gitignore
index d1c8be9..5824a71 100644
--- a/.gitignore
+++ b/.gitignore
@@ -3,6 +3,7 @@
 /passt.avx2
 /pasta
 /pasta.avx2
+/passt-repair
 /qrap
 /pasta.1
 /seccomp.h
diff --git a/Makefile b/Makefile
index 464eef1..6ab8d24 100644
--- a/Makefile
+++ b/Makefile
@@ -42,9 +42,10 @@ PASST_SRCS = arch.c arp.c checksum.c conf.c dhcp.c dhcpv6.c flow.c fwd.c \
 	tcp_buf.c tcp_splice.c tcp_vu.c udp.c udp_flow.c udp_vu.c util.c \
 	vhost_user.c virtio.c vu_common.c
 QRAP_SRCS = qrap.c
-SRCS = $(PASST_SRCS) $(QRAP_SRCS)
+PASST_REPAIR_SRCS = passt-repair.c
+SRCS = $(PASST_SRCS) $(QRAP_SRCS) $(PASST_REPAIR_SRCS)
 
-MANPAGES = passt.1 pasta.1 qrap.1
+MANPAGES = passt.1 pasta.1 qrap.1 passt-repair.1
 
 PASST_HEADERS = arch.h arp.h checksum.h conf.h dhcp.h dhcpv6.h flow.h fwd.h \
 	flow_table.h icmp.h icmp_flow.h inany.h iov.h ip.h isolation.h \
@@ -72,9 +73,9 @@ mandir		?= $(datarootdir)/man
 man1dir		?= $(mandir)/man1
 
 ifeq ($(TARGET_ARCH),x86_64)
-BIN := passt passt.avx2 pasta pasta.avx2 qrap
+BIN := passt passt.avx2 pasta pasta.avx2 qrap passt-repair
 else
-BIN := passt pasta qrap
+BIN := passt pasta qrap passt-repair
 endif
 
 all: $(BIN) $(MANPAGES) docs
@@ -83,7 +84,10 @@ static: FLAGS += -static -DGLIBC_NO_STATIC_NSS
 static: clean all
 
 seccomp.h: seccomp.sh $(PASST_SRCS) $(PASST_HEADERS)
-	@ EXTRA_SYSCALLS="$(EXTRA_SYSCALLS)" ARCH="$(TARGET_ARCH)" CC="$(CC)" ./seccomp.sh $(PASST_SRCS) $(PASST_HEADERS)
+	@ EXTRA_SYSCALLS="$(EXTRA_SYSCALLS)" ARCH="$(TARGET_ARCH)" CC="$(CC)" ./seccomp.sh seccomp.h $(PASST_SRCS) $(PASST_HEADERS)
+
+seccomp_repair.h: seccomp.sh $(PASST_REPAIR_SRCS)
+	@ ARCH="$(TARGET_ARCH)" CC="$(CC)" ./seccomp.sh seccomp_repair.h $(PASST_REPAIR_SRCS)
 
 passt: $(PASST_SRCS) $(HEADERS)
 	$(CC) $(FLAGS) $(CFLAGS) $(CPPFLAGS) $(PASST_SRCS) -o passt $(LDFLAGS)
@@ -101,6 +105,9 @@ pasta.avx2 pasta.1 pasta: pasta%: passt%
 qrap: $(QRAP_SRCS) passt.h
 	$(CC) $(FLAGS) $(CFLAGS) $(CPPFLAGS) -DARCH=\"$(TARGET_ARCH)\" $(QRAP_SRCS) -o qrap $(LDFLAGS)
 
+passt-repair: $(PASST_REPAIR_SRCS) seccomp_repair.h
+	$(CC) $(FLAGS) $(CFLAGS) $(CPPFLAGS) $(PASST_REPAIR_SRCS) -o passt-repair $(LDFLAGS)
+
 valgrind: EXTRA_SYSCALLS += rt_sigprocmask rt_sigtimedwait rt_sigaction	\
 			    rt_sigreturn getpid gettid kill clock_gettime mmap \
 			    mmap2 munmap open unlink gettimeofday futex statx \
diff --git a/contrib/apparmor/usr.bin.passt-repair b/contrib/apparmor/usr.bin.passt-repair
new file mode 100644
index 0000000..901189d
--- /dev/null
+++ b/contrib/apparmor/usr.bin.passt-repair
@@ -0,0 +1,29 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+#
+# PASST - Plug A Simple Socket Transport
+#  for qemu/UNIX domain socket mode
+#
+# PASTA - Pack A Subtle Tap Abstraction
+#  for network namespace/tap device mode
+#
+# contrib/apparmor/usr.bin.passt-repair - AppArmor profile for passt-repair(1)
+#
+# Copyright (c) 2025 Red Hat GmbH
+# Author: Stefano Brivio <sbrivio@redhat.com>
+
+abi <abi/3.0>,
+
+#include <tunables/global>
+
+profile passt-repair /usr/bin/passt-repair {
+  #include <abstractions/base>
+  /** rw,			# passt's ".repair" socket might be anywhere
+  unix (connect, receive, send) type=stream,
+
+  capability dac_override,	# connect to passt's socket as root
+  capability net_admin,		# currently needed for TCP_REPAIR socket option
+  capability net_raw,		# what TCP_REPAIR should require instead
+
+  network unix stream,		# connect and use UNIX domain socket
+  network inet stream,		# use TCP sockets
+}
diff --git a/contrib/fedora/passt.spec b/contrib/fedora/passt.spec
index 7950fb9..6a83f8b 100644
--- a/contrib/fedora/passt.spec
+++ b/contrib/fedora/passt.spec
@@ -108,9 +108,11 @@ fi
 %{_bindir}/passt
 %{_bindir}/pasta
 %{_bindir}/qrap
+%{_bindir}/passt-repair
 %{_mandir}/man1/passt.1*
 %{_mandir}/man1/pasta.1*
 %{_mandir}/man1/qrap.1*
+%{_mandir}/man1/passt-repair.1*
 %ifarch x86_64
 %{_bindir}/passt.avx2
 %{_mandir}/man1/passt.avx2.1*
diff --git a/contrib/selinux/passt-repair.fc b/contrib/selinux/passt-repair.fc
new file mode 100644
index 0000000..bcd526e
--- /dev/null
+++ b/contrib/selinux/passt-repair.fc
@@ -0,0 +1,11 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+#
+# PASST - Plug A Simple Socket Transport
+#  for qemu/UNIX domain socket mode
+#
+# contrib/selinux/passt-repair.fc - SELinux: File Context for passt-repair
+#
+# Copyright (c) 2025 Red Hat GmbH
+# Author: Stefano Brivio <sbrivio@redhat.com>
+
+/usr/bin/passt-repair		system_u:object_r:passt_repair_exec_t:s0
diff --git a/contrib/selinux/passt-repair.te b/contrib/selinux/passt-repair.te
new file mode 100644
index 0000000..e3ffbcd
--- /dev/null
+++ b/contrib/selinux/passt-repair.te
@@ -0,0 +1,58 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+#
+# PASST - Plug A Simple Socket Transport
+#  for qemu/UNIX domain socket mode
+#
+# contrib/selinux/passt-repair.te - SELinux: Type Enforcement for passt-repair
+#
+# Copyright (c) 2025 Red Hat GmbH
+# Author: Stefano Brivio <sbrivio@redhat.com>
+
+policy_module(passt-repair, 0.1)
+
+require {
+	type unconfined_t;
+	type passt_t;
+	role unconfined_r;
+	class process transition;
+
+	class file { read execute execute_no_trans entrypoint open map };
+	class capability { dac_override net_admin net_raw };
+	class chr_file { append open getattr read write ioctl };
+
+	class unix_stream_socket { create connect sendto };
+	class sock_file { read write };
+
+	class tcp_socket { read setopt write };
+
+	type console_device_t;
+	type user_devpts_t;
+	type user_tmp_t;
+}
+
+type passt_repair_t;
+domain_type(passt_repair_t);
+type passt_repair_exec_t;
+files_type(passt_repair_exec_t);
+
+role unconfined_r types passt_repair_t;
+
+allow passt_repair_t passt_repair_exec_t:file { read execute execute_no_trans entrypoint open map };
+type_transition unconfined_t passt_repair_exec_t:process passt_repair_t;
+allow unconfined_t passt_repair_t:process transition;
+
+allow passt_repair_t self:capability { dac_override net_admin net_raw };
+
+allow passt_repair_t console_device_t:chr_file { append open getattr read write ioctl };
+allow passt_repair_t user_devpts_t:chr_file { append open getattr read write ioctl };
+
+allow passt_repair_t unconfined_t:unix_stream_socket { connectto read write };
+allow passt_repair_t passt_t:unix_stream_socket { connectto read write };
+allow passt_repair_t user_tmp_t:unix_stream_socket { connectto read write };
+
+allow passt_repair_t unconfined_t:sock_file { read write };
+allow passt_repair_t passt_t:sock_file { read write };
+allow passt_repair_t user_tmp_t:sock_file { read write };
+
+allow passt_repair_t unconfined_t:tcp_socket { read setopt write };
+allow passt_repair_t passt_t:tcp_socket { read setopt write };
diff --git a/hooks/pre-push b/hooks/pre-push
index 33a2052..8dbfa5f 100755
--- a/hooks/pre-push
+++ b/hooks/pre-push
@@ -56,6 +56,7 @@ cd ..
 make pkgs
 scp passt passt.avx2 passt.1 qrap qrap.1	"${USER_HOST}:${BIN}"
 scp pasta pasta.avx2 pasta.1			"${USER_HOST}:${BIN}"
+scp passt-repair passt-repair.1			"${USER_HOST}:${BIN}"
 
 ssh "${USER_HOST}" 				"rm -f ${BIN}/*.deb"
 ssh "${USER_HOST}"				"rm -f ${BIN}/*.rpm"
diff --git a/passt-repair.1 b/passt-repair.1
new file mode 100644
index 0000000..8d07c97
--- /dev/null
+++ b/passt-repair.1
@@ -0,0 +1,70 @@
+.\" SPDX-License-Identifier: GPL-2.0-or-later
+.\" Copyright (c) 2025 Red Hat GmbH
+.\" Author: Stefano Brivio <sbrivio@redhat.com>
+.TH passt-repair 1
+
+.SH NAME
+.B passt-repair
+\- Helper setting TCP_REPAIR socket options for \fBpasst\fR(1)
+
+.SH SYNOPSIS
+.B passt-repair
+\fIPATH\fR
+
+.SH DESCRIPTION
+
+.B passt-repair
+is a privileged helper setting and clearing repair mode on TCP sockets on behalf
+of \fBpasst\fR(1), as instructed via single-byte commands over a UNIX domain
+socket, specified by \fIPATH\fR.
+
+It can be used to migrate TCP connections between guests without granting
+additional capabilities to \fBpasst\fR(1) itself: to migrate TCP connections,
+\fBpasst\fR(1) leverages repair mode, which needs the \fBCAP_NET_ADMIN\fR
+capability (see \fBcapabilities\fR(7)) to be set or cleared.
+
+.SH PROTOCOL
+
+\fBpasst-repair\fR(1) connects to \fBpasst\fR(1) using the socket specified via
+\fI--repair-path\fR option in \fBpasst\fR(1) itself. By default, the name is the
+same as the UNIX domain socket used for guest communication, suffixed by
+\fI.repair\fR.
+
+The messages consist of one 8-bit signed integer that can be \fITCP_REPAIR_ON\fR
+(1), \fITCP_REPAIR_OFF\fR (2), or \fITCP_REPAIR_OFF_WP\fR (-1), as defined by
+the Linux kernel user API, and one to SCM_MAX_FD (253) sockets as SCM_RIGHTS
+(see \fBunix\fR(7)) ancillary message, sent by the server, \fBpasst\fR(1).
+
+The client, \fBpasst-repair\fR(1), replies with the same byte (and no ancillary
+message) to indicate success, and closes the connection on failure.
+
+The server closes the connection on error or completion.
+
+.SH NOTES
+
+\fBpasst-repair\fR(1) can be granted the \fBCAP_NET_ADMIN\fR capability
+(preferred, as it limits privileges to the strictly necessary ones), or it can
+be run as root.
+
+.SH AUTHOR
+
+Stefano Brivio <sbrivio@redhat.com>.
+
+.SH REPORTING BUGS
+
+Please report issues on the bug tracker at https://bugs.passt.top/, or
+send a message to the passt-user@passt.top mailing list, see
+https://lists.passt.top/.
+
+.SH COPYRIGHT
+
+Copyright (c) 2025 Red Hat GmbH.
+
+\fBpasst-repair\fR is free software: you can redistribute them and/or modify
+them under the terms of the GNU General Public License as published by the Free
+Software Foundation, either version 2 of the License, or (at your option) any
+later version. 
+
+.SH SEE ALSO
+
+\fBpasst\fR(1), \fBqemu\fR(1), \fBcapabilities\fR(7), \fBunix\fR(7).
diff --git a/passt-repair.c b/passt-repair.c
new file mode 100644
index 0000000..767a821
--- /dev/null
+++ b/passt-repair.c
@@ -0,0 +1,154 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+/* PASST - Plug A Simple Socket Transport
+ *  for qemu/UNIX domain socket mode
+ *
+ * PASTA - Pack A Subtle Tap Abstraction
+ *  for network namespace/tap device mode
+ *
+ * passt-repair.c - Privileged helper to set/clear TCP_REPAIR on sockets
+ *
+ * Copyright (c) 2025 Red Hat GmbH
+ * Author: Stefano Brivio <sbrivio@redhat.com>
+ *
+ * Connect to passt via UNIX domain socket, receive sockets via SCM_RIGHTS along
+ * with byte commands mapping to TCP_REPAIR values, and switch repair mode on or
+ * off. Reply by echoing the command. Exit on EOF.
+ */
+
+#include <sys/prctl.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <sys/un.h>
+#include <errno.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <limits.h>
+#include <unistd.h>
+#include <netdb.h>
+
+#include <netinet/tcp.h>
+
+#include <linux/audit.h>
+#include <linux/capability.h>
+#include <linux/filter.h>
+#include <linux/seccomp.h>
+
+#include "seccomp_repair.h"
+
+#define SCM_MAX_FD 253 /* From Linux kernel (include/net/scm.h), not in UAPI */
+
+/**
+ * main() - Entry point and whole program with loop
+ * @argc:	Argument count, must be 2
+ * @argv:	Argument: path of UNIX domain socket to connect to
+ *
+ * Return: 0 on success (EOF), 1 on error, 2 on usage error
+ *
+ * #syscalls:repair connect setsockopt write exit_group
+ * #syscalls:repair socket s390x:socketcall i686:socketcall
+ * #syscalls:repair recvfrom recvmsg arm:recv ppc64le:recv
+ * #syscalls:repair sendto sendmsg arm:send ppc64le:send
+ */
+int main(int argc, char **argv)
+{
+	char buf[CMSG_SPACE(sizeof(int) * SCM_MAX_FD)]
+	     __attribute__ ((aligned(__alignof__(struct cmsghdr))));
+	struct sockaddr_un a = { AF_UNIX, "" };
+	int fds[SCM_MAX_FD], s, ret, i, n;
+	struct sock_fprog prog;
+	int8_t cmd = INT8_MAX;
+	struct cmsghdr *cmsg;
+	struct msghdr msg;
+	struct iovec iov;
+
+	prctl(PR_SET_DUMPABLE, 0);
+
+	prog.len = (unsigned short)sizeof(filter_repair) /
+				   sizeof(filter_repair[0]);
+	prog.filter = filter_repair;
+	if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0) ||
+	    prctl(PR_SET_SECCOMP, SECCOMP_MODE_FILTER, &prog)) {
+		fprintf(stderr, "Failed to apply seccomp filter");
+		return 1;
+	}
+
+	iov = (struct iovec){ &cmd, sizeof(cmd) };
+	msg = (struct msghdr){ NULL, 0, &iov, 1, buf, sizeof(buf), 0 };
+	cmsg = CMSG_FIRSTHDR(&msg);
+
+	if (argc != 2) {
+		fprintf(stderr, "Usage: %s PATH\n", argv[0]);
+		return 2;
+	}
+
+	ret = snprintf(a.sun_path, sizeof(a.sun_path), "%s", argv[1]);
+	if (ret <= 0 || ret >= (int)sizeof(a.sun_path)) {
+		fprintf(stderr, "Invalid socket path: %s\n", argv[1]);
+		return 2;
+	}
+
+	if ((s = socket(AF_UNIX, SOCK_STREAM, 0)) < 0) {
+		perror("Failed to create AF_UNIX socket");
+		return 1;
+	}
+
+	if (connect(s, (struct sockaddr *)&a, sizeof(a))) {
+		fprintf(stderr, "Failed to connect to %s: %s\n", argv[1],
+			strerror(errno));
+		return 1;
+	}
+
+loop:
+	ret = recvmsg(s, &msg, 0);
+	if (ret < 0) {
+		perror("Failed to receive message");
+		return 1;
+	}
+
+	if (!ret)	/* Done */
+		return 0;
+
+	if (!cmsg ||
+	    cmsg->cmsg_len < CMSG_LEN(sizeof(int)) ||
+	    cmsg->cmsg_len > CMSG_LEN(sizeof(int) * SCM_MAX_FD) ||
+	    cmsg->cmsg_type != SCM_RIGHTS) {
+		fprintf(stderr, "No/bad ancillary data from peer\n");
+		return 1;
+	}
+
+	n = cmsg->cmsg_len / CMSG_LEN(sizeof(int));
+	memcpy(fds, CMSG_DATA(cmsg), sizeof(int) * n);
+
+	if (cmd != TCP_REPAIR_ON && cmd != TCP_REPAIR_OFF &&
+	    cmd != TCP_REPAIR_OFF_NO_WP) {
+		fprintf(stderr, "Unsupported command 0x%04x\n", cmd);
+		return 1;
+	}
+
+	for (i = 0; i < n; i++) {
+		int o = cmd;
+
+		if (setsockopt(fds[i], SOL_TCP, TCP_REPAIR, &o, sizeof(o))) {
+			fprintf(stderr,
+				"Setting TCP_REPAIR to %i on socket %i: %s", o,
+				fds[i], strerror(errno));
+			return 1;
+		}
+
+		/* Close _our_ copy */
+		close(fds[i]);
+
+		/* Confirm setting by echoing the command back */
+		if (send(s, &cmd, sizeof(cmd), 0) < 0) {
+			fprintf(stderr, "Reply to command %i: %s\n",
+				o, strerror(errno));
+			return 1;
+		}
+	}
+
+	goto loop;
+
+	return 0;
+}
diff --git a/seccomp.sh b/seccomp.sh
index 6499c58..4c521ae 100755
--- a/seccomp.sh
+++ b/seccomp.sh
@@ -14,8 +14,10 @@
 # Author: Stefano Brivio <sbrivio@redhat.com>
 
 TMP="$(mktemp)"
-IN="$@"
 OUT="$(mktemp)"
+OUT_FINAL="${1}"
+shift
+IN="$@"
 
 [ -z "${ARCH}" ] && ARCH="$(uname -m)"
 [ -z "${CC}" ] && CC="cc"
@@ -268,4 +270,4 @@ for __p in ${__profiles}; do
 	gen_profile "${__p}" ${__calls}
 done
 
-mv "${OUT}" seccomp.h
+mv "${OUT}" "${OUT_FINAL}"
-- 
@@ -14,8 +14,10 @@
 # Author: Stefano Brivio <sbrivio@redhat.com>
 
 TMP="$(mktemp)"
-IN="$@"
 OUT="$(mktemp)"
+OUT_FINAL="${1}"
+shift
+IN="$@"
 
 [ -z "${ARCH}" ] && ARCH="$(uname -m)"
 [ -z "${CC}" ] && CC="cc"
@@ -268,4 +270,4 @@ for __p in ${__profiles}; do
 	gen_profile "${__p}" ${__calls}
 done
 
-mv "${OUT}" seccomp.h
+mv "${OUT}" "${OUT_FINAL}"
-- 
2.43.0

Re: [PATCH] Introduce passt-repair

[David Gibson]

[-- Attachment #1: Type: text/plain, Size: 1039 bytes --]

On Tue, Feb 04, 2025 at 01:05:14AM +0100, Stefano Brivio wrote:
> A privileged helper to set/clear TCP_REPAIR on sockets on behalf of
> passt. Not used yet.
> 
> Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
> ---
> Note: I'm merging this right now, but posting anyway to ease review.

[snip]
> +/**
> + * main() - Entry point and whole program with loop
> + * @argc:	Argument count, must be 2
> + * @argv:	Argument: path of UNIX domain socket to connect to
> + *
> + * Return: 0 on success (EOF), 1 on error, 2 on usage error
> + *
> + * #syscalls:repair connect setsockopt write exit_group
> + * #syscalls:repair socket s390x:socketcall i686:socketcall
> + * #syscalls:repair recvfrom recvmsg arm:recv ppc64le:recv
> + * #syscalls:repair sendto sendmsg arm:send ppc64le:send
> + */

You are missing close() :(.

-- 
David Gibson (he or they)	| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you, not the other way
				| around.
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]