Index: head/UPDATING =================================================================== --- head/UPDATING (revision 287221) +++ head/UPDATING (revision 287222) @@ -1,1383 +1,1388 @@ Updating Information for FreeBSD current users. This file is maintained and copyrighted by M. Warner Losh . See end of file for further details. For commonly done items, please see the COMMON ITEMS: section later in the file. These instructions assume that you basically know what you are doing. If not, then please consult the FreeBSD handbook: http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/makeworld.html Items affecting the ports and packages system can be found in /usr/ports/UPDATING. Please read that file before running portupgrade. NOTE: FreeBSD has switched from gcc to clang. If you have trouble bootstrapping from older versions of FreeBSD, try WITHOUT_CLANG and WITH_GCC to bootstrap to the tip of head, and then rebuild without this option. The bootstrap process from older version of current across the gcc/clang cutover is a bit fragile. NOTE TO PEOPLE WHO THINK THAT FreeBSD 11.x IS SLOW: FreeBSD 11.x has many debugging features turned on, in both the kernel and userland. These features attempt to detect incorrect use of system primitives, and encourage loud failure through extra sanity checking and fail stop semantics. They also substantially impact system performance. If you want to do performance measurement, benchmarking, and optimization, you'll want to turn them off. This includes various WITNESS- related kernel options, INVARIANTS, malloc debugging flags in userland, and various verbose features in the kernel. Many developers choose to disable these features on build machines to maximize performance. (To completely disable malloc debugging, define MALLOC_PRODUCTION in /etc/make.conf, or to merely disable the most expensive debugging functionality run "ln -s 'abort:false,junk:false' /etc/malloc.conf".) +20150827: + pf no longer supports 'scrub fragment crop' or 'scrub fragment drop-ovl' + These configurations are now automatically interpreted as + 'scrub fragment reassemble'. + 20150817: Kernel-loadable modules for the random(4) device are back. To use them, the kernel must have device random options RANDOM_LOADABLE kldload(8) can then be used to load random_fortuna.ko or random_yarrow.ko. Please note that due to the indirect function calls that the loadable modules need to provide, the build-in variants will be slightly more efficient. The random(4) kernel option RANDOM_DUMMY has been retired due to unpopularity. It was not all that useful anyway. 20150813: The WITHOUT_ELFTOOLCHAIN_TOOLS src.conf(5) knob has been retired. Control over building the ELF Tool Chain tools is now provided by the WITHOUT_TOOLCHAIN knob. 20150810: The polarity of Pulse Per Second (PPS) capture events with the uart(4) driver has been corrected. Prior to this change the PPS "assert" event corresponded to the trailing edge of a positive PPS pulse and the "clear" event was the leading edge of the next pulse. As the width of a PPS pulse in a typical GPS receiver is on the order of 1 millisecond, most users will not notice any significant difference with this change. Anyone who has compensated for the historical polarity reversal by configuring a negative offset equal to the pulse width will need to remove that workaround. 20150809: The default group assigned to /dev/dri entries has been changed from 'wheel' to 'video' with the id of '44'. If you want to have access to the dri devices please add yourself to the video group with: # pw groupmod video -m $USER 20150806: The menu.rc and loader.rc files will now be replaced during upgrades. Please migrate local changes to menu.rc.local and loader.rc.local instead. 20150805: GNU Binutils versions of addr2line, c++filt, nm, readelf, size, strings and strip have been removed. The src.conf(5) knob WITHOUT_ELFTOOLCHAIN_TOOLS no longer provides the binutils tools. 20150728: As ZFS requires more kernel stack pages than is the default on some architectures e.g. i386, it now warns if KSTACK_PAGES is less than ZFS_MIN_KSTACK_PAGES (which is 4 at the time of writing). Please consider using 'options KSTACK_PAGES=X' where X is greater than or equal to ZFS_MIN_KSTACK_PAGES i.e. 4 in such configurations. 20150706: sendmail has been updated to 8.15.2. Starting with FreeBSD 11.0 and sendmail 8.15, sendmail uses uncompressed IPv6 addresses by default, i.e., they will not contain "::". For example, instead of ::1, it will be 0:0:0:0:0:0:0:1. This permits a zero subnet to have a more specific match, such as different map entries for IPv6:0:0 vs IPv6:0. This change requires that configuration data (including maps, files, classes, custom ruleset, etc.) must use the same format, so make certain such configuration data is upgrading. As a very simple check search for patterns like 'IPv6:[0-9a-fA-F:]*::' and 'IPv6::'. To return to the old behavior, set the m4 option confUSE_COMPRESSED_IPV6_ADDRESSES or the cf option UseCompressedIPv6Addresses. 20150630: The default kernel entropy-processing algorithm is now Fortuna, replacing Yarrow. Assuming you have 'device random' in your kernel config file, the configurations allow a kernel option to override this default. You may choose *ONE* of: options RANDOM_YARROW # Legacy /dev/random algorithm. options RANDOM_DUMMY # Blocking-only driver. If you have neither, you get Fortuna. For most people, read no further, Fortuna will give a /dev/random that works like it always used to, and the difference will be irrelevant. If you remove 'device random', you get *NO* kernel-processed entopy at all. This may be acceptable to folks building embedded systems, but has complications. Carry on reading, and it is assumed you know what you need. *PLEASE* read random(4) and random(9) if you are in the habit of tweeking kernel configs, and/or if you are a member of the embedded community, wanting specific and not-usual behaviour from your security subsystems. NOTE!! If you use RANDOM_DUMMY and/or have no 'device random', you will NOT have a functioning /dev/random, and many cryptographic features will not work, including SSH. You may also find strange behaviour from the random(3) set of library functions, in particular sranddev(3), srandomdev(3) and arc4random(3). The reason for this is that the KERN_ARND sysctl only returns entropy if it thinks it has some to share, and with RANDOM_DUMMY or no 'device random' this will never happen. 20150623: An additional fix for the issue described in the 20150614 sendmail entry below has been been committed in revision 284717. 20150616: FreeBSD's old make (fmake) has been removed from the system. It is available as the devel/fmake port or via pkg install fmake. 20150615: The fix for the issue described in the 20150614 sendmail entry below has been been committed in revision 284436. The work around described in that entry is no longer needed unless the default setting is overridden by a confDH_PARAMETERS configuration setting of '5' or pointing to a 512 bit DH parameter file. 20150614: ALLOW_DEPRECATED_ATF_TOOLS/ATFFILE support has been removed from atf.test.mk (included from bsd.test.mk). Please upgrade devel/atf and devel/kyua to version 0.20+ and adjust any calling code to work with Kyuafile and kyua. 20150614: The import of openssl to address the FreeBSD-SA-15:10.openssl security advisory includes a change which rejects handshakes with DH parameters below 768 bits. sendmail releases prior to 8.15.2 (not yet released), defaulted to a 512 bit DH parameter setting for client connections. To work around this interoperability, sendmail can be configured to use a 2048 bit DH parameter by: 1. Edit /etc/mail/`hostname`.mc 2. If a setting for confDH_PARAMETERS does not exist or exists and is set to a string beginning with '5', replace it with '2'. 3. If a setting for confDH_PARAMETERS exists and is set to a file path, create a new file with: openssl dhparam -out /path/to/file 2048 4. Rebuild the .cf file: cd /etc/mail/; make; make install 5. Restart sendmail: cd /etc/mail/; make restart A sendmail patch is coming, at which time this file will be updated. 20150604: Generation of legacy formatted entries have been disabled by default in pwd_mkdb(8), as all base system consumers of the legacy formatted entries were converted to use the new format by default when the new, machine independent format have been added and supported since FreeBSD 5.x. Please see the pwd_mkdb(8) manual page for further details. 20150525: Clang and llvm have been upgraded to 3.6.1 release. Please see the 20141231 entry below for information about prerequisites and upgrading, if you are not already using 3.5.0 or higher. 20150521: TI platform code switched to using vendor DTS files and this update may break existing systems running on Beaglebone, Beaglebone Black, and Pandaboard: - dtb files should be regenerated/reinstalled. Filenames are the same but content is different now - GPIO addressing was changed, now each GPIO bank (32 pins per bank) has its own /dev/gpiocX device, e.g. pin 121 on /dev/gpioc0 in old addressing scheme is now pin 25 on /dev/gpioc3. - Pandaboard: /etc/ttys should be updated, serial console device is now /dev/ttyu2, not /dev/ttyu0 20150501: soelim(1) from gnu/usr.bin/groff has been replaced by usr.bin/soelim. If you need the GNU extension from groff soelim(1), install groff from package: pkg install groff, or via ports: textproc/groff. 20150423: chmod, chflags, chown and chgrp now affect symlinks in -R mode as defined in symlink(7); previously symlinks were silently ignored. 20150415: The const qualifier has been removed from iconv(3) to comply with POSIX. The ports tree is aware of this from r384038 onwards. 20150416: Libraries specified by LIBADD in Makefiles must have a corresponding DPADD_ variable to ensure correct dependencies. This is now enforced in src.libnames.mk. 20150324: From legacy ata(4) driver was removed support for SATA controllers supported by more functional drivers ahci(4), siis(4) and mvs(4). Kernel modules ataahci and ataadaptec were removed completely, replaced by ahci and mvs modules respectively. 20150315: Clang, llvm and lldb have been upgraded to 3.6.0 release. Please see the 20141231 entry below for information about prerequisites and upgrading, if you are not already using 3.5.0 or higher. 20150307: The 32-bit PowerPC kernel has been changed to a position-independent executable. This can only be booted with a version of loader(8) newer than January 31, 2015, so make sure to update both world and kernel before rebooting. 20150217: If you are running a -CURRENT kernel since r273872 (Oct 30th, 2014), but before r278950, the RNG was not seeded properly. Immediately upgrade the kernel to r278950 or later and regenerate any keys (e.g. ssh keys or openssl keys) that were generated w/ a kernel from that range. This does not affect programs that directly used /dev/random or /dev/urandom. All userland uses of arc4random(3) are affected. 20150210: The autofs(4) ABI was changed in order to restore binary compatibility with 10.1-RELEASE. The automountd(8) daemon needs to be rebuilt to work with the new kernel. 20150131: The powerpc64 kernel has been changed to a position-independent executable. This can only be booted with a new version of loader(8), so make sure to update both world and kernel before rebooting. 20150118: Clang and llvm have been upgraded to 3.5.1 release. This is a bugfix only release, no new features have been added. Please see the 20141231 entry below for information about prerequisites and upgrading, if you are not already using 3.5.0. 20150107: ELF tools addr2line, elfcopy (strip), nm, size, and strings are now taken from the ELF Tool Chain project rather than GNU binutils. They should be drop-in replacements, with the addition of arm64 support. The WITHOUT_ELFTOOLCHAIN_TOOLS= knob may be used to obtain the binutils tools, if necessary. See 20150805 for updated information. 20150105: The default Unbound configuration now enables remote control using a local socket. Users who have already enabled the local_unbound service should regenerate their configuration by running "service local_unbound setup" as root. 20150102: The GNU texinfo and GNU info pages have been removed. To be able to view GNU info pages please install texinfo from ports. 20141231: Clang, llvm and lldb have been upgraded to 3.5.0 release. As of this release, a prerequisite for building clang, llvm and lldb is a C++11 capable compiler and C++11 standard library. This means that to be able to successfully build the cross-tools stage of buildworld, with clang as the bootstrap compiler, your system compiler or cross compiler should either be clang 3.3 or later, or gcc 4.8 or later, and your system C++ library should be libc++, or libdstdc++ from gcc 4.8 or later. On any standard FreeBSD 10.x or 11.x installation, where clang and libc++ are on by default (that is, on x86 or arm), this should work out of the box. On 9.x installations where clang is enabled by default, e.g. on x86 and powerpc, libc++ will not be enabled by default, so libc++ should be built (with clang) and installed first. If both clang and libc++ are missing, build clang first, then use it to build libc++. On 8.x and earlier installations, upgrade to 9.x first, and then follow the instructions for 9.x above. Sparc64 and mips users are unaffected, as they still use gcc 4.2.1 by default, and do not build clang. Many embedded systems are resource constrained, and will not be able to build clang in a reasonable time, or in some cases at all. In those cases, cross building bootable systems on amd64 is a workaround. This new version of clang introduces a number of new warnings, of which the following are most likely to appear: -Wabsolute-value This warns in two cases, for both C and C++: * When the code is trying to take the absolute value of an unsigned quantity, which is effectively a no-op, and almost never what was intended. The code should be fixed, if at all possible. If you are sure that the unsigned quantity can be safely cast to signed, without loss of information or undefined behavior, you can add an explicit cast, or disable the warning. * When the code is trying to take an absolute value, but the called abs() variant is for the wrong type, which can lead to truncation. If you want to disable the warning instead of fixing the code, please make sure that truncation will not occur, or it might lead to unwanted side-effects. -Wtautological-undefined-compare and -Wundefined-bool-conversion These warn when C++ code is trying to compare 'this' against NULL, while 'this' should never be NULL in well-defined C++ code. However, there is some legacy (pre C++11) code out there, which actively abuses this feature, which was less strictly defined in previous C++ versions. Squid and openjdk do this, for example. The warning can be turned off for C++98 and earlier, but compiling the code in C++11 mode might result in unexpected behavior; for example, the parts of the program that are unreachable could be optimized away. 20141222: The old NFS client and server (kernel options NFSCLIENT, NFSSERVER) kernel sources have been removed. The .h files remain, since some utilities include them. This will need to be fixed later. If "mount -t oldnfs ..." is attempted, it will fail. If the "-o" option on mountd(8), nfsd(8) or nfsstat(1) is used, the utilities will report errors. 20141121: The handling of LOCAL_LIB_DIRS has been altered to skip addition of directories to top level SUBDIR variable when their parent directory is included in LOCAL_DIRS. Users with build systems with such hierarchies and without SUBDIR entries in the parent directory Makefiles should add them or add the directories to LOCAL_DIRS. 20141109: faith(4) and faithd(8) have been removed from the base system. Faith has been obsolete for a very long time. 20141104: vt(4), the new console driver, is enabled by default. It brings support for Unicode and double-width characters, as well as support for UEFI and integration with the KMS kernel video drivers. You may need to update your console settings in /etc/rc.conf, most probably the keymap. During boot, /etc/rc.d/syscons will indicate what you need to do. vt(4) still has issues and lacks some features compared to syscons(4). See the wiki for up-to-date information: https://wiki.freebsd.org/Newcons If you want to keep using syscons(4), you can do so by adding the following line to /boot/loader.conf: kern.vty=sc 20141102: pjdfstest has been integrated into kyua as an opt-in test suite. Please see share/doc/pjdfstest/README for more details on how to execute it. 20141009: gperf has been removed from the base system for architectures that use clang. Ports that require gperf will obtain it from the devel/gperf port. 20140923: pjdfstest has been moved from tools/regression/pjdfstest to contrib/pjdfstest . 20140922: At svn r271982, The default linux compat kernel ABI has been adjusted to 2.6.18 in support of the linux-c6 compat ports infrastructure update. If you wish to continue using the linux-f10 compat ports, add compat.linux.osrelease=2.6.16 to your local sysctl.conf. Users are encouraged to update their linux-compat packages to linux-c6 during their next update cycle. 20140729: The ofwfb driver, used to provide a graphics console on PowerPC when using vt(4), no longer allows mmap() of all physical memory. This will prevent Xorg on PowerPC with some ATI graphics cards from initializing properly unless x11-servers/xorg-server is updated to 1.12.4_8 or newer. 20140723: The xdev targets have been converted to using TARGET and TARGET_ARCH instead of XDEV and XDEV_ARCH. 20140719: The default unbound configuration has been modified to address issues with reverse lookups on networks that use private address ranges. If you use the local_unbound service, run "service local_unbound setup" as root to regenerate your configuration, then "service local_unbound reload" to load the new configuration. 20140709: The GNU texinfo and GNU info pages are not built and installed anymore, WITH_INFO knob has been added to allow to built and install them again. UPDATE: see 20150102 entry on texinfo's removal 20140708: The GNU readline library is now an INTERNALLIB - that is, it is statically linked into consumers (GDB and variants) in the base system, and the shared library is no longer installed. The devel/readline port is available for third party software that requires readline. 20140702: The Itanium architecture (ia64) has been removed from the list of known architectures. This is the first step in the removal of the architecture. 20140701: Commit r268115 has added NFSv4.1 server support, merged from projects/nfsv4.1-server. Since this includes changes to the internal interfaces between the NFS related modules, a full build of the kernel and modules will be necessary. __FreeBSD_version has been bumped. 20140629: The WITHOUT_VT_SUPPORT kernel config knob has been renamed WITHOUT_VT. (The other _SUPPORT knobs have a consistent meaning which differs from the behaviour controlled by this knob.) 20140619: Maximal length of the serial number in CTL was increased from 16 to 64 chars, that breaks ABI. All CTL-related tools, such as ctladm and ctld, need to be rebuilt to work with a new kernel. 20140606: The libatf-c and libatf-c++ major versions were downgraded to 0 and 1 respectively to match the upstream numbers. They were out of sync because, when they were originally added to FreeBSD, the upstream versions were not respected. These libraries are private and not yet built by default, so renumbering them should be a non-issue. However, unclean source trees will yield broken test programs once the operator executes "make delete-old-libs" after a "make installworld". Additionally, the atf-sh binary was made private by moving it into /usr/libexec/. Already-built shell test programs will keep the path to the old binary so they will break after "make delete-old" is run. If you are using WITH_TESTS=yes (not the default), wipe the object tree and rebuild from scratch to prevent spurious test failures. This is only needed once: the misnumbered libraries and misplaced binaries have been added to OptionalObsoleteFiles.inc so they will be removed during a clean upgrade. 20140512: Clang and llvm have been upgraded to 3.4.1 release. 20140508: We bogusly installed src.opts.mk in /usr/share/mk. This file should be removed to avoid issues in the future (and has been added to ObsoleteFiles.inc). 20140505: /etc/src.conf now affects only builds of the FreeBSD src tree. In the past, it affected all builds that used the bsd.*.mk files. The old behavior was a bug, but people may have relied upon it. To get this behavior back, you can .include /etc/src.conf from /etc/make.conf (which is still global and isn't changed). This also changes the behavior of incremental builds inside the tree of individual directories. Set MAKESYSPATH to ".../share/mk" to do that. Although this has survived make universe and some upgrade scenarios, other upgrade scenarios may have broken. At least one form of temporary breakage was fixed with MAKESYSPATH settings for buildworld as well... In cases where MAKESYSPATH isn't working with this setting, you'll need to set it to the full path to your tree. One side effect of all this cleaning up is that bsd.compiler.mk is no longer implicitly included by bsd.own.mk. If you wish to use COMPILER_TYPE, you must now explicitly include bsd.compiler.mk as well. 20140430: The lindev device has been removed since /dev/full has been made a standard device. __FreeBSD_version has been bumped. 20140424: The knob WITHOUT_VI was added to the base system, which controls building ex(1), vi(1), etc. Older releases of FreeBSD required ex(1) in order to reorder files share/termcap and didn't build ex(1) as a build tool, so building/installing with WITH_VI is highly advised for build hosts for older releases. This issue has been fixed in stable/9 and stable/10 in r277022 and r276991, respectively. 20140418: The YES_HESIOD knob has been removed. It has been obsolete for a decade. Please move to using WITH_HESIOD instead or your builds will silently lack HESIOD. 20140405: The uart(4) driver has been changed with respect to its handling of the low-level console. Previously the uart(4) driver prevented any process from changing the baudrate or the CLOCAL and HUPCL control flags. By removing the restrictions, operators can make changes to the serial console port without having to reboot. However, when getty(8) is started on the serial device that is associated with the low-level console, a misconfigured terminal line in /etc/ttys will now have a real impact. Before upgrading the kernel, make sure that /etc/ttys has the serial console device configured as 3wire without baudrate to preserve the previous behaviour. E.g: ttyu0 "/usr/libexec/getty 3wire" vt100 on secure 20140306: Support for libwrap (TCP wrappers) in rpcbind was disabled by default to improve performance. To re-enable it, if needed, run rpcbind with command line option -W. 20140226: Switched back to the GPL dtc compiler due to updates in the upstream dts files not being supported by the BSDL dtc compiler. You will need to rebuild your kernel toolchain to pick up the new compiler. Core dumps may result while building dtb files during a kernel build if you fail to do so. Set WITHOUT_GPL_DTC if you require the BSDL compiler. 20140216: Clang and llvm have been upgraded to 3.4 release. 20140216: The nve(4) driver has been removed. Please use the nfe(4) driver for NVIDIA nForce MCP Ethernet adapters instead. 20140212: An ABI incompatibility crept into the libc++ 3.4 import in r261283. This could cause certain C++ applications using shared libraries built against the previous version of libc++ to crash. The incompatibility has now been fixed, but any C++ applications or shared libraries built between r261283 and r261801 should be recompiled. 20140204: OpenSSH will now ignore errors caused by kernel lacking of Capsicum capability mode support. Please note that enabling the feature in kernel is still highly recommended. 20140131: OpenSSH is now built with sandbox support, and will use sandbox as the default privilege separation method. This requires Capsicum capability mode support in kernel. 20140128: The libelf and libdwarf libraries have been updated to newer versions from upstream. Shared library version numbers for these two libraries were bumped. Any ports or binaries requiring these two libraries should be recompiled. __FreeBSD_version is bumped to 1100006. 20140110: If a Makefile in a tests/ directory was auto-generating a Kyuafile instead of providing an explicit one, this would prevent such Makefile from providing its own Kyuafile in the future during NO_CLEAN builds. This has been fixed in the Makefiles but manual intervention is needed to clean an objdir if you use NO_CLEAN: # find /usr/obj -name Kyuafile | xargs rm -f 20131213: The behavior of gss_pseudo_random() for the krb5 mechanism has changed, for applications requesting a longer random string than produced by the underlying enctype's pseudo-random() function. In particular, the random string produced from a session key of enctype aes256-cts-hmac-sha1-96 or aes256-cts-hmac-sha1-96 will be different at the 17th octet and later, after this change. The counter used in the PRF+ construction is now encoded as a big-endian integer in accordance with RFC 4402. __FreeBSD_version is bumped to 1100004. 20131108: The WITHOUT_ATF build knob has been removed and its functionality has been subsumed into the more generic WITHOUT_TESTS. If you were using the former to disable the build of the ATF libraries, you should change your settings to use the latter. 20131025: The default version of mtree is nmtree which is obtained from NetBSD. The output is generally the same, but may vary slightly. If you found you need identical output adding "-F freebsd9" to the command line should do the trick. For the time being, the old mtree is available as fmtree. 20131014: libbsdyml has been renamed to libyaml and moved to /usr/lib/private. This will break ports-mgmt/pkg. Rebuild the port, or upgrade to pkg 1.1.4_8 and verify bsdyml not linked in, before running "make delete-old-libs": # make -C /usr/ports/ports-mgmt/pkg build deinstall install clean or # pkg install pkg; ldd /usr/local/sbin/pkg | grep bsdyml 20131010: The stable/10 branch has been created in subversion from head revision r256279. 20131010: The rc.d/jail script has been updated to support jail(8) configuration file. The "jail__*" rc.conf(5) variables for per-jail configuration are automatically converted to /var/run/jail..conf before the jail(8) utility is invoked. This is transparently backward compatible. See below about some incompatibilities and rc.conf(5) manual page for more details. These variables are now deprecated in favor of jail(8) configuration file. One can use "rc.d/jail config " command to generate a jail(8) configuration file in /var/run/jail..conf without running the jail(8) utility. The default pathname of the configuration file is /etc/jail.conf and can be specified by using $jail_conf or $jail__conf variables. Please note that jail_devfs_ruleset accepts an integer at this moment. Please consider to rewrite the ruleset name with an integer. 20130930: BIND has been removed from the base system. If all you need is a local resolver, simply enable and start the local_unbound service instead. Otherwise, several versions of BIND are available in the ports tree. The dns/bind99 port is one example. With this change, nslookup(1) and dig(1) are no longer in the base system. Users should instead use host(1) and drill(1) which are in the base system. Alternatively, nslookup and dig can be obtained by installing the dns/bind-tools port. 20130916: With the addition of unbound(8), a new unbound user is now required during installworld. "mergemaster -p" can be used to add the user prior to installworld, as documented in the handbook. 20130911: OpenSSH is now built with DNSSEC support, and will by default silently trust signed SSHFP records. This can be controlled with the VerifyHostKeyDNS client configuration setting. DNSSEC support can be disabled entirely with the WITHOUT_LDNS option in src.conf. 20130906: The GNU Compiler Collection and C++ standard library (libstdc++) are no longer built by default on platforms where clang is the system compiler. You can enable them with the WITH_GCC and WITH_GNUCXX options in src.conf. 20130905: The PROCDESC kernel option is now part of the GENERIC kernel configuration and is required for the rwhod(8) to work. If you are using custom kernel configuration, you should include 'options PROCDESC'. 20130905: The API and ABI related to the Capsicum framework was modified in backward incompatible way. The userland libraries and programs have to be recompiled to work with the new kernel. This includes the following libraries and programs, but the whole buildworld is advised: libc, libprocstat, dhclient, tcpdump, hastd, hastctl, kdump, procstat, rwho, rwhod, uniq. 20130903: AES-NI intrinsic support has been added to gcc. The AES-NI module has been updated to use this support. A new gcc is required to build the aesni module on both i386 and amd64. 20130821: The PADLOCK_RNG and RDRAND_RNG kernel options are now devices. Thus "device padlock_rng" and "device rdrand_rng" should be used instead of "options PADLOCK_RNG" & "options RDRAND_RNG". 20130813: WITH_ICONV has been split into two feature sets. WITH_ICONV now enables just the iconv* functionality and is now on by default. WITH_LIBICONV_COMPAT enables the libiconv api and link time compatability. Set WITHOUT_ICONV to build the old way. If you have been using WITH_ICONV before, you will very likely need to turn on WITH_LIBICONV_COMPAT. 20130806: INVARIANTS option now enables DEBUG for code with OpenSolaris and Illumos origin, including ZFS. If you have INVARIANTS in your kernel configuration, then there is no need to set DEBUG or ZFS_DEBUG explicitly. DEBUG used to enable witness(9) tracking of OpenSolaris (mostly ZFS) locks if WITNESS option was set. Because that generated a lot of witness(9) reports and all of them were believed to be false positives, this is no longer done. New option OPENSOLARIS_WITNESS can be used to achieve the previous behavior. 20130806: Timer values in IPv6 data structures now use time_uptime instead of time_second. Although this is not a user-visible functional change, userland utilities which directly use them---ndp(8), rtadvd(8), and rtsold(8) in the base system---need to be updated to r253970 or later. 20130802: find -delete can now delete the pathnames given as arguments, instead of only files found below them or if the pathname did not contain any slashes. Formerly, the following error message would result: find: -delete: : relative path potentially not safe Deleting the pathnames given as arguments can be prevented without error messages using -mindepth 1 or by changing directory and passing "." as argument to find. This works in the old as well as the new version of find. 20130726: Behavior of devfs rules path matching has been changed. Pattern is now always matched against fully qualified devfs path and slash characters must be explicitly matched by slashes in pattern (FNM_PATHNAME). Rulesets involving devfs subdirectories must be reviewed. 20130716: The default ARM ABI has changed to the ARM EABI. The old ABI is incompatible with the ARM EABI and all programs and modules will need to be rebuilt to work with a new kernel. To keep using the old ABI ensure the WITHOUT_ARM_EABI knob is set. NOTE: Support for the old ABI will be removed in the future and users are advised to upgrade. 20130709: pkg_install has been disconnected from the build if you really need it you should add WITH_PKGTOOLS in your src.conf(5). 20130709: Most of network statistics structures were changed to be able keep 64-bits counters. Thus all tools, that work with networking statistics, must be rebuilt (netstat(1), bsnmpd(1), etc.) 20130629: Fix targets that run multiple make's to use && rather than ; so that subsequent steps depend on success of previous. NOTE: if building 'universe' with -j* on stable/8 or stable/9 it would be better to start the build using bmake, to avoid overloading the machine. 20130618: Fix a bug that allowed a tracing process (e.g. gdb) to write to a memory-mapped file in the traced process's address space even if neither the traced process nor the tracing process had write access to that file. 20130615: CVS has been removed from the base system. An exact copy of the code is available from the devel/cvs port. 20130613: Some people report the following error after the switch to bmake: make: illegal option -- J usage: make [-BPSXeiknpqrstv] [-C directory] [-D variable] ... *** [buildworld] Error code 2 this likely due to an old instance of make in ${MAKEPATH} (${MAKEOBJDIRPREFIX}${.CURDIR}/make.${MACHINE}) which src/Makefile will use that blindly, if it exists, so if you see the above error: rm -rf `make -V MAKEPATH` should resolve it. 20130516: Use bmake by default. Whereas before one could choose to build with bmake via -DWITH_BMAKE one must now use -DWITHOUT_BMAKE to use the old make. The goal is to remove these knobs for 10-RELEASE. It is worth noting that bmake (like gmake) treats the command line as the unit of failure, rather than statements within the command line. Thus '(cd some/where && dosomething)' is safer than 'cd some/where; dosomething'. The '()' allows consistent behavior in parallel build. 20130429: Fix a bug that allows NFS clients to issue READDIR on files. 20130426: The WITHOUT_IDEA option has been removed because the IDEA patent expired. 20130426: The sysctl which controls TRIM support under ZFS has been renamed from vfs.zfs.trim_disable -> vfs.zfs.trim.enabled and has been enabled by default. 20130425: The mergemaster command now uses the default MAKEOBJDIRPREFIX rather than creating it's own in the temporary directory in order allow access to bootstrapped versions of tools such as install and mtree. When upgrading from version of FreeBSD where the install command does not support -l, you will need to install a new mergemaster command if mergemaster -p is required. This can be accomplished with the command (cd src/usr.sbin/mergemaster && make install). 20130404: Legacy ATA stack, disabled and replaced by new CAM-based one since FreeBSD 9.0, completely removed from the sources. Kernel modules atadisk and atapi*, user-level tools atacontrol and burncd are removed. Kernel option `options ATA_CAM` is now permanently enabled and removed. 20130319: SOCK_CLOEXEC and SOCK_NONBLOCK flags have been added to socket(2) and socketpair(2). Software, in particular Kerberos, may automatically detect and use these during building. The resulting binaries will not work on older kernels. 20130308: CTL_DISABLE has also been added to the sparc64 GENERIC (for further information, see the respective 20130304 entry). 20130304: Recent commits to callout(9) changed the size of struct callout, so the KBI is probably heavily disturbed. Also, some functions in callout(9)/sleep(9)/sleepqueue(9)/condvar(9) KPIs were replaced by macros. Every kernel module using it won't load, so rebuild is requested. The ctl device has been re-enabled in GENERIC for i386 and amd64, but does not initialize by default (because of the new CTL_DISABLE option) to save memory. To re-enable it, remove the CTL_DISABLE option from the kernel config file or set kern.cam.ctl.disable=0 in /boot/loader.conf. 20130301: The ctl device has been disabled in GENERIC for i386 and amd64. This was done due to the extra memory being allocated at system initialisation time by the ctl driver which was only used if a CAM target device was created. This makes a FreeBSD system unusable on 128MB or less of RAM. 20130208: A new compression method (lz4) has been merged to -HEAD. Please refer to zpool-features(7) for more information. Please refer to the "ZFS notes" section of this file for information on upgrading boot ZFS pools. 20130129: A BSD-licensed patch(1) variant has been added and is installed as bsdpatch, being the GNU version the default patch. To inverse the logic and use the BSD-licensed one as default, while having the GNU version installed as gnupatch, rebuild and install world with the WITH_BSD_PATCH knob set. 20130121: Due to the use of the new -l option to install(1) during build and install, you must take care not to directly set the INSTALL make variable in your /etc/make.conf, /etc/src.conf, or on the command line. If you wish to use the -C flag for all installs you may be able to add INSTALL+=-C to /etc/make.conf or /etc/src.conf. 20130118: The install(1) option -M has changed meaning and now takes an argument that is a file or path to append logs to. In the unlikely event that -M was the last option on the command line and the command line contained at least two files and a target directory the first file will have logs appended to it. The -M option served little practical purpose in the last decade so its use is expected to be extremely rare. 20121223: After switching to Clang as the default compiler some users of ZFS on i386 systems started to experience stack overflow kernel panics. Please consider using 'options KSTACK_PAGES=4' in such configurations. 20121222: GEOM_LABEL now mangles label names read from file system metadata. Mangling affect labels containing spaces, non-printable characters, '%' or '"'. Device names in /etc/fstab and other places may need to be updated. 20121217: By default, only the 10 most recent kernel dumps will be saved. To restore the previous behaviour (no limit on the number of kernel dumps stored in the dump directory) add the following line to /etc/rc.conf: savecore_flags="" 20121201: With the addition of auditdistd(8), a new auditdistd user is now required during installworld. "mergemaster -p" can be used to add the user prior to installworld, as documented in the handbook. 20121117: The sin6_scope_id member variable in struct sockaddr_in6 is now filled by the kernel before passing the structure to the userland via sysctl or routing socket. This means the KAME-specific embedded scope id in sin6_addr.s6_addr[2] is always cleared in userland application. This behavior can be controlled by net.inet6.ip6.deembed_scopeid. __FreeBSD_version is bumped to 1000025. 20121105: On i386 and amd64 systems WITH_CLANG_IS_CC is now the default. This means that the world and kernel will be compiled with clang and that clang will be installed as /usr/bin/cc, /usr/bin/c++, and /usr/bin/cpp. To disable this behavior and revert to building with gcc, compile with WITHOUT_CLANG_IS_CC. Really old versions of current may need to bootstrap WITHOUT_CLANG first if the clang build fails (its compatibility window doesn't extend to the 9 stable branch point). 20121102: The IPFIREWALL_FORWARD kernel option has been removed. Its functionality now turned on by default. 20121023: The ZERO_COPY_SOCKET kernel option has been removed and split into SOCKET_SEND_COW and SOCKET_RECV_PFLIP. NB: SOCKET_SEND_COW uses the VM page based copy-on-write mechanism which is not safe and may result in kernel crashes. NB: The SOCKET_RECV_PFLIP mechanism is useless as no current driver supports disposeable external page sized mbuf storage. Proper replacements for both zero-copy mechanisms are under consideration and will eventually lead to complete removal of the two kernel options. 20121023: The IPv4 network stack has been converted to network byte order. The following modules need to be recompiled together with kernel: carp(4), divert(4), gif(4), siftr(4), gre(4), pf(4), ipfw(4), ng_ipfw(4), stf(4). 20121022: Support for non-MPSAFE filesystems was removed from VFS. The VFS_VERSION was bumped, all filesystem modules shall be recompiled. 20121018: All the non-MPSAFE filesystems have been disconnected from the build. The full list includes: codafs, hpfs, ntfs, nwfs, portalfs, smbfs, xfs. 20121016: The interface cloning API and ABI has changed. The following modules need to be recompiled together with kernel: ipfw(4), pfsync(4), pflog(4), usb(4), wlan(4), stf(4), vlan(4), disc(4), edsc(4), if_bridge(4), gif(4), tap(4), faith(4), epair(4), enc(4), tun(4), if_lagg(4), gre(4). 20121015: The sdhci driver was split in two parts: sdhci (generic SD Host Controller logic) and sdhci_pci (actual hardware driver). No kernel config modifications are required, but if you load sdhc as a module you must switch to sdhci_pci instead. 20121014: Import the FUSE kernel and userland support into base system. 20121013: The GNU sort(1) program has been removed since the BSD-licensed sort(1) has been the default for quite some time and no serious problems have been reported. The corresponding WITH_GNU_SORT knob has also gone. 20121006: The pfil(9) API/ABI for AF_INET family has been changed. Packet filtering modules: pf(4), ipfw(4), ipfilter(4) need to be recompiled with new kernel. 20121001: The net80211(4) ABI has been changed to allow for improved driver PS-POLL and power-save support. All wireless drivers need to be recompiled to work with the new kernel. 20120913: The random(4) support for the VIA hardware random number generator (`PADLOCK') is no longer enabled unconditionally. Add the padlock_rng device in the custom kernel config if needed. The GENERIC kernels on i386 and amd64 do include the device, so the change only affects the custom kernel configurations. 20120908: The pf(4) packet filter ABI has been changed. pfctl(8) and snmp_pf module need to be recompiled to work with new kernel. 20120828: A new ZFS feature flag "com.delphix:empty_bpobj" has been merged to -HEAD. Pools that have empty_bpobj in active state can not be imported read-write with ZFS implementations that do not support this feature. For more information read the zpool-features(5) manual page. 20120727: The sparc64 ZFS loader has been changed to no longer try to auto- detect ZFS providers based on diskN aliases but now requires these to be explicitly listed in the OFW boot-device environment variable. 20120712: The OpenSSL has been upgraded to 1.0.1c. Any binaries requiring libcrypto.so.6 or libssl.so.6 must be recompiled. Also, there are configuration changes. Make sure to merge /etc/ssl/openssl.cnf. 20120712: The following sysctls and tunables have been renamed for consistency with other variables: kern.cam.da.da_send_ordered -> kern.cam.da.send_ordered kern.cam.ada.ada_send_ordered -> kern.cam.ada.send_ordered 20120628: The sort utility has been replaced with BSD sort. For now, GNU sort is also available as "gnusort" or the default can be set back to GNU sort by setting WITH_GNU_SORT. In this case, BSD sort will be installed as "bsdsort". 20120611: A new version of ZFS (pool version 5000) has been merged to -HEAD. Starting with this version the old system of ZFS pool versioning is superseded by "feature flags". This concept enables forward compatibility against certain future changes in functionality of ZFS pools. The first read-only compatible "feature flag" for ZFS pools is named "com.delphix:async_destroy". For more information read the new zpool-features(5) manual page. Please refer to the "ZFS notes" section of this file for information on upgrading boot ZFS pools. 20120417: The malloc(3) implementation embedded in libc now uses sources imported as contrib/jemalloc. The most disruptive API change is to /etc/malloc.conf. If your system has an old-style /etc/malloc.conf, delete it prior to installworld, and optionally re-create it using the new format after rebooting. See malloc.conf(5) for details (specifically the TUNING section and the "opt.*" entries in the MALLCTL NAMESPACE section). 20120328: Big-endian MIPS TARGET_ARCH values no longer end in "eb". mips64eb is now spelled mips64. mipsn32eb is now spelled mipsn32. mipseb is now spelled mips. This is to aid compatibility with third-party software that expects this naming scheme in uname(3). Little-endian settings are unchanged. If you are updating a big-endian mips64 machine from before this change, you may need to set MACHINE_ARCH=mips64 in your environment before the new build system will recognize your machine. 20120306: Disable by default the option VFS_ALLOW_NONMPSAFE for all supported platforms. 20120229: Now unix domain sockets behave "as expected" on nullfs(5). Previously nullfs(5) did not pass through all behaviours to the underlying layer, as a result if we bound to a socket on the lower layer we could connect only to the lower path; if we bound to the upper layer we could connect only to the upper path. The new behavior is one can connect to both the lower and the upper paths regardless what layer path one binds to. 20120211: The getifaddrs upgrade path broken with 20111215 has been restored. If you have upgraded in between 20111215 and 20120209 you need to recompile libc again with your kernel. You still need to recompile world to be able to configure CARP but this restriction already comes from 20111215. 20120114: The set_rcvar() function has been removed from /etc/rc.subr. All base and ports rc.d scripts have been updated, so if you have a port installed with a script in /usr/local/etc/rc.d you can either hand-edit the rcvar= line, or reinstall the port. An easy way to handle the mass-update of /etc/rc.d: rm /etc/rc.d/* && mergemaster -i 20120109: panic(9) now stops other CPUs in the SMP systems, disables interrupts on the current CPU and prevents other threads from running. This behavior can be reverted using the kern.stop_scheduler_on_panic tunable/sysctl. The new behavior can be incompatible with kern.sync_on_panic. 20111215: The carp(4) facility has been changed significantly. Configuration of the CARP protocol via ifconfig(8) has changed, as well as format of CARP events submitted to devd(8) has changed. See manual pages for more information. The arpbalance feature of carp(4) is currently not supported anymore. Size of struct in_aliasreq, struct in6_aliasreq has changed. User utilities using SIOCAIFADDR, SIOCAIFADDR_IN6, e.g. ifconfig(8), need to be recompiled. 20111122: The acpi_wmi(4) status device /dev/wmistat has been renamed to /dev/wmistat0. 20111108: The option VFS_ALLOW_NONMPSAFE option has been added in order to explicitely support non-MPSAFE filesystems. It is on by default for all supported platform at this present time. 20111101: The broken amd(4) driver has been replaced with esp(4) in the amd64, i386 and pc98 GENERIC kernel configuration files. 20110930: sysinstall has been removed 20110923: The stable/9 branch created in subversion. This corresponds to the RELENG_9 branch in CVS. COMMON ITEMS: General Notes ------------- Avoid using make -j when upgrading. While generally safe, there are sometimes problems using -j to upgrade. If your upgrade fails with -j, please try again without -j. From time to time in the past there have been problems using -j with buildworld and/or installworld. This is especially true when upgrading between "distant" versions (eg one that cross a major release boundary or several minor releases, or when several months have passed on the -current branch). Sometimes, obscure build problems are the result of environment poisoning. This can happen because the make utility reads its environment when searching for values for global variables. To run your build attempts in an "environmental clean room", prefix all make commands with 'env -i '. See the env(1) manual page for more details. When upgrading from one major version to another it is generally best to upgrade to the latest code in the currently installed branch first, then do an upgrade to the new branch. This is the best-tested upgrade path, and has the highest probability of being successful. Please try this approach before reporting problems with a major version upgrade. When upgrading a live system, having a root shell around before installing anything can help undo problems. Not having a root shell around can lead to problems if pam has changed too much from your starting point to allow continued authentication after the upgrade. This file should be read as a log of events. When a later event changes information of a prior event, the prior event should not be deleted. Instead, a pointer to the entry with the new information should be placed in the old entry. Readers of this file should also sanity check older entries before relying on them blindly. Authors of new entries should write them with this in mind. ZFS notes --------- When upgrading the boot ZFS pool to a new version, always follow these two steps: 1.) recompile and reinstall the ZFS boot loader and boot block (this is part of "make buildworld" and "make installworld") 2.) update the ZFS boot block on your boot drive The following example updates the ZFS boot block on the first partition (freebsd-boot) of a GPT partitioned drive ada0: "gpart bootcode -p /boot/gptzfsboot -i 1 ada0" Non-boot pools do not need these updates. To build a kernel ----------------- If you are updating from a prior version of FreeBSD (even one just a few days old), you should follow this procedure. It is the most failsafe as it uses a /usr/obj tree with a fresh mini-buildworld, make kernel-toolchain make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE To test a kernel once --------------------- If you just want to boot a kernel once (because you are not sure if it works, or if you want to boot a known bad kernel to provide debugging information) run make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel nextboot -k testkernel To just build a kernel when you know that it won't mess you up -------------------------------------------------------------- This assumes you are already running a CURRENT system. Replace ${arch} with the architecture of your machine (e.g. "i386", "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc). cd src/sys/${arch}/conf config KERNEL_NAME_HERE cd ../compile/KERNEL_NAME_HERE make depend make make install If this fails, go to the "To build a kernel" section. To rebuild everything and install it on the current system. ----------------------------------------------------------- # Note: sometimes if you are running current you gotta do more than # is listed here if you are upgrading from a really old current. make buildworld make kernel KERNCONF=YOUR_KERNEL_HERE [1] [3] mergemaster -Fp [5] make installworld mergemaster -Fi [4] make delete-old [6] To cross-install current onto a separate partition -------------------------------------------------- # In this approach we use a separate partition to hold # current's root, 'usr', and 'var' directories. A partition # holding "/", "/usr" and "/var" should be about 2GB in # size. make buildworld make buildkernel KERNCONF=YOUR_KERNEL_HERE make installworld DESTDIR=${CURRENT_ROOT} -DDB_FROM_SRC make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT} cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd To upgrade in-place from stable to current ---------------------------------------------- make buildworld [9] make kernel KERNCONF=YOUR_KERNEL_HERE [8] [1] [3] mergemaster -Fp [5] make installworld mergemaster -Fi [4] make delete-old [6] Make sure that you've read the UPDATING file to understand the tweaks to various things you need. At this point in the life cycle of current, things change often and you are on your own to cope. The defaults can also change, so please read ALL of the UPDATING entries. Also, if you are tracking -current, you must be subscribed to freebsd-current@freebsd.org. Make sure that before you update your sources that you have read and understood all the recent messages there. If in doubt, please track -stable which has much fewer pitfalls. [1] If you have third party modules, such as vmware, you should disable them at this point so they don't crash your system on reboot. [3] From the bootblocks, boot -s, and then do fsck -p mount -u / mount -a cd src adjkerntz -i # if CMOS is wall time Also, when doing a major release upgrade, it is required that you boot into single user mode to do the installworld. [4] Note: This step is non-optional. Failure to do this step can result in a significant reduction in the functionality of the system. Attempting to do it by hand is not recommended and those that pursue this avenue should read this file carefully, as well as the archives of freebsd-current and freebsd-hackers mailing lists for potential gotchas. The -U option is also useful to consider. See mergemaster(8) for more information. [5] Usually this step is a noop. However, from time to time you may need to do this if you get unknown user in the following step. It never hurts to do it all the time. You may need to install a new mergemaster (cd src/usr.sbin/mergemaster && make install) after the buildworld before this step if you last updated from current before 20130425 or from -stable before 20130430. [6] This only deletes old files and directories. Old libraries can be deleted by "make delete-old-libs", but you have to make sure that no program is using those libraries anymore. [8] In order to have a kernel that can run the 4.x binaries needed to do an installworld, you must include the COMPAT_FREEBSD4 option in your kernel. Failure to do so may leave you with a system that is hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is required to run the 5.x binaries on more recent kernels. And so on for COMPAT_FREEBSD6 and COMPAT_FREEBSD7. Make sure that you merge any new devices from GENERIC since the last time you updated your kernel config file. [9] When checking out sources, you must include the -P flag to have cvs prune empty directories. If CPUTYPE is defined in your /etc/make.conf, make sure to use the "?=" instead of the "=" assignment operator, so that buildworld can override the CPUTYPE if it needs to. MAKEOBJDIRPREFIX must be defined in an environment variable, and not on the command line, or in /etc/make.conf. buildworld will warn if it is improperly defined. FORMAT: This file contains a list, in reverse chronological order, of major breakages in tracking -current. It is not guaranteed to be a complete list of such breakages, and only contains entries since September 23, 2011. If you need to see UPDATING entries from before that date, you will need to fetch an UPDATING file from an older FreeBSD release. Copyright information: Copyright 1998-2009 M. Warner Losh. All Rights Reserved. Redistribution, publication, translation and use, with or without modification, in full or in part, in any form or format of this document are permitted without further permission from the author. THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. Contact Warner Losh if you have any questions about your use of this document. $FreeBSD$ Index: head/sbin/pfctl/parse.y =================================================================== --- head/sbin/pfctl/parse.y (revision 287221) +++ head/sbin/pfctl/parse.y (revision 287222) @@ -1,6173 +1,6173 @@ /* $OpenBSD: parse.y,v 1.554 2008/10/17 12:59:53 henning Exp $ */ /* * Copyright (c) 2001 Markus Friedl. All rights reserved. * Copyright (c) 2001 Daniel Hartmeier. All rights reserved. * Copyright (c) 2001 Theo de Raadt. All rights reserved. * Copyright (c) 2002,2003 Henning Brauer. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ %{ #include __FBSDID("$FreeBSD$"); #include #include #include #ifdef __FreeBSD__ #include #endif #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include "pfctl_parser.h" #include "pfctl.h" static struct pfctl *pf = NULL; static int debug = 0; static int rulestate = 0; static u_int16_t returnicmpdefault = (ICMP_UNREACH << 8) | ICMP_UNREACH_PORT; static u_int16_t returnicmp6default = (ICMP6_DST_UNREACH << 8) | ICMP6_DST_UNREACH_NOPORT; static int blockpolicy = PFRULE_DROP; static int require_order = 1; static int default_statelock; TAILQ_HEAD(files, file) files = TAILQ_HEAD_INITIALIZER(files); static struct file { TAILQ_ENTRY(file) entry; FILE *stream; char *name; int lineno; int errors; } *file; struct file *pushfile(const char *, int); int popfile(void); int check_file_secrecy(int, const char *); int yyparse(void); int yylex(void); int yyerror(const char *, ...); int kw_cmp(const void *, const void *); int lookup(char *); int lgetc(int); int lungetc(int); int findeol(void); TAILQ_HEAD(symhead, sym) symhead = TAILQ_HEAD_INITIALIZER(symhead); struct sym { TAILQ_ENTRY(sym) entry; int used; int persist; char *nam; char *val; }; int symset(const char *, const char *, int); char *symget(const char *); int atoul(char *, u_long *); enum { PFCTL_STATE_NONE, PFCTL_STATE_OPTION, PFCTL_STATE_SCRUB, PFCTL_STATE_QUEUE, PFCTL_STATE_NAT, PFCTL_STATE_FILTER }; struct node_proto { u_int8_t proto; struct node_proto *next; struct node_proto *tail; }; struct node_port { u_int16_t port[2]; u_int8_t op; struct node_port *next; struct node_port *tail; }; struct node_uid { uid_t uid[2]; u_int8_t op; struct node_uid *next; struct node_uid *tail; }; struct node_gid { gid_t gid[2]; u_int8_t op; struct node_gid *next; struct node_gid *tail; }; struct node_icmp { u_int8_t code; u_int8_t type; u_int8_t proto; struct node_icmp *next; struct node_icmp *tail; }; enum { PF_STATE_OPT_MAX, PF_STATE_OPT_NOSYNC, PF_STATE_OPT_SRCTRACK, PF_STATE_OPT_MAX_SRC_STATES, PF_STATE_OPT_MAX_SRC_CONN, PF_STATE_OPT_MAX_SRC_CONN_RATE, PF_STATE_OPT_MAX_SRC_NODES, PF_STATE_OPT_OVERLOAD, PF_STATE_OPT_STATELOCK, PF_STATE_OPT_TIMEOUT, PF_STATE_OPT_SLOPPY, }; enum { PF_SRCTRACK_NONE, PF_SRCTRACK, PF_SRCTRACK_GLOBAL, PF_SRCTRACK_RULE }; struct node_state_opt { int type; union { u_int32_t max_states; u_int32_t max_src_states; u_int32_t max_src_conn; struct { u_int32_t limit; u_int32_t seconds; } max_src_conn_rate; struct { u_int8_t flush; char tblname[PF_TABLE_NAME_SIZE]; } overload; u_int32_t max_src_nodes; u_int8_t src_track; u_int32_t statelock; struct { int number; u_int32_t seconds; } timeout; } data; struct node_state_opt *next; struct node_state_opt *tail; }; struct peer { struct node_host *host; struct node_port *port; }; struct node_queue { char queue[PF_QNAME_SIZE]; char parent[PF_QNAME_SIZE]; char ifname[IFNAMSIZ]; int scheduler; struct node_queue *next; struct node_queue *tail; } *queues = NULL; struct node_qassign { char *qname; char *pqname; }; struct filter_opts { int marker; #define FOM_FLAGS 0x01 #define FOM_ICMP 0x02 #define FOM_TOS 0x04 #define FOM_KEEP 0x08 #define FOM_SRCTRACK 0x10 struct node_uid *uid; struct node_gid *gid; struct { u_int8_t b1; u_int8_t b2; u_int16_t w; u_int16_t w2; } flags; struct node_icmp *icmpspec; u_int32_t tos; u_int32_t prob; struct { int action; struct node_state_opt *options; } keep; int fragment; int allowopts; char *label; struct node_qassign queues; char *tag; char *match_tag; u_int8_t match_tag_not; u_int rtableid; struct { struct node_host *addr; u_int16_t port; } divert; } filter_opts; struct antispoof_opts { char *label; u_int rtableid; } antispoof_opts; struct scrub_opts { int marker; #define SOM_MINTTL 0x01 #define SOM_MAXMSS 0x02 #define SOM_FRAGCACHE 0x04 #define SOM_SETTOS 0x08 int nodf; int minttl; int maxmss; int settos; int fragcache; int randomid; int reassemble_tcp; char *match_tag; u_int8_t match_tag_not; u_int rtableid; } scrub_opts; struct queue_opts { int marker; #define QOM_BWSPEC 0x01 #define QOM_SCHEDULER 0x02 #define QOM_PRIORITY 0x04 #define QOM_TBRSIZE 0x08 #define QOM_QLIMIT 0x10 struct node_queue_bw queue_bwspec; struct node_queue_opt scheduler; int priority; int tbrsize; int qlimit; } queue_opts; struct table_opts { int flags; int init_addr; struct node_tinithead init_nodes; } table_opts; struct pool_opts { int marker; #define POM_TYPE 0x01 #define POM_STICKYADDRESS 0x02 u_int8_t opts; int type; int staticport; struct pf_poolhashkey *key; } pool_opts; struct codel_opts codel_opts; struct node_hfsc_opts hfsc_opts; struct node_fairq_opts fairq_opts; struct node_state_opt *keep_state_defaults = NULL; int disallow_table(struct node_host *, const char *); int disallow_urpf_failed(struct node_host *, const char *); int disallow_alias(struct node_host *, const char *); int rule_consistent(struct pf_rule *, int); int filter_consistent(struct pf_rule *, int); int nat_consistent(struct pf_rule *); int rdr_consistent(struct pf_rule *); int process_tabledef(char *, struct table_opts *); void expand_label_str(char *, size_t, const char *, const char *); void expand_label_if(const char *, char *, size_t, const char *); void expand_label_addr(const char *, char *, size_t, u_int8_t, struct node_host *); void expand_label_port(const char *, char *, size_t, struct node_port *); void expand_label_proto(const char *, char *, size_t, u_int8_t); void expand_label_nr(const char *, char *, size_t); void expand_label(char *, size_t, const char *, u_int8_t, struct node_host *, struct node_port *, struct node_host *, struct node_port *, u_int8_t); void expand_rule(struct pf_rule *, struct node_if *, struct node_host *, struct node_proto *, struct node_os *, struct node_host *, struct node_port *, struct node_host *, struct node_port *, struct node_uid *, struct node_gid *, struct node_icmp *, const char *); int expand_altq(struct pf_altq *, struct node_if *, struct node_queue *, struct node_queue_bw bwspec, struct node_queue_opt *); int expand_queue(struct pf_altq *, struct node_if *, struct node_queue *, struct node_queue_bw, struct node_queue_opt *); int expand_skip_interface(struct node_if *); int check_rulestate(int); int getservice(char *); int rule_label(struct pf_rule *, char *); int rt_tableid_max(void); void mv_rules(struct pf_ruleset *, struct pf_ruleset *); void decide_address_family(struct node_host *, sa_family_t *); void remove_invalid_hosts(struct node_host **, sa_family_t *); int invalid_redirect(struct node_host *, sa_family_t); u_int16_t parseicmpspec(char *, sa_family_t); TAILQ_HEAD(loadanchorshead, loadanchors) loadanchorshead = TAILQ_HEAD_INITIALIZER(loadanchorshead); struct loadanchors { TAILQ_ENTRY(loadanchors) entries; char *anchorname; char *filename; }; typedef struct { union { int64_t number; double probability; int i; char *string; u_int rtableid; struct { u_int8_t b1; u_int8_t b2; u_int16_t w; u_int16_t w2; } b; struct range { int a; int b; int t; } range; struct node_if *interface; struct node_proto *proto; struct node_icmp *icmp; struct node_host *host; struct node_os *os; struct node_port *port; struct node_uid *uid; struct node_gid *gid; struct node_state_opt *state_opt; struct peer peer; struct { struct peer src, dst; struct node_os *src_os; } fromto; struct { struct node_host *host; u_int8_t rt; u_int8_t pool_opts; sa_family_t af; struct pf_poolhashkey *key; } route; struct redirection { struct node_host *host; struct range rport; } *redirection; struct { int action; struct node_state_opt *options; } keep_state; struct { u_int8_t log; u_int8_t logif; u_int8_t quick; } logquick; struct { int neg; char *name; } tagged; struct pf_poolhashkey *hashkey; struct node_queue *queue; struct node_queue_opt queue_options; struct node_queue_bw queue_bwspec; struct node_qassign qassign; struct filter_opts filter_opts; struct antispoof_opts antispoof_opts; struct queue_opts queue_opts; struct scrub_opts scrub_opts; struct table_opts table_opts; struct pool_opts pool_opts; struct node_hfsc_opts hfsc_opts; struct node_fairq_opts fairq_opts; struct codel_opts codel_opts; } v; int lineno; } YYSTYPE; #define PPORT_RANGE 1 #define PPORT_STAR 2 int parseport(char *, struct range *r, int); #define DYNIF_MULTIADDR(addr) ((addr).type == PF_ADDR_DYNIFTL && \ (!((addr).iflags & PFI_AFLAG_NOALIAS) || \ !isdigit((addr).v.ifname[strlen((addr).v.ifname)-1]))) %} %token PASS BLOCK SCRUB RETURN IN OS OUT LOG QUICK ON FROM TO FLAGS %token RETURNRST RETURNICMP RETURNICMP6 PROTO INET INET6 ALL ANY ICMPTYPE %token ICMP6TYPE CODE KEEP MODULATE STATE PORT RDR NAT BINAT ARROW NODF %token MINTTL ERROR ALLOWOPTS FASTROUTE FILENAME ROUTETO DUPTO REPLYTO NO LABEL %token NOROUTE URPFFAILED FRAGMENT USER GROUP MAXMSS MAXIMUM TTL TOS DROP TABLE %token REASSEMBLE FRAGDROP FRAGCROP ANCHOR NATANCHOR RDRANCHOR BINATANCHOR %token SET OPTIMIZATION TIMEOUT LIMIT LOGINTERFACE BLOCKPOLICY RANDOMID %token REQUIREORDER SYNPROXY FINGERPRINTS NOSYNC DEBUG SKIP HOSTID %token ANTISPOOF FOR INCLUDE %token BITMASK RANDOM SOURCEHASH ROUNDROBIN STATICPORT PROBABILITY %token ALTQ CBQ CODEL PRIQ HFSC FAIRQ BANDWIDTH TBRSIZE LINKSHARE REALTIME %token UPPERLIMIT QUEUE PRIORITY QLIMIT HOGS BUCKETS RTABLE TARGET INTERVAL %token LOAD RULESET_OPTIMIZATION %token STICKYADDRESS MAXSRCSTATES MAXSRCNODES SOURCETRACK GLOBAL RULE %token MAXSRCCONN MAXSRCCONNRATE OVERLOAD FLUSH SLOPPY %token TAGGED TAG IFBOUND FLOATING STATEPOLICY STATEDEFAULTS ROUTE SETTOS %token DIVERTTO DIVERTREPLY %token STRING %token NUMBER %token PORTBINARY %type interface if_list if_item_not if_item %type number icmptype icmp6type uid gid %type tos not yesno %type probability %type no dir af fragcache optimizer %type sourcetrack flush unaryop statelock %type action nataction natpasslog scrubaction %type flags flag blockspec %type portplain portstar portrange %type hashkey %type proto proto_list proto_item %type protoval %type icmpspec %type icmp_list icmp_item %type icmp6_list icmp6_item %type reticmpspec reticmp6spec %type fromto %type ipportspec from to %type ipspec toipspec xhost host dynaddr host_list %type redir_host_list redirspec %type route_host route_host_list routespec %type os xos os_list %type portspec port_list port_item %type uids uid_list uid_item %type gids gid_list gid_item %type route %type redirection redirpool %type label stringall tag anchorname %type string varstring numberstring %type keep %type state_opt_spec state_opt_list state_opt_item %type logquick quick log logopts logopt %type antispoof_ifspc antispoof_iflst antispoof_if %type qname %type qassign qassign_list qassign_item %type scheduler %type cbqflags_list cbqflags_item %type priqflags_list priqflags_item %type hfscopts_list hfscopts_item hfsc_opts %type fairqopts_list fairqopts_item fairq_opts %type codelopts_list codelopts_item codel_opts %type bandwidth %type filter_opts filter_opt filter_opts_l %type antispoof_opts antispoof_opt antispoof_opts_l %type queue_opts queue_opt queue_opts_l %type scrub_opts scrub_opt scrub_opts_l %type table_opts table_opt table_opts_l %type pool_opts pool_opt pool_opts_l %type tagged %type rtable %% ruleset : /* empty */ | ruleset include '\n' | ruleset '\n' | ruleset option '\n' | ruleset scrubrule '\n' | ruleset natrule '\n' | ruleset binatrule '\n' | ruleset pfrule '\n' | ruleset anchorrule '\n' | ruleset loadrule '\n' | ruleset altqif '\n' | ruleset queuespec '\n' | ruleset varset '\n' | ruleset antispoof '\n' | ruleset tabledef '\n' | '{' fakeanchor '}' '\n'; | ruleset error '\n' { file->errors++; } ; include : INCLUDE STRING { struct file *nfile; if ((nfile = pushfile($2, 0)) == NULL) { yyerror("failed to include file %s", $2); free($2); YYERROR; } free($2); file = nfile; lungetc('\n'); } ; /* * apply to previouslys specified rule: must be careful to note * what that is: pf or nat or binat or rdr */ fakeanchor : fakeanchor '\n' | fakeanchor anchorrule '\n' | fakeanchor binatrule '\n' | fakeanchor natrule '\n' | fakeanchor pfrule '\n' | fakeanchor error '\n' ; optimizer : string { if (!strcmp($1, "none")) $$ = 0; else if (!strcmp($1, "basic")) $$ = PF_OPTIMIZE_BASIC; else if (!strcmp($1, "profile")) $$ = PF_OPTIMIZE_BASIC | PF_OPTIMIZE_PROFILE; else { yyerror("unknown ruleset-optimization %s", $1); YYERROR; } } ; option : SET OPTIMIZATION STRING { if (check_rulestate(PFCTL_STATE_OPTION)) { free($3); YYERROR; } if (pfctl_set_optimization(pf, $3) != 0) { yyerror("unknown optimization %s", $3); free($3); YYERROR; } free($3); } | SET RULESET_OPTIMIZATION optimizer { if (!(pf->opts & PF_OPT_OPTIMIZE)) { pf->opts |= PF_OPT_OPTIMIZE; pf->optimize = $3; } } | SET TIMEOUT timeout_spec | SET TIMEOUT '{' optnl timeout_list '}' | SET LIMIT limit_spec | SET LIMIT '{' optnl limit_list '}' | SET LOGINTERFACE stringall { if (check_rulestate(PFCTL_STATE_OPTION)) { free($3); YYERROR; } if (pfctl_set_logif(pf, $3) != 0) { yyerror("error setting loginterface %s", $3); free($3); YYERROR; } free($3); } | SET HOSTID number { if ($3 == 0 || $3 > UINT_MAX) { yyerror("hostid must be non-zero"); YYERROR; } if (pfctl_set_hostid(pf, $3) != 0) { yyerror("error setting hostid %08x", $3); YYERROR; } } | SET BLOCKPOLICY DROP { if (pf->opts & PF_OPT_VERBOSE) printf("set block-policy drop\n"); if (check_rulestate(PFCTL_STATE_OPTION)) YYERROR; blockpolicy = PFRULE_DROP; } | SET BLOCKPOLICY RETURN { if (pf->opts & PF_OPT_VERBOSE) printf("set block-policy return\n"); if (check_rulestate(PFCTL_STATE_OPTION)) YYERROR; blockpolicy = PFRULE_RETURN; } | SET REQUIREORDER yesno { if (pf->opts & PF_OPT_VERBOSE) printf("set require-order %s\n", $3 == 1 ? "yes" : "no"); require_order = $3; } | SET FINGERPRINTS STRING { if (pf->opts & PF_OPT_VERBOSE) printf("set fingerprints \"%s\"\n", $3); if (check_rulestate(PFCTL_STATE_OPTION)) { free($3); YYERROR; } if (!pf->anchor->name[0]) { if (pfctl_file_fingerprints(pf->dev, pf->opts, $3)) { yyerror("error loading " "fingerprints %s", $3); free($3); YYERROR; } } free($3); } | SET STATEPOLICY statelock { if (pf->opts & PF_OPT_VERBOSE) switch ($3) { case 0: printf("set state-policy floating\n"); break; case PFRULE_IFBOUND: printf("set state-policy if-bound\n"); break; } default_statelock = $3; } | SET DEBUG STRING { if (check_rulestate(PFCTL_STATE_OPTION)) { free($3); YYERROR; } if (pfctl_set_debug(pf, $3) != 0) { yyerror("error setting debuglevel %s", $3); free($3); YYERROR; } free($3); } | SET SKIP interface { if (expand_skip_interface($3) != 0) { yyerror("error setting skip interface(s)"); YYERROR; } } | SET STATEDEFAULTS state_opt_list { if (keep_state_defaults != NULL) { yyerror("cannot redefine state-defaults"); YYERROR; } keep_state_defaults = $3; } ; stringall : STRING { $$ = $1; } | ALL { if (($$ = strdup("all")) == NULL) { err(1, "stringall: strdup"); } } ; string : STRING string { if (asprintf(&$$, "%s %s", $1, $2) == -1) err(1, "string: asprintf"); free($1); free($2); } | STRING ; varstring : numberstring varstring { if (asprintf(&$$, "%s %s", $1, $2) == -1) err(1, "string: asprintf"); free($1); free($2); } | numberstring ; numberstring : NUMBER { char *s; if (asprintf(&s, "%lld", (long long)$1) == -1) { yyerror("string: asprintf"); YYERROR; } $$ = s; } | STRING ; varset : STRING '=' varstring { if (pf->opts & PF_OPT_VERBOSE) printf("%s = \"%s\"\n", $1, $3); if (symset($1, $3, 0) == -1) err(1, "cannot store variable %s", $1); free($1); free($3); } ; anchorname : STRING { $$ = $1; } | /* empty */ { $$ = NULL; } ; pfa_anchorlist : /* empty */ | pfa_anchorlist '\n' | pfa_anchorlist pfrule '\n' | pfa_anchorlist anchorrule '\n' ; pfa_anchor : '{' { char ta[PF_ANCHOR_NAME_SIZE]; struct pf_ruleset *rs; /* steping into a brace anchor */ pf->asd++; pf->bn++; pf->brace = 1; /* create a holding ruleset in the root */ snprintf(ta, PF_ANCHOR_NAME_SIZE, "_%d", pf->bn); rs = pf_find_or_create_ruleset(ta); if (rs == NULL) err(1, "pfa_anchor: pf_find_or_create_ruleset"); pf->astack[pf->asd] = rs->anchor; pf->anchor = rs->anchor; } '\n' pfa_anchorlist '}' { pf->alast = pf->anchor; pf->asd--; pf->anchor = pf->astack[pf->asd]; } | /* empty */ ; anchorrule : ANCHOR anchorname dir quick interface af proto fromto filter_opts pfa_anchor { struct pf_rule r; struct node_proto *proto; if (check_rulestate(PFCTL_STATE_FILTER)) { if ($2) free($2); YYERROR; } if ($2 && ($2[0] == '_' || strstr($2, "/_") != NULL)) { free($2); yyerror("anchor names beginning with '_' " "are reserved for internal use"); YYERROR; } memset(&r, 0, sizeof(r)); if (pf->astack[pf->asd + 1]) { /* move inline rules into relative location */ pf_anchor_setup(&r, &pf->astack[pf->asd]->ruleset, $2 ? $2 : pf->alast->name); if (r.anchor == NULL) err(1, "anchorrule: unable to " "create ruleset"); if (pf->alast != r.anchor) { if (r.anchor->match) { yyerror("inline anchor '%s' " "already exists", r.anchor->name); YYERROR; } mv_rules(&pf->alast->ruleset, &r.anchor->ruleset); } pf_remove_if_empty_ruleset(&pf->alast->ruleset); pf->alast = r.anchor; } else { if (!$2) { yyerror("anchors without explicit " "rules must specify a name"); YYERROR; } } r.direction = $3; r.quick = $4.quick; r.af = $6; r.prob = $9.prob; r.rtableid = $9.rtableid; if ($9.tag) if (strlcpy(r.tagname, $9.tag, PF_TAG_NAME_SIZE) >= PF_TAG_NAME_SIZE) { yyerror("tag too long, max %u chars", PF_TAG_NAME_SIZE - 1); YYERROR; } if ($9.match_tag) if (strlcpy(r.match_tagname, $9.match_tag, PF_TAG_NAME_SIZE) >= PF_TAG_NAME_SIZE) { yyerror("tag too long, max %u chars", PF_TAG_NAME_SIZE - 1); YYERROR; } r.match_tag_not = $9.match_tag_not; if (rule_label(&r, $9.label)) YYERROR; free($9.label); r.flags = $9.flags.b1; r.flagset = $9.flags.b2; if (($9.flags.b1 & $9.flags.b2) != $9.flags.b1) { yyerror("flags always false"); YYERROR; } if ($9.flags.b1 || $9.flags.b2 || $8.src_os) { for (proto = $7; proto != NULL && proto->proto != IPPROTO_TCP; proto = proto->next) ; /* nothing */ if (proto == NULL && $7 != NULL) { if ($9.flags.b1 || $9.flags.b2) yyerror( "flags only apply to tcp"); if ($8.src_os) yyerror( "OS fingerprinting only " "applies to tcp"); YYERROR; } } r.tos = $9.tos; if ($9.keep.action) { yyerror("cannot specify state handling " "on anchors"); YYERROR; } if ($9.match_tag) if (strlcpy(r.match_tagname, $9.match_tag, PF_TAG_NAME_SIZE) >= PF_TAG_NAME_SIZE) { yyerror("tag too long, max %u chars", PF_TAG_NAME_SIZE - 1); YYERROR; } r.match_tag_not = $9.match_tag_not; decide_address_family($8.src.host, &r.af); decide_address_family($8.dst.host, &r.af); expand_rule(&r, $5, NULL, $7, $8.src_os, $8.src.host, $8.src.port, $8.dst.host, $8.dst.port, $9.uid, $9.gid, $9.icmpspec, pf->astack[pf->asd + 1] ? pf->alast->name : $2); free($2); pf->astack[pf->asd + 1] = NULL; } | NATANCHOR string interface af proto fromto rtable { struct pf_rule r; if (check_rulestate(PFCTL_STATE_NAT)) { free($2); YYERROR; } memset(&r, 0, sizeof(r)); r.action = PF_NAT; r.af = $4; r.rtableid = $7; decide_address_family($6.src.host, &r.af); decide_address_family($6.dst.host, &r.af); expand_rule(&r, $3, NULL, $5, $6.src_os, $6.src.host, $6.src.port, $6.dst.host, $6.dst.port, 0, 0, 0, $2); free($2); } | RDRANCHOR string interface af proto fromto rtable { struct pf_rule r; if (check_rulestate(PFCTL_STATE_NAT)) { free($2); YYERROR; } memset(&r, 0, sizeof(r)); r.action = PF_RDR; r.af = $4; r.rtableid = $7; decide_address_family($6.src.host, &r.af); decide_address_family($6.dst.host, &r.af); if ($6.src.port != NULL) { yyerror("source port parameter not supported" " in rdr-anchor"); YYERROR; } if ($6.dst.port != NULL) { if ($6.dst.port->next != NULL) { yyerror("destination port list " "expansion not supported in " "rdr-anchor"); YYERROR; } else if ($6.dst.port->op != PF_OP_EQ) { yyerror("destination port operators" " not supported in rdr-anchor"); YYERROR; } r.dst.port[0] = $6.dst.port->port[0]; r.dst.port[1] = $6.dst.port->port[1]; r.dst.port_op = $6.dst.port->op; } expand_rule(&r, $3, NULL, $5, $6.src_os, $6.src.host, $6.src.port, $6.dst.host, $6.dst.port, 0, 0, 0, $2); free($2); } | BINATANCHOR string interface af proto fromto rtable { struct pf_rule r; if (check_rulestate(PFCTL_STATE_NAT)) { free($2); YYERROR; } memset(&r, 0, sizeof(r)); r.action = PF_BINAT; r.af = $4; r.rtableid = $7; if ($5 != NULL) { if ($5->next != NULL) { yyerror("proto list expansion" " not supported in binat-anchor"); YYERROR; } r.proto = $5->proto; free($5); } if ($6.src.host != NULL || $6.src.port != NULL || $6.dst.host != NULL || $6.dst.port != NULL) { yyerror("fromto parameter not supported" " in binat-anchor"); YYERROR; } decide_address_family($6.src.host, &r.af); decide_address_family($6.dst.host, &r.af); pfctl_add_rule(pf, &r, $2); free($2); } ; loadrule : LOAD ANCHOR string FROM string { struct loadanchors *loadanchor; if (strlen(pf->anchor->name) + 1 + strlen($3) >= MAXPATHLEN) { yyerror("anchorname %s too long, max %u\n", $3, MAXPATHLEN - 1); free($3); YYERROR; } loadanchor = calloc(1, sizeof(struct loadanchors)); if (loadanchor == NULL) err(1, "loadrule: calloc"); if ((loadanchor->anchorname = malloc(MAXPATHLEN)) == NULL) err(1, "loadrule: malloc"); if (pf->anchor->name[0]) snprintf(loadanchor->anchorname, MAXPATHLEN, "%s/%s", pf->anchor->name, $3); else strlcpy(loadanchor->anchorname, $3, MAXPATHLEN); if ((loadanchor->filename = strdup($5)) == NULL) err(1, "loadrule: strdup"); TAILQ_INSERT_TAIL(&loadanchorshead, loadanchor, entries); free($3); free($5); }; scrubaction : no SCRUB { $$.b2 = $$.w = 0; if ($1) $$.b1 = PF_NOSCRUB; else $$.b1 = PF_SCRUB; } ; scrubrule : scrubaction dir logquick interface af proto fromto scrub_opts { struct pf_rule r; if (check_rulestate(PFCTL_STATE_SCRUB)) YYERROR; memset(&r, 0, sizeof(r)); r.action = $1.b1; r.direction = $2; r.log = $3.log; r.logif = $3.logif; if ($3.quick) { yyerror("scrub rules do not support 'quick'"); YYERROR; } r.af = $5; if ($8.nodf) r.rule_flag |= PFRULE_NODF; if ($8.randomid) r.rule_flag |= PFRULE_RANDOMID; if ($8.reassemble_tcp) { if (r.direction != PF_INOUT) { yyerror("reassemble tcp rules can not " "specify direction"); YYERROR; } r.rule_flag |= PFRULE_REASSEMBLE_TCP; } if ($8.minttl) r.min_ttl = $8.minttl; if ($8.maxmss) r.max_mss = $8.maxmss; if ($8.marker & SOM_SETTOS) { r.rule_flag |= PFRULE_SET_TOS; r.set_tos = $8.settos; } if ($8.fragcache) r.rule_flag |= $8.fragcache; if ($8.match_tag) if (strlcpy(r.match_tagname, $8.match_tag, PF_TAG_NAME_SIZE) >= PF_TAG_NAME_SIZE) { yyerror("tag too long, max %u chars", PF_TAG_NAME_SIZE - 1); YYERROR; } r.match_tag_not = $8.match_tag_not; r.rtableid = $8.rtableid; expand_rule(&r, $4, NULL, $6, $7.src_os, $7.src.host, $7.src.port, $7.dst.host, $7.dst.port, NULL, NULL, NULL, ""); } ; scrub_opts : { bzero(&scrub_opts, sizeof scrub_opts); scrub_opts.rtableid = -1; } scrub_opts_l { $$ = scrub_opts; } | /* empty */ { bzero(&scrub_opts, sizeof scrub_opts); scrub_opts.rtableid = -1; $$ = scrub_opts; } ; scrub_opts_l : scrub_opts_l scrub_opt | scrub_opt ; scrub_opt : NODF { if (scrub_opts.nodf) { yyerror("no-df cannot be respecified"); YYERROR; } scrub_opts.nodf = 1; } | MINTTL NUMBER { if (scrub_opts.marker & SOM_MINTTL) { yyerror("min-ttl cannot be respecified"); YYERROR; } if ($2 < 0 || $2 > 255) { yyerror("illegal min-ttl value %d", $2); YYERROR; } scrub_opts.marker |= SOM_MINTTL; scrub_opts.minttl = $2; } | MAXMSS NUMBER { if (scrub_opts.marker & SOM_MAXMSS) { yyerror("max-mss cannot be respecified"); YYERROR; } if ($2 < 0 || $2 > 65535) { yyerror("illegal max-mss value %d", $2); YYERROR; } scrub_opts.marker |= SOM_MAXMSS; scrub_opts.maxmss = $2; } | SETTOS tos { if (scrub_opts.marker & SOM_SETTOS) { yyerror("set-tos cannot be respecified"); YYERROR; } scrub_opts.marker |= SOM_SETTOS; scrub_opts.settos = $2; } | fragcache { if (scrub_opts.marker & SOM_FRAGCACHE) { yyerror("fragcache cannot be respecified"); YYERROR; } scrub_opts.marker |= SOM_FRAGCACHE; scrub_opts.fragcache = $1; } | REASSEMBLE STRING { if (strcasecmp($2, "tcp") != 0) { yyerror("scrub reassemble supports only tcp, " "not '%s'", $2); free($2); YYERROR; } free($2); if (scrub_opts.reassemble_tcp) { yyerror("reassemble tcp cannot be respecified"); YYERROR; } scrub_opts.reassemble_tcp = 1; } | RANDOMID { if (scrub_opts.randomid) { yyerror("random-id cannot be respecified"); YYERROR; } scrub_opts.randomid = 1; } | RTABLE NUMBER { if ($2 < 0 || $2 > rt_tableid_max()) { yyerror("invalid rtable id"); YYERROR; } scrub_opts.rtableid = $2; } | not TAGGED string { scrub_opts.match_tag = $3; scrub_opts.match_tag_not = $1; } ; fragcache : FRAGMENT REASSEMBLE { $$ = 0; /* default */ } - | FRAGMENT FRAGCROP { $$ = PFRULE_FRAGCROP; } - | FRAGMENT FRAGDROP { $$ = PFRULE_FRAGDROP; } + | FRAGMENT FRAGCROP { $$ = 0; } + | FRAGMENT FRAGDROP { $$ = 0; } ; antispoof : ANTISPOOF logquick antispoof_ifspc af antispoof_opts { struct pf_rule r; struct node_host *h = NULL, *hh; struct node_if *i, *j; if (check_rulestate(PFCTL_STATE_FILTER)) YYERROR; for (i = $3; i; i = i->next) { bzero(&r, sizeof(r)); r.action = PF_DROP; r.direction = PF_IN; r.log = $2.log; r.logif = $2.logif; r.quick = $2.quick; r.af = $4; if (rule_label(&r, $5.label)) YYERROR; r.rtableid = $5.rtableid; j = calloc(1, sizeof(struct node_if)); if (j == NULL) err(1, "antispoof: calloc"); if (strlcpy(j->ifname, i->ifname, sizeof(j->ifname)) >= sizeof(j->ifname)) { free(j); yyerror("interface name too long"); YYERROR; } j->not = 1; if (i->dynamic) { h = calloc(1, sizeof(*h)); if (h == NULL) err(1, "address: calloc"); h->addr.type = PF_ADDR_DYNIFTL; set_ipmask(h, 128); if (strlcpy(h->addr.v.ifname, i->ifname, sizeof(h->addr.v.ifname)) >= sizeof(h->addr.v.ifname)) { free(h); yyerror( "interface name too long"); YYERROR; } hh = malloc(sizeof(*hh)); if (hh == NULL) err(1, "address: malloc"); bcopy(h, hh, sizeof(*hh)); h->addr.iflags = PFI_AFLAG_NETWORK; } else { h = ifa_lookup(j->ifname, PFI_AFLAG_NETWORK); hh = NULL; } if (h != NULL) expand_rule(&r, j, NULL, NULL, NULL, h, NULL, NULL, NULL, NULL, NULL, NULL, ""); if ((i->ifa_flags & IFF_LOOPBACK) == 0) { bzero(&r, sizeof(r)); r.action = PF_DROP; r.direction = PF_IN; r.log = $2.log; r.logif = $2.logif; r.quick = $2.quick; r.af = $4; if (rule_label(&r, $5.label)) YYERROR; r.rtableid = $5.rtableid; if (hh != NULL) h = hh; else h = ifa_lookup(i->ifname, 0); if (h != NULL) expand_rule(&r, NULL, NULL, NULL, NULL, h, NULL, NULL, NULL, NULL, NULL, NULL, ""); } else free(hh); } free($5.label); } ; antispoof_ifspc : FOR antispoof_if { $$ = $2; } | FOR '{' optnl antispoof_iflst '}' { $$ = $4; } ; antispoof_iflst : antispoof_if optnl { $$ = $1; } | antispoof_iflst comma antispoof_if optnl { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; antispoof_if : if_item { $$ = $1; } | '(' if_item ')' { $2->dynamic = 1; $$ = $2; } ; antispoof_opts : { bzero(&antispoof_opts, sizeof antispoof_opts); antispoof_opts.rtableid = -1; } antispoof_opts_l { $$ = antispoof_opts; } | /* empty */ { bzero(&antispoof_opts, sizeof antispoof_opts); antispoof_opts.rtableid = -1; $$ = antispoof_opts; } ; antispoof_opts_l : antispoof_opts_l antispoof_opt | antispoof_opt ; antispoof_opt : label { if (antispoof_opts.label) { yyerror("label cannot be redefined"); YYERROR; } antispoof_opts.label = $1; } | RTABLE NUMBER { if ($2 < 0 || $2 > rt_tableid_max()) { yyerror("invalid rtable id"); YYERROR; } antispoof_opts.rtableid = $2; } ; not : '!' { $$ = 1; } | /* empty */ { $$ = 0; } ; tabledef : TABLE '<' STRING '>' table_opts { struct node_host *h, *nh; struct node_tinit *ti, *nti; if (strlen($3) >= PF_TABLE_NAME_SIZE) { yyerror("table name too long, max %d chars", PF_TABLE_NAME_SIZE - 1); free($3); YYERROR; } if (pf->loadopt & PFCTL_FLAG_TABLE) if (process_tabledef($3, &$5)) { free($3); YYERROR; } free($3); for (ti = SIMPLEQ_FIRST(&$5.init_nodes); ti != SIMPLEQ_END(&$5.init_nodes); ti = nti) { if (ti->file) free(ti->file); for (h = ti->host; h != NULL; h = nh) { nh = h->next; free(h); } nti = SIMPLEQ_NEXT(ti, entries); free(ti); } } ; table_opts : { bzero(&table_opts, sizeof table_opts); SIMPLEQ_INIT(&table_opts.init_nodes); } table_opts_l { $$ = table_opts; } | /* empty */ { bzero(&table_opts, sizeof table_opts); SIMPLEQ_INIT(&table_opts.init_nodes); $$ = table_opts; } ; table_opts_l : table_opts_l table_opt | table_opt ; table_opt : STRING { if (!strcmp($1, "const")) table_opts.flags |= PFR_TFLAG_CONST; else if (!strcmp($1, "persist")) table_opts.flags |= PFR_TFLAG_PERSIST; else if (!strcmp($1, "counters")) table_opts.flags |= PFR_TFLAG_COUNTERS; else { yyerror("invalid table option '%s'", $1); free($1); YYERROR; } free($1); } | '{' optnl '}' { table_opts.init_addr = 1; } | '{' optnl host_list '}' { struct node_host *n; struct node_tinit *ti; for (n = $3; n != NULL; n = n->next) { switch (n->addr.type) { case PF_ADDR_ADDRMASK: continue; /* ok */ case PF_ADDR_RANGE: yyerror("address ranges are not " "permitted inside tables"); break; case PF_ADDR_DYNIFTL: yyerror("dynamic addresses are not " "permitted inside tables"); break; case PF_ADDR_TABLE: yyerror("tables cannot contain tables"); break; case PF_ADDR_NOROUTE: yyerror("\"no-route\" is not permitted " "inside tables"); break; case PF_ADDR_URPFFAILED: yyerror("\"urpf-failed\" is not " "permitted inside tables"); break; default: yyerror("unknown address type %d", n->addr.type); } YYERROR; } if (!(ti = calloc(1, sizeof(*ti)))) err(1, "table_opt: calloc"); ti->host = $3; SIMPLEQ_INSERT_TAIL(&table_opts.init_nodes, ti, entries); table_opts.init_addr = 1; } | FILENAME STRING { struct node_tinit *ti; if (!(ti = calloc(1, sizeof(*ti)))) err(1, "table_opt: calloc"); ti->file = $2; SIMPLEQ_INSERT_TAIL(&table_opts.init_nodes, ti, entries); table_opts.init_addr = 1; } ; altqif : ALTQ interface queue_opts QUEUE qassign { struct pf_altq a; if (check_rulestate(PFCTL_STATE_QUEUE)) YYERROR; memset(&a, 0, sizeof(a)); if ($3.scheduler.qtype == ALTQT_NONE) { yyerror("no scheduler specified!"); YYERROR; } a.scheduler = $3.scheduler.qtype; a.qlimit = $3.qlimit; a.tbrsize = $3.tbrsize; if ($5 == NULL && $3.scheduler.qtype != ALTQT_CODEL) { yyerror("no child queues specified"); YYERROR; } if (expand_altq(&a, $2, $5, $3.queue_bwspec, &$3.scheduler)) YYERROR; } ; queuespec : QUEUE STRING interface queue_opts qassign { struct pf_altq a; if (check_rulestate(PFCTL_STATE_QUEUE)) { free($2); YYERROR; } memset(&a, 0, sizeof(a)); if (strlcpy(a.qname, $2, sizeof(a.qname)) >= sizeof(a.qname)) { yyerror("queue name too long (max " "%d chars)", PF_QNAME_SIZE-1); free($2); YYERROR; } free($2); if ($4.tbrsize) { yyerror("cannot specify tbrsize for queue"); YYERROR; } if ($4.priority > 255) { yyerror("priority out of range: max 255"); YYERROR; } a.priority = $4.priority; a.qlimit = $4.qlimit; a.scheduler = $4.scheduler.qtype; if (expand_queue(&a, $3, $5, $4.queue_bwspec, &$4.scheduler)) { yyerror("errors in queue definition"); YYERROR; } } ; queue_opts : { bzero(&queue_opts, sizeof queue_opts); queue_opts.priority = DEFAULT_PRIORITY; queue_opts.qlimit = DEFAULT_QLIMIT; queue_opts.scheduler.qtype = ALTQT_NONE; queue_opts.queue_bwspec.bw_percent = 100; } queue_opts_l { $$ = queue_opts; } | /* empty */ { bzero(&queue_opts, sizeof queue_opts); queue_opts.priority = DEFAULT_PRIORITY; queue_opts.qlimit = DEFAULT_QLIMIT; queue_opts.scheduler.qtype = ALTQT_NONE; queue_opts.queue_bwspec.bw_percent = 100; $$ = queue_opts; } ; queue_opts_l : queue_opts_l queue_opt | queue_opt ; queue_opt : BANDWIDTH bandwidth { if (queue_opts.marker & QOM_BWSPEC) { yyerror("bandwidth cannot be respecified"); YYERROR; } queue_opts.marker |= QOM_BWSPEC; queue_opts.queue_bwspec = $2; } | PRIORITY NUMBER { if (queue_opts.marker & QOM_PRIORITY) { yyerror("priority cannot be respecified"); YYERROR; } if ($2 < 0 || $2 > 255) { yyerror("priority out of range: max 255"); YYERROR; } queue_opts.marker |= QOM_PRIORITY; queue_opts.priority = $2; } | QLIMIT NUMBER { if (queue_opts.marker & QOM_QLIMIT) { yyerror("qlimit cannot be respecified"); YYERROR; } if ($2 < 0 || $2 > 65535) { yyerror("qlimit out of range: max 65535"); YYERROR; } queue_opts.marker |= QOM_QLIMIT; queue_opts.qlimit = $2; } | scheduler { if (queue_opts.marker & QOM_SCHEDULER) { yyerror("scheduler cannot be respecified"); YYERROR; } queue_opts.marker |= QOM_SCHEDULER; queue_opts.scheduler = $1; } | TBRSIZE NUMBER { if (queue_opts.marker & QOM_TBRSIZE) { yyerror("tbrsize cannot be respecified"); YYERROR; } if ($2 < 0 || $2 > 65535) { yyerror("tbrsize too big: max 65535"); YYERROR; } queue_opts.marker |= QOM_TBRSIZE; queue_opts.tbrsize = $2; } ; bandwidth : STRING { double bps; char *cp; $$.bw_percent = 0; bps = strtod($1, &cp); if (cp != NULL) { if (!strcmp(cp, "b")) ; /* nothing */ else if (!strcmp(cp, "Kb")) bps *= 1000; else if (!strcmp(cp, "Mb")) bps *= 1000 * 1000; else if (!strcmp(cp, "Gb")) bps *= 1000 * 1000 * 1000; else if (!strcmp(cp, "%")) { if (bps < 0 || bps > 100) { yyerror("bandwidth spec " "out of range"); free($1); YYERROR; } $$.bw_percent = bps; bps = 0; } else { yyerror("unknown unit %s", cp); free($1); YYERROR; } } free($1); $$.bw_absolute = (u_int32_t)bps; } | NUMBER { if ($1 < 0 || $1 > UINT_MAX) { yyerror("bandwidth number too big"); YYERROR; } $$.bw_percent = 0; $$.bw_absolute = $1; } ; scheduler : CBQ { $$.qtype = ALTQT_CBQ; $$.data.cbq_opts.flags = 0; } | CBQ '(' cbqflags_list ')' { $$.qtype = ALTQT_CBQ; $$.data.cbq_opts.flags = $3; } | PRIQ { $$.qtype = ALTQT_PRIQ; $$.data.priq_opts.flags = 0; } | PRIQ '(' priqflags_list ')' { $$.qtype = ALTQT_PRIQ; $$.data.priq_opts.flags = $3; } | HFSC { $$.qtype = ALTQT_HFSC; bzero(&$$.data.hfsc_opts, sizeof(struct node_hfsc_opts)); } | HFSC '(' hfsc_opts ')' { $$.qtype = ALTQT_HFSC; $$.data.hfsc_opts = $3; } | FAIRQ { $$.qtype = ALTQT_FAIRQ; bzero(&$$.data.fairq_opts, sizeof(struct node_fairq_opts)); } | FAIRQ '(' fairq_opts ')' { $$.qtype = ALTQT_FAIRQ; $$.data.fairq_opts = $3; } | CODEL { $$.qtype = ALTQT_CODEL; bzero(&$$.data.codel_opts, sizeof(struct codel_opts)); } | CODEL '(' codel_opts ')' { $$.qtype = ALTQT_CODEL; $$.data.codel_opts = $3; } ; cbqflags_list : cbqflags_item { $$ |= $1; } | cbqflags_list comma cbqflags_item { $$ |= $3; } ; cbqflags_item : STRING { if (!strcmp($1, "default")) $$ = CBQCLF_DEFCLASS; else if (!strcmp($1, "borrow")) $$ = CBQCLF_BORROW; else if (!strcmp($1, "red")) $$ = CBQCLF_RED; else if (!strcmp($1, "ecn")) $$ = CBQCLF_RED|CBQCLF_ECN; else if (!strcmp($1, "rio")) $$ = CBQCLF_RIO; else if (!strcmp($1, "codel")) $$ = CBQCLF_CODEL; else { yyerror("unknown cbq flag \"%s\"", $1); free($1); YYERROR; } free($1); } ; priqflags_list : priqflags_item { $$ |= $1; } | priqflags_list comma priqflags_item { $$ |= $3; } ; priqflags_item : STRING { if (!strcmp($1, "default")) $$ = PRCF_DEFAULTCLASS; else if (!strcmp($1, "red")) $$ = PRCF_RED; else if (!strcmp($1, "ecn")) $$ = PRCF_RED|PRCF_ECN; else if (!strcmp($1, "rio")) $$ = PRCF_RIO; else if (!strcmp($1, "codel")) $$ = PRCF_CODEL; else { yyerror("unknown priq flag \"%s\"", $1); free($1); YYERROR; } free($1); } ; hfsc_opts : { bzero(&hfsc_opts, sizeof(struct node_hfsc_opts)); } hfscopts_list { $$ = hfsc_opts; } ; hfscopts_list : hfscopts_item | hfscopts_list comma hfscopts_item ; hfscopts_item : LINKSHARE bandwidth { if (hfsc_opts.linkshare.used) { yyerror("linkshare already specified"); YYERROR; } hfsc_opts.linkshare.m2 = $2; hfsc_opts.linkshare.used = 1; } | LINKSHARE '(' bandwidth comma NUMBER comma bandwidth ')' { if ($5 < 0 || $5 > INT_MAX) { yyerror("timing in curve out of range"); YYERROR; } if (hfsc_opts.linkshare.used) { yyerror("linkshare already specified"); YYERROR; } hfsc_opts.linkshare.m1 = $3; hfsc_opts.linkshare.d = $5; hfsc_opts.linkshare.m2 = $7; hfsc_opts.linkshare.used = 1; } | REALTIME bandwidth { if (hfsc_opts.realtime.used) { yyerror("realtime already specified"); YYERROR; } hfsc_opts.realtime.m2 = $2; hfsc_opts.realtime.used = 1; } | REALTIME '(' bandwidth comma NUMBER comma bandwidth ')' { if ($5 < 0 || $5 > INT_MAX) { yyerror("timing in curve out of range"); YYERROR; } if (hfsc_opts.realtime.used) { yyerror("realtime already specified"); YYERROR; } hfsc_opts.realtime.m1 = $3; hfsc_opts.realtime.d = $5; hfsc_opts.realtime.m2 = $7; hfsc_opts.realtime.used = 1; } | UPPERLIMIT bandwidth { if (hfsc_opts.upperlimit.used) { yyerror("upperlimit already specified"); YYERROR; } hfsc_opts.upperlimit.m2 = $2; hfsc_opts.upperlimit.used = 1; } | UPPERLIMIT '(' bandwidth comma NUMBER comma bandwidth ')' { if ($5 < 0 || $5 > INT_MAX) { yyerror("timing in curve out of range"); YYERROR; } if (hfsc_opts.upperlimit.used) { yyerror("upperlimit already specified"); YYERROR; } hfsc_opts.upperlimit.m1 = $3; hfsc_opts.upperlimit.d = $5; hfsc_opts.upperlimit.m2 = $7; hfsc_opts.upperlimit.used = 1; } | STRING { if (!strcmp($1, "default")) hfsc_opts.flags |= HFCF_DEFAULTCLASS; else if (!strcmp($1, "red")) hfsc_opts.flags |= HFCF_RED; else if (!strcmp($1, "ecn")) hfsc_opts.flags |= HFCF_RED|HFCF_ECN; else if (!strcmp($1, "rio")) hfsc_opts.flags |= HFCF_RIO; else if (!strcmp($1, "codel")) hfsc_opts.flags |= HFCF_CODEL; else { yyerror("unknown hfsc flag \"%s\"", $1); free($1); YYERROR; } free($1); } ; fairq_opts : { bzero(&fairq_opts, sizeof(struct node_fairq_opts)); } fairqopts_list { $$ = fairq_opts; } ; fairqopts_list : fairqopts_item | fairqopts_list comma fairqopts_item ; fairqopts_item : LINKSHARE bandwidth { if (fairq_opts.linkshare.used) { yyerror("linkshare already specified"); YYERROR; } fairq_opts.linkshare.m2 = $2; fairq_opts.linkshare.used = 1; } | LINKSHARE '(' bandwidth number bandwidth ')' { if (fairq_opts.linkshare.used) { yyerror("linkshare already specified"); YYERROR; } fairq_opts.linkshare.m1 = $3; fairq_opts.linkshare.d = $4; fairq_opts.linkshare.m2 = $5; fairq_opts.linkshare.used = 1; } | HOGS bandwidth { fairq_opts.hogs_bw = $2; } | BUCKETS number { fairq_opts.nbuckets = $2; } | STRING { if (!strcmp($1, "default")) fairq_opts.flags |= FARF_DEFAULTCLASS; else if (!strcmp($1, "red")) fairq_opts.flags |= FARF_RED; else if (!strcmp($1, "ecn")) fairq_opts.flags |= FARF_RED|FARF_ECN; else if (!strcmp($1, "rio")) fairq_opts.flags |= FARF_RIO; else if (!strcmp($1, "codel")) fairq_opts.flags |= FARF_CODEL; else { yyerror("unknown fairq flag \"%s\"", $1); free($1); YYERROR; } free($1); } ; codel_opts : { bzero(&codel_opts, sizeof(struct codel_opts)); } codelopts_list { $$ = codel_opts; } ; codelopts_list : codelopts_item | codelopts_list comma codelopts_item ; codelopts_item : INTERVAL number { if (codel_opts.interval) { yyerror("interval already specified"); YYERROR; } codel_opts.interval = $2; } | TARGET number { if (codel_opts.target) { yyerror("target already specified"); YYERROR; } codel_opts.target = $2; } | STRING { if (!strcmp($1, "ecn")) codel_opts.ecn = 1; else { yyerror("unknown codel option \"%s\"", $1); free($1); YYERROR; } free($1); } ; qassign : /* empty */ { $$ = NULL; } | qassign_item { $$ = $1; } | '{' optnl qassign_list '}' { $$ = $3; } ; qassign_list : qassign_item optnl { $$ = $1; } | qassign_list comma qassign_item optnl { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; qassign_item : STRING { $$ = calloc(1, sizeof(struct node_queue)); if ($$ == NULL) err(1, "qassign_item: calloc"); if (strlcpy($$->queue, $1, sizeof($$->queue)) >= sizeof($$->queue)) { yyerror("queue name '%s' too long (max " "%d chars)", $1, sizeof($$->queue)-1); free($1); free($$); YYERROR; } free($1); $$->next = NULL; $$->tail = $$; } ; pfrule : action dir logquick interface route af proto fromto filter_opts { struct pf_rule r; struct node_state_opt *o; struct node_proto *proto; int srctrack = 0; int statelock = 0; int adaptive = 0; int defaults = 0; if (check_rulestate(PFCTL_STATE_FILTER)) YYERROR; memset(&r, 0, sizeof(r)); r.action = $1.b1; switch ($1.b2) { case PFRULE_RETURNRST: r.rule_flag |= PFRULE_RETURNRST; r.return_ttl = $1.w; break; case PFRULE_RETURNICMP: r.rule_flag |= PFRULE_RETURNICMP; r.return_icmp = $1.w; r.return_icmp6 = $1.w2; break; case PFRULE_RETURN: r.rule_flag |= PFRULE_RETURN; r.return_icmp = $1.w; r.return_icmp6 = $1.w2; break; } r.direction = $2; r.log = $3.log; r.logif = $3.logif; r.quick = $3.quick; r.prob = $9.prob; r.rtableid = $9.rtableid; r.af = $6; if ($9.tag) if (strlcpy(r.tagname, $9.tag, PF_TAG_NAME_SIZE) >= PF_TAG_NAME_SIZE) { yyerror("tag too long, max %u chars", PF_TAG_NAME_SIZE - 1); YYERROR; } if ($9.match_tag) if (strlcpy(r.match_tagname, $9.match_tag, PF_TAG_NAME_SIZE) >= PF_TAG_NAME_SIZE) { yyerror("tag too long, max %u chars", PF_TAG_NAME_SIZE - 1); YYERROR; } r.match_tag_not = $9.match_tag_not; if (rule_label(&r, $9.label)) YYERROR; free($9.label); r.flags = $9.flags.b1; r.flagset = $9.flags.b2; if (($9.flags.b1 & $9.flags.b2) != $9.flags.b1) { yyerror("flags always false"); YYERROR; } if ($9.flags.b1 || $9.flags.b2 || $8.src_os) { for (proto = $7; proto != NULL && proto->proto != IPPROTO_TCP; proto = proto->next) ; /* nothing */ if (proto == NULL && $7 != NULL) { if ($9.flags.b1 || $9.flags.b2) yyerror( "flags only apply to tcp"); if ($8.src_os) yyerror( "OS fingerprinting only " "apply to tcp"); YYERROR; } #if 0 if (($9.flags.b1 & parse_flags("S")) == 0 && $8.src_os) { yyerror("OS fingerprinting requires " "the SYN TCP flag (flags S/SA)"); YYERROR; } #endif } r.tos = $9.tos; r.keep_state = $9.keep.action; o = $9.keep.options; /* 'keep state' by default on pass rules. */ if (!r.keep_state && !r.action && !($9.marker & FOM_KEEP)) { r.keep_state = PF_STATE_NORMAL; o = keep_state_defaults; defaults = 1; } while (o) { struct node_state_opt *p = o; switch (o->type) { case PF_STATE_OPT_MAX: if (r.max_states) { yyerror("state option 'max' " "multiple definitions"); YYERROR; } r.max_states = o->data.max_states; break; case PF_STATE_OPT_NOSYNC: if (r.rule_flag & PFRULE_NOSYNC) { yyerror("state option 'sync' " "multiple definitions"); YYERROR; } r.rule_flag |= PFRULE_NOSYNC; break; case PF_STATE_OPT_SRCTRACK: if (srctrack) { yyerror("state option " "'source-track' " "multiple definitions"); YYERROR; } srctrack = o->data.src_track; r.rule_flag |= PFRULE_SRCTRACK; break; case PF_STATE_OPT_MAX_SRC_STATES: if (r.max_src_states) { yyerror("state option " "'max-src-states' " "multiple definitions"); YYERROR; } if (o->data.max_src_states == 0) { yyerror("'max-src-states' must " "be > 0"); YYERROR; } r.max_src_states = o->data.max_src_states; r.rule_flag |= PFRULE_SRCTRACK; break; case PF_STATE_OPT_OVERLOAD: if (r.overload_tblname[0]) { yyerror("multiple 'overload' " "table definitions"); YYERROR; } if (strlcpy(r.overload_tblname, o->data.overload.tblname, PF_TABLE_NAME_SIZE) >= PF_TABLE_NAME_SIZE) { yyerror("state option: " "strlcpy"); YYERROR; } r.flush = o->data.overload.flush; break; case PF_STATE_OPT_MAX_SRC_CONN: if (r.max_src_conn) { yyerror("state option " "'max-src-conn' " "multiple definitions"); YYERROR; } if (o->data.max_src_conn == 0) { yyerror("'max-src-conn' " "must be > 0"); YYERROR; } r.max_src_conn = o->data.max_src_conn; r.rule_flag |= PFRULE_SRCTRACK | PFRULE_RULESRCTRACK; break; case PF_STATE_OPT_MAX_SRC_CONN_RATE: if (r.max_src_conn_rate.limit) { yyerror("state option " "'max-src-conn-rate' " "multiple definitions"); YYERROR; } if (!o->data.max_src_conn_rate.limit || !o->data.max_src_conn_rate.seconds) { yyerror("'max-src-conn-rate' " "values must be > 0"); YYERROR; } if (o->data.max_src_conn_rate.limit > PF_THRESHOLD_MAX) { yyerror("'max-src-conn-rate' " "maximum rate must be < %u", PF_THRESHOLD_MAX); YYERROR; } r.max_src_conn_rate.limit = o->data.max_src_conn_rate.limit; r.max_src_conn_rate.seconds = o->data.max_src_conn_rate.seconds; r.rule_flag |= PFRULE_SRCTRACK | PFRULE_RULESRCTRACK; break; case PF_STATE_OPT_MAX_SRC_NODES: if (r.max_src_nodes) { yyerror("state option " "'max-src-nodes' " "multiple definitions"); YYERROR; } if (o->data.max_src_nodes == 0) { yyerror("'max-src-nodes' must " "be > 0"); YYERROR; } r.max_src_nodes = o->data.max_src_nodes; r.rule_flag |= PFRULE_SRCTRACK | PFRULE_RULESRCTRACK; break; case PF_STATE_OPT_STATELOCK: if (statelock) { yyerror("state locking option: " "multiple definitions"); YYERROR; } statelock = 1; r.rule_flag |= o->data.statelock; break; case PF_STATE_OPT_SLOPPY: if (r.rule_flag & PFRULE_STATESLOPPY) { yyerror("state sloppy option: " "multiple definitions"); YYERROR; } r.rule_flag |= PFRULE_STATESLOPPY; break; case PF_STATE_OPT_TIMEOUT: if (o->data.timeout.number == PFTM_ADAPTIVE_START || o->data.timeout.number == PFTM_ADAPTIVE_END) adaptive = 1; if (r.timeout[o->data.timeout.number]) { yyerror("state timeout %s " "multiple definitions", pf_timeouts[o->data. timeout.number].name); YYERROR; } r.timeout[o->data.timeout.number] = o->data.timeout.seconds; } o = o->next; if (!defaults) free(p); } /* 'flags S/SA' by default on stateful rules */ if (!r.action && !r.flags && !r.flagset && !$9.fragment && !($9.marker & FOM_FLAGS) && r.keep_state) { r.flags = parse_flags("S"); r.flagset = parse_flags("SA"); } if (!adaptive && r.max_states) { r.timeout[PFTM_ADAPTIVE_START] = (r.max_states / 10) * 6; r.timeout[PFTM_ADAPTIVE_END] = (r.max_states / 10) * 12; } if (r.rule_flag & PFRULE_SRCTRACK) { if (srctrack == PF_SRCTRACK_GLOBAL && r.max_src_nodes) { yyerror("'max-src-nodes' is " "incompatible with " "'source-track global'"); YYERROR; } if (srctrack == PF_SRCTRACK_GLOBAL && r.max_src_conn) { yyerror("'max-src-conn' is " "incompatible with " "'source-track global'"); YYERROR; } if (srctrack == PF_SRCTRACK_GLOBAL && r.max_src_conn_rate.seconds) { yyerror("'max-src-conn-rate' is " "incompatible with " "'source-track global'"); YYERROR; } if (r.timeout[PFTM_SRC_NODE] < r.max_src_conn_rate.seconds) r.timeout[PFTM_SRC_NODE] = r.max_src_conn_rate.seconds; r.rule_flag |= PFRULE_SRCTRACK; if (srctrack == PF_SRCTRACK_RULE) r.rule_flag |= PFRULE_RULESRCTRACK; } if (r.keep_state && !statelock) r.rule_flag |= default_statelock; if ($9.fragment) r.rule_flag |= PFRULE_FRAGMENT; r.allow_opts = $9.allowopts; decide_address_family($8.src.host, &r.af); decide_address_family($8.dst.host, &r.af); if ($5.rt) { if (!r.direction) { yyerror("direction must be explicit " "with rules that specify routing"); YYERROR; } r.rt = $5.rt; r.rpool.opts = $5.pool_opts; if ($5.key != NULL) memcpy(&r.rpool.key, $5.key, sizeof(struct pf_poolhashkey)); } if (r.rt && r.rt != PF_FASTROUTE) { decide_address_family($5.host, &r.af); remove_invalid_hosts(&$5.host, &r.af); if ($5.host == NULL) { yyerror("no routing address with " "matching address family found."); YYERROR; } if ((r.rpool.opts & PF_POOL_TYPEMASK) == PF_POOL_NONE && ($5.host->next != NULL || $5.host->addr.type == PF_ADDR_TABLE || DYNIF_MULTIADDR($5.host->addr))) r.rpool.opts |= PF_POOL_ROUNDROBIN; if ((r.rpool.opts & PF_POOL_TYPEMASK) != PF_POOL_ROUNDROBIN && disallow_table($5.host, "tables are only " "supported in round-robin routing pools")) YYERROR; if ((r.rpool.opts & PF_POOL_TYPEMASK) != PF_POOL_ROUNDROBIN && disallow_alias($5.host, "interface (%s) " "is only supported in round-robin " "routing pools")) YYERROR; if ($5.host->next != NULL) { if ((r.rpool.opts & PF_POOL_TYPEMASK) != PF_POOL_ROUNDROBIN) { yyerror("r.rpool.opts must " "be PF_POOL_ROUNDROBIN"); YYERROR; } } } if ($9.queues.qname != NULL) { if (strlcpy(r.qname, $9.queues.qname, sizeof(r.qname)) >= sizeof(r.qname)) { yyerror("rule qname too long (max " "%d chars)", sizeof(r.qname)-1); YYERROR; } free($9.queues.qname); } if ($9.queues.pqname != NULL) { if (strlcpy(r.pqname, $9.queues.pqname, sizeof(r.pqname)) >= sizeof(r.pqname)) { yyerror("rule pqname too long (max " "%d chars)", sizeof(r.pqname)-1); YYERROR; } free($9.queues.pqname); } #ifdef __FreeBSD__ r.divert.port = $9.divert.port; #else if ((r.divert.port = $9.divert.port)) { if (r.direction == PF_OUT) { if ($9.divert.addr) { yyerror("address specified " "for outgoing divert"); YYERROR; } bzero(&r.divert.addr, sizeof(r.divert.addr)); } else { if (!$9.divert.addr) { yyerror("no address specified " "for incoming divert"); YYERROR; } if ($9.divert.addr->af != r.af) { yyerror("address family " "mismatch for divert"); YYERROR; } r.divert.addr = $9.divert.addr->addr.v.a.addr; } } #endif expand_rule(&r, $4, $5.host, $7, $8.src_os, $8.src.host, $8.src.port, $8.dst.host, $8.dst.port, $9.uid, $9.gid, $9.icmpspec, ""); } ; filter_opts : { bzero(&filter_opts, sizeof filter_opts); filter_opts.rtableid = -1; } filter_opts_l { $$ = filter_opts; } | /* empty */ { bzero(&filter_opts, sizeof filter_opts); filter_opts.rtableid = -1; $$ = filter_opts; } ; filter_opts_l : filter_opts_l filter_opt | filter_opt ; filter_opt : USER uids { if (filter_opts.uid) $2->tail->next = filter_opts.uid; filter_opts.uid = $2; } | GROUP gids { if (filter_opts.gid) $2->tail->next = filter_opts.gid; filter_opts.gid = $2; } | flags { if (filter_opts.marker & FOM_FLAGS) { yyerror("flags cannot be redefined"); YYERROR; } filter_opts.marker |= FOM_FLAGS; filter_opts.flags.b1 |= $1.b1; filter_opts.flags.b2 |= $1.b2; filter_opts.flags.w |= $1.w; filter_opts.flags.w2 |= $1.w2; } | icmpspec { if (filter_opts.marker & FOM_ICMP) { yyerror("icmp-type cannot be redefined"); YYERROR; } filter_opts.marker |= FOM_ICMP; filter_opts.icmpspec = $1; } | TOS tos { if (filter_opts.marker & FOM_TOS) { yyerror("tos cannot be redefined"); YYERROR; } filter_opts.marker |= FOM_TOS; filter_opts.tos = $2; } | keep { if (filter_opts.marker & FOM_KEEP) { yyerror("modulate or keep cannot be redefined"); YYERROR; } filter_opts.marker |= FOM_KEEP; filter_opts.keep.action = $1.action; filter_opts.keep.options = $1.options; } | FRAGMENT { filter_opts.fragment = 1; } | ALLOWOPTS { filter_opts.allowopts = 1; } | label { if (filter_opts.label) { yyerror("label cannot be redefined"); YYERROR; } filter_opts.label = $1; } | qname { if (filter_opts.queues.qname) { yyerror("queue cannot be redefined"); YYERROR; } filter_opts.queues = $1; } | TAG string { filter_opts.tag = $2; } | not TAGGED string { filter_opts.match_tag = $3; filter_opts.match_tag_not = $1; } | PROBABILITY probability { double p; p = floor($2 * UINT_MAX + 0.5); if (p < 0.0 || p > UINT_MAX) { yyerror("invalid probability: %lf", p); YYERROR; } filter_opts.prob = (u_int32_t)p; if (filter_opts.prob == 0) filter_opts.prob = 1; } | RTABLE NUMBER { if ($2 < 0 || $2 > rt_tableid_max()) { yyerror("invalid rtable id"); YYERROR; } filter_opts.rtableid = $2; } | DIVERTTO portplain { #ifdef __FreeBSD__ filter_opts.divert.port = $2.a; if (!filter_opts.divert.port) { yyerror("invalid divert port: %u", ntohs($2.a)); YYERROR; } #endif } | DIVERTTO STRING PORT portplain { #ifndef __FreeBSD__ if ((filter_opts.divert.addr = host($2)) == NULL) { yyerror("could not parse divert address: %s", $2); free($2); YYERROR; } #else if ($2) #endif free($2); filter_opts.divert.port = $4.a; if (!filter_opts.divert.port) { yyerror("invalid divert port: %u", ntohs($4.a)); YYERROR; } } | DIVERTREPLY { #ifdef __FreeBSD__ yyerror("divert-reply has no meaning in FreeBSD pf(4)"); YYERROR; #else filter_opts.divert.port = 1; /* some random value */ #endif } ; probability : STRING { char *e; double p = strtod($1, &e); if (*e == '%') { p *= 0.01; e++; } if (*e) { yyerror("invalid probability: %s", $1); free($1); YYERROR; } free($1); $$ = p; } | NUMBER { $$ = (double)$1; } ; action : PASS { $$.b1 = PF_PASS; $$.b2 = $$.w = 0; } | BLOCK blockspec { $$ = $2; $$.b1 = PF_DROP; } ; blockspec : /* empty */ { $$.b2 = blockpolicy; $$.w = returnicmpdefault; $$.w2 = returnicmp6default; } | DROP { $$.b2 = PFRULE_DROP; $$.w = 0; $$.w2 = 0; } | RETURNRST { $$.b2 = PFRULE_RETURNRST; $$.w = 0; $$.w2 = 0; } | RETURNRST '(' TTL NUMBER ')' { if ($4 < 0 || $4 > 255) { yyerror("illegal ttl value %d", $4); YYERROR; } $$.b2 = PFRULE_RETURNRST; $$.w = $4; $$.w2 = 0; } | RETURNICMP { $$.b2 = PFRULE_RETURNICMP; $$.w = returnicmpdefault; $$.w2 = returnicmp6default; } | RETURNICMP6 { $$.b2 = PFRULE_RETURNICMP; $$.w = returnicmpdefault; $$.w2 = returnicmp6default; } | RETURNICMP '(' reticmpspec ')' { $$.b2 = PFRULE_RETURNICMP; $$.w = $3; $$.w2 = returnicmpdefault; } | RETURNICMP6 '(' reticmp6spec ')' { $$.b2 = PFRULE_RETURNICMP; $$.w = returnicmpdefault; $$.w2 = $3; } | RETURNICMP '(' reticmpspec comma reticmp6spec ')' { $$.b2 = PFRULE_RETURNICMP; $$.w = $3; $$.w2 = $5; } | RETURN { $$.b2 = PFRULE_RETURN; $$.w = returnicmpdefault; $$.w2 = returnicmp6default; } ; reticmpspec : STRING { if (!($$ = parseicmpspec($1, AF_INET))) { free($1); YYERROR; } free($1); } | NUMBER { u_int8_t icmptype; if ($1 < 0 || $1 > 255) { yyerror("invalid icmp code %lu", $1); YYERROR; } icmptype = returnicmpdefault >> 8; $$ = (icmptype << 8 | $1); } ; reticmp6spec : STRING { if (!($$ = parseicmpspec($1, AF_INET6))) { free($1); YYERROR; } free($1); } | NUMBER { u_int8_t icmptype; if ($1 < 0 || $1 > 255) { yyerror("invalid icmp code %lu", $1); YYERROR; } icmptype = returnicmp6default >> 8; $$ = (icmptype << 8 | $1); } ; dir : /* empty */ { $$ = PF_INOUT; } | IN { $$ = PF_IN; } | OUT { $$ = PF_OUT; } ; quick : /* empty */ { $$.quick = 0; } | QUICK { $$.quick = 1; } ; logquick : /* empty */ { $$.log = 0; $$.quick = 0; $$.logif = 0; } | log { $$ = $1; $$.quick = 0; } | QUICK { $$.quick = 1; $$.log = 0; $$.logif = 0; } | log QUICK { $$ = $1; $$.quick = 1; } | QUICK log { $$ = $2; $$.quick = 1; } ; log : LOG { $$.log = PF_LOG; $$.logif = 0; } | LOG '(' logopts ')' { $$.log = PF_LOG | $3.log; $$.logif = $3.logif; } ; logopts : logopt { $$ = $1; } | logopts comma logopt { $$.log = $1.log | $3.log; $$.logif = $3.logif; if ($$.logif == 0) $$.logif = $1.logif; } ; logopt : ALL { $$.log = PF_LOG_ALL; $$.logif = 0; } | USER { $$.log = PF_LOG_SOCKET_LOOKUP; $$.logif = 0; } | GROUP { $$.log = PF_LOG_SOCKET_LOOKUP; $$.logif = 0; } | TO string { const char *errstr; u_int i; $$.log = 0; if (strncmp($2, "pflog", 5)) { yyerror("%s: should be a pflog interface", $2); free($2); YYERROR; } i = strtonum($2 + 5, 0, 255, &errstr); if (errstr) { yyerror("%s: %s", $2, errstr); free($2); YYERROR; } free($2); $$.logif = i; } ; interface : /* empty */ { $$ = NULL; } | ON if_item_not { $$ = $2; } | ON '{' optnl if_list '}' { $$ = $4; } ; if_list : if_item_not optnl { $$ = $1; } | if_list comma if_item_not optnl { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; if_item_not : not if_item { $$ = $2; $$->not = $1; } ; if_item : STRING { struct node_host *n; $$ = calloc(1, sizeof(struct node_if)); if ($$ == NULL) err(1, "if_item: calloc"); if (strlcpy($$->ifname, $1, sizeof($$->ifname)) >= sizeof($$->ifname)) { free($1); free($$); yyerror("interface name too long"); YYERROR; } if ((n = ifa_exists($1)) != NULL) $$->ifa_flags = n->ifa_flags; free($1); $$->not = 0; $$->next = NULL; $$->tail = $$; } ; af : /* empty */ { $$ = 0; } | INET { $$ = AF_INET; } | INET6 { $$ = AF_INET6; } ; proto : /* empty */ { $$ = NULL; } | PROTO proto_item { $$ = $2; } | PROTO '{' optnl proto_list '}' { $$ = $4; } ; proto_list : proto_item optnl { $$ = $1; } | proto_list comma proto_item optnl { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; proto_item : protoval { u_int8_t pr; pr = (u_int8_t)$1; if (pr == 0) { yyerror("proto 0 cannot be used"); YYERROR; } $$ = calloc(1, sizeof(struct node_proto)); if ($$ == NULL) err(1, "proto_item: calloc"); $$->proto = pr; $$->next = NULL; $$->tail = $$; } ; protoval : STRING { struct protoent *p; p = getprotobyname($1); if (p == NULL) { yyerror("unknown protocol %s", $1); free($1); YYERROR; } $$ = p->p_proto; free($1); } | NUMBER { if ($1 < 0 || $1 > 255) { yyerror("protocol outside range"); YYERROR; } } ; fromto : ALL { $$.src.host = NULL; $$.src.port = NULL; $$.dst.host = NULL; $$.dst.port = NULL; $$.src_os = NULL; } | from os to { $$.src = $1; $$.src_os = $2; $$.dst = $3; } ; os : /* empty */ { $$ = NULL; } | OS xos { $$ = $2; } | OS '{' optnl os_list '}' { $$ = $4; } ; xos : STRING { $$ = calloc(1, sizeof(struct node_os)); if ($$ == NULL) err(1, "os: calloc"); $$->os = $1; $$->tail = $$; } ; os_list : xos optnl { $$ = $1; } | os_list comma xos optnl { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; from : /* empty */ { $$.host = NULL; $$.port = NULL; } | FROM ipportspec { $$ = $2; } ; to : /* empty */ { $$.host = NULL; $$.port = NULL; } | TO ipportspec { if (disallow_urpf_failed($2.host, "\"urpf-failed\" is " "not permitted in a destination address")) YYERROR; $$ = $2; } ; ipportspec : ipspec { $$.host = $1; $$.port = NULL; } | ipspec PORT portspec { $$.host = $1; $$.port = $3; } | PORT portspec { $$.host = NULL; $$.port = $2; } ; optnl : '\n' optnl | ; ipspec : ANY { $$ = NULL; } | xhost { $$ = $1; } | '{' optnl host_list '}' { $$ = $3; } ; toipspec : TO ipspec { $$ = $2; } | /* empty */ { $$ = NULL; } ; host_list : ipspec optnl { $$ = $1; } | host_list comma ipspec optnl { if ($3 == NULL) $$ = $1; else if ($1 == NULL) $$ = $3; else { $1->tail->next = $3; $1->tail = $3->tail; $$ = $1; } } ; xhost : not host { struct node_host *n; for (n = $2; n != NULL; n = n->next) n->not = $1; $$ = $2; } | not NOROUTE { $$ = calloc(1, sizeof(struct node_host)); if ($$ == NULL) err(1, "xhost: calloc"); $$->addr.type = PF_ADDR_NOROUTE; $$->next = NULL; $$->not = $1; $$->tail = $$; } | not URPFFAILED { $$ = calloc(1, sizeof(struct node_host)); if ($$ == NULL) err(1, "xhost: calloc"); $$->addr.type = PF_ADDR_URPFFAILED; $$->next = NULL; $$->not = $1; $$->tail = $$; } ; host : STRING { if (($$ = host($1)) == NULL) { /* error. "any" is handled elsewhere */ free($1); yyerror("could not parse host specification"); YYERROR; } free($1); } | STRING '-' STRING { struct node_host *b, *e; if ((b = host($1)) == NULL || (e = host($3)) == NULL) { free($1); free($3); yyerror("could not parse host specification"); YYERROR; } if (b->af != e->af || b->addr.type != PF_ADDR_ADDRMASK || e->addr.type != PF_ADDR_ADDRMASK || unmask(&b->addr.v.a.mask, b->af) != (b->af == AF_INET ? 32 : 128) || unmask(&e->addr.v.a.mask, e->af) != (e->af == AF_INET ? 32 : 128) || b->next != NULL || b->not || e->next != NULL || e->not) { free(b); free(e); free($1); free($3); yyerror("invalid address range"); YYERROR; } memcpy(&b->addr.v.a.mask, &e->addr.v.a.addr, sizeof(b->addr.v.a.mask)); b->addr.type = PF_ADDR_RANGE; $$ = b; free(e); free($1); free($3); } | STRING '/' NUMBER { char *buf; if (asprintf(&buf, "%s/%lld", $1, (long long)$3) == -1) err(1, "host: asprintf"); free($1); if (($$ = host(buf)) == NULL) { /* error. "any" is handled elsewhere */ free(buf); yyerror("could not parse host specification"); YYERROR; } free(buf); } | NUMBER '/' NUMBER { char *buf; /* ie. for 10/8 parsing */ #ifdef __FreeBSD__ if (asprintf(&buf, "%lld/%lld", (long long)$1, (long long)$3) == -1) #else if (asprintf(&buf, "%lld/%lld", $1, $3) == -1) #endif err(1, "host: asprintf"); if (($$ = host(buf)) == NULL) { /* error. "any" is handled elsewhere */ free(buf); yyerror("could not parse host specification"); YYERROR; } free(buf); } | dynaddr | dynaddr '/' NUMBER { struct node_host *n; if ($3 < 0 || $3 > 128) { yyerror("bit number too big"); YYERROR; } $$ = $1; for (n = $1; n != NULL; n = n->next) set_ipmask(n, $3); } | '<' STRING '>' { if (strlen($2) >= PF_TABLE_NAME_SIZE) { yyerror("table name '%s' too long", $2); free($2); YYERROR; } $$ = calloc(1, sizeof(struct node_host)); if ($$ == NULL) err(1, "host: calloc"); $$->addr.type = PF_ADDR_TABLE; if (strlcpy($$->addr.v.tblname, $2, sizeof($$->addr.v.tblname)) >= sizeof($$->addr.v.tblname)) errx(1, "host: strlcpy"); free($2); $$->next = NULL; $$->tail = $$; } ; number : NUMBER | STRING { u_long ulval; if (atoul($1, &ulval) == -1) { yyerror("%s is not a number", $1); free($1); YYERROR; } else $$ = ulval; free($1); } ; dynaddr : '(' STRING ')' { int flags = 0; char *p, *op; op = $2; if (!isalpha(op[0])) { yyerror("invalid interface name '%s'", op); free(op); YYERROR; } while ((p = strrchr($2, ':')) != NULL) { if (!strcmp(p+1, "network")) flags |= PFI_AFLAG_NETWORK; else if (!strcmp(p+1, "broadcast")) flags |= PFI_AFLAG_BROADCAST; else if (!strcmp(p+1, "peer")) flags |= PFI_AFLAG_PEER; else if (!strcmp(p+1, "0")) flags |= PFI_AFLAG_NOALIAS; else { yyerror("interface %s has bad modifier", $2); free(op); YYERROR; } *p = '\0'; } if (flags & (flags - 1) & PFI_AFLAG_MODEMASK) { free(op); yyerror("illegal combination of " "interface modifiers"); YYERROR; } $$ = calloc(1, sizeof(struct node_host)); if ($$ == NULL) err(1, "address: calloc"); $$->af = 0; set_ipmask($$, 128); $$->addr.type = PF_ADDR_DYNIFTL; $$->addr.iflags = flags; if (strlcpy($$->addr.v.ifname, $2, sizeof($$->addr.v.ifname)) >= sizeof($$->addr.v.ifname)) { free(op); free($$); yyerror("interface name too long"); YYERROR; } free(op); $$->next = NULL; $$->tail = $$; } ; portspec : port_item { $$ = $1; } | '{' optnl port_list '}' { $$ = $3; } ; port_list : port_item optnl { $$ = $1; } | port_list comma port_item optnl { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; port_item : portrange { $$ = calloc(1, sizeof(struct node_port)); if ($$ == NULL) err(1, "port_item: calloc"); $$->port[0] = $1.a; $$->port[1] = $1.b; if ($1.t) $$->op = PF_OP_RRG; else $$->op = PF_OP_EQ; $$->next = NULL; $$->tail = $$; } | unaryop portrange { if ($2.t) { yyerror("':' cannot be used with an other " "port operator"); YYERROR; } $$ = calloc(1, sizeof(struct node_port)); if ($$ == NULL) err(1, "port_item: calloc"); $$->port[0] = $2.a; $$->port[1] = $2.b; $$->op = $1; $$->next = NULL; $$->tail = $$; } | portrange PORTBINARY portrange { if ($1.t || $3.t) { yyerror("':' cannot be used with an other " "port operator"); YYERROR; } $$ = calloc(1, sizeof(struct node_port)); if ($$ == NULL) err(1, "port_item: calloc"); $$->port[0] = $1.a; $$->port[1] = $3.a; $$->op = $2; $$->next = NULL; $$->tail = $$; } ; portplain : numberstring { if (parseport($1, &$$, 0) == -1) { free($1); YYERROR; } free($1); } ; portrange : numberstring { if (parseport($1, &$$, PPORT_RANGE) == -1) { free($1); YYERROR; } free($1); } ; uids : uid_item { $$ = $1; } | '{' optnl uid_list '}' { $$ = $3; } ; uid_list : uid_item optnl { $$ = $1; } | uid_list comma uid_item optnl { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; uid_item : uid { $$ = calloc(1, sizeof(struct node_uid)); if ($$ == NULL) err(1, "uid_item: calloc"); $$->uid[0] = $1; $$->uid[1] = $1; $$->op = PF_OP_EQ; $$->next = NULL; $$->tail = $$; } | unaryop uid { if ($2 == UID_MAX && $1 != PF_OP_EQ && $1 != PF_OP_NE) { yyerror("user unknown requires operator = or " "!="); YYERROR; } $$ = calloc(1, sizeof(struct node_uid)); if ($$ == NULL) err(1, "uid_item: calloc"); $$->uid[0] = $2; $$->uid[1] = $2; $$->op = $1; $$->next = NULL; $$->tail = $$; } | uid PORTBINARY uid { if ($1 == UID_MAX || $3 == UID_MAX) { yyerror("user unknown requires operator = or " "!="); YYERROR; } $$ = calloc(1, sizeof(struct node_uid)); if ($$ == NULL) err(1, "uid_item: calloc"); $$->uid[0] = $1; $$->uid[1] = $3; $$->op = $2; $$->next = NULL; $$->tail = $$; } ; uid : STRING { if (!strcmp($1, "unknown")) $$ = UID_MAX; else { struct passwd *pw; if ((pw = getpwnam($1)) == NULL) { yyerror("unknown user %s", $1); free($1); YYERROR; } $$ = pw->pw_uid; } free($1); } | NUMBER { if ($1 < 0 || $1 >= UID_MAX) { yyerror("illegal uid value %lu", $1); YYERROR; } $$ = $1; } ; gids : gid_item { $$ = $1; } | '{' optnl gid_list '}' { $$ = $3; } ; gid_list : gid_item optnl { $$ = $1; } | gid_list comma gid_item optnl { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; gid_item : gid { $$ = calloc(1, sizeof(struct node_gid)); if ($$ == NULL) err(1, "gid_item: calloc"); $$->gid[0] = $1; $$->gid[1] = $1; $$->op = PF_OP_EQ; $$->next = NULL; $$->tail = $$; } | unaryop gid { if ($2 == GID_MAX && $1 != PF_OP_EQ && $1 != PF_OP_NE) { yyerror("group unknown requires operator = or " "!="); YYERROR; } $$ = calloc(1, sizeof(struct node_gid)); if ($$ == NULL) err(1, "gid_item: calloc"); $$->gid[0] = $2; $$->gid[1] = $2; $$->op = $1; $$->next = NULL; $$->tail = $$; } | gid PORTBINARY gid { if ($1 == GID_MAX || $3 == GID_MAX) { yyerror("group unknown requires operator = or " "!="); YYERROR; } $$ = calloc(1, sizeof(struct node_gid)); if ($$ == NULL) err(1, "gid_item: calloc"); $$->gid[0] = $1; $$->gid[1] = $3; $$->op = $2; $$->next = NULL; $$->tail = $$; } ; gid : STRING { if (!strcmp($1, "unknown")) $$ = GID_MAX; else { struct group *grp; if ((grp = getgrnam($1)) == NULL) { yyerror("unknown group %s", $1); free($1); YYERROR; } $$ = grp->gr_gid; } free($1); } | NUMBER { if ($1 < 0 || $1 >= GID_MAX) { yyerror("illegal gid value %lu", $1); YYERROR; } $$ = $1; } ; flag : STRING { int f; if ((f = parse_flags($1)) < 0) { yyerror("bad flags %s", $1); free($1); YYERROR; } free($1); $$.b1 = f; } ; flags : FLAGS flag '/' flag { $$.b1 = $2.b1; $$.b2 = $4.b1; } | FLAGS '/' flag { $$.b1 = 0; $$.b2 = $3.b1; } | FLAGS ANY { $$.b1 = 0; $$.b2 = 0; } ; icmpspec : ICMPTYPE icmp_item { $$ = $2; } | ICMPTYPE '{' optnl icmp_list '}' { $$ = $4; } | ICMP6TYPE icmp6_item { $$ = $2; } | ICMP6TYPE '{' optnl icmp6_list '}' { $$ = $4; } ; icmp_list : icmp_item optnl { $$ = $1; } | icmp_list comma icmp_item optnl { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; icmp6_list : icmp6_item optnl { $$ = $1; } | icmp6_list comma icmp6_item optnl { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; icmp_item : icmptype { $$ = calloc(1, sizeof(struct node_icmp)); if ($$ == NULL) err(1, "icmp_item: calloc"); $$->type = $1; $$->code = 0; $$->proto = IPPROTO_ICMP; $$->next = NULL; $$->tail = $$; } | icmptype CODE STRING { const struct icmpcodeent *p; if ((p = geticmpcodebyname($1-1, $3, AF_INET)) == NULL) { yyerror("unknown icmp-code %s", $3); free($3); YYERROR; } free($3); $$ = calloc(1, sizeof(struct node_icmp)); if ($$ == NULL) err(1, "icmp_item: calloc"); $$->type = $1; $$->code = p->code + 1; $$->proto = IPPROTO_ICMP; $$->next = NULL; $$->tail = $$; } | icmptype CODE NUMBER { if ($3 < 0 || $3 > 255) { yyerror("illegal icmp-code %lu", $3); YYERROR; } $$ = calloc(1, sizeof(struct node_icmp)); if ($$ == NULL) err(1, "icmp_item: calloc"); $$->type = $1; $$->code = $3 + 1; $$->proto = IPPROTO_ICMP; $$->next = NULL; $$->tail = $$; } ; icmp6_item : icmp6type { $$ = calloc(1, sizeof(struct node_icmp)); if ($$ == NULL) err(1, "icmp_item: calloc"); $$->type = $1; $$->code = 0; $$->proto = IPPROTO_ICMPV6; $$->next = NULL; $$->tail = $$; } | icmp6type CODE STRING { const struct icmpcodeent *p; if ((p = geticmpcodebyname($1-1, $3, AF_INET6)) == NULL) { yyerror("unknown icmp6-code %s", $3); free($3); YYERROR; } free($3); $$ = calloc(1, sizeof(struct node_icmp)); if ($$ == NULL) err(1, "icmp_item: calloc"); $$->type = $1; $$->code = p->code + 1; $$->proto = IPPROTO_ICMPV6; $$->next = NULL; $$->tail = $$; } | icmp6type CODE NUMBER { if ($3 < 0 || $3 > 255) { yyerror("illegal icmp-code %lu", $3); YYERROR; } $$ = calloc(1, sizeof(struct node_icmp)); if ($$ == NULL) err(1, "icmp_item: calloc"); $$->type = $1; $$->code = $3 + 1; $$->proto = IPPROTO_ICMPV6; $$->next = NULL; $$->tail = $$; } ; icmptype : STRING { const struct icmptypeent *p; if ((p = geticmptypebyname($1, AF_INET)) == NULL) { yyerror("unknown icmp-type %s", $1); free($1); YYERROR; } $$ = p->type + 1; free($1); } | NUMBER { if ($1 < 0 || $1 > 255) { yyerror("illegal icmp-type %lu", $1); YYERROR; } $$ = $1 + 1; } ; icmp6type : STRING { const struct icmptypeent *p; if ((p = geticmptypebyname($1, AF_INET6)) == NULL) { yyerror("unknown icmp6-type %s", $1); free($1); YYERROR; } $$ = p->type + 1; free($1); } | NUMBER { if ($1 < 0 || $1 > 255) { yyerror("illegal icmp6-type %lu", $1); YYERROR; } $$ = $1 + 1; } ; tos : STRING { if (!strcmp($1, "lowdelay")) $$ = IPTOS_LOWDELAY; else if (!strcmp($1, "throughput")) $$ = IPTOS_THROUGHPUT; else if (!strcmp($1, "reliability")) $$ = IPTOS_RELIABILITY; else if ($1[0] == '0' && $1[1] == 'x') $$ = strtoul($1, NULL, 16); else $$ = 0; /* flag bad argument */ if (!$$ || $$ > 255) { yyerror("illegal tos value %s", $1); free($1); YYERROR; } free($1); } | NUMBER { $$ = $1; if (!$$ || $$ > 255) { yyerror("illegal tos value %s", $1); YYERROR; } } ; sourcetrack : SOURCETRACK { $$ = PF_SRCTRACK; } | SOURCETRACK GLOBAL { $$ = PF_SRCTRACK_GLOBAL; } | SOURCETRACK RULE { $$ = PF_SRCTRACK_RULE; } ; statelock : IFBOUND { $$ = PFRULE_IFBOUND; } | FLOATING { $$ = 0; } ; keep : NO STATE { $$.action = 0; $$.options = NULL; } | KEEP STATE state_opt_spec { $$.action = PF_STATE_NORMAL; $$.options = $3; } | MODULATE STATE state_opt_spec { $$.action = PF_STATE_MODULATE; $$.options = $3; } | SYNPROXY STATE state_opt_spec { $$.action = PF_STATE_SYNPROXY; $$.options = $3; } ; flush : /* empty */ { $$ = 0; } | FLUSH { $$ = PF_FLUSH; } | FLUSH GLOBAL { $$ = PF_FLUSH | PF_FLUSH_GLOBAL; } ; state_opt_spec : '(' state_opt_list ')' { $$ = $2; } | /* empty */ { $$ = NULL; } ; state_opt_list : state_opt_item { $$ = $1; } | state_opt_list comma state_opt_item { $1->tail->next = $3; $1->tail = $3; $$ = $1; } ; state_opt_item : MAXIMUM NUMBER { if ($2 < 0 || $2 > UINT_MAX) { yyerror("only positive values permitted"); YYERROR; } $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); $$->type = PF_STATE_OPT_MAX; $$->data.max_states = $2; $$->next = NULL; $$->tail = $$; } | NOSYNC { $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); $$->type = PF_STATE_OPT_NOSYNC; $$->next = NULL; $$->tail = $$; } | MAXSRCSTATES NUMBER { if ($2 < 0 || $2 > UINT_MAX) { yyerror("only positive values permitted"); YYERROR; } $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); $$->type = PF_STATE_OPT_MAX_SRC_STATES; $$->data.max_src_states = $2; $$->next = NULL; $$->tail = $$; } | MAXSRCCONN NUMBER { if ($2 < 0 || $2 > UINT_MAX) { yyerror("only positive values permitted"); YYERROR; } $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); $$->type = PF_STATE_OPT_MAX_SRC_CONN; $$->data.max_src_conn = $2; $$->next = NULL; $$->tail = $$; } | MAXSRCCONNRATE NUMBER '/' NUMBER { if ($2 < 0 || $2 > UINT_MAX || $4 < 0 || $4 > UINT_MAX) { yyerror("only positive values permitted"); YYERROR; } $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); $$->type = PF_STATE_OPT_MAX_SRC_CONN_RATE; $$->data.max_src_conn_rate.limit = $2; $$->data.max_src_conn_rate.seconds = $4; $$->next = NULL; $$->tail = $$; } | OVERLOAD '<' STRING '>' flush { if (strlen($3) >= PF_TABLE_NAME_SIZE) { yyerror("table name '%s' too long", $3); free($3); YYERROR; } $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); if (strlcpy($$->data.overload.tblname, $3, PF_TABLE_NAME_SIZE) >= PF_TABLE_NAME_SIZE) errx(1, "state_opt_item: strlcpy"); free($3); $$->type = PF_STATE_OPT_OVERLOAD; $$->data.overload.flush = $5; $$->next = NULL; $$->tail = $$; } | MAXSRCNODES NUMBER { if ($2 < 0 || $2 > UINT_MAX) { yyerror("only positive values permitted"); YYERROR; } $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); $$->type = PF_STATE_OPT_MAX_SRC_NODES; $$->data.max_src_nodes = $2; $$->next = NULL; $$->tail = $$; } | sourcetrack { $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); $$->type = PF_STATE_OPT_SRCTRACK; $$->data.src_track = $1; $$->next = NULL; $$->tail = $$; } | statelock { $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); $$->type = PF_STATE_OPT_STATELOCK; $$->data.statelock = $1; $$->next = NULL; $$->tail = $$; } | SLOPPY { $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); $$->type = PF_STATE_OPT_SLOPPY; $$->next = NULL; $$->tail = $$; } | STRING NUMBER { int i; if ($2 < 0 || $2 > UINT_MAX) { yyerror("only positive values permitted"); YYERROR; } for (i = 0; pf_timeouts[i].name && strcmp(pf_timeouts[i].name, $1); ++i) ; /* nothing */ if (!pf_timeouts[i].name) { yyerror("illegal timeout name %s", $1); free($1); YYERROR; } if (strchr(pf_timeouts[i].name, '.') == NULL) { yyerror("illegal state timeout %s", $1); free($1); YYERROR; } free($1); $$ = calloc(1, sizeof(struct node_state_opt)); if ($$ == NULL) err(1, "state_opt_item: calloc"); $$->type = PF_STATE_OPT_TIMEOUT; $$->data.timeout.number = pf_timeouts[i].timeout; $$->data.timeout.seconds = $2; $$->next = NULL; $$->tail = $$; } ; label : LABEL STRING { $$ = $2; } ; qname : QUEUE STRING { $$.qname = $2; $$.pqname = NULL; } | QUEUE '(' STRING ')' { $$.qname = $3; $$.pqname = NULL; } | QUEUE '(' STRING comma STRING ')' { $$.qname = $3; $$.pqname = $5; } ; no : /* empty */ { $$ = 0; } | NO { $$ = 1; } ; portstar : numberstring { if (parseport($1, &$$, PPORT_RANGE|PPORT_STAR) == -1) { free($1); YYERROR; } free($1); } ; redirspec : host { $$ = $1; } | '{' optnl redir_host_list '}' { $$ = $3; } ; redir_host_list : host optnl { $$ = $1; } | redir_host_list comma host optnl { $1->tail->next = $3; $1->tail = $3->tail; $$ = $1; } ; redirpool : /* empty */ { $$ = NULL; } | ARROW redirspec { $$ = calloc(1, sizeof(struct redirection)); if ($$ == NULL) err(1, "redirection: calloc"); $$->host = $2; $$->rport.a = $$->rport.b = $$->rport.t = 0; } | ARROW redirspec PORT portstar { $$ = calloc(1, sizeof(struct redirection)); if ($$ == NULL) err(1, "redirection: calloc"); $$->host = $2; $$->rport = $4; } ; hashkey : /* empty */ { $$ = calloc(1, sizeof(struct pf_poolhashkey)); if ($$ == NULL) err(1, "hashkey: calloc"); $$->key32[0] = arc4random(); $$->key32[1] = arc4random(); $$->key32[2] = arc4random(); $$->key32[3] = arc4random(); } | string { if (!strncmp($1, "0x", 2)) { if (strlen($1) != 34) { free($1); yyerror("hex key must be 128 bits " "(32 hex digits) long"); YYERROR; } $$ = calloc(1, sizeof(struct pf_poolhashkey)); if ($$ == NULL) err(1, "hashkey: calloc"); if (sscanf($1, "0x%8x%8x%8x%8x", &$$->key32[0], &$$->key32[1], &$$->key32[2], &$$->key32[3]) != 4) { free($$); free($1); yyerror("invalid hex key"); YYERROR; } } else { MD5_CTX context; $$ = calloc(1, sizeof(struct pf_poolhashkey)); if ($$ == NULL) err(1, "hashkey: calloc"); MD5Init(&context); MD5Update(&context, (unsigned char *)$1, strlen($1)); MD5Final((unsigned char *)$$, &context); HTONL($$->key32[0]); HTONL($$->key32[1]); HTONL($$->key32[2]); HTONL($$->key32[3]); } free($1); } ; pool_opts : { bzero(&pool_opts, sizeof pool_opts); } pool_opts_l { $$ = pool_opts; } | /* empty */ { bzero(&pool_opts, sizeof pool_opts); $$ = pool_opts; } ; pool_opts_l : pool_opts_l pool_opt | pool_opt ; pool_opt : BITMASK { if (pool_opts.type) { yyerror("pool type cannot be redefined"); YYERROR; } pool_opts.type = PF_POOL_BITMASK; } | RANDOM { if (pool_opts.type) { yyerror("pool type cannot be redefined"); YYERROR; } pool_opts.type = PF_POOL_RANDOM; } | SOURCEHASH hashkey { if (pool_opts.type) { yyerror("pool type cannot be redefined"); YYERROR; } pool_opts.type = PF_POOL_SRCHASH; pool_opts.key = $2; } | ROUNDROBIN { if (pool_opts.type) { yyerror("pool type cannot be redefined"); YYERROR; } pool_opts.type = PF_POOL_ROUNDROBIN; } | STATICPORT { if (pool_opts.staticport) { yyerror("static-port cannot be redefined"); YYERROR; } pool_opts.staticport = 1; } | STICKYADDRESS { if (filter_opts.marker & POM_STICKYADDRESS) { yyerror("sticky-address cannot be redefined"); YYERROR; } pool_opts.marker |= POM_STICKYADDRESS; pool_opts.opts |= PF_POOL_STICKYADDR; } ; redirection : /* empty */ { $$ = NULL; } | ARROW host { $$ = calloc(1, sizeof(struct redirection)); if ($$ == NULL) err(1, "redirection: calloc"); $$->host = $2; $$->rport.a = $$->rport.b = $$->rport.t = 0; } | ARROW host PORT portstar { $$ = calloc(1, sizeof(struct redirection)); if ($$ == NULL) err(1, "redirection: calloc"); $$->host = $2; $$->rport = $4; } ; natpasslog : /* empty */ { $$.b1 = $$.b2 = 0; $$.w2 = 0; } | PASS { $$.b1 = 1; $$.b2 = 0; $$.w2 = 0; } | PASS log { $$.b1 = 1; $$.b2 = $2.log; $$.w2 = $2.logif; } | log { $$.b1 = 0; $$.b2 = $1.log; $$.w2 = $1.logif; } ; nataction : no NAT natpasslog { if ($1 && $3.b1) { yyerror("\"pass\" not valid with \"no\""); YYERROR; } if ($1) $$.b1 = PF_NONAT; else $$.b1 = PF_NAT; $$.b2 = $3.b1; $$.w = $3.b2; $$.w2 = $3.w2; } | no RDR natpasslog { if ($1 && $3.b1) { yyerror("\"pass\" not valid with \"no\""); YYERROR; } if ($1) $$.b1 = PF_NORDR; else $$.b1 = PF_RDR; $$.b2 = $3.b1; $$.w = $3.b2; $$.w2 = $3.w2; } ; natrule : nataction interface af proto fromto tag tagged rtable redirpool pool_opts { struct pf_rule r; if (check_rulestate(PFCTL_STATE_NAT)) YYERROR; memset(&r, 0, sizeof(r)); r.action = $1.b1; r.natpass = $1.b2; r.log = $1.w; r.logif = $1.w2; r.af = $3; if (!r.af) { if ($5.src.host && $5.src.host->af && !$5.src.host->ifindex) r.af = $5.src.host->af; else if ($5.dst.host && $5.dst.host->af && !$5.dst.host->ifindex) r.af = $5.dst.host->af; } if ($6 != NULL) if (strlcpy(r.tagname, $6, PF_TAG_NAME_SIZE) >= PF_TAG_NAME_SIZE) { yyerror("tag too long, max %u chars", PF_TAG_NAME_SIZE - 1); YYERROR; } if ($7.name) if (strlcpy(r.match_tagname, $7.name, PF_TAG_NAME_SIZE) >= PF_TAG_NAME_SIZE) { yyerror("tag too long, max %u chars", PF_TAG_NAME_SIZE - 1); YYERROR; } r.match_tag_not = $7.neg; r.rtableid = $8; if (r.action == PF_NONAT || r.action == PF_NORDR) { if ($9 != NULL) { yyerror("translation rule with 'no' " "does not need '->'"); YYERROR; } } else { if ($9 == NULL || $9->host == NULL) { yyerror("translation rule requires '-> " "address'"); YYERROR; } if (!r.af && ! $9->host->ifindex) r.af = $9->host->af; remove_invalid_hosts(&$9->host, &r.af); if (invalid_redirect($9->host, r.af)) YYERROR; if (check_netmask($9->host, r.af)) YYERROR; r.rpool.proxy_port[0] = ntohs($9->rport.a); switch (r.action) { case PF_RDR: if (!$9->rport.b && $9->rport.t && $5.dst.port != NULL) { r.rpool.proxy_port[1] = ntohs($9->rport.a) + (ntohs( $5.dst.port->port[1]) - ntohs( $5.dst.port->port[0])); } else r.rpool.proxy_port[1] = ntohs($9->rport.b); break; case PF_NAT: r.rpool.proxy_port[1] = ntohs($9->rport.b); if (!r.rpool.proxy_port[0] && !r.rpool.proxy_port[1]) { r.rpool.proxy_port[0] = PF_NAT_PROXY_PORT_LOW; r.rpool.proxy_port[1] = PF_NAT_PROXY_PORT_HIGH; } else if (!r.rpool.proxy_port[1]) r.rpool.proxy_port[1] = r.rpool.proxy_port[0]; break; default: break; } r.rpool.opts = $10.type; if ((r.rpool.opts & PF_POOL_TYPEMASK) == PF_POOL_NONE && ($9->host->next != NULL || $9->host->addr.type == PF_ADDR_TABLE || DYNIF_MULTIADDR($9->host->addr))) r.rpool.opts = PF_POOL_ROUNDROBIN; if ((r.rpool.opts & PF_POOL_TYPEMASK) != PF_POOL_ROUNDROBIN && disallow_table($9->host, "tables are only " "supported in round-robin redirection " "pools")) YYERROR; if ((r.rpool.opts & PF_POOL_TYPEMASK) != PF_POOL_ROUNDROBIN && disallow_alias($9->host, "interface (%s) " "is only supported in round-robin " "redirection pools")) YYERROR; if ($9->host->next != NULL) { if ((r.rpool.opts & PF_POOL_TYPEMASK) != PF_POOL_ROUNDROBIN) { yyerror("only round-robin " "valid for multiple " "redirection addresses"); YYERROR; } } } if ($10.key != NULL) memcpy(&r.rpool.key, $10.key, sizeof(struct pf_poolhashkey)); if ($10.opts) r.rpool.opts |= $10.opts; if ($10.staticport) { if (r.action != PF_NAT) { yyerror("the 'static-port' option is " "only valid with nat rules"); YYERROR; } if (r.rpool.proxy_port[0] != PF_NAT_PROXY_PORT_LOW && r.rpool.proxy_port[1] != PF_NAT_PROXY_PORT_HIGH) { yyerror("the 'static-port' option can't" " be used when specifying a port" " range"); YYERROR; } r.rpool.proxy_port[0] = 0; r.rpool.proxy_port[1] = 0; } expand_rule(&r, $2, $9 == NULL ? NULL : $9->host, $4, $5.src_os, $5.src.host, $5.src.port, $5.dst.host, $5.dst.port, 0, 0, 0, ""); free($9); } ; binatrule : no BINAT natpasslog interface af proto FROM host toipspec tag tagged rtable redirection { struct pf_rule binat; struct pf_pooladdr *pa; if (check_rulestate(PFCTL_STATE_NAT)) YYERROR; if (disallow_urpf_failed($9, "\"urpf-failed\" is not " "permitted as a binat destination")) YYERROR; memset(&binat, 0, sizeof(binat)); if ($1 && $3.b1) { yyerror("\"pass\" not valid with \"no\""); YYERROR; } if ($1) binat.action = PF_NOBINAT; else binat.action = PF_BINAT; binat.natpass = $3.b1; binat.log = $3.b2; binat.logif = $3.w2; binat.af = $5; if (!binat.af && $8 != NULL && $8->af) binat.af = $8->af; if (!binat.af && $9 != NULL && $9->af) binat.af = $9->af; if (!binat.af && $13 != NULL && $13->host) binat.af = $13->host->af; if (!binat.af) { yyerror("address family (inet/inet6) " "undefined"); YYERROR; } if ($4 != NULL) { memcpy(binat.ifname, $4->ifname, sizeof(binat.ifname)); binat.ifnot = $4->not; free($4); } if ($10 != NULL) if (strlcpy(binat.tagname, $10, PF_TAG_NAME_SIZE) >= PF_TAG_NAME_SIZE) { yyerror("tag too long, max %u chars", PF_TAG_NAME_SIZE - 1); YYERROR; } if ($11.name) if (strlcpy(binat.match_tagname, $11.name, PF_TAG_NAME_SIZE) >= PF_TAG_NAME_SIZE) { yyerror("tag too long, max %u chars", PF_TAG_NAME_SIZE - 1); YYERROR; } binat.match_tag_not = $11.neg; binat.rtableid = $12; if ($6 != NULL) { binat.proto = $6->proto; free($6); } if ($8 != NULL && disallow_table($8, "invalid use of " "table <%s> as the source address of a binat rule")) YYERROR; if ($8 != NULL && disallow_alias($8, "invalid use of " "interface (%s) as the source address of a binat " "rule")) YYERROR; if ($13 != NULL && $13->host != NULL && disallow_table( $13->host, "invalid use of table <%s> as the " "redirect address of a binat rule")) YYERROR; if ($13 != NULL && $13->host != NULL && disallow_alias( $13->host, "invalid use of interface (%s) as the " "redirect address of a binat rule")) YYERROR; if ($8 != NULL) { if ($8->next) { yyerror("multiple binat ip addresses"); YYERROR; } if ($8->addr.type == PF_ADDR_DYNIFTL) $8->af = binat.af; if ($8->af != binat.af) { yyerror("binat ip versions must match"); YYERROR; } if (check_netmask($8, binat.af)) YYERROR; memcpy(&binat.src.addr, &$8->addr, sizeof(binat.src.addr)); free($8); } if ($9 != NULL) { if ($9->next) { yyerror("multiple binat ip addresses"); YYERROR; } if ($9->af != binat.af && $9->af) { yyerror("binat ip versions must match"); YYERROR; } if (check_netmask($9, binat.af)) YYERROR; memcpy(&binat.dst.addr, &$9->addr, sizeof(binat.dst.addr)); binat.dst.neg = $9->not; free($9); } if (binat.action == PF_NOBINAT) { if ($13 != NULL) { yyerror("'no binat' rule does not need" " '->'"); YYERROR; } } else { if ($13 == NULL || $13->host == NULL) { yyerror("'binat' rule requires" " '-> address'"); YYERROR; } remove_invalid_hosts(&$13->host, &binat.af); if (invalid_redirect($13->host, binat.af)) YYERROR; if ($13->host->next != NULL) { yyerror("binat rule must redirect to " "a single address"); YYERROR; } if (check_netmask($13->host, binat.af)) YYERROR; if (!PF_AZERO(&binat.src.addr.v.a.mask, binat.af) && !PF_AEQ(&binat.src.addr.v.a.mask, &$13->host->addr.v.a.mask, binat.af)) { yyerror("'binat' source mask and " "redirect mask must be the same"); YYERROR; } TAILQ_INIT(&binat.rpool.list); pa = calloc(1, sizeof(struct pf_pooladdr)); if (pa == NULL) err(1, "binat: calloc"); pa->addr = $13->host->addr; pa->ifname[0] = 0; TAILQ_INSERT_TAIL(&binat.rpool.list, pa, entries); free($13); } pfctl_add_rule(pf, &binat, ""); } ; tag : /* empty */ { $$ = NULL; } | TAG STRING { $$ = $2; } ; tagged : /* empty */ { $$.neg = 0; $$.name = NULL; } | not TAGGED string { $$.neg = $1; $$.name = $3; } ; rtable : /* empty */ { $$ = -1; } | RTABLE NUMBER { if ($2 < 0 || $2 > rt_tableid_max()) { yyerror("invalid rtable id"); YYERROR; } $$ = $2; } ; route_host : STRING { $$ = calloc(1, sizeof(struct node_host)); if ($$ == NULL) err(1, "route_host: calloc"); $$->ifname = $1; set_ipmask($$, 128); $$->next = NULL; $$->tail = $$; } | '(' STRING host ')' { $$ = $3; $$->ifname = $2; } ; route_host_list : route_host optnl { $$ = $1; } | route_host_list comma route_host optnl { if ($1->af == 0) $1->af = $3->af; if ($1->af != $3->af) { yyerror("all pool addresses must be in the " "same address family"); YYERROR; } $1->tail->next = $3; $1->tail = $3->tail; $$ = $1; } ; routespec : route_host { $$ = $1; } | '{' optnl route_host_list '}' { $$ = $3; } ; route : /* empty */ { $$.host = NULL; $$.rt = 0; $$.pool_opts = 0; } | FASTROUTE { $$.host = NULL; $$.rt = PF_FASTROUTE; $$.pool_opts = 0; } | ROUTETO routespec pool_opts { $$.host = $2; $$.rt = PF_ROUTETO; $$.pool_opts = $3.type | $3.opts; if ($3.key != NULL) $$.key = $3.key; } | REPLYTO routespec pool_opts { $$.host = $2; $$.rt = PF_REPLYTO; $$.pool_opts = $3.type | $3.opts; if ($3.key != NULL) $$.key = $3.key; } | DUPTO routespec pool_opts { $$.host = $2; $$.rt = PF_DUPTO; $$.pool_opts = $3.type | $3.opts; if ($3.key != NULL) $$.key = $3.key; } ; timeout_spec : STRING NUMBER { if (check_rulestate(PFCTL_STATE_OPTION)) { free($1); YYERROR; } if ($2 < 0 || $2 > UINT_MAX) { yyerror("only positive values permitted"); YYERROR; } if (pfctl_set_timeout(pf, $1, $2, 0) != 0) { yyerror("unknown timeout %s", $1); free($1); YYERROR; } free($1); } ; timeout_list : timeout_list comma timeout_spec optnl | timeout_spec optnl ; limit_spec : STRING NUMBER { if (check_rulestate(PFCTL_STATE_OPTION)) { free($1); YYERROR; } if ($2 < 0 || $2 > UINT_MAX) { yyerror("only positive values permitted"); YYERROR; } if (pfctl_set_limit(pf, $1, $2) != 0) { yyerror("unable to set limit %s %u", $1, $2); free($1); YYERROR; } free($1); } ; limit_list : limit_list comma limit_spec optnl | limit_spec optnl ; comma : ',' | /* empty */ ; yesno : NO { $$ = 0; } | STRING { if (!strcmp($1, "yes")) $$ = 1; else { yyerror("invalid value '%s', expected 'yes' " "or 'no'", $1); free($1); YYERROR; } free($1); } ; unaryop : '=' { $$ = PF_OP_EQ; } | '!' '=' { $$ = PF_OP_NE; } | '<' '=' { $$ = PF_OP_LE; } | '<' { $$ = PF_OP_LT; } | '>' '=' { $$ = PF_OP_GE; } | '>' { $$ = PF_OP_GT; } ; %% int yyerror(const char *fmt, ...) { va_list ap; file->errors++; va_start(ap, fmt); fprintf(stderr, "%s:%d: ", file->name, yylval.lineno); vfprintf(stderr, fmt, ap); fprintf(stderr, "\n"); va_end(ap); return (0); } int disallow_table(struct node_host *h, const char *fmt) { for (; h != NULL; h = h->next) if (h->addr.type == PF_ADDR_TABLE) { yyerror(fmt, h->addr.v.tblname); return (1); } return (0); } int disallow_urpf_failed(struct node_host *h, const char *fmt) { for (; h != NULL; h = h->next) if (h->addr.type == PF_ADDR_URPFFAILED) { yyerror(fmt); return (1); } return (0); } int disallow_alias(struct node_host *h, const char *fmt) { for (; h != NULL; h = h->next) if (DYNIF_MULTIADDR(h->addr)) { yyerror(fmt, h->addr.v.tblname); return (1); } return (0); } int rule_consistent(struct pf_rule *r, int anchor_call) { int problems = 0; switch (r->action) { case PF_PASS: case PF_DROP: case PF_SCRUB: case PF_NOSCRUB: problems = filter_consistent(r, anchor_call); break; case PF_NAT: case PF_NONAT: problems = nat_consistent(r); break; case PF_RDR: case PF_NORDR: problems = rdr_consistent(r); break; case PF_BINAT: case PF_NOBINAT: default: break; } return (problems); } int filter_consistent(struct pf_rule *r, int anchor_call) { int problems = 0; if (r->proto != IPPROTO_TCP && r->proto != IPPROTO_UDP && (r->src.port_op || r->dst.port_op)) { yyerror("port only applies to tcp/udp"); problems++; } if (r->proto != IPPROTO_ICMP && r->proto != IPPROTO_ICMPV6 && (r->type || r->code)) { yyerror("icmp-type/code only applies to icmp"); problems++; } if (!r->af && (r->type || r->code)) { yyerror("must indicate address family with icmp-type/code"); problems++; } if (r->overload_tblname[0] && r->max_src_conn == 0 && r->max_src_conn_rate.seconds == 0) { yyerror("'overload' requires 'max-src-conn' " "or 'max-src-conn-rate'"); problems++; } if ((r->proto == IPPROTO_ICMP && r->af == AF_INET6) || (r->proto == IPPROTO_ICMPV6 && r->af == AF_INET)) { yyerror("proto %s doesn't match address family %s", r->proto == IPPROTO_ICMP ? "icmp" : "icmp6", r->af == AF_INET ? "inet" : "inet6"); problems++; } if (r->allow_opts && r->action != PF_PASS) { yyerror("allow-opts can only be specified for pass rules"); problems++; } if (r->rule_flag & PFRULE_FRAGMENT && (r->src.port_op || r->dst.port_op || r->flagset || r->type || r->code)) { yyerror("fragments can be filtered only on IP header fields"); problems++; } if (r->rule_flag & PFRULE_RETURNRST && r->proto != IPPROTO_TCP) { yyerror("return-rst can only be applied to TCP rules"); problems++; } if (r->max_src_nodes && !(r->rule_flag & PFRULE_RULESRCTRACK)) { yyerror("max-src-nodes requires 'source-track rule'"); problems++; } if (r->action == PF_DROP && r->keep_state) { yyerror("keep state on block rules doesn't make sense"); problems++; } if (r->rule_flag & PFRULE_STATESLOPPY && (r->keep_state == PF_STATE_MODULATE || r->keep_state == PF_STATE_SYNPROXY)) { yyerror("sloppy state matching cannot be used with " "synproxy state or modulate state"); problems++; } return (-problems); } int nat_consistent(struct pf_rule *r) { return (0); /* yeah! */ } int rdr_consistent(struct pf_rule *r) { int problems = 0; if (r->proto != IPPROTO_TCP && r->proto != IPPROTO_UDP) { if (r->src.port_op) { yyerror("src port only applies to tcp/udp"); problems++; } if (r->dst.port_op) { yyerror("dst port only applies to tcp/udp"); problems++; } if (r->rpool.proxy_port[0]) { yyerror("rpool port only applies to tcp/udp"); problems++; } } if (r->dst.port_op && r->dst.port_op != PF_OP_EQ && r->dst.port_op != PF_OP_RRG) { yyerror("invalid port operator for rdr destination port"); problems++; } return (-problems); } int process_tabledef(char *name, struct table_opts *opts) { struct pfr_buffer ab; struct node_tinit *ti; bzero(&ab, sizeof(ab)); ab.pfrb_type = PFRB_ADDRS; SIMPLEQ_FOREACH(ti, &opts->init_nodes, entries) { if (ti->file) if (pfr_buf_load(&ab, ti->file, 0, append_addr)) { if (errno) yyerror("cannot load \"%s\": %s", ti->file, strerror(errno)); else yyerror("file \"%s\" contains bad data", ti->file); goto _error; } if (ti->host) if (append_addr_host(&ab, ti->host, 0, 0)) { yyerror("cannot create address buffer: %s", strerror(errno)); goto _error; } } if (pf->opts & PF_OPT_VERBOSE) print_tabledef(name, opts->flags, opts->init_addr, &opts->init_nodes); if (!(pf->opts & PF_OPT_NOACTION) && pfctl_define_table(name, opts->flags, opts->init_addr, pf->anchor->name, &ab, pf->anchor->ruleset.tticket)) { yyerror("cannot define table %s: %s", name, pfr_strerror(errno)); goto _error; } pf->tdirty = 1; pfr_buf_clear(&ab); return (0); _error: pfr_buf_clear(&ab); return (-1); } struct keywords { const char *k_name; int k_val; }; /* macro gore, but you should've seen the prior indentation nightmare... */ #define FREE_LIST(T,r) \ do { \ T *p, *node = r; \ while (node != NULL) { \ p = node; \ node = node->next; \ free(p); \ } \ } while (0) #define LOOP_THROUGH(T,n,r,C) \ do { \ T *n; \ if (r == NULL) { \ r = calloc(1, sizeof(T)); \ if (r == NULL) \ err(1, "LOOP: calloc"); \ r->next = NULL; \ } \ n = r; \ while (n != NULL) { \ do { \ C; \ } while (0); \ n = n->next; \ } \ } while (0) void expand_label_str(char *label, size_t len, const char *srch, const char *repl) { char *tmp; char *p, *q; if ((tmp = calloc(1, len)) == NULL) err(1, "expand_label_str: calloc"); p = q = label; while ((q = strstr(p, srch)) != NULL) { *q = '\0'; if ((strlcat(tmp, p, len) >= len) || (strlcat(tmp, repl, len) >= len)) errx(1, "expand_label: label too long"); q += strlen(srch); p = q; } if (strlcat(tmp, p, len) >= len) errx(1, "expand_label: label too long"); strlcpy(label, tmp, len); /* always fits */ free(tmp); } void expand_label_if(const char *name, char *label, size_t len, const char *ifname) { if (strstr(label, name) != NULL) { if (!*ifname) expand_label_str(label, len, name, "any"); else expand_label_str(label, len, name, ifname); } } void expand_label_addr(const char *name, char *label, size_t len, sa_family_t af, struct node_host *h) { char tmp[64], tmp_not[66]; if (strstr(label, name) != NULL) { switch (h->addr.type) { case PF_ADDR_DYNIFTL: snprintf(tmp, sizeof(tmp), "(%s)", h->addr.v.ifname); break; case PF_ADDR_TABLE: snprintf(tmp, sizeof(tmp), "<%s>", h->addr.v.tblname); break; case PF_ADDR_NOROUTE: snprintf(tmp, sizeof(tmp), "no-route"); break; case PF_ADDR_URPFFAILED: snprintf(tmp, sizeof(tmp), "urpf-failed"); break; case PF_ADDR_ADDRMASK: if (!af || (PF_AZERO(&h->addr.v.a.addr, af) && PF_AZERO(&h->addr.v.a.mask, af))) snprintf(tmp, sizeof(tmp), "any"); else { char a[48]; int bits; if (inet_ntop(af, &h->addr.v.a.addr, a, sizeof(a)) == NULL) snprintf(tmp, sizeof(tmp), "?"); else { bits = unmask(&h->addr.v.a.mask, af); if ((af == AF_INET && bits < 32) || (af == AF_INET6 && bits < 128)) snprintf(tmp, sizeof(tmp), "%s/%d", a, bits); else snprintf(tmp, sizeof(tmp), "%s", a); } } break; default: snprintf(tmp, sizeof(tmp), "?"); break; } if (h->not) { snprintf(tmp_not, sizeof(tmp_not), "! %s", tmp); expand_label_str(label, len, name, tmp_not); } else expand_label_str(label, len, name, tmp); } } void expand_label_port(const char *name, char *label, size_t len, struct node_port *port) { char a1[6], a2[6], op[13] = ""; if (strstr(label, name) != NULL) { snprintf(a1, sizeof(a1), "%u", ntohs(port->port[0])); snprintf(a2, sizeof(a2), "%u", ntohs(port->port[1])); if (!port->op) ; else if (port->op == PF_OP_IRG) snprintf(op, sizeof(op), "%s><%s", a1, a2); else if (port->op == PF_OP_XRG) snprintf(op, sizeof(op), "%s<>%s", a1, a2); else if (port->op == PF_OP_EQ) snprintf(op, sizeof(op), "%s", a1); else if (port->op == PF_OP_NE) snprintf(op, sizeof(op), "!=%s", a1); else if (port->op == PF_OP_LT) snprintf(op, sizeof(op), "<%s", a1); else if (port->op == PF_OP_LE) snprintf(op, sizeof(op), "<=%s", a1); else if (port->op == PF_OP_GT) snprintf(op, sizeof(op), ">%s", a1); else if (port->op == PF_OP_GE) snprintf(op, sizeof(op), ">=%s", a1); expand_label_str(label, len, name, op); } } void expand_label_proto(const char *name, char *label, size_t len, u_int8_t proto) { struct protoent *pe; char n[4]; if (strstr(label, name) != NULL) { pe = getprotobynumber(proto); if (pe != NULL) expand_label_str(label, len, name, pe->p_name); else { snprintf(n, sizeof(n), "%u", proto); expand_label_str(label, len, name, n); } } } void expand_label_nr(const char *name, char *label, size_t len) { char n[11]; if (strstr(label, name) != NULL) { snprintf(n, sizeof(n), "%u", pf->anchor->match); expand_label_str(label, len, name, n); } } void expand_label(char *label, size_t len, const char *ifname, sa_family_t af, struct node_host *src_host, struct node_port *src_port, struct node_host *dst_host, struct node_port *dst_port, u_int8_t proto) { expand_label_if("$if", label, len, ifname); expand_label_addr("$srcaddr", label, len, af, src_host); expand_label_addr("$dstaddr", label, len, af, dst_host); expand_label_port("$srcport", label, len, src_port); expand_label_port("$dstport", label, len, dst_port); expand_label_proto("$proto", label, len, proto); expand_label_nr("$nr", label, len); } int expand_altq(struct pf_altq *a, struct node_if *interfaces, struct node_queue *nqueues, struct node_queue_bw bwspec, struct node_queue_opt *opts) { struct pf_altq pa, pb; char qname[PF_QNAME_SIZE]; struct node_queue *n; struct node_queue_bw bw; int errs = 0; if ((pf->loadopt & PFCTL_FLAG_ALTQ) == 0) { FREE_LIST(struct node_if, interfaces); if (nqueues) FREE_LIST(struct node_queue, nqueues); return (0); } LOOP_THROUGH(struct node_if, interface, interfaces, memcpy(&pa, a, sizeof(struct pf_altq)); if (strlcpy(pa.ifname, interface->ifname, sizeof(pa.ifname)) >= sizeof(pa.ifname)) errx(1, "expand_altq: strlcpy"); if (interface->not) { yyerror("altq on ! is not supported"); errs++; } else { if (eval_pfaltq(pf, &pa, &bwspec, opts)) errs++; else if (pfctl_add_altq(pf, &pa)) errs++; if (pf->opts & PF_OPT_VERBOSE) { print_altq(&pf->paltq->altq, 0, &bwspec, opts); if (nqueues && nqueues->tail) { printf("queue { "); LOOP_THROUGH(struct node_queue, queue, nqueues, printf("%s ", queue->queue); ); printf("}"); } printf("\n"); } if (pa.scheduler == ALTQT_CBQ || pa.scheduler == ALTQT_HFSC) { /* now create a root queue */ memset(&pb, 0, sizeof(struct pf_altq)); if (strlcpy(qname, "root_", sizeof(qname)) >= sizeof(qname)) errx(1, "expand_altq: strlcpy"); if (strlcat(qname, interface->ifname, sizeof(qname)) >= sizeof(qname)) errx(1, "expand_altq: strlcat"); if (strlcpy(pb.qname, qname, sizeof(pb.qname)) >= sizeof(pb.qname)) errx(1, "expand_altq: strlcpy"); if (strlcpy(pb.ifname, interface->ifname, sizeof(pb.ifname)) >= sizeof(pb.ifname)) errx(1, "expand_altq: strlcpy"); pb.qlimit = pa.qlimit; pb.scheduler = pa.scheduler; bw.bw_absolute = pa.ifbandwidth; bw.bw_percent = 0; if (eval_pfqueue(pf, &pb, &bw, opts)) errs++; else if (pfctl_add_altq(pf, &pb)) errs++; } LOOP_THROUGH(struct node_queue, queue, nqueues, n = calloc(1, sizeof(struct node_queue)); if (n == NULL) err(1, "expand_altq: calloc"); if (pa.scheduler == ALTQT_CBQ || pa.scheduler == ALTQT_HFSC) if (strlcpy(n->parent, qname, sizeof(n->parent)) >= sizeof(n->parent)) errx(1, "expand_altq: strlcpy"); if (strlcpy(n->queue, queue->queue, sizeof(n->queue)) >= sizeof(n->queue)) errx(1, "expand_altq: strlcpy"); if (strlcpy(n->ifname, interface->ifname, sizeof(n->ifname)) >= sizeof(n->ifname)) errx(1, "expand_altq: strlcpy"); n->scheduler = pa.scheduler; n->next = NULL; n->tail = n; if (queues == NULL) queues = n; else { queues->tail->next = n; queues->tail = n; } ); } ); FREE_LIST(struct node_if, interfaces); if (nqueues) FREE_LIST(struct node_queue, nqueues); return (errs); } int expand_queue(struct pf_altq *a, struct node_if *interfaces, struct node_queue *nqueues, struct node_queue_bw bwspec, struct node_queue_opt *opts) { struct node_queue *n, *nq; struct pf_altq pa; u_int8_t found = 0; u_int8_t errs = 0; if ((pf->loadopt & PFCTL_FLAG_ALTQ) == 0) { FREE_LIST(struct node_queue, nqueues); return (0); } if (queues == NULL) { yyerror("queue %s has no parent", a->qname); FREE_LIST(struct node_queue, nqueues); return (1); } LOOP_THROUGH(struct node_if, interface, interfaces, LOOP_THROUGH(struct node_queue, tqueue, queues, if (!strncmp(a->qname, tqueue->queue, PF_QNAME_SIZE) && (interface->ifname[0] == 0 || (!interface->not && !strncmp(interface->ifname, tqueue->ifname, IFNAMSIZ)) || (interface->not && strncmp(interface->ifname, tqueue->ifname, IFNAMSIZ)))) { /* found ourself in queues */ found++; memcpy(&pa, a, sizeof(struct pf_altq)); if (pa.scheduler != ALTQT_NONE && pa.scheduler != tqueue->scheduler) { yyerror("exactly one scheduler type " "per interface allowed"); return (1); } pa.scheduler = tqueue->scheduler; /* scheduler dependent error checking */ switch (pa.scheduler) { case ALTQT_PRIQ: if (nqueues != NULL) { yyerror("priq queues cannot " "have child queues"); return (1); } if (bwspec.bw_absolute > 0 || bwspec.bw_percent < 100) { yyerror("priq doesn't take " "bandwidth"); return (1); } break; default: break; } if (strlcpy(pa.ifname, tqueue->ifname, sizeof(pa.ifname)) >= sizeof(pa.ifname)) errx(1, "expand_queue: strlcpy"); if (strlcpy(pa.parent, tqueue->parent, sizeof(pa.parent)) >= sizeof(pa.parent)) errx(1, "expand_queue: strlcpy"); if (eval_pfqueue(pf, &pa, &bwspec, opts)) errs++; else if (pfctl_add_altq(pf, &pa)) errs++; for (nq = nqueues; nq != NULL; nq = nq->next) { if (!strcmp(a->qname, nq->queue)) { yyerror("queue cannot have " "itself as child"); errs++; continue; } n = calloc(1, sizeof(struct node_queue)); if (n == NULL) err(1, "expand_queue: calloc"); if (strlcpy(n->parent, a->qname, sizeof(n->parent)) >= sizeof(n->parent)) errx(1, "expand_queue strlcpy"); if (strlcpy(n->queue, nq->queue, sizeof(n->queue)) >= sizeof(n->queue)) errx(1, "expand_queue strlcpy"); if (strlcpy(n->ifname, tqueue->ifname, sizeof(n->ifname)) >= sizeof(n->ifname)) errx(1, "expand_queue strlcpy"); n->scheduler = tqueue->scheduler; n->next = NULL; n->tail = n; if (queues == NULL) queues = n; else { queues->tail->next = n; queues->tail = n; } } if ((pf->opts & PF_OPT_VERBOSE) && ( (found == 1 && interface->ifname[0] == 0) || (found > 0 && interface->ifname[0] != 0))) { print_queue(&pf->paltq->altq, 0, &bwspec, interface->ifname[0] != 0, opts); if (nqueues && nqueues->tail) { printf("{ "); LOOP_THROUGH(struct node_queue, queue, nqueues, printf("%s ", queue->queue); ); printf("}"); } printf("\n"); } } ); ); FREE_LIST(struct node_queue, nqueues); FREE_LIST(struct node_if, interfaces); if (!found) { yyerror("queue %s has no parent", a->qname); errs++; } if (errs) return (1); else return (0); } void expand_rule(struct pf_rule *r, struct node_if *interfaces, struct node_host *rpool_hosts, struct node_proto *protos, struct node_os *src_oses, struct node_host *src_hosts, struct node_port *src_ports, struct node_host *dst_hosts, struct node_port *dst_ports, struct node_uid *uids, struct node_gid *gids, struct node_icmp *icmp_types, const char *anchor_call) { sa_family_t af = r->af; int added = 0, error = 0; char ifname[IF_NAMESIZE]; char label[PF_RULE_LABEL_SIZE]; char tagname[PF_TAG_NAME_SIZE]; char match_tagname[PF_TAG_NAME_SIZE]; struct pf_pooladdr *pa; struct node_host *h; u_int8_t flags, flagset, keep_state; if (strlcpy(label, r->label, sizeof(label)) >= sizeof(label)) errx(1, "expand_rule: strlcpy"); if (strlcpy(tagname, r->tagname, sizeof(tagname)) >= sizeof(tagname)) errx(1, "expand_rule: strlcpy"); if (strlcpy(match_tagname, r->match_tagname, sizeof(match_tagname)) >= sizeof(match_tagname)) errx(1, "expand_rule: strlcpy"); flags = r->flags; flagset = r->flagset; keep_state = r->keep_state; LOOP_THROUGH(struct node_if, interface, interfaces, LOOP_THROUGH(struct node_proto, proto, protos, LOOP_THROUGH(struct node_icmp, icmp_type, icmp_types, LOOP_THROUGH(struct node_host, src_host, src_hosts, LOOP_THROUGH(struct node_port, src_port, src_ports, LOOP_THROUGH(struct node_os, src_os, src_oses, LOOP_THROUGH(struct node_host, dst_host, dst_hosts, LOOP_THROUGH(struct node_port, dst_port, dst_ports, LOOP_THROUGH(struct node_uid, uid, uids, LOOP_THROUGH(struct node_gid, gid, gids, r->af = af; /* for link-local IPv6 address, interface must match up */ if ((r->af && src_host->af && r->af != src_host->af) || (r->af && dst_host->af && r->af != dst_host->af) || (src_host->af && dst_host->af && src_host->af != dst_host->af) || (src_host->ifindex && dst_host->ifindex && src_host->ifindex != dst_host->ifindex) || (src_host->ifindex && *interface->ifname && src_host->ifindex != if_nametoindex(interface->ifname)) || (dst_host->ifindex && *interface->ifname && dst_host->ifindex != if_nametoindex(interface->ifname))) continue; if (!r->af && src_host->af) r->af = src_host->af; else if (!r->af && dst_host->af) r->af = dst_host->af; if (*interface->ifname) strlcpy(r->ifname, interface->ifname, sizeof(r->ifname)); else if (if_indextoname(src_host->ifindex, ifname)) strlcpy(r->ifname, ifname, sizeof(r->ifname)); else if (if_indextoname(dst_host->ifindex, ifname)) strlcpy(r->ifname, ifname, sizeof(r->ifname)); else memset(r->ifname, '\0', sizeof(r->ifname)); if (strlcpy(r->label, label, sizeof(r->label)) >= sizeof(r->label)) errx(1, "expand_rule: strlcpy"); if (strlcpy(r->tagname, tagname, sizeof(r->tagname)) >= sizeof(r->tagname)) errx(1, "expand_rule: strlcpy"); if (strlcpy(r->match_tagname, match_tagname, sizeof(r->match_tagname)) >= sizeof(r->match_tagname)) errx(1, "expand_rule: strlcpy"); expand_label(r->label, PF_RULE_LABEL_SIZE, r->ifname, r->af, src_host, src_port, dst_host, dst_port, proto->proto); expand_label(r->tagname, PF_TAG_NAME_SIZE, r->ifname, r->af, src_host, src_port, dst_host, dst_port, proto->proto); expand_label(r->match_tagname, PF_TAG_NAME_SIZE, r->ifname, r->af, src_host, src_port, dst_host, dst_port, proto->proto); error += check_netmask(src_host, r->af); error += check_netmask(dst_host, r->af); r->ifnot = interface->not; r->proto = proto->proto; r->src.addr = src_host->addr; r->src.neg = src_host->not; r->src.port[0] = src_port->port[0]; r->src.port[1] = src_port->port[1]; r->src.port_op = src_port->op; r->dst.addr = dst_host->addr; r->dst.neg = dst_host->not; r->dst.port[0] = dst_port->port[0]; r->dst.port[1] = dst_port->port[1]; r->dst.port_op = dst_port->op; r->uid.op = uid->op; r->uid.uid[0] = uid->uid[0]; r->uid.uid[1] = uid->uid[1]; r->gid.op = gid->op; r->gid.gid[0] = gid->gid[0]; r->gid.gid[1] = gid->gid[1]; r->type = icmp_type->type; r->code = icmp_type->code; if ((keep_state == PF_STATE_MODULATE || keep_state == PF_STATE_SYNPROXY) && r->proto && r->proto != IPPROTO_TCP) r->keep_state = PF_STATE_NORMAL; else r->keep_state = keep_state; if (r->proto && r->proto != IPPROTO_TCP) { r->flags = 0; r->flagset = 0; } else { r->flags = flags; r->flagset = flagset; } if (icmp_type->proto && r->proto != icmp_type->proto) { yyerror("icmp-type mismatch"); error++; } if (src_os && src_os->os) { r->os_fingerprint = pfctl_get_fingerprint(src_os->os); if ((pf->opts & PF_OPT_VERBOSE2) && r->os_fingerprint == PF_OSFP_NOMATCH) fprintf(stderr, "warning: unknown '%s' OS fingerprint\n", src_os->os); } else { r->os_fingerprint = PF_OSFP_ANY; } TAILQ_INIT(&r->rpool.list); for (h = rpool_hosts; h != NULL; h = h->next) { pa = calloc(1, sizeof(struct pf_pooladdr)); if (pa == NULL) err(1, "expand_rule: calloc"); pa->addr = h->addr; if (h->ifname != NULL) { if (strlcpy(pa->ifname, h->ifname, sizeof(pa->ifname)) >= sizeof(pa->ifname)) errx(1, "expand_rule: strlcpy"); } else pa->ifname[0] = 0; TAILQ_INSERT_TAIL(&r->rpool.list, pa, entries); } if (rule_consistent(r, anchor_call[0]) < 0 || error) yyerror("skipping rule due to errors"); else { r->nr = pf->astack[pf->asd]->match++; pfctl_add_rule(pf, r, anchor_call); added++; } )))))))))); FREE_LIST(struct node_if, interfaces); FREE_LIST(struct node_proto, protos); FREE_LIST(struct node_host, src_hosts); FREE_LIST(struct node_port, src_ports); FREE_LIST(struct node_os, src_oses); FREE_LIST(struct node_host, dst_hosts); FREE_LIST(struct node_port, dst_ports); FREE_LIST(struct node_uid, uids); FREE_LIST(struct node_gid, gids); FREE_LIST(struct node_icmp, icmp_types); FREE_LIST(struct node_host, rpool_hosts); if (!added) yyerror("rule expands to no valid combination"); } int expand_skip_interface(struct node_if *interfaces) { int errs = 0; if (!interfaces || (!interfaces->next && !interfaces->not && !strcmp(interfaces->ifname, "none"))) { if (pf->opts & PF_OPT_VERBOSE) printf("set skip on none\n"); errs = pfctl_set_interface_flags(pf, "", PFI_IFLAG_SKIP, 0); return (errs); } if (pf->opts & PF_OPT_VERBOSE) printf("set skip on {"); LOOP_THROUGH(struct node_if, interface, interfaces, if (pf->opts & PF_OPT_VERBOSE) printf(" %s", interface->ifname); if (interface->not) { yyerror("skip on ! is not supported"); errs++; } else errs += pfctl_set_interface_flags(pf, interface->ifname, PFI_IFLAG_SKIP, 1); ); if (pf->opts & PF_OPT_VERBOSE) printf(" }\n"); FREE_LIST(struct node_if, interfaces); if (errs) return (1); else return (0); } #undef FREE_LIST #undef LOOP_THROUGH int check_rulestate(int desired_state) { if (require_order && (rulestate > desired_state)) { yyerror("Rules must be in order: options, normalization, " "queueing, translation, filtering"); return (1); } rulestate = desired_state; return (0); } int kw_cmp(const void *k, const void *e) { return (strcmp(k, ((const struct keywords *)e)->k_name)); } int lookup(char *s) { /* this has to be sorted always */ static const struct keywords keywords[] = { { "all", ALL}, { "allow-opts", ALLOWOPTS}, { "altq", ALTQ}, { "anchor", ANCHOR}, { "antispoof", ANTISPOOF}, { "any", ANY}, { "bandwidth", BANDWIDTH}, { "binat", BINAT}, { "binat-anchor", BINATANCHOR}, { "bitmask", BITMASK}, { "block", BLOCK}, { "block-policy", BLOCKPOLICY}, { "buckets", BUCKETS}, { "cbq", CBQ}, { "code", CODE}, { "codelq", CODEL}, { "crop", FRAGCROP}, { "debug", DEBUG}, { "divert-reply", DIVERTREPLY}, { "divert-to", DIVERTTO}, { "drop", DROP}, { "drop-ovl", FRAGDROP}, { "dup-to", DUPTO}, { "fairq", FAIRQ}, { "fastroute", FASTROUTE}, { "file", FILENAME}, { "fingerprints", FINGERPRINTS}, { "flags", FLAGS}, { "floating", FLOATING}, { "flush", FLUSH}, { "for", FOR}, { "fragment", FRAGMENT}, { "from", FROM}, { "global", GLOBAL}, { "group", GROUP}, { "hfsc", HFSC}, { "hogs", HOGS}, { "hostid", HOSTID}, { "icmp-type", ICMPTYPE}, { "icmp6-type", ICMP6TYPE}, { "if-bound", IFBOUND}, { "in", IN}, { "include", INCLUDE}, { "inet", INET}, { "inet6", INET6}, { "interval", INTERVAL}, { "keep", KEEP}, { "label", LABEL}, { "limit", LIMIT}, { "linkshare", LINKSHARE}, { "load", LOAD}, { "log", LOG}, { "loginterface", LOGINTERFACE}, { "max", MAXIMUM}, { "max-mss", MAXMSS}, { "max-src-conn", MAXSRCCONN}, { "max-src-conn-rate", MAXSRCCONNRATE}, { "max-src-nodes", MAXSRCNODES}, { "max-src-states", MAXSRCSTATES}, { "min-ttl", MINTTL}, { "modulate", MODULATE}, { "nat", NAT}, { "nat-anchor", NATANCHOR}, { "no", NO}, { "no-df", NODF}, { "no-route", NOROUTE}, { "no-sync", NOSYNC}, { "on", ON}, { "optimization", OPTIMIZATION}, { "os", OS}, { "out", OUT}, { "overload", OVERLOAD}, { "pass", PASS}, { "port", PORT}, { "priority", PRIORITY}, { "priq", PRIQ}, { "probability", PROBABILITY}, { "proto", PROTO}, { "qlimit", QLIMIT}, { "queue", QUEUE}, { "quick", QUICK}, { "random", RANDOM}, { "random-id", RANDOMID}, { "rdr", RDR}, { "rdr-anchor", RDRANCHOR}, { "realtime", REALTIME}, { "reassemble", REASSEMBLE}, { "reply-to", REPLYTO}, { "require-order", REQUIREORDER}, { "return", RETURN}, { "return-icmp", RETURNICMP}, { "return-icmp6", RETURNICMP6}, { "return-rst", RETURNRST}, { "round-robin", ROUNDROBIN}, { "route", ROUTE}, { "route-to", ROUTETO}, { "rtable", RTABLE}, { "rule", RULE}, { "ruleset-optimization", RULESET_OPTIMIZATION}, { "scrub", SCRUB}, { "set", SET}, { "set-tos", SETTOS}, { "skip", SKIP}, { "sloppy", SLOPPY}, { "source-hash", SOURCEHASH}, { "source-track", SOURCETRACK}, { "state", STATE}, { "state-defaults", STATEDEFAULTS}, { "state-policy", STATEPOLICY}, { "static-port", STATICPORT}, { "sticky-address", STICKYADDRESS}, { "synproxy", SYNPROXY}, { "table", TABLE}, { "tag", TAG}, { "tagged", TAGGED}, { "target", TARGET}, { "tbrsize", TBRSIZE}, { "timeout", TIMEOUT}, { "to", TO}, { "tos", TOS}, { "ttl", TTL}, { "upperlimit", UPPERLIMIT}, { "urpf-failed", URPFFAILED}, { "user", USER}, }; const struct keywords *p; p = bsearch(s, keywords, sizeof(keywords)/sizeof(keywords[0]), sizeof(keywords[0]), kw_cmp); if (p) { if (debug > 1) fprintf(stderr, "%s: %d\n", s, p->k_val); return (p->k_val); } else { if (debug > 1) fprintf(stderr, "string: %s\n", s); return (STRING); } } #define MAXPUSHBACK 128 char *parsebuf; int parseindex; char pushback_buffer[MAXPUSHBACK]; int pushback_index = 0; int lgetc(int quotec) { int c, next; if (parsebuf) { /* Read character from the parsebuffer instead of input. */ if (parseindex >= 0) { c = parsebuf[parseindex++]; if (c != '\0') return (c); parsebuf = NULL; } else parseindex++; } if (pushback_index) return (pushback_buffer[--pushback_index]); if (quotec) { if ((c = getc(file->stream)) == EOF) { yyerror("reached end of file while parsing quoted string"); if (popfile() == EOF) return (EOF); return (quotec); } return (c); } while ((c = getc(file->stream)) == '\\') { next = getc(file->stream); if (next != '\n') { c = next; break; } yylval.lineno = file->lineno; file->lineno++; } while (c == EOF) { if (popfile() == EOF) return (EOF); c = getc(file->stream); } return (c); } int lungetc(int c) { if (c == EOF) return (EOF); if (parsebuf) { parseindex--; if (parseindex >= 0) return (c); } if (pushback_index < MAXPUSHBACK-1) return (pushback_buffer[pushback_index++] = c); else return (EOF); } int findeol(void) { int c; parsebuf = NULL; /* skip to either EOF or the first real EOL */ while (1) { if (pushback_index) c = pushback_buffer[--pushback_index]; else c = lgetc(0); if (c == '\n') { file->lineno++; break; } if (c == EOF) break; } return (ERROR); } int yylex(void) { char buf[8096]; char *p, *val; int quotec, next, c; int token; top: p = buf; while ((c = lgetc(0)) == ' ' || c == '\t') ; /* nothing */ yylval.lineno = file->lineno; if (c == '#') while ((c = lgetc(0)) != '\n' && c != EOF) ; /* nothing */ if (c == '$' && parsebuf == NULL) { while (1) { if ((c = lgetc(0)) == EOF) return (0); if (p + 1 >= buf + sizeof(buf) - 1) { yyerror("string too long"); return (findeol()); } if (isalnum(c) || c == '_') { *p++ = (char)c; continue; } *p = '\0'; lungetc(c); break; } val = symget(buf); if (val == NULL) { yyerror("macro '%s' not defined", buf); return (findeol()); } parsebuf = val; parseindex = 0; goto top; } switch (c) { case '\'': case '"': quotec = c; while (1) { if ((c = lgetc(quotec)) == EOF) return (0); if (c == '\n') { file->lineno++; continue; } else if (c == '\\') { if ((next = lgetc(quotec)) == EOF) return (0); if (next == quotec || c == ' ' || c == '\t') c = next; else if (next == '\n') continue; else lungetc(next); } else if (c == quotec) { *p = '\0'; break; } if (p + 1 >= buf + sizeof(buf) - 1) { yyerror("string too long"); return (findeol()); } *p++ = (char)c; } yylval.v.string = strdup(buf); if (yylval.v.string == NULL) err(1, "yylex: strdup"); return (STRING); case '<': next = lgetc(0); if (next == '>') { yylval.v.i = PF_OP_XRG; return (PORTBINARY); } lungetc(next); break; case '>': next = lgetc(0); if (next == '<') { yylval.v.i = PF_OP_IRG; return (PORTBINARY); } lungetc(next); break; case '-': next = lgetc(0); if (next == '>') return (ARROW); lungetc(next); break; } #define allowed_to_end_number(x) \ (isspace(x) || x == ')' || x ==',' || x == '/' || x == '}' || x == '=') if (c == '-' || isdigit(c)) { do { *p++ = c; if ((unsigned)(p-buf) >= sizeof(buf)) { yyerror("string too long"); return (findeol()); } } while ((c = lgetc(0)) != EOF && isdigit(c)); lungetc(c); if (p == buf + 1 && buf[0] == '-') goto nodigits; if (c == EOF || allowed_to_end_number(c)) { const char *errstr = NULL; *p = '\0'; yylval.v.number = strtonum(buf, LLONG_MIN, LLONG_MAX, &errstr); if (errstr) { yyerror("\"%s\" invalid number: %s", buf, errstr); return (findeol()); } return (NUMBER); } else { nodigits: while (p > buf + 1) lungetc(*--p); c = *--p; if (c == '-') return (c); } } #define allowed_in_string(x) \ (isalnum(x) || (ispunct(x) && x != '(' && x != ')' && \ x != '{' && x != '}' && x != '<' && x != '>' && \ x != '!' && x != '=' && x != '/' && x != '#' && \ x != ',')) if (isalnum(c) || c == ':' || c == '_') { do { *p++ = c; if ((unsigned)(p-buf) >= sizeof(buf)) { yyerror("string too long"); return (findeol()); } } while ((c = lgetc(0)) != EOF && (allowed_in_string(c))); lungetc(c); *p = '\0'; if ((token = lookup(buf)) == STRING) if ((yylval.v.string = strdup(buf)) == NULL) err(1, "yylex: strdup"); return (token); } if (c == '\n') { yylval.lineno = file->lineno; file->lineno++; } if (c == EOF) return (0); return (c); } int check_file_secrecy(int fd, const char *fname) { struct stat st; if (fstat(fd, &st)) { warn("cannot stat %s", fname); return (-1); } if (st.st_uid != 0 && st.st_uid != getuid()) { warnx("%s: owner not root or current user", fname); return (-1); } if (st.st_mode & (S_IRWXG | S_IRWXO)) { warnx("%s: group/world readable/writeable", fname); return (-1); } return (0); } struct file * pushfile(const char *name, int secret) { struct file *nfile; if ((nfile = calloc(1, sizeof(struct file))) == NULL || (nfile->name = strdup(name)) == NULL) { warn("malloc"); return (NULL); } if (TAILQ_FIRST(&files) == NULL && strcmp(nfile->name, "-") == 0) { nfile->stream = stdin; free(nfile->name); if ((nfile->name = strdup("stdin")) == NULL) { warn("strdup"); free(nfile); return (NULL); } } else if ((nfile->stream = fopen(nfile->name, "r")) == NULL) { warn("%s", nfile->name); free(nfile->name); free(nfile); return (NULL); } else if (secret && check_file_secrecy(fileno(nfile->stream), nfile->name)) { fclose(nfile->stream); free(nfile->name); free(nfile); return (NULL); } nfile->lineno = 1; TAILQ_INSERT_TAIL(&files, nfile, entry); return (nfile); } int popfile(void) { struct file *prev; if ((prev = TAILQ_PREV(file, files, entry)) != NULL) { prev->errors += file->errors; TAILQ_REMOVE(&files, file, entry); fclose(file->stream); free(file->name); free(file); file = prev; return (0); } return (EOF); } int parse_config(char *filename, struct pfctl *xpf) { int errors = 0; struct sym *sym; pf = xpf; errors = 0; rulestate = PFCTL_STATE_NONE; returnicmpdefault = (ICMP_UNREACH << 8) | ICMP_UNREACH_PORT; returnicmp6default = (ICMP6_DST_UNREACH << 8) | ICMP6_DST_UNREACH_NOPORT; blockpolicy = PFRULE_DROP; require_order = 1; if ((file = pushfile(filename, 0)) == NULL) { warn("cannot open the main config file!"); return (-1); } yyparse(); errors = file->errors; popfile(); /* Free macros and check which have not been used. */ while ((sym = TAILQ_FIRST(&symhead))) { if ((pf->opts & PF_OPT_VERBOSE2) && !sym->used) fprintf(stderr, "warning: macro '%s' not " "used\n", sym->nam); free(sym->nam); free(sym->val); TAILQ_REMOVE(&symhead, sym, entry); free(sym); } return (errors ? -1 : 0); } int symset(const char *nam, const char *val, int persist) { struct sym *sym; for (sym = TAILQ_FIRST(&symhead); sym && strcmp(nam, sym->nam); sym = TAILQ_NEXT(sym, entry)) ; /* nothing */ if (sym != NULL) { if (sym->persist == 1) return (0); else { free(sym->nam); free(sym->val); TAILQ_REMOVE(&symhead, sym, entry); free(sym); } } if ((sym = calloc(1, sizeof(*sym))) == NULL) return (-1); sym->nam = strdup(nam); if (sym->nam == NULL) { free(sym); return (-1); } sym->val = strdup(val); if (sym->val == NULL) { free(sym->nam); free(sym); return (-1); } sym->used = 0; sym->persist = persist; TAILQ_INSERT_TAIL(&symhead, sym, entry); return (0); } int pfctl_cmdline_symset(char *s) { char *sym, *val; int ret; if ((val = strrchr(s, '=')) == NULL) return (-1); if ((sym = malloc(strlen(s) - strlen(val) + 1)) == NULL) err(1, "pfctl_cmdline_symset: malloc"); strlcpy(sym, s, strlen(s) - strlen(val) + 1); ret = symset(sym, val + 1, 1); free(sym); return (ret); } char * symget(const char *nam) { struct sym *sym; TAILQ_FOREACH(sym, &symhead, entry) if (strcmp(nam, sym->nam) == 0) { sym->used = 1; return (sym->val); } return (NULL); } void mv_rules(struct pf_ruleset *src, struct pf_ruleset *dst) { int i; struct pf_rule *r; for (i = 0; i < PF_RULESET_MAX; ++i) { while ((r = TAILQ_FIRST(src->rules[i].active.ptr)) != NULL) { TAILQ_REMOVE(src->rules[i].active.ptr, r, entries); TAILQ_INSERT_TAIL(dst->rules[i].active.ptr, r, entries); dst->anchor->match++; } src->anchor->match = 0; while ((r = TAILQ_FIRST(src->rules[i].inactive.ptr)) != NULL) { TAILQ_REMOVE(src->rules[i].inactive.ptr, r, entries); TAILQ_INSERT_TAIL(dst->rules[i].inactive.ptr, r, entries); } } } void decide_address_family(struct node_host *n, sa_family_t *af) { if (*af != 0 || n == NULL) return; *af = n->af; while ((n = n->next) != NULL) { if (n->af != *af) { *af = 0; return; } } } void remove_invalid_hosts(struct node_host **nh, sa_family_t *af) { struct node_host *n = *nh, *prev = NULL; while (n != NULL) { if (*af && n->af && n->af != *af) { /* unlink and free n */ struct node_host *next = n->next; /* adjust tail pointer */ if (n == (*nh)->tail) (*nh)->tail = prev; /* adjust previous node's next pointer */ if (prev == NULL) *nh = next; else prev->next = next; /* free node */ if (n->ifname != NULL) free(n->ifname); free(n); n = next; } else { if (n->af && !*af) *af = n->af; prev = n; n = n->next; } } } int invalid_redirect(struct node_host *nh, sa_family_t af) { if (!af) { struct node_host *n; /* tables and dyniftl are ok without an address family */ for (n = nh; n != NULL; n = n->next) { if (n->addr.type != PF_ADDR_TABLE && n->addr.type != PF_ADDR_DYNIFTL) { yyerror("address family not given and " "translation address expands to multiple " "address families"); return (1); } } } if (nh == NULL) { yyerror("no translation address with matching address family " "found."); return (1); } return (0); } int atoul(char *s, u_long *ulvalp) { u_long ulval; char *ep; errno = 0; ulval = strtoul(s, &ep, 0); if (s[0] == '\0' || *ep != '\0') return (-1); if (errno == ERANGE && ulval == ULONG_MAX) return (-1); *ulvalp = ulval; return (0); } int getservice(char *n) { struct servent *s; u_long ulval; if (atoul(n, &ulval) == 0) { if (ulval > 65535) { yyerror("illegal port value %lu", ulval); return (-1); } return (htons(ulval)); } else { s = getservbyname(n, "tcp"); if (s == NULL) s = getservbyname(n, "udp"); if (s == NULL) { yyerror("unknown port %s", n); return (-1); } return (s->s_port); } } int rule_label(struct pf_rule *r, char *s) { if (s) { if (strlcpy(r->label, s, sizeof(r->label)) >= sizeof(r->label)) { yyerror("rule label too long (max %d chars)", sizeof(r->label)-1); return (-1); } } return (0); } u_int16_t parseicmpspec(char *w, sa_family_t af) { const struct icmpcodeent *p; u_long ulval; u_int8_t icmptype; if (af == AF_INET) icmptype = returnicmpdefault >> 8; else icmptype = returnicmp6default >> 8; if (atoul(w, &ulval) == -1) { if ((p = geticmpcodebyname(icmptype, w, af)) == NULL) { yyerror("unknown icmp code %s", w); return (0); } ulval = p->code; } if (ulval > 255) { yyerror("invalid icmp code %lu", ulval); return (0); } return (icmptype << 8 | ulval); } int parseport(char *port, struct range *r, int extensions) { char *p = strchr(port, ':'); if (p == NULL) { if ((r->a = getservice(port)) == -1) return (-1); r->b = 0; r->t = PF_OP_NONE; return (0); } if ((extensions & PPORT_STAR) && !strcmp(p+1, "*")) { *p = 0; if ((r->a = getservice(port)) == -1) return (-1); r->b = 0; r->t = PF_OP_IRG; return (0); } if ((extensions & PPORT_RANGE)) { *p++ = 0; if ((r->a = getservice(port)) == -1 || (r->b = getservice(p)) == -1) return (-1); if (r->a == r->b) { r->b = 0; r->t = PF_OP_NONE; } else r->t = PF_OP_RRG; return (0); } return (-1); } int pfctl_load_anchors(int dev, struct pfctl *pf, struct pfr_buffer *trans) { struct loadanchors *la; TAILQ_FOREACH(la, &loadanchorshead, entries) { if (pf->opts & PF_OPT_VERBOSE) fprintf(stderr, "\nLoading anchor %s from %s\n", la->anchorname, la->filename); if (pfctl_rules(dev, la->filename, pf->opts, pf->optimize, la->anchorname, trans) == -1) return (-1); } return (0); } int rt_tableid_max(void) { #ifdef __FreeBSD__ int fibs; size_t l = sizeof(fibs); if (sysctlbyname("net.fibs", &fibs, &l, NULL, 0) == -1) fibs = 16; /* XXX RT_MAXFIBS, at least limit it some. */ /* * As the OpenBSD code only compares > and not >= we need to adjust * here given we only accept values of 0..n and want to avoid #ifdefs * in the grammer. */ return (fibs - 1); #else return (RT_TABLEID_MAX); #endif } Index: head/sbin/pfctl/pfctl_parser.c =================================================================== --- head/sbin/pfctl/pfctl_parser.c (revision 287221) +++ head/sbin/pfctl/pfctl_parser.c (revision 287222) @@ -1,1766 +1,1761 @@ /* $OpenBSD: pfctl_parser.c,v 1.240 2008/06/10 20:55:02 mcbride Exp $ */ /* * Copyright (c) 2001 Daniel Hartmeier * Copyright (c) 2002,2003 Henning Brauer * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * * - Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * - Redistributions in binary form must reproduce the above * copyright notice, this list of conditions and the following * disclaimer in the documentation and/or other materials provided * with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE * COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. * */ #include __FBSDID("$FreeBSD$"); #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include "pfctl_parser.h" #include "pfctl.h" void print_op (u_int8_t, const char *, const char *); void print_port (u_int8_t, u_int16_t, u_int16_t, const char *, int); void print_ugid (u_int8_t, unsigned, unsigned, const char *, unsigned); void print_flags (u_int8_t); void print_fromto(struct pf_rule_addr *, pf_osfp_t, struct pf_rule_addr *, u_int8_t, u_int8_t, int, int); int ifa_skip_if(const char *filter, struct node_host *p); struct node_host *ifa_grouplookup(const char *, int); struct node_host *host_if(const char *, int); struct node_host *host_v4(const char *, int); struct node_host *host_v6(const char *, int); struct node_host *host_dns(const char *, int, int); const char *tcpflags = "FSRPAUEW"; static const struct icmptypeent icmp_type[] = { { "echoreq", ICMP_ECHO }, { "echorep", ICMP_ECHOREPLY }, { "unreach", ICMP_UNREACH }, { "squench", ICMP_SOURCEQUENCH }, { "redir", ICMP_REDIRECT }, { "althost", ICMP_ALTHOSTADDR }, { "routeradv", ICMP_ROUTERADVERT }, { "routersol", ICMP_ROUTERSOLICIT }, { "timex", ICMP_TIMXCEED }, { "paramprob", ICMP_PARAMPROB }, { "timereq", ICMP_TSTAMP }, { "timerep", ICMP_TSTAMPREPLY }, { "inforeq", ICMP_IREQ }, { "inforep", ICMP_IREQREPLY }, { "maskreq", ICMP_MASKREQ }, { "maskrep", ICMP_MASKREPLY }, { "trace", ICMP_TRACEROUTE }, { "dataconv", ICMP_DATACONVERR }, { "mobredir", ICMP_MOBILE_REDIRECT }, { "ipv6-where", ICMP_IPV6_WHEREAREYOU }, { "ipv6-here", ICMP_IPV6_IAMHERE }, { "mobregreq", ICMP_MOBILE_REGREQUEST }, { "mobregrep", ICMP_MOBILE_REGREPLY }, { "skip", ICMP_SKIP }, { "photuris", ICMP_PHOTURIS } }; static const struct icmptypeent icmp6_type[] = { { "unreach", ICMP6_DST_UNREACH }, { "toobig", ICMP6_PACKET_TOO_BIG }, { "timex", ICMP6_TIME_EXCEEDED }, { "paramprob", ICMP6_PARAM_PROB }, { "echoreq", ICMP6_ECHO_REQUEST }, { "echorep", ICMP6_ECHO_REPLY }, { "groupqry", ICMP6_MEMBERSHIP_QUERY }, { "listqry", MLD_LISTENER_QUERY }, { "grouprep", ICMP6_MEMBERSHIP_REPORT }, { "listenrep", MLD_LISTENER_REPORT }, { "groupterm", ICMP6_MEMBERSHIP_REDUCTION }, { "listendone", MLD_LISTENER_DONE }, { "routersol", ND_ROUTER_SOLICIT }, { "routeradv", ND_ROUTER_ADVERT }, { "neighbrsol", ND_NEIGHBOR_SOLICIT }, { "neighbradv", ND_NEIGHBOR_ADVERT }, { "redir", ND_REDIRECT }, { "routrrenum", ICMP6_ROUTER_RENUMBERING }, { "wrureq", ICMP6_WRUREQUEST }, { "wrurep", ICMP6_WRUREPLY }, { "fqdnreq", ICMP6_FQDN_QUERY }, { "fqdnrep", ICMP6_FQDN_REPLY }, { "niqry", ICMP6_NI_QUERY }, { "nirep", ICMP6_NI_REPLY }, { "mtraceresp", MLD_MTRACE_RESP }, { "mtrace", MLD_MTRACE } }; static const struct icmpcodeent icmp_code[] = { { "net-unr", ICMP_UNREACH, ICMP_UNREACH_NET }, { "host-unr", ICMP_UNREACH, ICMP_UNREACH_HOST }, { "proto-unr", ICMP_UNREACH, ICMP_UNREACH_PROTOCOL }, { "port-unr", ICMP_UNREACH, ICMP_UNREACH_PORT }, { "needfrag", ICMP_UNREACH, ICMP_UNREACH_NEEDFRAG }, { "srcfail", ICMP_UNREACH, ICMP_UNREACH_SRCFAIL }, { "net-unk", ICMP_UNREACH, ICMP_UNREACH_NET_UNKNOWN }, { "host-unk", ICMP_UNREACH, ICMP_UNREACH_HOST_UNKNOWN }, { "isolate", ICMP_UNREACH, ICMP_UNREACH_ISOLATED }, { "net-prohib", ICMP_UNREACH, ICMP_UNREACH_NET_PROHIB }, { "host-prohib", ICMP_UNREACH, ICMP_UNREACH_HOST_PROHIB }, { "net-tos", ICMP_UNREACH, ICMP_UNREACH_TOSNET }, { "host-tos", ICMP_UNREACH, ICMP_UNREACH_TOSHOST }, { "filter-prohib", ICMP_UNREACH, ICMP_UNREACH_FILTER_PROHIB }, { "host-preced", ICMP_UNREACH, ICMP_UNREACH_HOST_PRECEDENCE }, { "cutoff-preced", ICMP_UNREACH, ICMP_UNREACH_PRECEDENCE_CUTOFF }, { "redir-net", ICMP_REDIRECT, ICMP_REDIRECT_NET }, { "redir-host", ICMP_REDIRECT, ICMP_REDIRECT_HOST }, { "redir-tos-net", ICMP_REDIRECT, ICMP_REDIRECT_TOSNET }, { "redir-tos-host", ICMP_REDIRECT, ICMP_REDIRECT_TOSHOST }, { "normal-adv", ICMP_ROUTERADVERT, ICMP_ROUTERADVERT_NORMAL }, { "common-adv", ICMP_ROUTERADVERT, ICMP_ROUTERADVERT_NOROUTE_COMMON }, { "transit", ICMP_TIMXCEED, ICMP_TIMXCEED_INTRANS }, { "reassemb", ICMP_TIMXCEED, ICMP_TIMXCEED_REASS }, { "badhead", ICMP_PARAMPROB, ICMP_PARAMPROB_ERRATPTR }, { "optmiss", ICMP_PARAMPROB, ICMP_PARAMPROB_OPTABSENT }, { "badlen", ICMP_PARAMPROB, ICMP_PARAMPROB_LENGTH }, { "unknown-ind", ICMP_PHOTURIS, ICMP_PHOTURIS_UNKNOWN_INDEX }, { "auth-fail", ICMP_PHOTURIS, ICMP_PHOTURIS_AUTH_FAILED }, { "decrypt-fail", ICMP_PHOTURIS, ICMP_PHOTURIS_DECRYPT_FAILED } }; static const struct icmpcodeent icmp6_code[] = { { "admin-unr", ICMP6_DST_UNREACH, ICMP6_DST_UNREACH_ADMIN }, { "noroute-unr", ICMP6_DST_UNREACH, ICMP6_DST_UNREACH_NOROUTE }, { "notnbr-unr", ICMP6_DST_UNREACH, ICMP6_DST_UNREACH_NOTNEIGHBOR }, { "beyond-unr", ICMP6_DST_UNREACH, ICMP6_DST_UNREACH_BEYONDSCOPE }, { "addr-unr", ICMP6_DST_UNREACH, ICMP6_DST_UNREACH_ADDR }, { "port-unr", ICMP6_DST_UNREACH, ICMP6_DST_UNREACH_NOPORT }, { "transit", ICMP6_TIME_EXCEEDED, ICMP6_TIME_EXCEED_TRANSIT }, { "reassemb", ICMP6_TIME_EXCEEDED, ICMP6_TIME_EXCEED_REASSEMBLY }, { "badhead", ICMP6_PARAM_PROB, ICMP6_PARAMPROB_HEADER }, { "nxthdr", ICMP6_PARAM_PROB, ICMP6_PARAMPROB_NEXTHEADER }, { "redironlink", ND_REDIRECT, ND_REDIRECT_ONLINK }, { "redirrouter", ND_REDIRECT, ND_REDIRECT_ROUTER } }; const struct pf_timeout pf_timeouts[] = { { "tcp.first", PFTM_TCP_FIRST_PACKET }, { "tcp.opening", PFTM_TCP_OPENING }, { "tcp.established", PFTM_TCP_ESTABLISHED }, { "tcp.closing", PFTM_TCP_CLOSING }, { "tcp.finwait", PFTM_TCP_FIN_WAIT }, { "tcp.closed", PFTM_TCP_CLOSED }, { "tcp.tsdiff", PFTM_TS_DIFF }, { "udp.first", PFTM_UDP_FIRST_PACKET }, { "udp.single", PFTM_UDP_SINGLE }, { "udp.multiple", PFTM_UDP_MULTIPLE }, { "icmp.first", PFTM_ICMP_FIRST_PACKET }, { "icmp.error", PFTM_ICMP_ERROR_REPLY }, { "other.first", PFTM_OTHER_FIRST_PACKET }, { "other.single", PFTM_OTHER_SINGLE }, { "other.multiple", PFTM_OTHER_MULTIPLE }, { "frag", PFTM_FRAG }, { "interval", PFTM_INTERVAL }, { "adaptive.start", PFTM_ADAPTIVE_START }, { "adaptive.end", PFTM_ADAPTIVE_END }, { "src.track", PFTM_SRC_NODE }, { NULL, 0 } }; const struct icmptypeent * geticmptypebynumber(u_int8_t type, sa_family_t af) { unsigned int i; if (af != AF_INET6) { for (i=0; i < (sizeof (icmp_type) / sizeof(icmp_type[0])); i++) { if (type == icmp_type[i].type) return (&icmp_type[i]); } } else { for (i=0; i < (sizeof (icmp6_type) / sizeof(icmp6_type[0])); i++) { if (type == icmp6_type[i].type) return (&icmp6_type[i]); } } return (NULL); } const struct icmptypeent * geticmptypebyname(char *w, sa_family_t af) { unsigned int i; if (af != AF_INET6) { for (i=0; i < (sizeof (icmp_type) / sizeof(icmp_type[0])); i++) { if (!strcmp(w, icmp_type[i].name)) return (&icmp_type[i]); } } else { for (i=0; i < (sizeof (icmp6_type) / sizeof(icmp6_type[0])); i++) { if (!strcmp(w, icmp6_type[i].name)) return (&icmp6_type[i]); } } return (NULL); } const struct icmpcodeent * geticmpcodebynumber(u_int8_t type, u_int8_t code, sa_family_t af) { unsigned int i; if (af != AF_INET6) { for (i=0; i < (sizeof (icmp_code) / sizeof(icmp_code[0])); i++) { if (type == icmp_code[i].type && code == icmp_code[i].code) return (&icmp_code[i]); } } else { for (i=0; i < (sizeof (icmp6_code) / sizeof(icmp6_code[0])); i++) { if (type == icmp6_code[i].type && code == icmp6_code[i].code) return (&icmp6_code[i]); } } return (NULL); } const struct icmpcodeent * geticmpcodebyname(u_long type, char *w, sa_family_t af) { unsigned int i; if (af != AF_INET6) { for (i=0; i < (sizeof (icmp_code) / sizeof(icmp_code[0])); i++) { if (type == icmp_code[i].type && !strcmp(w, icmp_code[i].name)) return (&icmp_code[i]); } } else { for (i=0; i < (sizeof (icmp6_code) / sizeof(icmp6_code[0])); i++) { if (type == icmp6_code[i].type && !strcmp(w, icmp6_code[i].name)) return (&icmp6_code[i]); } } return (NULL); } void print_op(u_int8_t op, const char *a1, const char *a2) { if (op == PF_OP_IRG) printf(" %s >< %s", a1, a2); else if (op == PF_OP_XRG) printf(" %s <> %s", a1, a2); else if (op == PF_OP_EQ) printf(" = %s", a1); else if (op == PF_OP_NE) printf(" != %s", a1); else if (op == PF_OP_LT) printf(" < %s", a1); else if (op == PF_OP_LE) printf(" <= %s", a1); else if (op == PF_OP_GT) printf(" > %s", a1); else if (op == PF_OP_GE) printf(" >= %s", a1); else if (op == PF_OP_RRG) printf(" %s:%s", a1, a2); } void print_port(u_int8_t op, u_int16_t p1, u_int16_t p2, const char *proto, int numeric) { char a1[6], a2[6]; struct servent *s; if (!numeric) s = getservbyport(p1, proto); else s = NULL; p1 = ntohs(p1); p2 = ntohs(p2); snprintf(a1, sizeof(a1), "%u", p1); snprintf(a2, sizeof(a2), "%u", p2); printf(" port"); if (s != NULL && (op == PF_OP_EQ || op == PF_OP_NE)) print_op(op, s->s_name, a2); else print_op(op, a1, a2); } void print_ugid(u_int8_t op, unsigned u1, unsigned u2, const char *t, unsigned umax) { char a1[11], a2[11]; snprintf(a1, sizeof(a1), "%u", u1); snprintf(a2, sizeof(a2), "%u", u2); printf(" %s", t); if (u1 == umax && (op == PF_OP_EQ || op == PF_OP_NE)) print_op(op, "unknown", a2); else print_op(op, a1, a2); } void print_flags(u_int8_t f) { int i; for (i = 0; tcpflags[i]; ++i) if (f & (1 << i)) printf("%c", tcpflags[i]); } void print_fromto(struct pf_rule_addr *src, pf_osfp_t osfp, struct pf_rule_addr *dst, sa_family_t af, u_int8_t proto, int verbose, int numeric) { char buf[PF_OSFP_LEN*3]; if (src->addr.type == PF_ADDR_ADDRMASK && dst->addr.type == PF_ADDR_ADDRMASK && PF_AZERO(&src->addr.v.a.addr, AF_INET6) && PF_AZERO(&src->addr.v.a.mask, AF_INET6) && PF_AZERO(&dst->addr.v.a.addr, AF_INET6) && PF_AZERO(&dst->addr.v.a.mask, AF_INET6) && !src->neg && !dst->neg && !src->port_op && !dst->port_op && osfp == PF_OSFP_ANY) printf(" all"); else { printf(" from "); if (src->neg) printf("! "); print_addr(&src->addr, af, verbose); if (src->port_op) print_port(src->port_op, src->port[0], src->port[1], proto == IPPROTO_TCP ? "tcp" : "udp", numeric); if (osfp != PF_OSFP_ANY) printf(" os \"%s\"", pfctl_lookup_fingerprint(osfp, buf, sizeof(buf))); printf(" to "); if (dst->neg) printf("! "); print_addr(&dst->addr, af, verbose); if (dst->port_op) print_port(dst->port_op, dst->port[0], dst->port[1], proto == IPPROTO_TCP ? "tcp" : "udp", numeric); } } void print_pool(struct pf_pool *pool, u_int16_t p1, u_int16_t p2, sa_family_t af, int id) { struct pf_pooladdr *pooladdr; if ((TAILQ_FIRST(&pool->list) != NULL) && TAILQ_NEXT(TAILQ_FIRST(&pool->list), entries) != NULL) printf("{ "); TAILQ_FOREACH(pooladdr, &pool->list, entries){ switch (id) { case PF_NAT: case PF_RDR: case PF_BINAT: print_addr(&pooladdr->addr, af, 0); break; case PF_PASS: if (PF_AZERO(&pooladdr->addr.v.a.addr, af)) printf("%s", pooladdr->ifname); else { printf("(%s ", pooladdr->ifname); print_addr(&pooladdr->addr, af, 0); printf(")"); } break; default: break; } if (TAILQ_NEXT(pooladdr, entries) != NULL) printf(", "); else if (TAILQ_NEXT(TAILQ_FIRST(&pool->list), entries) != NULL) printf(" }"); } switch (id) { case PF_NAT: if ((p1 != PF_NAT_PROXY_PORT_LOW || p2 != PF_NAT_PROXY_PORT_HIGH) && (p1 != 0 || p2 != 0)) { if (p1 == p2) printf(" port %u", p1); else printf(" port %u:%u", p1, p2); } break; case PF_RDR: if (p1) { printf(" port %u", p1); if (p2 && (p2 != p1)) printf(":%u", p2); } break; default: break; } switch (pool->opts & PF_POOL_TYPEMASK) { case PF_POOL_NONE: break; case PF_POOL_BITMASK: printf(" bitmask"); break; case PF_POOL_RANDOM: printf(" random"); break; case PF_POOL_SRCHASH: printf(" source-hash 0x%08x%08x%08x%08x", pool->key.key32[0], pool->key.key32[1], pool->key.key32[2], pool->key.key32[3]); break; case PF_POOL_ROUNDROBIN: printf(" round-robin"); break; } if (pool->opts & PF_POOL_STICKYADDR) printf(" sticky-address"); if (id == PF_NAT && p1 == 0 && p2 == 0) printf(" static-port"); } const char *pf_reasons[PFRES_MAX+1] = PFRES_NAMES; const char *pf_lcounters[LCNT_MAX+1] = LCNT_NAMES; const char *pf_fcounters[FCNT_MAX+1] = FCNT_NAMES; const char *pf_scounters[FCNT_MAX+1] = FCNT_NAMES; void print_status(struct pf_status *s, int opts) { char statline[80], *running; time_t runtime; int i; char buf[PF_MD5_DIGEST_LENGTH * 2 + 1]; static const char hex[] = "0123456789abcdef"; runtime = time(NULL) - s->since; running = s->running ? "Enabled" : "Disabled"; if (s->since) { unsigned int sec, min, hrs, day = runtime; sec = day % 60; day /= 60; min = day % 60; day /= 60; hrs = day % 24; day /= 24; snprintf(statline, sizeof(statline), "Status: %s for %u days %.2u:%.2u:%.2u", running, day, hrs, min, sec); } else snprintf(statline, sizeof(statline), "Status: %s", running); printf("%-44s", statline); switch (s->debug) { case PF_DEBUG_NONE: printf("%15s\n\n", "Debug: None"); break; case PF_DEBUG_URGENT: printf("%15s\n\n", "Debug: Urgent"); break; case PF_DEBUG_MISC: printf("%15s\n\n", "Debug: Misc"); break; case PF_DEBUG_NOISY: printf("%15s\n\n", "Debug: Loud"); break; } if (opts & PF_OPT_VERBOSE) { printf("Hostid: 0x%08x\n", ntohl(s->hostid)); for (i = 0; i < PF_MD5_DIGEST_LENGTH; i++) { buf[i + i] = hex[s->pf_chksum[i] >> 4]; buf[i + i + 1] = hex[s->pf_chksum[i] & 0x0f]; } buf[i + i] = '\0'; printf("Checksum: 0x%s\n\n", buf); } if (s->ifname[0] != 0) { printf("Interface Stats for %-16s %5s %16s\n", s->ifname, "IPv4", "IPv6"); printf(" %-25s %14llu %16llu\n", "Bytes In", (unsigned long long)s->bcounters[0][0], (unsigned long long)s->bcounters[1][0]); printf(" %-25s %14llu %16llu\n", "Bytes Out", (unsigned long long)s->bcounters[0][1], (unsigned long long)s->bcounters[1][1]); printf(" Packets In\n"); printf(" %-23s %14llu %16llu\n", "Passed", (unsigned long long)s->pcounters[0][0][PF_PASS], (unsigned long long)s->pcounters[1][0][PF_PASS]); printf(" %-23s %14llu %16llu\n", "Blocked", (unsigned long long)s->pcounters[0][0][PF_DROP], (unsigned long long)s->pcounters[1][0][PF_DROP]); printf(" Packets Out\n"); printf(" %-23s %14llu %16llu\n", "Passed", (unsigned long long)s->pcounters[0][1][PF_PASS], (unsigned long long)s->pcounters[1][1][PF_PASS]); printf(" %-23s %14llu %16llu\n\n", "Blocked", (unsigned long long)s->pcounters[0][1][PF_DROP], (unsigned long long)s->pcounters[1][1][PF_DROP]); } printf("%-27s %14s %16s\n", "State Table", "Total", "Rate"); printf(" %-25s %14u %14s\n", "current entries", s->states, ""); for (i = 0; i < FCNT_MAX; i++) { printf(" %-25s %14llu ", pf_fcounters[i], (unsigned long long)s->fcounters[i]); if (runtime > 0) printf("%14.1f/s\n", (double)s->fcounters[i] / (double)runtime); else printf("%14s\n", ""); } if (opts & PF_OPT_VERBOSE) { printf("Source Tracking Table\n"); printf(" %-25s %14u %14s\n", "current entries", s->src_nodes, ""); for (i = 0; i < SCNT_MAX; i++) { printf(" %-25s %14lld ", pf_scounters[i], #ifdef __FreeBSD__ (long long)s->scounters[i]); #else s->scounters[i]); #endif if (runtime > 0) printf("%14.1f/s\n", (double)s->scounters[i] / (double)runtime); else printf("%14s\n", ""); } } printf("Counters\n"); for (i = 0; i < PFRES_MAX; i++) { printf(" %-25s %14llu ", pf_reasons[i], (unsigned long long)s->counters[i]); if (runtime > 0) printf("%14.1f/s\n", (double)s->counters[i] / (double)runtime); else printf("%14s\n", ""); } if (opts & PF_OPT_VERBOSE) { printf("Limit Counters\n"); for (i = 0; i < LCNT_MAX; i++) { printf(" %-25s %14lld ", pf_lcounters[i], #ifdef __FreeBSD__ (unsigned long long)s->lcounters[i]); #else s->lcounters[i]); #endif if (runtime > 0) printf("%14.1f/s\n", (double)s->lcounters[i] / (double)runtime); else printf("%14s\n", ""); } } } void print_src_node(struct pf_src_node *sn, int opts) { struct pf_addr_wrap aw; int min, sec; memset(&aw, 0, sizeof(aw)); if (sn->af == AF_INET) aw.v.a.mask.addr32[0] = 0xffffffff; else memset(&aw.v.a.mask, 0xff, sizeof(aw.v.a.mask)); aw.v.a.addr = sn->addr; print_addr(&aw, sn->af, opts & PF_OPT_VERBOSE2); printf(" -> "); aw.v.a.addr = sn->raddr; print_addr(&aw, sn->af, opts & PF_OPT_VERBOSE2); printf(" ( states %u, connections %u, rate %u.%u/%us )\n", sn->states, sn->conn, sn->conn_rate.count / 1000, (sn->conn_rate.count % 1000) / 100, sn->conn_rate.seconds); if (opts & PF_OPT_VERBOSE) { sec = sn->creation % 60; sn->creation /= 60; min = sn->creation % 60; sn->creation /= 60; printf(" age %.2u:%.2u:%.2u", sn->creation, min, sec); if (sn->states == 0) { sec = sn->expire % 60; sn->expire /= 60; min = sn->expire % 60; sn->expire /= 60; printf(", expires in %.2u:%.2u:%.2u", sn->expire, min, sec); } printf(", %llu pkts, %llu bytes", #ifdef __FreeBSD__ (unsigned long long)(sn->packets[0] + sn->packets[1]), (unsigned long long)(sn->bytes[0] + sn->bytes[1])); #else sn->packets[0] + sn->packets[1], sn->bytes[0] + sn->bytes[1]); #endif switch (sn->ruletype) { case PF_NAT: if (sn->rule.nr != -1) printf(", nat rule %u", sn->rule.nr); break; case PF_RDR: if (sn->rule.nr != -1) printf(", rdr rule %u", sn->rule.nr); break; case PF_PASS: if (sn->rule.nr != -1) printf(", filter rule %u", sn->rule.nr); break; } printf("\n"); } } void print_rule(struct pf_rule *r, const char *anchor_call, int verbose, int numeric) { static const char *actiontypes[] = { "pass", "block", "scrub", "no scrub", "nat", "no nat", "binat", "no binat", "rdr", "no rdr" }; static const char *anchortypes[] = { "anchor", "anchor", "anchor", "anchor", "nat-anchor", "nat-anchor", "binat-anchor", "binat-anchor", "rdr-anchor", "rdr-anchor" }; int i, opts; if (verbose) printf("@%d ", r->nr); if (r->action > PF_NORDR) printf("action(%d)", r->action); else if (anchor_call[0]) { if (anchor_call[0] == '_') { printf("%s", anchortypes[r->action]); } else printf("%s \"%s\"", anchortypes[r->action], anchor_call); } else { printf("%s", actiontypes[r->action]); if (r->natpass) printf(" pass"); } if (r->action == PF_DROP) { if (r->rule_flag & PFRULE_RETURN) printf(" return"); else if (r->rule_flag & PFRULE_RETURNRST) { if (!r->return_ttl) printf(" return-rst"); else printf(" return-rst(ttl %d)", r->return_ttl); } else if (r->rule_flag & PFRULE_RETURNICMP) { const struct icmpcodeent *ic, *ic6; ic = geticmpcodebynumber(r->return_icmp >> 8, r->return_icmp & 255, AF_INET); ic6 = geticmpcodebynumber(r->return_icmp6 >> 8, r->return_icmp6 & 255, AF_INET6); switch (r->af) { case AF_INET: printf(" return-icmp"); if (ic == NULL) printf("(%u)", r->return_icmp & 255); else printf("(%s)", ic->name); break; case AF_INET6: printf(" return-icmp6"); if (ic6 == NULL) printf("(%u)", r->return_icmp6 & 255); else printf("(%s)", ic6->name); break; default: printf(" return-icmp"); if (ic == NULL) printf("(%u, ", r->return_icmp & 255); else printf("(%s, ", ic->name); if (ic6 == NULL) printf("%u)", r->return_icmp6 & 255); else printf("%s)", ic6->name); break; } } else printf(" drop"); } if (r->direction == PF_IN) printf(" in"); else if (r->direction == PF_OUT) printf(" out"); if (r->log) { printf(" log"); if (r->log & ~PF_LOG || r->logif) { int count = 0; printf(" ("); if (r->log & PF_LOG_ALL) printf("%sall", count++ ? ", " : ""); if (r->log & PF_LOG_SOCKET_LOOKUP) printf("%suser", count++ ? ", " : ""); if (r->logif) printf("%sto pflog%u", count++ ? ", " : "", r->logif); printf(")"); } } if (r->quick) printf(" quick"); if (r->ifname[0]) { if (r->ifnot) printf(" on ! %s", r->ifname); else printf(" on %s", r->ifname); } if (r->rt) { if (r->rt == PF_ROUTETO) printf(" route-to"); else if (r->rt == PF_REPLYTO) printf(" reply-to"); else if (r->rt == PF_DUPTO) printf(" dup-to"); else if (r->rt == PF_FASTROUTE) printf(" fastroute"); if (r->rt != PF_FASTROUTE) { printf(" "); print_pool(&r->rpool, 0, 0, r->af, PF_PASS); } } if (r->af) { if (r->af == AF_INET) printf(" inet"); else printf(" inet6"); } if (r->proto) { struct protoent *p; if ((p = getprotobynumber(r->proto)) != NULL) printf(" proto %s", p->p_name); else printf(" proto %u", r->proto); } print_fromto(&r->src, r->os_fingerprint, &r->dst, r->af, r->proto, verbose, numeric); if (r->uid.op) print_ugid(r->uid.op, r->uid.uid[0], r->uid.uid[1], "user", UID_MAX); if (r->gid.op) print_ugid(r->gid.op, r->gid.gid[0], r->gid.gid[1], "group", GID_MAX); if (r->flags || r->flagset) { printf(" flags "); print_flags(r->flags); printf("/"); print_flags(r->flagset); } else if (r->action == PF_PASS && (!r->proto || r->proto == IPPROTO_TCP) && !(r->rule_flag & PFRULE_FRAGMENT) && !anchor_call[0] && r->keep_state) printf(" flags any"); if (r->type) { const struct icmptypeent *it; it = geticmptypebynumber(r->type-1, r->af); if (r->af != AF_INET6) printf(" icmp-type"); else printf(" icmp6-type"); if (it != NULL) printf(" %s", it->name); else printf(" %u", r->type-1); if (r->code) { const struct icmpcodeent *ic; ic = geticmpcodebynumber(r->type-1, r->code-1, r->af); if (ic != NULL) printf(" code %s", ic->name); else printf(" code %u", r->code-1); } } if (r->tos) printf(" tos 0x%2.2x", r->tos); if (!r->keep_state && r->action == PF_PASS && !anchor_call[0]) printf(" no state"); else if (r->keep_state == PF_STATE_NORMAL) printf(" keep state"); else if (r->keep_state == PF_STATE_MODULATE) printf(" modulate state"); else if (r->keep_state == PF_STATE_SYNPROXY) printf(" synproxy state"); if (r->prob) { char buf[20]; snprintf(buf, sizeof(buf), "%f", r->prob*100.0/(UINT_MAX+1.0)); for (i = strlen(buf)-1; i > 0; i--) { if (buf[i] == '0') buf[i] = '\0'; else { if (buf[i] == '.') buf[i] = '\0'; break; } } printf(" probability %s%%", buf); } opts = 0; if (r->max_states || r->max_src_nodes || r->max_src_states) opts = 1; if (r->rule_flag & PFRULE_NOSYNC) opts = 1; if (r->rule_flag & PFRULE_SRCTRACK) opts = 1; if (r->rule_flag & PFRULE_IFBOUND) opts = 1; if (r->rule_flag & PFRULE_STATESLOPPY) opts = 1; for (i = 0; !opts && i < PFTM_MAX; ++i) if (r->timeout[i]) opts = 1; if (opts) { printf(" ("); if (r->max_states) { printf("max %u", r->max_states); opts = 0; } if (r->rule_flag & PFRULE_NOSYNC) { if (!opts) printf(", "); printf("no-sync"); opts = 0; } if (r->rule_flag & PFRULE_SRCTRACK) { if (!opts) printf(", "); printf("source-track"); if (r->rule_flag & PFRULE_RULESRCTRACK) printf(" rule"); else printf(" global"); opts = 0; } if (r->max_src_states) { if (!opts) printf(", "); printf("max-src-states %u", r->max_src_states); opts = 0; } if (r->max_src_conn) { if (!opts) printf(", "); printf("max-src-conn %u", r->max_src_conn); opts = 0; } if (r->max_src_conn_rate.limit) { if (!opts) printf(", "); printf("max-src-conn-rate %u/%u", r->max_src_conn_rate.limit, r->max_src_conn_rate.seconds); opts = 0; } if (r->max_src_nodes) { if (!opts) printf(", "); printf("max-src-nodes %u", r->max_src_nodes); opts = 0; } if (r->overload_tblname[0]) { if (!opts) printf(", "); printf("overload <%s>", r->overload_tblname); if (r->flush) printf(" flush"); if (r->flush & PF_FLUSH_GLOBAL) printf(" global"); } if (r->rule_flag & PFRULE_IFBOUND) { if (!opts) printf(", "); printf("if-bound"); opts = 0; } if (r->rule_flag & PFRULE_STATESLOPPY) { if (!opts) printf(", "); printf("sloppy"); opts = 0; } for (i = 0; i < PFTM_MAX; ++i) if (r->timeout[i]) { int j; if (!opts) printf(", "); opts = 0; for (j = 0; pf_timeouts[j].name != NULL; ++j) if (pf_timeouts[j].timeout == i) break; printf("%s %u", pf_timeouts[j].name == NULL ? "inv.timeout" : pf_timeouts[j].name, r->timeout[i]); } printf(")"); } if (r->rule_flag & PFRULE_FRAGMENT) printf(" fragment"); if (r->rule_flag & PFRULE_NODF) printf(" no-df"); if (r->rule_flag & PFRULE_RANDOMID) printf(" random-id"); if (r->min_ttl) printf(" min-ttl %d", r->min_ttl); if (r->max_mss) printf(" max-mss %d", r->max_mss); if (r->rule_flag & PFRULE_SET_TOS) printf(" set-tos 0x%2.2x", r->set_tos); if (r->allow_opts) printf(" allow-opts"); if (r->action == PF_SCRUB) { if (r->rule_flag & PFRULE_REASSEMBLE_TCP) printf(" reassemble tcp"); - if (r->rule_flag & PFRULE_FRAGDROP) - printf(" fragment drop-ovl"); - else if (r->rule_flag & PFRULE_FRAGCROP) - printf(" fragment crop"); - else - printf(" fragment reassemble"); + printf(" fragment reassemble"); } if (r->label[0]) printf(" label \"%s\"", r->label); if (r->qname[0] && r->pqname[0]) printf(" queue(%s, %s)", r->qname, r->pqname); else if (r->qname[0]) printf(" queue %s", r->qname); if (r->tagname[0]) printf(" tag %s", r->tagname); if (r->match_tagname[0]) { if (r->match_tag_not) printf(" !"); printf(" tagged %s", r->match_tagname); } if (r->rtableid != -1) printf(" rtable %u", r->rtableid); if (r->divert.port) { #ifdef __FreeBSD__ printf(" divert-to %u", ntohs(r->divert.port)); #else if (PF_AZERO(&r->divert.addr, r->af)) { printf(" divert-reply"); } else { /* XXX cut&paste from print_addr */ char buf[48]; printf(" divert-to "); if (inet_ntop(r->af, &r->divert.addr, buf, sizeof(buf)) == NULL) printf("?"); else printf("%s", buf); printf(" port %u", ntohs(r->divert.port)); } #endif } if (!anchor_call[0] && (r->action == PF_NAT || r->action == PF_BINAT || r->action == PF_RDR)) { printf(" -> "); print_pool(&r->rpool, r->rpool.proxy_port[0], r->rpool.proxy_port[1], r->af, r->action); } } void print_tabledef(const char *name, int flags, int addrs, struct node_tinithead *nodes) { struct node_tinit *ti, *nti; struct node_host *h; printf("table <%s>", name); if (flags & PFR_TFLAG_CONST) printf(" const"); if (flags & PFR_TFLAG_PERSIST) printf(" persist"); if (flags & PFR_TFLAG_COUNTERS) printf(" counters"); SIMPLEQ_FOREACH(ti, nodes, entries) { if (ti->file) { printf(" file \"%s\"", ti->file); continue; } printf(" {"); for (;;) { for (h = ti->host; h != NULL; h = h->next) { printf(h->not ? " !" : " "); print_addr(&h->addr, h->af, 0); } nti = SIMPLEQ_NEXT(ti, entries); if (nti != NULL && nti->file == NULL) ti = nti; /* merge lists */ else break; } printf(" }"); } if (addrs && SIMPLEQ_EMPTY(nodes)) printf(" { }"); printf("\n"); } int parse_flags(char *s) { char *p, *q; u_int8_t f = 0; for (p = s; *p; p++) { if ((q = strchr(tcpflags, *p)) == NULL) return -1; else f |= 1 << (q - tcpflags); } return (f ? f : PF_TH_ALL); } void set_ipmask(struct node_host *h, u_int8_t b) { struct pf_addr *m, *n; int i, j = 0; m = &h->addr.v.a.mask; memset(m, 0, sizeof(*m)); while (b >= 32) { m->addr32[j++] = 0xffffffff; b -= 32; } for (i = 31; i > 31-b; --i) m->addr32[j] |= (1 << i); if (b) m->addr32[j] = htonl(m->addr32[j]); /* Mask off bits of the address that will never be used. */ n = &h->addr.v.a.addr; if (h->addr.type == PF_ADDR_ADDRMASK) for (i = 0; i < 4; i++) n->addr32[i] = n->addr32[i] & m->addr32[i]; } int check_netmask(struct node_host *h, sa_family_t af) { struct node_host *n = NULL; struct pf_addr *m; for (n = h; n != NULL; n = n->next) { if (h->addr.type == PF_ADDR_TABLE) continue; m = &h->addr.v.a.mask; /* fix up netmask for dynaddr */ if (af == AF_INET && h->addr.type == PF_ADDR_DYNIFTL && unmask(m, AF_INET6) > 32) set_ipmask(n, 32); /* netmasks > 32 bit are invalid on v4 */ if (af == AF_INET && (m->addr32[1] || m->addr32[2] || m->addr32[3])) { fprintf(stderr, "netmask %u invalid for IPv4 address\n", unmask(m, AF_INET6)); return (1); } } return (0); } /* interface lookup routines */ struct node_host *iftab; void ifa_load(void) { struct ifaddrs *ifap, *ifa; struct node_host *n = NULL, *h = NULL; if (getifaddrs(&ifap) < 0) err(1, "getifaddrs"); for (ifa = ifap; ifa; ifa = ifa->ifa_next) { if (!(ifa->ifa_addr->sa_family == AF_INET || ifa->ifa_addr->sa_family == AF_INET6 || ifa->ifa_addr->sa_family == AF_LINK)) continue; n = calloc(1, sizeof(struct node_host)); if (n == NULL) err(1, "address: calloc"); n->af = ifa->ifa_addr->sa_family; n->ifa_flags = ifa->ifa_flags; #ifdef __KAME__ if (n->af == AF_INET6 && IN6_IS_ADDR_LINKLOCAL(&((struct sockaddr_in6 *) ifa->ifa_addr)->sin6_addr) && ((struct sockaddr_in6 *)ifa->ifa_addr)->sin6_scope_id == 0) { struct sockaddr_in6 *sin6; sin6 = (struct sockaddr_in6 *)ifa->ifa_addr; sin6->sin6_scope_id = sin6->sin6_addr.s6_addr[2] << 8 | sin6->sin6_addr.s6_addr[3]; sin6->sin6_addr.s6_addr[2] = 0; sin6->sin6_addr.s6_addr[3] = 0; } #endif n->ifindex = 0; if (n->af == AF_INET) { memcpy(&n->addr.v.a.addr, &((struct sockaddr_in *) ifa->ifa_addr)->sin_addr.s_addr, sizeof(struct in_addr)); memcpy(&n->addr.v.a.mask, &((struct sockaddr_in *) ifa->ifa_netmask)->sin_addr.s_addr, sizeof(struct in_addr)); if (ifa->ifa_broadaddr != NULL) memcpy(&n->bcast, &((struct sockaddr_in *) ifa->ifa_broadaddr)->sin_addr.s_addr, sizeof(struct in_addr)); if (ifa->ifa_dstaddr != NULL) memcpy(&n->peer, &((struct sockaddr_in *) ifa->ifa_dstaddr)->sin_addr.s_addr, sizeof(struct in_addr)); } else if (n->af == AF_INET6) { memcpy(&n->addr.v.a.addr, &((struct sockaddr_in6 *) ifa->ifa_addr)->sin6_addr.s6_addr, sizeof(struct in6_addr)); memcpy(&n->addr.v.a.mask, &((struct sockaddr_in6 *) ifa->ifa_netmask)->sin6_addr.s6_addr, sizeof(struct in6_addr)); if (ifa->ifa_broadaddr != NULL) memcpy(&n->bcast, &((struct sockaddr_in6 *) ifa->ifa_broadaddr)->sin6_addr.s6_addr, sizeof(struct in6_addr)); if (ifa->ifa_dstaddr != NULL) memcpy(&n->peer, &((struct sockaddr_in6 *) ifa->ifa_dstaddr)->sin6_addr.s6_addr, sizeof(struct in6_addr)); n->ifindex = ((struct sockaddr_in6 *) ifa->ifa_addr)->sin6_scope_id; } if ((n->ifname = strdup(ifa->ifa_name)) == NULL) err(1, "ifa_load: strdup"); n->next = NULL; n->tail = n; if (h == NULL) h = n; else { h->tail->next = n; h->tail = n; } } iftab = h; freeifaddrs(ifap); } int get_socket_domain(void) { int sdom; sdom = AF_UNSPEC; #ifdef WITH_INET6 if (sdom == AF_UNSPEC && feature_present("inet6")) sdom = AF_INET6; #endif #ifdef WITH_INET if (sdom == AF_UNSPEC && feature_present("inet")) sdom = AF_INET; #endif if (sdom == AF_UNSPEC) sdom = AF_LINK; return (sdom); } struct node_host * ifa_exists(const char *ifa_name) { struct node_host *n; struct ifgroupreq ifgr; int s; if (iftab == NULL) ifa_load(); /* check wether this is a group */ if ((s = socket(get_socket_domain(), SOCK_DGRAM, 0)) == -1) err(1, "socket"); bzero(&ifgr, sizeof(ifgr)); strlcpy(ifgr.ifgr_name, ifa_name, sizeof(ifgr.ifgr_name)); if (ioctl(s, SIOCGIFGMEMB, (caddr_t)&ifgr) == 0) { /* fake a node_host */ if ((n = calloc(1, sizeof(*n))) == NULL) err(1, "calloc"); if ((n->ifname = strdup(ifa_name)) == NULL) err(1, "strdup"); close(s); return (n); } close(s); for (n = iftab; n; n = n->next) { if (n->af == AF_LINK && !strncmp(n->ifname, ifa_name, IFNAMSIZ)) return (n); } return (NULL); } struct node_host * ifa_grouplookup(const char *ifa_name, int flags) { struct ifg_req *ifg; struct ifgroupreq ifgr; int s, len; struct node_host *n, *h = NULL; if ((s = socket(get_socket_domain(), SOCK_DGRAM, 0)) == -1) err(1, "socket"); bzero(&ifgr, sizeof(ifgr)); strlcpy(ifgr.ifgr_name, ifa_name, sizeof(ifgr.ifgr_name)); if (ioctl(s, SIOCGIFGMEMB, (caddr_t)&ifgr) == -1) { close(s); return (NULL); } len = ifgr.ifgr_len; if ((ifgr.ifgr_groups = calloc(1, len)) == NULL) err(1, "calloc"); if (ioctl(s, SIOCGIFGMEMB, (caddr_t)&ifgr) == -1) err(1, "SIOCGIFGMEMB"); for (ifg = ifgr.ifgr_groups; ifg && len >= sizeof(struct ifg_req); ifg++) { len -= sizeof(struct ifg_req); if ((n = ifa_lookup(ifg->ifgrq_member, flags)) == NULL) continue; if (h == NULL) h = n; else { h->tail->next = n; h->tail = n->tail; } } free(ifgr.ifgr_groups); close(s); return (h); } struct node_host * ifa_lookup(const char *ifa_name, int flags) { struct node_host *p = NULL, *h = NULL, *n = NULL; int got4 = 0, got6 = 0; const char *last_if = NULL; if ((h = ifa_grouplookup(ifa_name, flags)) != NULL) return (h); if (!strncmp(ifa_name, "self", IFNAMSIZ)) ifa_name = NULL; if (iftab == NULL) ifa_load(); for (p = iftab; p; p = p->next) { if (ifa_skip_if(ifa_name, p)) continue; if ((flags & PFI_AFLAG_BROADCAST) && p->af != AF_INET) continue; if ((flags & PFI_AFLAG_BROADCAST) && !(p->ifa_flags & IFF_BROADCAST)) continue; if ((flags & PFI_AFLAG_PEER) && !(p->ifa_flags & IFF_POINTOPOINT)) continue; if ((flags & PFI_AFLAG_NETWORK) && p->ifindex > 0) continue; if (last_if == NULL || strcmp(last_if, p->ifname)) got4 = got6 = 0; last_if = p->ifname; if ((flags & PFI_AFLAG_NOALIAS) && p->af == AF_INET && got4) continue; if ((flags & PFI_AFLAG_NOALIAS) && p->af == AF_INET6 && got6) continue; if (p->af == AF_INET) got4 = 1; else got6 = 1; n = calloc(1, sizeof(struct node_host)); if (n == NULL) err(1, "address: calloc"); n->af = p->af; if (flags & PFI_AFLAG_BROADCAST) memcpy(&n->addr.v.a.addr, &p->bcast, sizeof(struct pf_addr)); else if (flags & PFI_AFLAG_PEER) memcpy(&n->addr.v.a.addr, &p->peer, sizeof(struct pf_addr)); else memcpy(&n->addr.v.a.addr, &p->addr.v.a.addr, sizeof(struct pf_addr)); if (flags & PFI_AFLAG_NETWORK) set_ipmask(n, unmask(&p->addr.v.a.mask, n->af)); else { if (n->af == AF_INET) { if (p->ifa_flags & IFF_LOOPBACK && p->ifa_flags & IFF_LINK1) memcpy(&n->addr.v.a.mask, &p->addr.v.a.mask, sizeof(struct pf_addr)); else set_ipmask(n, 32); } else set_ipmask(n, 128); } n->ifindex = p->ifindex; n->next = NULL; n->tail = n; if (h == NULL) h = n; else { h->tail->next = n; h->tail = n; } } return (h); } int ifa_skip_if(const char *filter, struct node_host *p) { int n; if (p->af != AF_INET && p->af != AF_INET6) return (1); if (filter == NULL || !*filter) return (0); if (!strcmp(p->ifname, filter)) return (0); /* exact match */ n = strlen(filter); if (n < 1 || n >= IFNAMSIZ) return (1); /* sanity check */ if (filter[n-1] >= '0' && filter[n-1] <= '9') return (1); /* only do exact match in that case */ if (strncmp(p->ifname, filter, n)) return (1); /* prefix doesn't match */ return (p->ifname[n] < '0' || p->ifname[n] > '9'); } struct node_host * host(const char *s) { struct node_host *h = NULL; int mask, v4mask, v6mask, cont = 1; char *p, *q, *ps; if ((p = strrchr(s, '/')) != NULL) { mask = strtol(p+1, &q, 0); if (!q || *q || mask > 128 || q == (p+1)) { fprintf(stderr, "invalid netmask '%s'\n", p); return (NULL); } if ((ps = malloc(strlen(s) - strlen(p) + 1)) == NULL) err(1, "host: malloc"); strlcpy(ps, s, strlen(s) - strlen(p) + 1); v4mask = v6mask = mask; } else { if ((ps = strdup(s)) == NULL) err(1, "host: strdup"); v4mask = 32; v6mask = 128; mask = -1; } /* interface with this name exists? */ if (cont && (h = host_if(ps, mask)) != NULL) cont = 0; /* IPv4 address? */ if (cont && (h = host_v4(s, mask)) != NULL) cont = 0; /* IPv6 address? */ if (cont && (h = host_v6(ps, v6mask)) != NULL) cont = 0; /* dns lookup */ if (cont && (h = host_dns(ps, v4mask, v6mask)) != NULL) cont = 0; free(ps); if (h == NULL || cont == 1) { fprintf(stderr, "no IP address found for %s\n", s); return (NULL); } return (h); } struct node_host * host_if(const char *s, int mask) { struct node_host *n, *h = NULL; char *p, *ps; int flags = 0; if ((ps = strdup(s)) == NULL) err(1, "host_if: strdup"); while ((p = strrchr(ps, ':')) != NULL) { if (!strcmp(p+1, "network")) flags |= PFI_AFLAG_NETWORK; else if (!strcmp(p+1, "broadcast")) flags |= PFI_AFLAG_BROADCAST; else if (!strcmp(p+1, "peer")) flags |= PFI_AFLAG_PEER; else if (!strcmp(p+1, "0")) flags |= PFI_AFLAG_NOALIAS; else { free(ps); return (NULL); } *p = '\0'; } if (flags & (flags - 1) & PFI_AFLAG_MODEMASK) { /* Yep! */ fprintf(stderr, "illegal combination of interface modifiers\n"); free(ps); return (NULL); } if ((flags & (PFI_AFLAG_NETWORK|PFI_AFLAG_BROADCAST)) && mask > -1) { fprintf(stderr, "network or broadcast lookup, but " "extra netmask given\n"); free(ps); return (NULL); } if (ifa_exists(ps) || !strncmp(ps, "self", IFNAMSIZ)) { /* interface with this name exists */ h = ifa_lookup(ps, flags); for (n = h; n != NULL && mask > -1; n = n->next) set_ipmask(n, mask); } free(ps); return (h); } struct node_host * host_v4(const char *s, int mask) { struct node_host *h = NULL; struct in_addr ina; int bits = 32; memset(&ina, 0, sizeof(struct in_addr)); if (strrchr(s, '/') != NULL) { if ((bits = inet_net_pton(AF_INET, s, &ina, sizeof(ina))) == -1) return (NULL); } else { if (inet_pton(AF_INET, s, &ina) != 1) return (NULL); } h = calloc(1, sizeof(struct node_host)); if (h == NULL) err(1, "address: calloc"); h->ifname = NULL; h->af = AF_INET; h->addr.v.a.addr.addr32[0] = ina.s_addr; set_ipmask(h, bits); h->next = NULL; h->tail = h; return (h); } struct node_host * host_v6(const char *s, int mask) { struct addrinfo hints, *res; struct node_host *h = NULL; memset(&hints, 0, sizeof(hints)); hints.ai_family = AF_INET6; hints.ai_socktype = SOCK_DGRAM; /*dummy*/ hints.ai_flags = AI_NUMERICHOST; if (getaddrinfo(s, "0", &hints, &res) == 0) { h = calloc(1, sizeof(struct node_host)); if (h == NULL) err(1, "address: calloc"); h->ifname = NULL; h->af = AF_INET6; memcpy(&h->addr.v.a.addr, &((struct sockaddr_in6 *)res->ai_addr)->sin6_addr, sizeof(h->addr.v.a.addr)); h->ifindex = ((struct sockaddr_in6 *)res->ai_addr)->sin6_scope_id; set_ipmask(h, mask); freeaddrinfo(res); h->next = NULL; h->tail = h; } return (h); } struct node_host * host_dns(const char *s, int v4mask, int v6mask) { struct addrinfo hints, *res0, *res; struct node_host *n, *h = NULL; int error, noalias = 0; int got4 = 0, got6 = 0; char *p, *ps; if ((ps = strdup(s)) == NULL) err(1, "host_dns: strdup"); if ((p = strrchr(ps, ':')) != NULL && !strcmp(p, ":0")) { noalias = 1; *p = '\0'; } memset(&hints, 0, sizeof(hints)); hints.ai_family = PF_UNSPEC; hints.ai_socktype = SOCK_STREAM; /* DUMMY */ error = getaddrinfo(ps, NULL, &hints, &res0); if (error) { free(ps); return (h); } for (res = res0; res; res = res->ai_next) { if (res->ai_family != AF_INET && res->ai_family != AF_INET6) continue; if (noalias) { if (res->ai_family == AF_INET) { if (got4) continue; got4 = 1; } else { if (got6) continue; got6 = 1; } } n = calloc(1, sizeof(struct node_host)); if (n == NULL) err(1, "host_dns: calloc"); n->ifname = NULL; n->af = res->ai_family; if (res->ai_family == AF_INET) { memcpy(&n->addr.v.a.addr, &((struct sockaddr_in *) res->ai_addr)->sin_addr.s_addr, sizeof(struct in_addr)); set_ipmask(n, v4mask); } else { memcpy(&n->addr.v.a.addr, &((struct sockaddr_in6 *) res->ai_addr)->sin6_addr.s6_addr, sizeof(struct in6_addr)); n->ifindex = ((struct sockaddr_in6 *) res->ai_addr)->sin6_scope_id; set_ipmask(n, v6mask); } n->next = NULL; n->tail = n; if (h == NULL) h = n; else { h->tail->next = n; h->tail = n; } } freeaddrinfo(res0); free(ps); return (h); } /* * convert a hostname to a list of addresses and put them in the given buffer. * test: * if set to 1, only simple addresses are accepted (no netblock, no "!"). */ int append_addr(struct pfr_buffer *b, char *s, int test) { char *r; struct node_host *h, *n; int rv, not = 0; for (r = s; *r == '!'; r++) not = !not; if ((n = host(r)) == NULL) { errno = 0; return (-1); } rv = append_addr_host(b, n, test, not); do { h = n; n = n->next; free(h); } while (n != NULL); return (rv); } /* * same as previous function, but with a pre-parsed input and the ability * to "negate" the result. Does not free the node_host list. * not: * setting it to 1 is equivalent to adding "!" in front of parameter s. */ int append_addr_host(struct pfr_buffer *b, struct node_host *n, int test, int not) { int bits; struct pfr_addr addr; do { bzero(&addr, sizeof(addr)); addr.pfra_not = n->not ^ not; addr.pfra_af = n->af; addr.pfra_net = unmask(&n->addr.v.a.mask, n->af); switch (n->af) { case AF_INET: addr.pfra_ip4addr.s_addr = n->addr.v.a.addr.addr32[0]; bits = 32; break; case AF_INET6: memcpy(&addr.pfra_ip6addr, &n->addr.v.a.addr.v6, sizeof(struct in6_addr)); bits = 128; break; default: errno = EINVAL; return (-1); } if ((test && (not || addr.pfra_net != bits)) || addr.pfra_net > bits) { errno = EINVAL; return (-1); } if (pfr_buf_add(b, &addr)) return (-1); } while ((n = n->next) != NULL); return (0); } int pfctl_add_trans(struct pfr_buffer *buf, int rs_num, const char *anchor) { struct pfioc_trans_e trans; bzero(&trans, sizeof(trans)); trans.rs_num = rs_num; if (strlcpy(trans.anchor, anchor, sizeof(trans.anchor)) >= sizeof(trans.anchor)) errx(1, "pfctl_add_trans: strlcpy"); return pfr_buf_add(buf, &trans); } u_int32_t pfctl_get_ticket(struct pfr_buffer *buf, int rs_num, const char *anchor) { struct pfioc_trans_e *p; PFRB_FOREACH(p, buf) if (rs_num == p->rs_num && !strcmp(anchor, p->anchor)) return (p->ticket); errx(1, "pfctl_get_ticket: assertion failed"); } int pfctl_trans(int dev, struct pfr_buffer *buf, u_long cmd, int from) { struct pfioc_trans trans; bzero(&trans, sizeof(trans)); trans.size = buf->pfrb_size - from; trans.esize = sizeof(struct pfioc_trans_e); trans.array = ((struct pfioc_trans_e *)buf->pfrb_caddr) + from; return ioctl(dev, cmd, &trans); } Index: head/share/man/man5/pf.conf.5 =================================================================== --- head/share/man/man5/pf.conf.5 (revision 287221) +++ head/share/man/man5/pf.conf.5 (revision 287222) @@ -1,3061 +1,3033 @@ .\" $FreeBSD$ .\" $OpenBSD: pf.conf.5,v 1.406 2009/01/31 19:37:12 sobrado Exp $ .\" .\" Copyright (c) 2002, Daniel Hartmeier .\" All rights reserved. .\" .\" Redistribution and use in source and binary forms, with or without .\" modification, are permitted provided that the following conditions .\" are met: .\" .\" - Redistributions of source code must retain the above copyright .\" notice, this list of conditions and the following disclaimer. .\" - Redistributions in binary form must reproduce the above .\" copyright notice, this list of conditions and the following .\" disclaimer in the documentation and/or other materials provided .\" with the distribution. .\" .\" THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS .\" "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT .\" LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS .\" FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE .\" COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, .\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, .\" BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; .\" LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER .\" CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN .\" ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE .\" POSSIBILITY OF SUCH DAMAGE. .\" .Dd July 25, 2015 .Dt PF.CONF 5 .Os .Sh NAME .Nm pf.conf .Nd packet filter configuration file .Sh DESCRIPTION The .Xr pf 4 packet filter modifies, drops or passes packets according to rules or definitions specified in .Nm pf.conf . .Sh STATEMENT ORDER There are seven types of statements in .Nm pf.conf : .Bl -tag -width xxxx .It Cm Macros User-defined variables may be defined and used later, simplifying the configuration file. Macros must be defined before they are referenced in .Nm pf.conf . .It Cm Tables Tables provide a mechanism for increasing the performance and flexibility of rules with large numbers of source or destination addresses. .It Cm Options Options tune the behaviour of the packet filtering engine. .It Cm Traffic Normalization Li (e.g. Em scrub ) Traffic normalization protects internal machines against inconsistencies in Internet protocols and implementations. .It Cm Queueing Queueing provides rule-based bandwidth control. .It Cm Translation Li (Various forms of NAT) Translation rules specify how addresses are to be mapped or redirected to other addresses. .It Cm Packet Filtering Packet filtering provides rule-based blocking or passing of packets. .El .Pp With the exception of .Cm macros and .Cm tables , the types of statements should be grouped and appear in .Nm pf.conf in the order shown above, as this matches the operation of the underlying packet filtering engine. By default .Xr pfctl 8 enforces this order (see .Ar set require-order below). .Pp Comments can be put anywhere in the file using a hash mark .Pq Sq # , and extend to the end of the current line. .Pp Additional configuration files can be included with the .Ic include keyword, for example: .Bd -literal -offset indent include "/etc/pf/sub.filter.conf" .Ed .Sh MACROS Macros can be defined that will later be expanded in context. Macro names must start with a letter, and may contain letters, digits and underscores. Macro names may not be reserved words (for example .Ar pass , .Ar in , .Ar out ) . Macros are not expanded inside quotes. .Pp For example, .Bd -literal -offset indent ext_if = \&"kue0\&" all_ifs = \&"{\&" $ext_if lo0 \&"}\&" pass out on $ext_if from any to any pass in on $ext_if proto tcp from any to any port 25 .Ed .Sh TABLES Tables are named structures which can hold a collection of addresses and networks. Lookups against tables in .Xr pf 4 are relatively fast, making a single rule with tables much more efficient, in terms of processor usage and memory consumption, than a large number of rules which differ only in IP address (either created explicitly or automatically by rule expansion). .Pp Tables can be used as the source or destination of filter rules, .Ar scrub rules or translation rules such as .Ar nat or .Ar rdr (see below for details on the various rule types). Tables can also be used for the redirect address of .Ar nat and .Ar rdr rules and in the routing options of filter rules, but only for .Ar round-robin pools. .Pp Tables can be defined with any of the following .Xr pfctl 8 mechanisms. As with macros, reserved words may not be used as table names. .Bl -tag -width "manually" .It Ar manually Persistent tables can be manually created with the .Ar add or .Ar replace option of .Xr pfctl 8 , before or after the ruleset has been loaded. .It Pa pf.conf Table definitions can be placed directly in this file, and loaded at the same time as other rules are loaded, atomically. Table definitions inside .Nm pf.conf use the .Ar table statement, and are especially useful to define non-persistent tables. The contents of a pre-existing table defined without a list of addresses to initialize it is not altered when .Nm pf.conf is loaded. A table initialized with the empty list, .Li { } , will be cleared on load. .El .Pp Tables may be defined with the following attributes: .Bl -tag -width persist .It Ar persist The .Ar persist flag forces the kernel to keep the table even when no rules refer to it. If the flag is not set, the kernel will automatically remove the table when the last rule referring to it is flushed. .It Ar const The .Ar const flag prevents the user from altering the contents of the table once it has been created. Without that flag, .Xr pfctl 8 can be used to add or remove addresses from the table at any time, even when running with .Xr securelevel 7 = 2. .It Ar counters The .Ar counters flag enables per-address packet and byte counters which can be displayed with .Xr pfctl 8 . .El .Pp For example, .Bd -literal -offset indent table \*(Ltprivate\*(Gt const { 10/8, 172.16/12, 192.168/16 } table \*(Ltbadhosts\*(Gt persist block on fxp0 from { \*(Ltprivate\*(Gt, \*(Ltbadhosts\*(Gt } to any .Ed .Pp creates a table called private, to hold RFC 1918 private network blocks, and a table called badhosts, which is initially empty. A filter rule is set up to block all traffic coming from addresses listed in either table. The private table cannot have its contents changed and the badhosts table will exist even when no active filter rules reference it. Addresses may later be added to the badhosts table, so that traffic from these hosts can be blocked by using .Bd -literal -offset indent # pfctl -t badhosts -Tadd 204.92.77.111 .Ed .Pp A table can also be initialized with an address list specified in one or more external files, using the following syntax: .Bd -literal -offset indent table \*(Ltspam\*(Gt persist file \&"/etc/spammers\&" file \&"/etc/openrelays\&" block on fxp0 from \*(Ltspam\*(Gt to any .Ed .Pp The files .Pa /etc/spammers and .Pa /etc/openrelays list IP addresses, one per line. Any lines beginning with a # are treated as comments and ignored. In addition to being specified by IP address, hosts may also be specified by their hostname. When the resolver is called to add a hostname to a table, .Em all resulting IPv4 and IPv6 addresses are placed into the table. IP addresses can also be entered in a table by specifying a valid interface name, a valid interface group or the .Em self keyword, in which case all addresses assigned to the interface(s) will be added to the table. .Sh OPTIONS .Xr pf 4 may be tuned for various situations using the .Ar set command. .Bl -tag -width xxxx .It Ar set timeout .Pp .Bl -tag -width "src.track" -compact .It Ar interval Interval between purging expired states and fragments. .It Ar frag Seconds before an unassembled fragment is expired. .It Ar src.track Length of time to retain a source tracking entry after the last state expires. .El .Pp When a packet matches a stateful connection, the seconds to live for the connection will be updated to that of the .Ar proto.modifier which corresponds to the connection state. Each packet which matches this state will reset the TTL. Tuning these values may improve the performance of the firewall at the risk of dropping valid idle connections. .Pp .Bl -tag -width xxxx -compact .It Ar tcp.first The state after the first packet. .It Ar tcp.opening The state before the destination host ever sends a packet. .It Ar tcp.established The fully established state. .It Ar tcp.closing The state after the first FIN has been sent. .It Ar tcp.finwait The state after both FINs have been exchanged and the connection is closed. Some hosts (notably web servers on Solaris) send TCP packets even after closing the connection. Increasing .Ar tcp.finwait (and possibly .Ar tcp.closing ) can prevent blocking of such packets. .It Ar tcp.closed The state after one endpoint sends an RST. .El .Pp ICMP and UDP are handled in a fashion similar to TCP, but with a much more limited set of states: .Pp .Bl -tag -width xxxx -compact .It Ar udp.first The state after the first packet. .It Ar udp.single The state if the source host sends more than one packet but the destination host has never sent one back. .It Ar udp.multiple The state if both hosts have sent packets. .It Ar icmp.first The state after the first packet. .It Ar icmp.error The state after an ICMP error came back in response to an ICMP packet. .El .Pp Other protocols are handled similarly to UDP: .Pp .Bl -tag -width xxxx -compact .It Ar other.first .It Ar other.single .It Ar other.multiple .El .Pp Timeout values can be reduced adaptively as the number of state table entries grows. .Pp .Bl -tag -width xxxx -compact .It Ar adaptive.start When the number of state entries exceeds this value, adaptive scaling begins. All timeout values are scaled linearly with factor (adaptive.end - number of states) / (adaptive.end - adaptive.start). .It Ar adaptive.end When reaching this number of state entries, all timeout values become zero, effectively purging all state entries immediately. This value is used to define the scale factor, it should not actually be reached (set a lower state limit, see below). .El .Pp Adaptive timeouts are enabled by default, with an adaptive.start value equal to 60% of the state limit, and an adaptive.end value equal to 120% of the state limit. They can be disabled by setting both adaptive.start and adaptive.end to 0. .Pp The adaptive timeout values can be defined both globally and for each rule. When used on a per-rule basis, the values relate to the number of states created by the rule, otherwise to the total number of states. .Pp For example: .Bd -literal -offset indent set timeout tcp.first 120 set timeout tcp.established 86400 set timeout { adaptive.start 6000, adaptive.end 12000 } set limit states 10000 .Ed .Pp With 9000 state table entries, the timeout values are scaled to 50% (tcp.first 60, tcp.established 43200). .It Ar set loginterface Enable collection of packet and byte count statistics for the given interface or interface group. These statistics can be viewed using .Bd -literal -offset indent # pfctl -s info .Ed .Pp In this example .Xr pf 4 collects statistics on the interface named dc0: .Bd -literal -offset indent set loginterface dc0 .Ed .Pp One can disable the loginterface using: .Bd -literal -offset indent set loginterface none .Ed .It Ar set limit Sets hard limits on the memory pools used by the packet filter. See .Xr zone 9 for an explanation of memory pools. .Pp For example, .Bd -literal -offset indent set limit states 20000 .Ed .Pp sets the maximum number of entries in the memory pool used by state table entries (generated by .Ar pass rules which do not specify .Ar no state ) to 20000. Using .Bd -literal -offset indent set limit frags 20000 .Ed .Pp sets the maximum number of entries in the memory pool used for fragment reassembly (generated by .Ar scrub rules) to 20000. Using .Bd -literal -offset indent set limit src-nodes 2000 .Ed .Pp sets the maximum number of entries in the memory pool used for tracking source IP addresses (generated by the .Ar sticky-address and .Ar src.track options) to 2000. Using .Bd -literal -offset indent set limit tables 1000 set limit table-entries 100000 .Ed .Pp sets limits on the memory pools used by tables. The first limits the number of tables that can exist to 1000. The second limits the overall number of addresses that can be stored in tables to 100000. .Pp Various limits can be combined on a single line: .Bd -literal -offset indent set limit { states 20000, frags 20000, src-nodes 2000 } .Ed .It Ar set ruleset-optimization .Bl -tag -width xxxxxxxx -compact .It Ar none Disable the ruleset optimizer. .It Ar basic Enable basic ruleset optimization. This is the default behaviour. Basic ruleset optimization does four things to improve the performance of ruleset evaluations: .Pp .Bl -enum -compact .It remove duplicate rules .It remove rules that are a subset of another rule .It combine multiple rules into a table when advantageous .It re-order the rules to improve evaluation performance .El .Pp .It Ar profile Uses the currently loaded ruleset as a feedback profile to tailor the ordering of quick rules to actual network traffic. .El .Pp It is important to note that the ruleset optimizer will modify the ruleset to improve performance. A side effect of the ruleset modification is that per-rule accounting statistics will have different meanings than before. If per-rule accounting is important for billing purposes or whatnot, either the ruleset optimizer should not be used or a label field should be added to all of the accounting rules to act as optimization barriers. .Pp Optimization can also be set as a command-line argument to .Xr pfctl 8 , overriding the settings in .Nm . .It Ar set optimization Optimize state timeouts for one of the following network environments: .Pp .Bl -tag -width xxxx -compact .It Ar normal A normal network environment. Suitable for almost all networks. .It Ar high-latency A high-latency environment (such as a satellite connection). .It Ar satellite Alias for .Ar high-latency . .It Ar aggressive Aggressively expire connections. This can greatly reduce the memory usage of the firewall at the cost of dropping idle connections early. .It Ar conservative Extremely conservative settings. Avoid dropping legitimate connections at the expense of greater memory utilization (possibly much greater on a busy network) and slightly increased processor utilization. .El .Pp For example: .Bd -literal -offset indent set optimization aggressive .Ed .It Ar set block-policy The .Ar block-policy option sets the default behaviour for the packet .Ar block action: .Pp .Bl -tag -width xxxxxxxx -compact .It Ar drop Packet is silently dropped. .It Ar return A TCP RST is returned for blocked TCP packets, an ICMP UNREACHABLE is returned for blocked UDP packets, and all other packets are silently dropped. .El .Pp For example: .Bd -literal -offset indent set block-policy return .Ed .It Ar set state-policy The .Ar state-policy option sets the default behaviour for states: .Pp .Bl -tag -width group-bound -compact .It Ar if-bound States are bound to interface. .It Ar floating States can match packets on any interfaces (the default). .El .Pp For example: .Bd -literal -offset indent set state-policy if-bound .Ed .It Ar set state-defaults The .Ar state-defaults option sets the state options for states created from rules without an explicit .Ar keep state . For example: .Bd -literal -offset indent set state-defaults pflow, no-sync .Ed .It Ar set hostid The 32-bit .Ar hostid identifies this firewall's state table entries to other firewalls in a .Xr pfsync 4 failover cluster. By default the hostid is set to a pseudo-random value, however it may be desirable to manually configure it, for example to more easily identify the source of state table entries. .Bd -literal -offset indent set hostid 1 .Ed .Pp The hostid may be specified in either decimal or hexadecimal. .It Ar set require-order By default .Xr pfctl 8 enforces an ordering of the statement types in the ruleset to: .Em options , .Em normalization , .Em queueing , .Em translation , .Em filtering . Setting this option to .Ar no disables this enforcement. There may be non-trivial and non-obvious implications to an out of order ruleset. Consider carefully before disabling the order enforcement. .It Ar set fingerprints Load fingerprints of known operating systems from the given filename. By default fingerprints of known operating systems are automatically loaded from .Xr pf.os 5 in .Pa /etc but can be overridden via this option. Setting this option may leave a small period of time where the fingerprints referenced by the currently active ruleset are inconsistent until the new ruleset finishes loading. .Pp For example: .Pp .Dl set fingerprints \&"/etc/pf.os.devel\&" .It Ar set skip on Aq Ar ifspec List interfaces for which packets should not be filtered. Packets passing in or out on such interfaces are passed as if pf was disabled, i.e. pf does not process them in any way. This can be useful on loopback and other virtual interfaces, when packet filtering is not desired and can have unexpected effects. For example: .Pp .Dl set skip on lo0 .It Ar set debug Set the debug .Ar level to one of the following: .Pp .Bl -tag -width xxxxxxxxxxxx -compact .It Ar none Don't generate debug messages. .It Ar urgent Generate debug messages only for serious errors. .It Ar misc Generate debug messages for various errors. .It Ar loud Generate debug messages for common conditions. .El .El .Sh TRAFFIC NORMALIZATION Traffic normalization is used to sanitize packet content in such a way that there are no ambiguities in packet interpretation on the receiving side. The normalizer does IP fragment reassembly to prevent attacks that confuse intrusion detection systems by sending overlapping IP fragments. Packet normalization is invoked with the .Ar scrub directive. .Pp .Ar scrub has the following options: .Bl -tag -width xxxx .It Ar no-df Clears the .Ar dont-fragment bit from a matching IP packet. Some operating systems are known to generate fragmented packets with the .Ar dont-fragment bit set. This is particularly true with NFS. .Ar Scrub will drop such fragmented .Ar dont-fragment packets unless .Ar no-df is specified. .Pp Unfortunately some operating systems also generate their .Ar dont-fragment packets with a zero IP identification field. Clearing the .Ar dont-fragment bit on packets with a zero IP ID may cause deleterious results if an upstream router later fragments the packet. Using the .Ar random-id modifier (see below) is recommended in combination with the .Ar no-df modifier to ensure unique IP identifiers. .It Ar min-ttl Aq Ar number Enforces a minimum TTL for matching IP packets. .It Ar max-mss Aq Ar number Enforces a maximum MSS for matching TCP packets. .It Xo Ar set-tos Aq Ar string .No \*(Ba Aq Ar number .Xc Enforces a .Em TOS for matching IP packets. .Em TOS may be given as one of .Ar lowdelay , .Ar throughput , .Ar reliability , or as either hex or decimal. .It Ar random-id Replaces the IP identification field with random values to compensate for predictable values generated by many hosts. This option only applies to packets that are not fragmented after the optional fragment reassembly. .It Ar fragment reassemble Using .Ar scrub rules, fragments can be reassembled by normalization. In this case, fragments are buffered until they form a complete packet, and only the completed packet is passed on to the filter. The advantage is that filter rules have to deal only with complete packets, and can ignore fragments. The drawback of caching fragments is the additional memory cost. -But the full reassembly method is the only method that currently works -with NAT. -This is the default behavior of a -.Ar scrub -rule if no fragmentation modifier is supplied. -.It Ar fragment crop -The default fragment reassembly method is expensive, hence the option -to crop is provided. -In this case, -.Xr pf 4 -will track the fragments and cache a small range descriptor. -Duplicate fragments are dropped and overlaps are cropped. -Thus data will only occur once on the wire with ambiguities resolving to -the first occurrence. -Unlike the -.Ar fragment reassemble -modifier, fragments are not buffered, they are passed as soon as they -are received. -The -.Ar fragment crop -reassembly mechanism does not yet work with NAT. -.It Ar fragment drop-ovl -This option is similar to the -.Ar fragment crop -modifier except that all overlapping or duplicate fragments will be -dropped, and all further corresponding fragments will be -dropped as well. .It Ar reassemble tcp Statefully normalizes TCP connections. .Ar scrub reassemble tcp rules may not have the direction (in/out) specified. .Ar reassemble tcp performs the following normalizations: .Pp .Bl -tag -width timeout -compact .It ttl Neither side of the connection is allowed to reduce their IP TTL. An attacker may send a packet such that it reaches the firewall, affects the firewall state, and expires before reaching the destination host. .Ar reassemble tcp will raise the TTL of all packets back up to the highest value seen on the connection. .It timestamp modulation Modern TCP stacks will send a timestamp on every TCP packet and echo the other endpoint's timestamp back to them. Many operating systems will merely start the timestamp at zero when first booted, and increment it several times a second. The uptime of the host can be deduced by reading the timestamp and multiplying by a constant. Also observing several different timestamps can be used to count hosts behind a NAT device. And spoofing TCP packets into a connection requires knowing or guessing valid timestamps. Timestamps merely need to be monotonically increasing and not derived off a guessable base time. .Ar reassemble tcp will cause .Ar scrub to modulate the TCP timestamps with a random number. .It extended PAWS checks There is a problem with TCP on long fat pipes, in that a packet might get delayed for longer than it takes the connection to wrap its 32-bit sequence space. In such an occurrence, the old packet would be indistinguishable from a new packet and would be accepted as such. The solution to this is called PAWS: Protection Against Wrapped Sequence numbers. It protects against it by making sure the timestamp on each packet does not go backwards. .Ar reassemble tcp also makes sure the timestamp on the packet does not go forward more than the RFC allows. By doing this, .Xr pf 4 artificially extends the security of TCP sequence numbers by 10 to 18 bits when the host uses appropriately randomized timestamps, since a blind attacker would have to guess the timestamp as well. .El .El .Pp For example, .Bd -literal -offset indent scrub in on $ext_if all fragment reassemble .Ed .Pp The .Ar no option prefixed to a scrub rule causes matching packets to remain unscrubbed, much in the same way as .Ar drop quick works in the packet filter (see below). This mechanism should be used when it is necessary to exclude specific packets from broader scrub rules. .Sh QUEUEING The ALTQ system is currently not available in the GENERIC kernel nor as loadable modules. In order to use the herein after called queueing options one has to use a custom built kernel. Please refer to .Xr altq 4 to learn about the related kernel options. .Pp Packets can be assigned to queues for the purpose of bandwidth control. At least two declarations are required to configure queues, and later any packet filtering rule can reference the defined queues by name. During the filtering component of .Nm pf.conf , the last referenced .Ar queue name is where any packets from .Ar pass rules will be queued, while for .Ar block rules it specifies where any resulting ICMP or TCP RST packets should be queued. The .Ar scheduler defines the algorithm used to decide which packets get delayed, dropped, or sent out immediately. There are three .Ar schedulers currently supported. .Bl -tag -width xxxx .It Ar cbq Class Based Queueing. .Ar Queues attached to an interface build a tree, thus each .Ar queue can have further child .Ar queues . Each queue can have a .Ar priority and a .Ar bandwidth assigned. .Ar Priority mainly controls the time packets take to get sent out, while .Ar bandwidth has primarily effects on throughput. .Ar cbq achieves both partitioning and sharing of link bandwidth by hierarchically structured classes. Each class has its own .Ar queue and is assigned its share of .Ar bandwidth . A child class can borrow bandwidth from its parent class as long as excess bandwidth is available (see the option .Ar borrow , below). .It Ar priq Priority Queueing. .Ar Queues are flat attached to the interface, thus, .Ar queues cannot have further child .Ar queues . Each .Ar queue has a unique .Ar priority assigned, ranging from 0 to 15. Packets in the .Ar queue with the highest .Ar priority are processed first. .It Ar hfsc Hierarchical Fair Service Curve. .Ar Queues attached to an interface build a tree, thus each .Ar queue can have further child .Ar queues . Each queue can have a .Ar priority and a .Ar bandwidth assigned. .Ar Priority mainly controls the time packets take to get sent out, while .Ar bandwidth primarily affects throughput. .Ar hfsc supports both link-sharing and guaranteed real-time services. It employs a service curve based QoS model, and its unique feature is an ability to decouple .Ar delay and .Ar bandwidth allocation. .El .Pp The interfaces on which queueing should be activated are declared using the .Ar altq on declaration. .Ar altq on has the following keywords: .Bl -tag -width xxxx .It Aq Ar interface Queueing is enabled on the named interface. .It Aq Ar scheduler Specifies which queueing scheduler to use. Currently supported values are .Ar cbq for Class Based Queueing, .Ar priq for Priority Queueing and .Ar hfsc for the Hierarchical Fair Service Curve scheduler. .It Ar bandwidth Aq Ar bw The maximum bitrate for all queues on an interface may be specified using the .Ar bandwidth keyword. The value can be specified as an absolute value or as a percentage of the interface bandwidth. When using an absolute value, the suffixes .Ar b , .Ar Kb , .Ar Mb , and .Ar Gb are used to represent bits, kilobits, megabits, and gigabits per second, respectively. The value must not exceed the interface bandwidth. If .Ar bandwidth is not specified, the interface bandwidth is used (but take note that some interfaces do not know their bandwidth, or can adapt their bandwidth rates). .It Ar qlimit Aq Ar limit The maximum number of packets held in the queue. The default is 50. .It Ar tbrsize Aq Ar size Adjusts the size, in bytes, of the token bucket regulator. If not specified, heuristics based on the interface bandwidth are used to determine the size. .It Ar queue Aq Ar list Defines a list of subqueues to create on an interface. .El .Pp In the following example, the interface dc0 should queue up to 5Mbps in four second-level queues using Class Based Queueing. Those four queues will be shown in a later example. .Bd -literal -offset indent altq on dc0 cbq bandwidth 5Mb queue { std, http, mail, ssh } .Ed .Pp Once interfaces are activated for queueing using the .Ar altq directive, a sequence of .Ar queue directives may be defined. The name associated with a .Ar queue must match a queue defined in the .Ar altq directive (e.g. mail), or, except for the .Ar priq .Ar scheduler , in a parent .Ar queue declaration. The following keywords can be used: .Bl -tag -width xxxx .It Ar on Aq Ar interface Specifies the interface the queue operates on. If not given, it operates on all matching interfaces. .It Ar bandwidth Aq Ar bw Specifies the maximum bitrate to be processed by the queue. This value must not exceed the value of the parent .Ar queue and can be specified as an absolute value or a percentage of the parent queue's bandwidth. If not specified, defaults to 100% of the parent queue's bandwidth. The .Ar priq scheduler does not support bandwidth specification. .It Ar priority Aq Ar level Between queues a priority level can be set. For .Ar cbq and .Ar hfsc , the range is 0 to 7 and for .Ar priq , the range is 0 to 15. The default for all is 1. .Ar Priq queues with a higher priority are always served first. .Ar Cbq and .Ar Hfsc queues with a higher priority are preferred in the case of overload. .It Ar qlimit Aq Ar limit The maximum number of packets held in the queue. The default is 50. .El .Pp The .Ar scheduler can get additional parameters with .Xo Aq Ar scheduler .Pf ( Aq Ar parameters ) . .Xc Parameters are as follows: .Bl -tag -width Fl .It Ar default Packets not matched by another queue are assigned to this one. Exactly one default queue is required. .It Ar red Enable RED (Random Early Detection) on this queue. RED drops packets with a probability proportional to the average queue length. .It Ar rio Enables RIO on this queue. RIO is RED with IN/OUT, thus running RED two times more than RIO would achieve the same effect. RIO is currently not supported in the GENERIC kernel. .It Ar ecn Enables ECN (Explicit Congestion Notification) on this queue. ECN implies RED. .El .Pp The .Ar cbq .Ar scheduler supports an additional option: .Bl -tag -width Fl .It Ar borrow The queue can borrow bandwidth from the parent. .El .Pp The .Ar hfsc .Ar scheduler supports some additional options: .Bl -tag -width Fl .It Ar realtime Aq Ar sc The minimum required bandwidth for the queue. .It Ar upperlimit Aq Ar sc The maximum allowed bandwidth for the queue. .It Ar linkshare Aq Ar sc The bandwidth share of a backlogged queue. .El .Pp .Aq Ar sc is an acronym for .Ar service curve . .Pp The format for service curve specifications is .Ar ( m1 , d , m2 ) . .Ar m2 controls the bandwidth assigned to the queue. .Ar m1 and .Ar d are optional and can be used to control the initial bandwidth assignment. For the first .Ar d milliseconds the queue gets the bandwidth given as .Ar m1 , afterwards the value given in .Ar m2 . .Pp Furthermore, with .Ar cbq and .Ar hfsc , child queues can be specified as in an .Ar altq declaration, thus building a tree of queues using a part of their parent's bandwidth. .Pp Packets can be assigned to queues based on filter rules by using the .Ar queue keyword. Normally only one .Ar queue is specified; when a second one is specified it will instead be used for packets which have a .Em TOS of .Em lowdelay and for TCP ACKs with no data payload. .Pp To continue the previous example, the examples below would specify the four referenced queues, plus a few child queues. Interactive .Xr ssh 1 sessions get priority over bulk transfers like .Xr scp 1 and .Xr sftp 1 . The queues may then be referenced by filtering rules (see .Sx PACKET FILTERING below). .Bd -literal queue std bandwidth 10% cbq(default) queue http bandwidth 60% priority 2 cbq(borrow red) \e { employees, developers } queue developers bandwidth 75% cbq(borrow) queue employees bandwidth 15% queue mail bandwidth 10% priority 0 cbq(borrow ecn) queue ssh bandwidth 20% cbq(borrow) { ssh_interactive, ssh_bulk } queue ssh_interactive bandwidth 50% priority 7 cbq(borrow) queue ssh_bulk bandwidth 50% priority 0 cbq(borrow) block return out on dc0 inet all queue std pass out on dc0 inet proto tcp from $developerhosts to any port 80 \e queue developers pass out on dc0 inet proto tcp from $employeehosts to any port 80 \e queue employees pass out on dc0 inet proto tcp from any to any port 22 \e queue(ssh_bulk, ssh_interactive) pass out on dc0 inet proto tcp from any to any port 25 \e queue mail .Ed .Sh TRANSLATION Translation rules modify either the source or destination address of the packets associated with a stateful connection. A stateful connection is automatically created to track packets matching such a rule as long as they are not blocked by the filtering section of .Nm pf.conf . The translation engine modifies the specified address and/or port in the packet, recalculates IP, TCP and UDP checksums as necessary, and passes it to the packet filter for evaluation. .Pp Since translation occurs before filtering the filter engine will see packets as they look after any addresses and ports have been translated. Filter rules will therefore have to filter based on the translated address and port number. Packets that match a translation rule are only automatically passed if the .Ar pass modifier is given, otherwise they are still subject to .Ar block and .Ar pass rules. .Pp The state entry created permits .Xr pf 4 to keep track of the original address for traffic associated with that state and correctly direct return traffic for that connection. .Pp Various types of translation are possible with pf: .Bl -tag -width xxxx .It Ar binat A .Ar binat rule specifies a bidirectional mapping between an external IP netblock and an internal IP netblock. .It Ar nat A .Ar nat rule specifies that IP addresses are to be changed as the packet traverses the given interface. This technique allows one or more IP addresses on the translating host to support network traffic for a larger range of machines on an "inside" network. Although in theory any IP address can be used on the inside, it is strongly recommended that one of the address ranges defined by RFC 1918 be used. These netblocks are: .Bd -literal 10.0.0.0 - 10.255.255.255 (all of net 10, i.e., 10/8) 172.16.0.0 - 172.31.255.255 (i.e., 172.16/12) 192.168.0.0 - 192.168.255.255 (i.e., 192.168/16) .Ed .It Pa rdr The packet is redirected to another destination and possibly a different port. .Ar rdr rules can optionally specify port ranges instead of single ports. rdr ... port 2000:2999 -\*(Gt ... port 4000 redirects ports 2000 to 2999 (inclusive) to port 4000. rdr ... port 2000:2999 -\*(Gt ... port 4000:* redirects port 2000 to 4000, 2001 to 4001, ..., 2999 to 4999. .El .Pp In addition to modifying the address, some translation rules may modify source or destination ports for .Xr tcp 4 or .Xr udp 4 connections; implicitly in the case of .Ar nat rules and explicitly in the case of .Ar rdr rules. Port numbers are never translated with a .Ar binat rule. .Pp Evaluation order of the translation rules is dependent on the type of the translation rules and of the direction of a packet. .Ar binat rules are always evaluated first. Then either the .Ar rdr rules are evaluated on an inbound packet or the .Ar nat rules on an outbound packet. Rules of the same type are evaluated in the same order in which they appear in the ruleset. The first matching rule decides what action is taken. .Pp The .Ar no option prefixed to a translation rule causes packets to remain untranslated, much in the same way as .Ar drop quick works in the packet filter (see below). If no rule matches the packet it is passed to the filter engine unmodified. .Pp Translation rules apply only to packets that pass through the specified interface, and if no interface is specified, translation is applied to packets on all interfaces. For instance, redirecting port 80 on an external interface to an internal web server will only work for connections originating from the outside. Connections to the address of the external interface from local hosts will not be redirected, since such packets do not actually pass through the external interface. Redirections cannot reflect packets back through the interface they arrive on, they can only be redirected to hosts connected to different interfaces or to the firewall itself. .Pp Note that redirecting external incoming connections to the loopback address, as in .Bd -literal -offset indent rdr on ne3 inet proto tcp to port smtp -\*(Gt 127.0.0.1 port spamd .Ed .Pp will effectively allow an external host to connect to daemons bound solely to the loopback address, circumventing the traditional blocking of such connections on a real interface. Unless this effect is desired, any of the local non-loopback addresses should be used as redirection target instead, which allows external connections only to daemons bound to this address or not bound to any address. .Pp See .Sx TRANSLATION EXAMPLES below. .Sh PACKET FILTERING .Xr pf 4 has the ability to .Ar block and .Ar pass packets based on attributes of their layer 3 (see .Xr ip 4 and .Xr ip6 4 ) and layer 4 (see .Xr icmp 4 , .Xr icmp6 4 , .Xr tcp 4 , .Xr udp 4 ) headers. In addition, packets may also be assigned to queues for the purpose of bandwidth control. .Pp For each packet processed by the packet filter, the filter rules are evaluated in sequential order, from first to last. The last matching rule decides what action is taken. If no rule matches the packet, the default action is to pass the packet. .Pp The following actions can be used in the filter: .Bl -tag -width xxxx .It Ar block The packet is blocked. There are a number of ways in which a .Ar block rule can behave when blocking a packet. The default behaviour is to .Ar drop packets silently, however this can be overridden or made explicit either globally, by setting the .Ar block-policy option, or on a per-rule basis with one of the following options: .Pp .Bl -tag -width xxxx -compact .It Ar drop The packet is silently dropped. .It Ar return-rst This applies only to .Xr tcp 4 packets, and issues a TCP RST which closes the connection. .It Ar return-icmp .It Ar return-icmp6 This causes ICMP messages to be returned for packets which match the rule. By default this is an ICMP UNREACHABLE message, however this can be overridden by specifying a message as a code or number. .It Ar return This causes a TCP RST to be returned for .Xr tcp 4 packets and an ICMP UNREACHABLE for UDP and other packets. .El .Pp Options returning ICMP packets currently have no effect if .Xr pf 4 operates on a .Xr if_bridge 4 , as the code to support this feature has not yet been implemented. .Pp The simplest mechanism to block everything by default and only pass packets that match explicit rules is specify a first filter rule of: .Bd -literal -offset indent block all .Ed .It Ar pass The packet is passed; state is created unless the .Ar no state option is specified. .El .Pp By default .Xr pf 4 filters packets statefully; the first time a packet matches a .Ar pass rule, a state entry is created; for subsequent packets the filter checks whether the packet matches any state. If it does, the packet is passed without evaluation of any rules. After the connection is closed or times out, the state entry is automatically removed. .Pp This has several advantages. For TCP connections, comparing a packet to a state involves checking its sequence numbers, as well as TCP timestamps if a .Ar scrub reassemble tcp rule applies to the connection. If these values are outside the narrow windows of expected values, the packet is dropped. This prevents spoofing attacks, such as when an attacker sends packets with a fake source address/port but does not know the connection's sequence numbers. Similarly, .Xr pf 4 knows how to match ICMP replies to states. For example, .Bd -literal -offset indent pass out inet proto icmp all icmp-type echoreq .Ed .Pp allows echo requests (such as those created by .Xr ping 8 ) out statefully, and matches incoming echo replies correctly to states. .Pp Also, looking up states is usually faster than evaluating rules. If there are 50 rules, all of them are evaluated sequentially in O(n). Even with 50000 states, only 16 comparisons are needed to match a state, since states are stored in a binary search tree that allows searches in O(log2 n). .Pp Furthermore, correct handling of ICMP error messages is critical to many protocols, particularly TCP. .Xr pf 4 matches ICMP error messages to the correct connection, checks them against connection parameters, and passes them if appropriate. For example if an ICMP source quench message referring to a stateful TCP connection arrives, it will be matched to the state and get passed. .Pp Finally, state tracking is required for .Ar nat , binat No and Ar rdr rules, in order to track address and port translations and reverse the translation on returning packets. .Pp .Xr pf 4 will also create state for other protocols which are effectively stateless by nature. UDP packets are matched to states using only host addresses and ports, and other protocols are matched to states using only the host addresses. .Pp If stateless filtering of individual packets is desired, the .Ar no state keyword can be used to specify that state will not be created if this is the last matching rule. A number of parameters can also be set to affect how .Xr pf 4 handles state tracking. See .Sx STATEFUL TRACKING OPTIONS below for further details. .Sh PARAMETERS The rule parameters specify the packets to which a rule applies. A packet always comes in on, or goes out through, one interface. Most parameters are optional. If a parameter is specified, the rule only applies to packets with matching attributes. Certain parameters can be expressed as lists, in which case .Xr pfctl 8 generates all needed rule combinations. .Bl -tag -width xxxx .It Ar in No or Ar out This rule applies to incoming or outgoing packets. If neither .Ar in nor .Ar out are specified, the rule will match packets in both directions. .It Ar log In addition to the action specified, a log message is generated. Only the packet that establishes the state is logged, unless the .Ar no state option is specified. The logged packets are sent to a .Xr pflog 4 interface, by default .Ar pflog0 . This interface is monitored by the .Xr pflogd 8 logging daemon, which dumps the logged packets to the file .Pa /var/log/pflog in .Xr pcap 3 binary format. .It Ar log (all) Used to force logging of all packets for a connection. This is not necessary when .Ar no state is explicitly specified. As with .Ar log , packets are logged to .Xr pflog 4 . .It Ar log (user) Logs the .Ux user ID of the user that owns the socket and the PID of the process that has the socket open where the packet is sourced from or destined to (depending on which socket is local). This is in addition to the normal information logged. .Pp Only the first packet logged via .Ar log (all, user) will have the user credentials logged when using stateful matching. .It Ar log (to Aq Ar interface ) Send logs to the specified .Xr pflog 4 interface instead of .Ar pflog0 . .It Ar quick If a packet matches a rule which has the .Ar quick option set, this rule is considered the last matching rule, and evaluation of subsequent rules is skipped. .It Ar on Aq Ar interface This rule applies only to packets coming in on, or going out through, this particular interface or interface group. For more information on interface groups, see the .Ic group keyword in .Xr ifconfig 8 . .It Aq Ar af This rule applies only to packets of this address family. Supported values are .Ar inet and .Ar inet6 . .It Ar proto Aq Ar protocol This rule applies only to packets of this protocol. Common protocols are .Xr icmp 4 , .Xr icmp6 4 , .Xr tcp 4 , and .Xr udp 4 . For a list of all the protocol name to number mappings used by .Xr pfctl 8 , see the file .Pa /etc/protocols . .It Xo .Ar from Aq Ar source .Ar port Aq Ar source .Ar os Aq Ar source .Ar to Aq Ar dest .Ar port Aq Ar dest .Xc This rule applies only to packets with the specified source and destination addresses and ports. .Pp Addresses can be specified in CIDR notation (matching netblocks), as symbolic host names, interface names or interface group names, or as any of the following keywords: .Pp .Bl -tag -width xxxxxxxxxxxxxx -compact .It Ar any Any address. .It Ar no-route Any address which is not currently routable. .It Ar urpf-failed Any source address that fails a unicast reverse path forwarding (URPF) check, i.e. packets coming in on an interface other than that which holds the route back to the packet's source address. .It Aq Ar table Any address that matches the given table. .El .Pp Ranges of addresses are specified by using the .Sq - operator. For instance: .Dq 10.1.1.10 - 10.1.1.12 means all addresses from 10.1.1.10 to 10.1.1.12, hence addresses 10.1.1.10, 10.1.1.11, and 10.1.1.12. .Pp Interface names and interface group names can have modifiers appended: .Pp .Bl -tag -width xxxxxxxxxxxx -compact .It Ar :network Translates to the network(s) attached to the interface. .It Ar :broadcast Translates to the interface's broadcast address(es). .It Ar :peer Translates to the point-to-point interface's peer address(es). .It Ar :0 Do not include interface aliases. .El .Pp Host names may also have the .Ar :0 option appended to restrict the name resolution to the first of each v4 and v6 address found. .Pp Host name resolution and interface to address translation are done at ruleset load-time. When the address of an interface (or host name) changes (under DHCP or PPP, for instance), the ruleset must be reloaded for the change to be reflected in the kernel. Surrounding the interface name (and optional modifiers) in parentheses changes this behaviour. When the interface name is surrounded by parentheses, the rule is automatically updated whenever the interface changes its address. The ruleset does not need to be reloaded. This is especially useful with .Ar nat . .Pp Ports can be specified either by number or by name. For example, port 80 can be specified as .Em www . For a list of all port name to number mappings used by .Xr pfctl 8 , see the file .Pa /etc/services . .Pp Ports and ranges of ports are specified by using these operators: .Bd -literal -offset indent = (equal) != (unequal) \*(Lt (less than) \*(Le (less than or equal) \*(Gt (greater than) \*(Ge (greater than or equal) : (range including boundaries) \*(Gt\*(Lt (range excluding boundaries) \*(Lt\*(Gt (except range) .Ed .Pp .Sq \*(Gt\*(Lt , .Sq \*(Lt\*(Gt and .Sq \&: are binary operators (they take two arguments). For instance: .Bl -tag -width Fl .It Ar port 2000:2004 means .Sq all ports \*(Ge 2000 and \*(Le 2004 , hence ports 2000, 2001, 2002, 2003 and 2004. .It Ar port 2000 \*(Gt\*(Lt 2004 means .Sq all ports \*(Gt 2000 and \*(Lt 2004 , hence ports 2001, 2002 and 2003. .It Ar port 2000 \*(Lt\*(Gt 2004 means .Sq all ports \*(Lt 2000 or \*(Gt 2004 , hence ports 1-1999 and 2005-65535. .El .Pp The operating system of the source host can be specified in the case of TCP rules with the .Ar OS modifier. See the .Sx OPERATING SYSTEM FINGERPRINTING section for more information. .Pp The host, port and OS specifications are optional, as in the following examples: .Bd -literal -offset indent pass in all pass in from any to any pass in proto tcp from any port \*(Le 1024 to any pass in proto tcp from any to any port 25 pass in proto tcp from 10.0.0.0/8 port \*(Gt 1024 \e to ! 10.1.2.3 port != ssh pass in proto tcp from any os "OpenBSD" .Ed .It Ar all This is equivalent to "from any to any". .It Ar group Aq Ar group Similar to .Ar user , this rule only applies to packets of sockets owned by the specified group. .It Ar user Aq Ar user This rule only applies to packets of sockets owned by the specified user. For outgoing connections initiated from the firewall, this is the user that opened the connection. For incoming connections to the firewall itself, this is the user that listens on the destination port. For forwarded connections, where the firewall is not a connection endpoint, the user and group are .Em unknown . .Pp All packets, both outgoing and incoming, of one connection are associated with the same user and group. Only TCP and UDP packets can be associated with users; for other protocols these parameters are ignored. .Pp User and group refer to the effective (as opposed to the real) IDs, in case the socket is created by a setuid/setgid process. User and group IDs are stored when a socket is created; when a process creates a listening socket as root (for instance, by binding to a privileged port) and subsequently changes to another user ID (to drop privileges), the credentials will remain root. .Pp User and group IDs can be specified as either numbers or names. The syntax is similar to the one for ports. The value .Em unknown matches packets of forwarded connections. .Em unknown can only be used with the operators .Cm = and .Cm != . Other constructs like .Cm user \*(Ge unknown are invalid. Forwarded packets with unknown user and group ID match only rules that explicitly compare against .Em unknown with the operators .Cm = or .Cm != . For instance .Cm user \*(Ge 0 does not match forwarded packets. The following example allows only selected users to open outgoing connections: .Bd -literal -offset indent block out proto { tcp, udp } all pass out proto { tcp, udp } all user { \*(Lt 1000, dhartmei } .Ed .It Xo Ar flags Aq Ar a .Pf / Ns Aq Ar b .No \*(Ba / Ns Aq Ar b .No \*(Ba any .Xc This rule only applies to TCP packets that have the flags .Aq Ar a set out of set .Aq Ar b . Flags not specified in .Aq Ar b are ignored. For stateful connections, the default is .Ar flags S/SA . To indicate that flags should not be checked at all, specify .Ar flags any . The flags are: (F)IN, (S)YN, (R)ST, (P)USH, (A)CK, (U)RG, (E)CE, and C(W)R. .Bl -tag -width Fl .It Ar flags S/S Flag SYN is set. The other flags are ignored. .It Ar flags S/SA This is the default setting for stateful connections. Out of SYN and ACK, exactly SYN may be set. SYN, SYN+PSH and SYN+RST match, but SYN+ACK, ACK and ACK+RST do not. This is more restrictive than the previous example. .It Ar flags /SFRA If the first set is not specified, it defaults to none. All of SYN, FIN, RST and ACK must be unset. .El .Pp Because .Ar flags S/SA is applied by default (unless .Ar no state is specified), only the initial SYN packet of a TCP handshake will create a state for a TCP connection. It is possible to be less restrictive, and allow state creation from intermediate .Pq non-SYN packets, by specifying .Ar flags any . This will cause .Xr pf 4 to synchronize to existing connections, for instance if one flushes the state table. However, states created from such intermediate packets may be missing connection details such as the TCP window scaling factor. States which modify the packet flow, such as those affected by .Ar nat , binat No or Ar rdr rules, .Ar modulate No or Ar synproxy state options, or scrubbed with .Ar reassemble tcp will also not be recoverable from intermediate packets. Such connections will stall and time out. .It Xo Ar icmp-type Aq Ar type .Ar code Aq Ar code .Xc .It Xo Ar icmp6-type Aq Ar type .Ar code Aq Ar code .Xc This rule only applies to ICMP or ICMPv6 packets with the specified type and code. Text names for ICMP types and codes are listed in .Xr icmp 4 and .Xr icmp6 4 . This parameter is only valid for rules that cover protocols ICMP or ICMP6. The protocol and the ICMP type indicator .Po .Ar icmp-type or .Ar icmp6-type .Pc must match. .It Xo Ar tos Aq Ar string .No \*(Ba Aq Ar number .Xc This rule applies to packets with the specified .Em TOS bits set. .Em TOS may be given as one of .Ar lowdelay , .Ar throughput , .Ar reliability , or as either hex or decimal. .Pp For example, the following rules are identical: .Bd -literal -offset indent pass all tos lowdelay pass all tos 0x10 pass all tos 16 .Ed .It Ar allow-opts By default, IPv4 packets with IP options or IPv6 packets with routing extension headers are blocked. When .Ar allow-opts is specified for a .Ar pass rule, packets that pass the filter based on that rule (last matching) do so even if they contain IP options or routing extension headers. For packets that match state, the rule that initially created the state is used. The implicit .Ar pass rule that is used when a packet does not match any rules does not allow IP options. .It Ar label Aq Ar string Adds a label (name) to the rule, which can be used to identify the rule. For instance, pfctl -s labels shows per-rule statistics for rules that have labels. .Pp The following macros can be used in labels: .Pp .Bl -tag -width $srcaddr -compact -offset indent .It Ar $if The interface. .It Ar $srcaddr The source IP address. .It Ar $dstaddr The destination IP address. .It Ar $srcport The source port specification. .It Ar $dstport The destination port specification. .It Ar $proto The protocol name. .It Ar $nr The rule number. .El .Pp For example: .Bd -literal -offset indent ips = \&"{ 1.2.3.4, 1.2.3.5 }\&" pass in proto tcp from any to $ips \e port \*(Gt 1023 label \&"$dstaddr:$dstport\&" .Ed .Pp expands to .Bd -literal -offset indent pass in inet proto tcp from any to 1.2.3.4 \e port \*(Gt 1023 label \&"1.2.3.4:\*(Gt1023\&" pass in inet proto tcp from any to 1.2.3.5 \e port \*(Gt 1023 label \&"1.2.3.5:\*(Gt1023\&" .Ed .Pp The macro expansion for the .Ar label directive occurs only at configuration file parse time, not during runtime. .It Xo Ar queue Aq Ar queue .No \*(Ba ( Aq Ar queue , .Aq Ar queue ) .Xc Packets matching this rule will be assigned to the specified queue. If two queues are given, packets which have a .Em TOS of .Em lowdelay and TCP ACKs with no data payload will be assigned to the second one. See .Sx QUEUEING for setup details. .Pp For example: .Bd -literal -offset indent pass in proto tcp to port 25 queue mail pass in proto tcp to port 22 queue(ssh_bulk, ssh_prio) .Ed .It Ar tag Aq Ar string Packets matching this rule will be tagged with the specified string. The tag acts as an internal marker that can be used to identify these packets later on. This can be used, for example, to provide trust between interfaces and to determine if packets have been processed by translation rules. Tags are .Qq sticky , meaning that the packet will be tagged even if the rule is not the last matching rule. Further matching rules can replace the tag with a new one but will not remove a previously applied tag. A packet is only ever assigned one tag at a time. Packet tagging can be done during .Ar nat , .Ar rdr , or .Ar binat rules in addition to filter rules. Tags take the same macros as labels (see above). .It Ar tagged Aq Ar string Used with filter, translation or scrub rules to specify that packets must already be tagged with the given tag in order to match the rule. Inverse tag matching can also be done by specifying the .Cm !\& operator before the .Ar tagged keyword. .It Ar rtable Aq Ar number Used to select an alternate routing table for the routing lookup. Only effective before the route lookup happened, i.e. when filtering inbound. .It Xo Ar divert-to Aq Ar host .Ar port Aq Ar port .Xc Used to redirect packets to a local socket bound to .Ar host and .Ar port . The packets will not be modified, so .Xr getsockname 2 on the socket will return the original destination address of the packet. .It Ar divert-reply Used to receive replies for sockets that are bound to addresses which are not local to the machine. See .Xr setsockopt 2 for information on how to bind these sockets. .It Ar probability Aq Ar number A probability attribute can be attached to a rule, with a value set between 0 and 1, bounds not included. In that case, the rule will be honoured using the given probability value only. For example, the following rule will drop 20% of incoming ICMP packets: .Bd -literal -offset indent block in proto icmp probability 20% .Ed .El .Sh ROUTING If a packet matches a rule with a route option set, the packet filter will route the packet according to the type of route option. When such a rule creates state, the route option is also applied to all packets matching the same connection. .Bl -tag -width xxxx .It Ar fastroute The .Ar fastroute option does a normal route lookup to find the next hop for the packet. .It Ar route-to The .Ar route-to option routes the packet to the specified interface with an optional address for the next hop. When a .Ar route-to rule creates state, only packets that pass in the same direction as the filter rule specifies will be routed in this way. Packets passing in the opposite direction (replies) are not affected and are routed normally. .It Ar reply-to The .Ar reply-to option is similar to .Ar route-to , but routes packets that pass in the opposite direction (replies) to the specified interface. Opposite direction is only defined in the context of a state entry, and .Ar reply-to is useful only in rules that create state. It can be used on systems with multiple external connections to route all outgoing packets of a connection through the interface the incoming connection arrived through (symmetric routing enforcement). .It Ar dup-to The .Ar dup-to option creates a duplicate of the packet and routes it like .Ar route-to . The original packet gets routed as it normally would. .El .Sh POOL OPTIONS For .Ar nat and .Ar rdr rules, (as well as for the .Ar route-to , .Ar reply-to and .Ar dup-to rule options) for which there is a single redirection address which has a subnet mask smaller than 32 for IPv4 or 128 for IPv6 (more than one IP address), a variety of different methods for assigning this address can be used: .Bl -tag -width xxxx .It Ar bitmask The .Ar bitmask option applies the network portion of the redirection address to the address to be modified (source with .Ar nat , destination with .Ar rdr ) . .It Ar random The .Ar random option selects an address at random within the defined block of addresses. .It Ar source-hash The .Ar source-hash option uses a hash of the source address to determine the redirection address, ensuring that the redirection address is always the same for a given source. An optional key can be specified after this keyword either in hex or as a string; by default .Xr pfctl 8 randomly generates a key for source-hash every time the ruleset is reloaded. .It Ar round-robin The .Ar round-robin option loops through the redirection address(es). .Pp When more than one redirection address is specified, .Ar round-robin is the only permitted pool type. .It Ar static-port With .Ar nat rules, the .Ar static-port option prevents .Xr pf 4 from modifying the source port on TCP and UDP packets. .El .Pp Additionally, the .Ar sticky-address option can be specified to help ensure that multiple connections from the same source are mapped to the same redirection address. This option can be used with the .Ar random and .Ar round-robin pool options. Note that by default these associations are destroyed as soon as there are no longer states which refer to them; in order to make the mappings last beyond the lifetime of the states, increase the global options with .Ar set timeout src.track . See .Sx STATEFUL TRACKING OPTIONS for more ways to control the source tracking. .Sh STATE MODULATION Much of the security derived from TCP is attributable to how well the initial sequence numbers (ISNs) are chosen. Some popular stack implementations choose .Em very poor ISNs and thus are normally susceptible to ISN prediction exploits. By applying a .Ar modulate state rule to a TCP connection, .Xr pf 4 will create a high quality random sequence number for each connection endpoint. .Pp The .Ar modulate state directive implicitly keeps state on the rule and is only applicable to TCP connections. .Pp For instance: .Bd -literal -offset indent block all pass out proto tcp from any to any modulate state pass in proto tcp from any to any port 25 flags S/SFRA modulate state .Ed .Pp Note that modulated connections will not recover when the state table is lost (firewall reboot, flushing the state table, etc...). .Xr pf 4 will not be able to infer a connection again after the state table flushes the connection's modulator. When the state is lost, the connection may be left dangling until the respective endpoints time out the connection. It is possible on a fast local network for the endpoints to start an ACK storm while trying to resynchronize after the loss of the modulator. The default .Ar flags settings (or a more strict equivalent) should be used on .Ar modulate state rules to prevent ACK storms. .Pp Note that alternative methods are available to prevent loss of the state table and allow for firewall failover. See .Xr carp 4 and .Xr pfsync 4 for further information. .Sh SYN PROXY By default, .Xr pf 4 passes packets that are part of a .Xr tcp 4 handshake between the endpoints. The .Ar synproxy state option can be used to cause .Xr pf 4 itself to complete the handshake with the active endpoint, perform a handshake with the passive endpoint, and then forward packets between the endpoints. .Pp No packets are sent to the passive endpoint before the active endpoint has completed the handshake, hence so-called SYN floods with spoofed source addresses will not reach the passive endpoint, as the sender can't complete the handshake. .Pp The proxy is transparent to both endpoints, they each see a single connection from/to the other endpoint. .Xr pf 4 chooses random initial sequence numbers for both handshakes. Once the handshakes are completed, the sequence number modulators (see previous section) are used to translate further packets of the connection. .Ar synproxy state includes .Ar modulate state . .Pp Rules with .Ar synproxy will not work if .Xr pf 4 operates on a .Xr bridge 4 . .Pp Example: .Bd -literal -offset indent pass in proto tcp from any to any port www synproxy state .Ed .Sh STATEFUL TRACKING OPTIONS A number of options related to stateful tracking can be applied on a per-rule basis. .Ar keep state , .Ar modulate state and .Ar synproxy state support these options, and .Ar keep state must be specified explicitly to apply options to a rule. .Pp .Bl -tag -width xxxx -compact .It Ar max Aq Ar number Limits the number of concurrent states the rule may create. When this limit is reached, further packets that would create state will not match this rule until existing states time out. .It Ar no-sync Prevent state changes for states created by this rule from appearing on the .Xr pfsync 4 interface. .It Xo Aq Ar timeout .Aq Ar seconds .Xc Changes the timeout values used for states created by this rule. For a list of all valid timeout names, see .Sx OPTIONS above. .It Ar sloppy Uses a sloppy TCP connection tracker that does not check sequence numbers at all, which makes insertion and ICMP teardown attacks way easier. This is intended to be used in situations where one does not see all packets of a connection, e.g. in asymmetric routing situations. Cannot be used with modulate or synproxy state. .It Ar pflow States created by this rule are exported on the .Xr pflow 4 interface. .El .Pp Multiple options can be specified, separated by commas: .Bd -literal -offset indent pass in proto tcp from any to any \e port www keep state \e (max 100, source-track rule, max-src-nodes 75, \e max-src-states 3, tcp.established 60, tcp.closing 5) .Ed .Pp When the .Ar source-track keyword is specified, the number of states per source IP is tracked. .Pp .Bl -tag -width xxxx -compact .It Ar source-track rule The maximum number of states created by this rule is limited by the rule's .Ar max-src-nodes and .Ar max-src-states options. Only state entries created by this particular rule count toward the rule's limits. .It Ar source-track global The number of states created by all rules that use this option is limited. Each rule can specify different .Ar max-src-nodes and .Ar max-src-states options, however state entries created by any participating rule count towards each individual rule's limits. .El .Pp The following limits can be set: .Pp .Bl -tag -width xxxx -compact .It Ar max-src-nodes Aq Ar number Limits the maximum number of source addresses which can simultaneously have state table entries. .It Ar max-src-states Aq Ar number Limits the maximum number of simultaneous state entries that a single source address can create with this rule. .El .Pp For stateful TCP connections, limits on established connections (connections which have completed the TCP 3-way handshake) can also be enforced per source IP. .Pp .Bl -tag -width xxxx -compact .It Ar max-src-conn Aq Ar number Limits the maximum number of simultaneous TCP connections which have completed the 3-way handshake that a single host can make. .It Xo Ar max-src-conn-rate Aq Ar number .No / Aq Ar seconds .Xc Limit the rate of new connections over a time interval. The connection rate is an approximation calculated as a moving average. .El .Pp Because the 3-way handshake ensures that the source address is not being spoofed, more aggressive action can be taken based on these limits. With the .Ar overload Aq Ar table state option, source IP addresses which hit either of the limits on established connections will be added to the named table. This table can be used in the ruleset to block further activity from the offending host, redirect it to a tarpit process, or restrict its bandwidth. .Pp The optional .Ar flush keyword kills all states created by the matching rule which originate from the host which exceeds these limits. The .Ar global modifier to the flush command kills all states originating from the offending host, regardless of which rule created the state. .Pp For example, the following rules will protect the webserver against hosts making more than 100 connections in 10 seconds. Any host which connects faster than this rate will have its address added to the .Aq bad_hosts table and have all states originating from it flushed. Any new packets arriving from this host will be dropped unconditionally by the block rule. .Bd -literal -offset indent block quick from \*(Ltbad_hosts\*(Gt pass in on $ext_if proto tcp to $webserver port www keep state \e (max-src-conn-rate 100/10, overload \*(Ltbad_hosts\*(Gt flush global) .Ed .Sh OPERATING SYSTEM FINGERPRINTING Passive OS Fingerprinting is a mechanism to inspect nuances of a TCP connection's initial SYN packet and guess at the host's operating system. Unfortunately these nuances are easily spoofed by an attacker so the fingerprint is not useful in making security decisions. But the fingerprint is typically accurate enough to make policy decisions upon. .Pp The fingerprints may be specified by operating system class, by version, or by subtype/patchlevel. The class of an operating system is typically the vendor or genre and would be .Ox for the .Xr pf 4 firewall itself. The version of the oldest available .Ox release on the main FTP site would be 2.6 and the fingerprint would be written .Pp .Dl \&"OpenBSD 2.6\&" .Pp The subtype of an operating system is typically used to describe the patchlevel if that patch led to changes in the TCP stack behavior. In the case of .Ox , the only subtype is for a fingerprint that was normalized by the .Ar no-df scrub option and would be specified as .Pp .Dl \&"OpenBSD 3.3 no-df\&" .Pp Fingerprints for most popular operating systems are provided by .Xr pf.os 5 . Once .Xr pf 4 is running, a complete list of known operating system fingerprints may be listed by running: .Pp .Dl # pfctl -so .Pp Filter rules can enforce policy at any level of operating system specification assuming a fingerprint is present. Policy could limit traffic to approved operating systems or even ban traffic from hosts that aren't at the latest service pack. .Pp The .Ar unknown class can also be used as the fingerprint which will match packets for which no operating system fingerprint is known. .Pp Examples: .Bd -literal -offset indent pass out proto tcp from any os OpenBSD block out proto tcp from any os Doors block out proto tcp from any os "Doors PT" block out proto tcp from any os "Doors PT SP3" block out from any os "unknown" pass on lo0 proto tcp from any os "OpenBSD 3.3 lo0" .Ed .Pp Operating system fingerprinting is limited only to the TCP SYN packet. This means that it will not work on other protocols and will not match a currently established connection. .Pp Caveat: operating system fingerprints are occasionally wrong. There are three problems: an attacker can trivially craft his packets to appear as any operating system he chooses; an operating system patch could change the stack behavior and no fingerprints will match it until the database is updated; and multiple operating systems may have the same fingerprint. .Sh BLOCKING SPOOFED TRAFFIC "Spoofing" is the faking of IP addresses, typically for malicious purposes. The .Ar antispoof directive expands to a set of filter rules which will block all traffic with a source IP from the network(s) directly connected to the specified interface(s) from entering the system through any other interface. .Pp For example, the line .Bd -literal -offset indent antispoof for lo0 .Ed .Pp expands to .Bd -literal -offset indent block drop in on ! lo0 inet from 127.0.0.1/8 to any block drop in on ! lo0 inet6 from ::1 to any .Ed .Pp For non-loopback interfaces, there are additional rules to block incoming packets with a source IP address identical to the interface's IP(s). For example, assuming the interface wi0 had an IP address of 10.0.0.1 and a netmask of 255.255.255.0, the line .Bd -literal -offset indent antispoof for wi0 inet .Ed .Pp expands to .Bd -literal -offset indent block drop in on ! wi0 inet from 10.0.0.0/24 to any block drop in inet from 10.0.0.1 to any .Ed .Pp Caveat: Rules created by the .Ar antispoof directive interfere with packets sent over loopback interfaces to local addresses. One should pass these explicitly. .Sh FRAGMENT HANDLING The size of IP datagrams (packets) can be significantly larger than the maximum transmission unit (MTU) of the network. In cases when it is necessary or more efficient to send such large packets, the large packet will be fragmented into many smaller packets that will each fit onto the wire. Unfortunately for a firewalling device, only the first logical fragment will contain the necessary header information for the subprotocol that allows .Xr pf 4 to filter on things such as TCP ports or to perform NAT. .Pp Besides the use of .Ar scrub rules as described in .Sx TRAFFIC NORMALIZATION above, there are three options for handling fragments in the packet filter. .Pp One alternative is to filter individual fragments with filter rules. If no .Ar scrub rule applies to a fragment, it is passed to the filter. Filter rules with matching IP header parameters decide whether the fragment is passed or blocked, in the same way as complete packets are filtered. Without reassembly, fragments can only be filtered based on IP header fields (source/destination address, protocol), since subprotocol header fields are not available (TCP/UDP port numbers, ICMP code/type). The .Ar fragment option can be used to restrict filter rules to apply only to fragments, but not complete packets. Filter rules without the .Ar fragment option still apply to fragments, if they only specify IP header fields. For instance, the rule .Bd -literal -offset indent pass in proto tcp from any to any port 80 .Ed .Pp never applies to a fragment, even if the fragment is part of a TCP packet with destination port 80, because without reassembly this information is not available for each fragment. This also means that fragments cannot create new or match existing state table entries, which makes stateful filtering and address translation (NAT, redirection) for fragments impossible. .Pp It's also possible to reassemble only certain fragments by specifying source or destination addresses or protocols as parameters in .Ar scrub rules. .Pp In most cases, the benefits of reassembly outweigh the additional memory cost, and it's recommended to use .Ar scrub rules to reassemble all fragments via the .Ar fragment reassemble modifier. .Pp The memory allocated for fragment caching can be limited using .Xr pfctl 8 . Once this limit is reached, fragments that would have to be cached are dropped until other entries time out. The timeout value can also be adjusted. .Pp When forwarding reassembled IPv6 packets, pf refragments them with the original maximum fragment size. This allows the sender to determine the optimal fragment size by path MTU discovery. .Sh ANCHORS Besides the main ruleset, .Xr pfctl 8 can load rulesets into .Ar anchor attachment points. An .Ar anchor is a container that can hold rules, address tables, and other anchors. .Pp An .Ar anchor has a name which specifies the path where .Xr pfctl 8 can be used to access the anchor to perform operations on it, such as attaching child anchors to it or loading rules into it. Anchors may be nested, with components separated by .Sq / characters, similar to how file system hierarchies are laid out. The main ruleset is actually the default anchor, so filter and translation rules, for example, may also be contained in any anchor. .Pp An anchor can reference another .Ar anchor attachment point using the following kinds of rules: .Bl -tag -width xxxx .It Ar nat-anchor Aq Ar name Evaluates the .Ar nat rules in the specified .Ar anchor . .It Ar rdr-anchor Aq Ar name Evaluates the .Ar rdr rules in the specified .Ar anchor . .It Ar binat-anchor Aq Ar name Evaluates the .Ar binat rules in the specified .Ar anchor . .It Ar anchor Aq Ar name Evaluates the filter rules in the specified .Ar anchor . .It Xo Ar load anchor .Aq Ar name .Ar from Aq Ar file .Xc Loads the rules from the specified file into the anchor .Ar name . .El .Pp When evaluation of the main ruleset reaches an .Ar anchor rule, .Xr pf 4 will proceed to evaluate all rules specified in that anchor. .Pp Matching filter and translation rules marked with the .Ar quick option are final and abort the evaluation of the rules in other anchors and the main ruleset. If the .Ar anchor itself is marked with the .Ar quick option, ruleset evaluation will terminate when the anchor is exited if the packet is matched by any rule within the anchor. .Pp .Ar anchor rules are evaluated relative to the anchor in which they are contained. For example, all .Ar anchor rules specified in the main ruleset will reference anchor attachment points underneath the main ruleset, and .Ar anchor rules specified in a file loaded from a .Ar load anchor rule will be attached under that anchor point. .Pp Rules may be contained in .Ar anchor attachment points which do not contain any rules when the main ruleset is loaded, and later such anchors can be manipulated through .Xr pfctl 8 without reloading the main ruleset or other anchors. For example, .Bd -literal -offset indent ext_if = \&"kue0\&" block on $ext_if all anchor spam pass out on $ext_if all pass in on $ext_if proto tcp from any \e to $ext_if port smtp .Ed .Pp blocks all packets on the external interface by default, then evaluates all rules in the .Ar anchor named "spam", and finally passes all outgoing connections and incoming connections to port 25. .Bd -literal -offset indent # echo \&"block in quick from 1.2.3.4 to any\&" \&| \e pfctl -a spam -f - .Ed .Pp This loads a single rule into the .Ar anchor , which blocks all packets from a specific address. .Pp The anchor can also be populated by adding a .Ar load anchor rule after the .Ar anchor rule: .Bd -literal -offset indent anchor spam load anchor spam from "/etc/pf-spam.conf" .Ed .Pp When .Xr pfctl 8 loads .Nm pf.conf , it will also load all the rules from the file .Pa /etc/pf-spam.conf into the anchor. .Pp Optionally, .Ar anchor rules can specify packet filtering parameters using the same syntax as filter rules. When parameters are used, the .Ar anchor rule is only evaluated for matching packets. This allows conditional evaluation of anchors, like: .Bd -literal -offset indent block on $ext_if all anchor spam proto tcp from any to any port smtp pass out on $ext_if all pass in on $ext_if proto tcp from any to $ext_if port smtp .Ed .Pp The rules inside .Ar anchor spam are only evaluated for .Ar tcp packets with destination port 25. Hence, .Bd -literal -offset indent # echo \&"block in quick from 1.2.3.4 to any" \&| \e pfctl -a spam -f - .Ed .Pp will only block connections from 1.2.3.4 to port 25. .Pp Anchors may end with the asterisk .Pq Sq * character, which signifies that all anchors attached at that point should be evaluated in the alphabetical ordering of their anchor name. For example, .Bd -literal -offset indent anchor "spam/*" .Ed .Pp will evaluate each rule in each anchor attached to the .Li spam anchor. Note that it will only evaluate anchors that are directly attached to the .Li spam anchor, and will not descend to evaluate anchors recursively. .Pp Since anchors are evaluated relative to the anchor in which they are contained, there is a mechanism for accessing the parent and ancestor anchors of a given anchor. Similar to file system path name resolution, if the sequence .Dq .. appears as an anchor path component, the parent anchor of the current anchor in the path evaluation at that point will become the new current anchor. As an example, consider the following: .Bd -literal -offset indent # echo ' anchor "spam/allowed" ' | pfctl -f - # echo -e ' anchor "../banned" \en pass' | \e pfctl -a spam/allowed -f - .Ed .Pp Evaluation of the main ruleset will lead into the .Li spam/allowed anchor, which will evaluate the rules in the .Li spam/banned anchor, if any, before finally evaluating the .Ar pass rule. .Pp Filter rule .Ar anchors can also be loaded inline in the ruleset within a brace ('{' '}') delimited block. Brace delimited blocks may contain rules or other brace-delimited blocks. When anchors are loaded this way the anchor name becomes optional. .Bd -literal -offset indent anchor "external" on egress { block anchor out { pass proto tcp from any to port { 25, 80, 443 } } pass in proto tcp to any port 22 } .Ed .Pp Since the parser specification for anchor names is a string, any reference to an anchor name containing .Sq / characters will require double quote .Pq Sq \&" characters around the anchor name. .Sh TRANSLATION EXAMPLES This example maps incoming requests on port 80 to port 8080, on which a daemon is running (because, for example, it is not run as root, and therefore lacks permission to bind to port 80). .Bd -literal # use a macro for the interface name, so it can be changed easily ext_if = \&"ne3\&" # map daemon on 8080 to appear to be on 80 rdr on $ext_if proto tcp from any to any port 80 -\*(Gt 127.0.0.1 port 8080 .Ed .Pp If the .Ar pass modifier is given, packets matching the translation rule are passed without inspecting the filter rules: .Bd -literal rdr pass on $ext_if proto tcp from any to any port 80 -\*(Gt 127.0.0.1 \e port 8080 .Ed .Pp In the example below, vlan12 is configured as 192.168.168.1; the machine translates all packets coming from 192.168.168.0/24 to 204.92.77.111 when they are going out any interface except vlan12. This has the net effect of making traffic from the 192.168.168.0/24 network appear as though it is the Internet routable address 204.92.77.111 to nodes behind any interface on the router except for the nodes on vlan12. (Thus, 192.168.168.1 can talk to the 192.168.168.0/24 nodes.) .Bd -literal nat on ! vlan12 from 192.168.168.0/24 to any -\*(Gt 204.92.77.111 .Ed .Pp In the example below, the machine sits between a fake internal 144.19.74.* network, and a routable external IP of 204.92.77.100. The .Ar no nat rule excludes protocol AH from being translated. .Bd -literal # NO NAT no nat on $ext_if proto ah from 144.19.74.0/24 to any nat on $ext_if from 144.19.74.0/24 to any -\*(Gt 204.92.77.100 .Ed .Pp In the example below, packets bound for one specific server, as well as those generated by the sysadmins are not proxied; all other connections are. .Bd -literal # NO RDR no rdr on $int_if proto { tcp, udp } from any to $server port 80 no rdr on $int_if proto { tcp, udp } from $sysadmins to any port 80 rdr on $int_if proto { tcp, udp } from any to any port 80 -\*(Gt 127.0.0.1 \e port 80 .Ed .Pp This longer example uses both a NAT and a redirection. The external interface has the address 157.161.48.183. On localhost, we are running .Xr ftp-proxy 8 , waiting for FTP sessions to be redirected to it. The three mandatory anchors for .Xr ftp-proxy 8 are omitted from this example; see the .Xr ftp-proxy 8 manpage. .Bd -literal # NAT # Translate outgoing packets' source addresses (any protocol). # In this case, any address but the gateway's external address is mapped. nat on $ext_if inet from ! ($ext_if) to any -\*(Gt ($ext_if) # NAT PROXYING # Map outgoing packets' source port to an assigned proxy port instead of # an arbitrary port. # In this case, proxy outgoing isakmp with port 500 on the gateway. nat on $ext_if inet proto udp from any port = isakmp to any -\*(Gt ($ext_if) \e port 500 # BINAT # Translate outgoing packets' source address (any protocol). # Translate incoming packets' destination address to an internal machine # (bidirectional). binat on $ext_if from 10.1.2.150 to any -\*(Gt $ext_if # RDR # Translate incoming packets' destination addresses. # As an example, redirect a TCP and UDP port to an internal machine. rdr on $ext_if inet proto tcp from any to ($ext_if) port 8080 \e -\*(Gt 10.1.2.151 port 22 rdr on $ext_if inet proto udp from any to ($ext_if) port 8080 \e -\*(Gt 10.1.2.151 port 53 # RDR # Translate outgoing ftp control connections to send them to localhost # for proxying with ftp-proxy(8) running on port 8021. rdr on $int_if proto tcp from any to any port 21 -\*(Gt 127.0.0.1 port 8021 .Ed .Pp In this example, a NAT gateway is set up to translate internal addresses using a pool of public addresses (192.0.2.16/28) and to redirect incoming web server connections to a group of web servers on the internal network. .Bd -literal # NAT LOAD BALANCE # Translate outgoing packets' source addresses using an address pool. # A given source address is always translated to the same pool address by # using the source-hash keyword. nat on $ext_if inet from any to any -\*(Gt 192.0.2.16/28 source-hash # RDR ROUND ROBIN # Translate incoming web server connections to a group of web servers on # the internal network. rdr on $ext_if proto tcp from any to any port 80 \e -\*(Gt { 10.1.2.155, 10.1.2.160, 10.1.2.161 } round-robin .Ed .Sh FILTER EXAMPLES .Bd -literal # The external interface is kue0 # (157.161.48.183, the only routable address) # and the private network is 10.0.0.0/8, for which we are doing NAT. # use a macro for the interface name, so it can be changed easily ext_if = \&"kue0\&" # normalize all incoming traffic scrub in on $ext_if all fragment reassemble # block and log everything by default block return log on $ext_if all # block anything coming from source we have no back routes for block in from no-route to any # block packets whose ingress interface does not match the one in # the route back to their source address block in from urpf-failed to any # block and log outgoing packets that do not have our address as source, # they are either spoofed or something is misconfigured (NAT disabled, # for instance), we want to be nice and do not send out garbage. block out log quick on $ext_if from ! 157.161.48.183 to any # silently drop broadcasts (cable modem noise) block in quick on $ext_if from any to 255.255.255.255 # block and log incoming packets from reserved address space and invalid # addresses, they are either spoofed or misconfigured, we cannot reply to # them anyway (hence, no return-rst). block in log quick on $ext_if from { 10.0.0.0/8, 172.16.0.0/12, \e 192.168.0.0/16, 255.255.255.255/32 } to any # ICMP # pass out/in certain ICMP queries and keep state (ping) # state matching is done on host addresses and ICMP id (not type/code), # so replies (like 0/0 for 8/0) will match queries # ICMP error messages (which always refer to a TCP/UDP packet) are # handled by the TCP/UDP states pass on $ext_if inet proto icmp all icmp-type 8 code 0 # UDP # pass out all UDP connections and keep state pass out on $ext_if proto udp all # pass in certain UDP connections and keep state (DNS) pass in on $ext_if proto udp from any to any port domain # TCP # pass out all TCP connections and modulate state pass out on $ext_if proto tcp all modulate state # pass in certain TCP connections and keep state (SSH, SMTP, DNS, IDENT) pass in on $ext_if proto tcp from any to any port { ssh, smtp, domain, \e auth } # Do not allow Windows 9x SMTP connections since they are typically # a viral worm. Alternately we could limit these OSes to 1 connection each. block in on $ext_if proto tcp from any os {"Windows 95", "Windows 98"} \e to any port smtp # IPv6 # pass in/out all IPv6 traffic: note that we have to enable this in two # different ways, on both our physical interface and our tunnel pass quick on gif0 inet6 pass quick on $ext_if proto ipv6 # Packet Tagging # three interfaces: $int_if, $ext_if, and $wifi_if (wireless). NAT is # being done on $ext_if for all outgoing packets. tag packets in on # $int_if and pass those tagged packets out on $ext_if. all other # outgoing packets (i.e., packets from the wireless network) are only # permitted to access port 80. pass in on $int_if from any to any tag INTNET pass in on $wifi_if from any to any block out on $ext_if from any to any pass out quick on $ext_if tagged INTNET pass out on $ext_if proto tcp from any to any port 80 # tag incoming packets as they are redirected to spamd(8). use the tag # to pass those packets through the packet filter. rdr on $ext_if inet proto tcp from \*(Ltspammers\*(Gt to port smtp \e tag SPAMD -\*(Gt 127.0.0.1 port spamd block in on $ext_if pass in on $ext_if inet proto tcp tagged SPAMD .Ed .Sh GRAMMAR Syntax for .Nm in BNF: .Bd -literal line = ( option | pf-rule | nat-rule | binat-rule | rdr-rule | antispoof-rule | altq-rule | queue-rule | trans-anchors | anchor-rule | anchor-close | load-anchor | table-rule | include ) option = "set" ( [ "timeout" ( timeout | "{" timeout-list "}" ) ] | [ "ruleset-optimization" [ "none" | "basic" | "profile" ]] | [ "optimization" [ "default" | "normal" | "high-latency" | "satellite" | "aggressive" | "conservative" ] ] [ "limit" ( limit-item | "{" limit-list "}" ) ] | [ "loginterface" ( interface-name | "none" ) ] | [ "block-policy" ( "drop" | "return" ) ] | [ "state-policy" ( "if-bound" | "floating" ) ] [ "state-defaults" state-opts ] [ "require-order" ( "yes" | "no" ) ] [ "fingerprints" filename ] | [ "skip on" ifspec ] | [ "debug" ( "none" | "urgent" | "misc" | "loud" ) ] ) pf-rule = action [ ( "in" | "out" ) ] [ "log" [ "(" logopts ")"] ] [ "quick" ] [ "on" ifspec ] [ "fastroute" | route ] [ af ] [ protospec ] hosts [ filteropt-list ] logopts = logopt [ "," logopts ] logopt = "all" | "user" | "to" interface-name filteropt-list = filteropt-list filteropt | filteropt filteropt = user | group | flags | icmp-type | icmp6-type | "tos" tos | ( "no" | "keep" | "modulate" | "synproxy" ) "state" [ "(" state-opts ")" ] | "fragment" | "no-df" | "min-ttl" number | "set-tos" tos | "max-mss" number | "random-id" | "reassemble tcp" | fragmentation | "allow-opts" | "label" string | "tag" string | [ ! ] "tagged" string | "queue" ( string | "(" string [ [ "," ] string ] ")" ) | "rtable" number | "probability" number"%" nat-rule = [ "no" ] "nat" [ "pass" [ "log" [ "(" logopts ")" ] ] ] [ "on" ifspec ] [ af ] [ protospec ] hosts [ "tag" string ] [ "tagged" string ] [ "-\*(Gt" ( redirhost | "{" redirhost-list "}" ) [ portspec ] [ pooltype ] [ "static-port" ] ] binat-rule = [ "no" ] "binat" [ "pass" [ "log" [ "(" logopts ")" ] ] ] [ "on" interface-name ] [ af ] [ "proto" ( proto-name | proto-number ) ] "from" address [ "/" mask-bits ] "to" ipspec [ "tag" string ] [ "tagged" string ] [ "-\*(Gt" address [ "/" mask-bits ] ] rdr-rule = [ "no" ] "rdr" [ "pass" [ "log" [ "(" logopts ")" ] ] ] [ "on" ifspec ] [ af ] [ protospec ] hosts [ "tag" string ] [ "tagged" string ] [ "-\*(Gt" ( redirhost | "{" redirhost-list "}" ) [ portspec ] [ pooltype ] ] antispoof-rule = "antispoof" [ "log" ] [ "quick" ] "for" ifspec [ af ] [ "label" string ] table-rule = "table" "\*(Lt" string "\*(Gt" [ tableopts-list ] tableopts-list = tableopts-list tableopts | tableopts tableopts = "persist" | "const" | "counters" | "file" string | "{" [ tableaddr-list ] "}" tableaddr-list = tableaddr-list [ "," ] tableaddr-spec | tableaddr-spec tableaddr-spec = [ "!" ] tableaddr [ "/" mask-bits ] tableaddr = hostname | ifspec | "self" | ipv4-dotted-quad | ipv6-coloned-hex altq-rule = "altq on" interface-name queueopts-list "queue" subqueue queue-rule = "queue" string [ "on" interface-name ] queueopts-list subqueue anchor-rule = "anchor" [ string ] [ ( "in" | "out" ) ] [ "on" ifspec ] [ af ] [ protospec ] [ hosts ] [ filteropt-list ] [ "{" ] anchor-close = "}" trans-anchors = ( "nat-anchor" | "rdr-anchor" | "binat-anchor" ) string [ "on" ifspec ] [ af ] [ "proto" ] [ protospec ] [ hosts ] load-anchor = "load anchor" string "from" filename queueopts-list = queueopts-list queueopts | queueopts queueopts = [ "bandwidth" bandwidth-spec ] | [ "qlimit" number ] | [ "tbrsize" number ] | [ "priority" number ] | [ schedulers ] schedulers = ( cbq-def | priq-def | hfsc-def ) bandwidth-spec = "number" ( "b" | "Kb" | "Mb" | "Gb" | "%" ) action = "pass" | "block" [ return ] | [ "no" ] "scrub" return = "drop" | "return" | "return-rst" [ "( ttl" number ")" ] | "return-icmp" [ "(" icmpcode [ [ "," ] icmp6code ] ")" ] | "return-icmp6" [ "(" icmp6code ")" ] icmpcode = ( icmp-code-name | icmp-code-number ) icmp6code = ( icmp6-code-name | icmp6-code-number ) ifspec = ( [ "!" ] ( interface-name | interface-group ) ) | "{" interface-list "}" interface-list = [ "!" ] ( interface-name | interface-group ) [ [ "," ] interface-list ] route = ( "route-to" | "reply-to" | "dup-to" ) ( routehost | "{" routehost-list "}" ) [ pooltype ] af = "inet" | "inet6" protospec = "proto" ( proto-name | proto-number | "{" proto-list "}" ) proto-list = ( proto-name | proto-number ) [ [ "," ] proto-list ] hosts = "all" | "from" ( "any" | "no-route" | "urpf-failed" | "self" | host | "{" host-list "}" ) [ port ] [ os ] "to" ( "any" | "no-route" | "self" | host | "{" host-list "}" ) [ port ] ipspec = "any" | host | "{" host-list "}" host = [ "!" ] ( address [ "/" mask-bits ] | "\*(Lt" string "\*(Gt" ) redirhost = address [ "/" mask-bits ] routehost = "(" interface-name [ address [ "/" mask-bits ] ] ")" address = ( interface-name | interface-group | "(" ( interface-name | interface-group ) ")" | hostname | ipv4-dotted-quad | ipv6-coloned-hex ) host-list = host [ [ "," ] host-list ] redirhost-list = redirhost [ [ "," ] redirhost-list ] routehost-list = routehost [ [ "," ] routehost-list ] port = "port" ( unary-op | binary-op | "{" op-list "}" ) portspec = "port" ( number | name ) [ ":" ( "*" | number | name ) ] os = "os" ( os-name | "{" os-list "}" ) user = "user" ( unary-op | binary-op | "{" op-list "}" ) group = "group" ( unary-op | binary-op | "{" op-list "}" ) unary-op = [ "=" | "!=" | "\*(Lt" | "\*(Le" | "\*(Gt" | "\*(Ge" ] ( name | number ) binary-op = number ( "\*(Lt\*(Gt" | "\*(Gt\*(Lt" | ":" ) number op-list = ( unary-op | binary-op ) [ [ "," ] op-list ] os-name = operating-system-name os-list = os-name [ [ "," ] os-list ] flags = "flags" ( [ flag-set ] "/" flag-set | "any" ) flag-set = [ "F" ] [ "S" ] [ "R" ] [ "P" ] [ "A" ] [ "U" ] [ "E" ] [ "W" ] icmp-type = "icmp-type" ( icmp-type-code | "{" icmp-list "}" ) icmp6-type = "icmp6-type" ( icmp-type-code | "{" icmp-list "}" ) icmp-type-code = ( icmp-type-name | icmp-type-number ) [ "code" ( icmp-code-name | icmp-code-number ) ] icmp-list = icmp-type-code [ [ "," ] icmp-list ] tos = ( "lowdelay" | "throughput" | "reliability" | [ "0x" ] number ) state-opts = state-opt [ [ "," ] state-opts ] state-opt = ( "max" number | "no-sync" | timeout | "sloppy" | "pflow" | "source-track" [ ( "rule" | "global" ) ] | "max-src-nodes" number | "max-src-states" number | "max-src-conn" number | "max-src-conn-rate" number "/" number | "overload" "\*(Lt" string "\*(Gt" [ "flush" ] | "if-bound" | "floating" ) -fragmentation = [ "fragment reassemble" | "fragment crop" | - "fragment drop-ovl" ] +fragmentation = [ "fragment reassemble" ] timeout-list = timeout [ [ "," ] timeout-list ] timeout = ( "tcp.first" | "tcp.opening" | "tcp.established" | "tcp.closing" | "tcp.finwait" | "tcp.closed" | "udp.first" | "udp.single" | "udp.multiple" | "icmp.first" | "icmp.error" | "other.first" | "other.single" | "other.multiple" | "frag" | "interval" | "src.track" | "adaptive.start" | "adaptive.end" ) number limit-list = limit-item [ [ "," ] limit-list ] limit-item = ( "states" | "frags" | "src-nodes" ) number pooltype = ( "bitmask" | "random" | "source-hash" [ ( hex-key | string-key ) ] | "round-robin" ) [ sticky-address ] subqueue = string | "{" queue-list "}" queue-list = string [ [ "," ] string ] cbq-def = "cbq" [ "(" cbq-opt [ [ "," ] cbq-opt ] ")" ] priq-def = "priq" [ "(" priq-opt [ [ "," ] priq-opt ] ")" ] hfsc-def = "hfsc" [ "(" hfsc-opt [ [ "," ] hfsc-opt ] ")" ] cbq-opt = ( "default" | "borrow" | "red" | "ecn" | "rio" ) priq-opt = ( "default" | "red" | "ecn" | "rio" ) hfsc-opt = ( "default" | "red" | "ecn" | "rio" | linkshare-sc | realtime-sc | upperlimit-sc ) linkshare-sc = "linkshare" sc-spec realtime-sc = "realtime" sc-spec upperlimit-sc = "upperlimit" sc-spec sc-spec = ( bandwidth-spec | "(" bandwidth-spec number bandwidth-spec ")" ) include = "include" filename .Ed .Sh FILES .Bl -tag -width "/etc/protocols" -compact .It Pa /etc/hosts Host name database. .It Pa /etc/pf.conf Default location of the ruleset file. .It Pa /etc/pf.os Default location of OS fingerprints. .It Pa /etc/protocols Protocol name database. .It Pa /etc/services Service name database. .El .Sh SEE ALSO .Xr altq 4 , .Xr carp 4 , .Xr icmp 4 , .Xr icmp6 4 , .Xr ip 4 , .Xr ip6 4 , .Xr pf 4 , .Xr pflow 4 , .Xr pfsync 4 , .Xr tcp 4 , .Xr udp 4 , .Xr hosts 5 , .Xr pf.os 5 , .Xr protocols 5 , .Xr services 5 , .Xr ftp-proxy 8 , .Xr pfctl 8 , .Xr pflogd 8 .Sh HISTORY The .Nm file format first appeared in .Ox 3.0 . Index: head/sys/net/pfvar.h =================================================================== --- head/sys/net/pfvar.h (revision 287221) +++ head/sys/net/pfvar.h (revision 287222) @@ -1,1743 +1,1741 @@ /* * Copyright (c) 2001 Daniel Hartmeier * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * * - Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * - Redistributions in binary form must reproduce the above * copyright notice, this list of conditions and the following * disclaimer in the documentation and/or other materials provided * with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE * COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. * * $OpenBSD: pfvar.h,v 1.282 2009/01/29 15:12:28 pyr Exp $ * $FreeBSD$ */ #ifndef _NET_PFVAR_H_ #define _NET_PFVAR_H_ #include #include #include #include #include #include #include #include #include #include struct pf_addr { union { struct in_addr v4; struct in6_addr v6; u_int8_t addr8[16]; u_int16_t addr16[8]; u_int32_t addr32[4]; } pfa; /* 128-bit address */ #define v4 pfa.v4 #define v6 pfa.v6 #define addr8 pfa.addr8 #define addr16 pfa.addr16 #define addr32 pfa.addr32 }; #define PFI_AFLAG_NETWORK 0x01 #define PFI_AFLAG_BROADCAST 0x02 #define PFI_AFLAG_PEER 0x04 #define PFI_AFLAG_MODEMASK 0x07 #define PFI_AFLAG_NOALIAS 0x08 struct pf_addr_wrap { union { struct { struct pf_addr addr; struct pf_addr mask; } a; char ifname[IFNAMSIZ]; char tblname[PF_TABLE_NAME_SIZE]; } v; union { struct pfi_dynaddr *dyn; struct pfr_ktable *tbl; int dyncnt; int tblcnt; } p; u_int8_t type; /* PF_ADDR_* */ u_int8_t iflags; /* PFI_AFLAG_* */ }; #ifdef _KERNEL struct pfi_dynaddr { TAILQ_ENTRY(pfi_dynaddr) entry; struct pf_addr pfid_addr4; struct pf_addr pfid_mask4; struct pf_addr pfid_addr6; struct pf_addr pfid_mask6; struct pfr_ktable *pfid_kt; struct pfi_kif *pfid_kif; int pfid_net; /* mask or 128 */ int pfid_acnt4; /* address count IPv4 */ int pfid_acnt6; /* address count IPv6 */ sa_family_t pfid_af; /* rule af */ u_int8_t pfid_iflags; /* PFI_AFLAG_* */ }; /* * Address manipulation macros */ #define HTONL(x) (x) = htonl((__uint32_t)(x)) #define HTONS(x) (x) = htons((__uint16_t)(x)) #define NTOHL(x) (x) = ntohl((__uint32_t)(x)) #define NTOHS(x) (x) = ntohs((__uint16_t)(x)) #define PF_NAME "pf" #define PF_HASHROW_ASSERT(h) mtx_assert(&(h)->lock, MA_OWNED) #define PF_HASHROW_LOCK(h) mtx_lock(&(h)->lock) #define PF_HASHROW_UNLOCK(h) mtx_unlock(&(h)->lock) #define PF_STATE_LOCK(s) \ do { \ struct pf_idhash *_ih = &V_pf_idhash[PF_IDHASH(s)]; \ PF_HASHROW_LOCK(_ih); \ } while (0) #define PF_STATE_UNLOCK(s) \ do { \ struct pf_idhash *_ih = &V_pf_idhash[PF_IDHASH((s))]; \ PF_HASHROW_UNLOCK(_ih); \ } while (0) #ifdef INVARIANTS #define PF_STATE_LOCK_ASSERT(s) \ do { \ struct pf_idhash *_ih = &V_pf_idhash[PF_IDHASH(s)]; \ PF_HASHROW_ASSERT(_ih); \ } while (0) #else /* !INVARIANTS */ #define PF_STATE_LOCK_ASSERT(s) do {} while (0) #endif /* INVARIANTS */ extern struct mtx pf_unlnkdrules_mtx; #define PF_UNLNKDRULES_LOCK() mtx_lock(&pf_unlnkdrules_mtx) #define PF_UNLNKDRULES_UNLOCK() mtx_unlock(&pf_unlnkdrules_mtx) extern struct rwlock pf_rules_lock; #define PF_RULES_RLOCK() rw_rlock(&pf_rules_lock) #define PF_RULES_RUNLOCK() rw_runlock(&pf_rules_lock) #define PF_RULES_WLOCK() rw_wlock(&pf_rules_lock) #define PF_RULES_WUNLOCK() rw_wunlock(&pf_rules_lock) #define PF_RULES_ASSERT() rw_assert(&pf_rules_lock, RA_LOCKED) #define PF_RULES_RASSERT() rw_assert(&pf_rules_lock, RA_RLOCKED) #define PF_RULES_WASSERT() rw_assert(&pf_rules_lock, RA_WLOCKED) #define PF_MODVER 1 #define PFLOG_MODVER 1 #define PFSYNC_MODVER 1 #define PFLOG_MINVER 1 #define PFLOG_PREFVER PFLOG_MODVER #define PFLOG_MAXVER 1 #define PFSYNC_MINVER 1 #define PFSYNC_PREFVER PFSYNC_MODVER #define PFSYNC_MAXVER 1 #ifdef INET #ifndef INET6 #define PF_INET_ONLY #endif /* ! INET6 */ #endif /* INET */ #ifdef INET6 #ifndef INET #define PF_INET6_ONLY #endif /* ! INET */ #endif /* INET6 */ #ifdef INET #ifdef INET6 #define PF_INET_INET6 #endif /* INET6 */ #endif /* INET */ #else #define PF_INET_INET6 #endif /* _KERNEL */ /* Both IPv4 and IPv6 */ #ifdef PF_INET_INET6 #define PF_AEQ(a, b, c) \ ((c == AF_INET && (a)->addr32[0] == (b)->addr32[0]) || \ (c == AF_INET6 && (a)->addr32[3] == (b)->addr32[3] && \ (a)->addr32[2] == (b)->addr32[2] && \ (a)->addr32[1] == (b)->addr32[1] && \ (a)->addr32[0] == (b)->addr32[0])) \ #define PF_ANEQ(a, b, c) \ ((a)->addr32[0] != (b)->addr32[0] || \ (a)->addr32[1] != (b)->addr32[1] || \ (a)->addr32[2] != (b)->addr32[2] || \ (a)->addr32[3] != (b)->addr32[3]) \ #define PF_AZERO(a, c) \ ((c == AF_INET && !(a)->addr32[0]) || \ (c == AF_INET6 && !(a)->addr32[0] && !(a)->addr32[1] && \ !(a)->addr32[2] && !(a)->addr32[3] )) \ #define PF_MATCHA(n, a, m, b, f) \ pf_match_addr(n, a, m, b, f) #define PF_ACPY(a, b, f) \ pf_addrcpy(a, b, f) #define PF_AINC(a, f) \ pf_addr_inc(a, f) #define PF_POOLMASK(a, b, c, d, f) \ pf_poolmask(a, b, c, d, f) #else /* Just IPv6 */ #ifdef PF_INET6_ONLY #define PF_AEQ(a, b, c) \ ((a)->addr32[3] == (b)->addr32[3] && \ (a)->addr32[2] == (b)->addr32[2] && \ (a)->addr32[1] == (b)->addr32[1] && \ (a)->addr32[0] == (b)->addr32[0]) \ #define PF_ANEQ(a, b, c) \ ((a)->addr32[3] != (b)->addr32[3] || \ (a)->addr32[2] != (b)->addr32[2] || \ (a)->addr32[1] != (b)->addr32[1] || \ (a)->addr32[0] != (b)->addr32[0]) \ #define PF_AZERO(a, c) \ (!(a)->addr32[0] && \ !(a)->addr32[1] && \ !(a)->addr32[2] && \ !(a)->addr32[3] ) \ #define PF_MATCHA(n, a, m, b, f) \ pf_match_addr(n, a, m, b, f) #define PF_ACPY(a, b, f) \ pf_addrcpy(a, b, f) #define PF_AINC(a, f) \ pf_addr_inc(a, f) #define PF_POOLMASK(a, b, c, d, f) \ pf_poolmask(a, b, c, d, f) #else /* Just IPv4 */ #ifdef PF_INET_ONLY #define PF_AEQ(a, b, c) \ ((a)->addr32[0] == (b)->addr32[0]) #define PF_ANEQ(a, b, c) \ ((a)->addr32[0] != (b)->addr32[0]) #define PF_AZERO(a, c) \ (!(a)->addr32[0]) #define PF_MATCHA(n, a, m, b, f) \ pf_match_addr(n, a, m, b, f) #define PF_ACPY(a, b, f) \ (a)->v4.s_addr = (b)->v4.s_addr #define PF_AINC(a, f) \ do { \ (a)->addr32[0] = htonl(ntohl((a)->addr32[0]) + 1); \ } while (0) #define PF_POOLMASK(a, b, c, d, f) \ do { \ (a)->addr32[0] = ((b)->addr32[0] & (c)->addr32[0]) | \ (((c)->addr32[0] ^ 0xffffffff ) & (d)->addr32[0]); \ } while (0) #endif /* PF_INET_ONLY */ #endif /* PF_INET6_ONLY */ #endif /* PF_INET_INET6 */ /* * XXX callers not FIB-aware in our version of pf yet. * OpenBSD fixed it later it seems, 2010/05/07 13:33:16 claudio. */ #define PF_MISMATCHAW(aw, x, af, neg, ifp, rtid) \ ( \ (((aw)->type == PF_ADDR_NOROUTE && \ pf_routable((x), (af), NULL, (rtid))) || \ (((aw)->type == PF_ADDR_URPFFAILED && (ifp) != NULL && \ pf_routable((x), (af), (ifp), (rtid))) || \ ((aw)->type == PF_ADDR_TABLE && \ !pfr_match_addr((aw)->p.tbl, (x), (af))) || \ ((aw)->type == PF_ADDR_DYNIFTL && \ !pfi_match_addr((aw)->p.dyn, (x), (af))) || \ ((aw)->type == PF_ADDR_RANGE && \ !pf_match_addr_range(&(aw)->v.a.addr, \ &(aw)->v.a.mask, (x), (af))) || \ ((aw)->type == PF_ADDR_ADDRMASK && \ !PF_AZERO(&(aw)->v.a.mask, (af)) && \ !PF_MATCHA(0, &(aw)->v.a.addr, \ &(aw)->v.a.mask, (x), (af))))) != \ (neg) \ ) struct pf_rule_uid { uid_t uid[2]; u_int8_t op; }; struct pf_rule_gid { uid_t gid[2]; u_int8_t op; }; struct pf_rule_addr { struct pf_addr_wrap addr; u_int16_t port[2]; u_int8_t neg; u_int8_t port_op; }; struct pf_pooladdr { struct pf_addr_wrap addr; TAILQ_ENTRY(pf_pooladdr) entries; char ifname[IFNAMSIZ]; struct pfi_kif *kif; }; TAILQ_HEAD(pf_palist, pf_pooladdr); struct pf_poolhashkey { union { u_int8_t key8[16]; u_int16_t key16[8]; u_int32_t key32[4]; } pfk; /* 128-bit hash key */ #define key8 pfk.key8 #define key16 pfk.key16 #define key32 pfk.key32 }; struct pf_pool { struct pf_palist list; struct pf_pooladdr *cur; struct pf_poolhashkey key; struct pf_addr counter; int tblidx; u_int16_t proxy_port[2]; u_int8_t opts; }; /* A packed Operating System description for fingerprinting */ typedef u_int32_t pf_osfp_t; #define PF_OSFP_ANY ((pf_osfp_t)0) #define PF_OSFP_UNKNOWN ((pf_osfp_t)-1) #define PF_OSFP_NOMATCH ((pf_osfp_t)-2) struct pf_osfp_entry { SLIST_ENTRY(pf_osfp_entry) fp_entry; pf_osfp_t fp_os; int fp_enflags; #define PF_OSFP_EXPANDED 0x001 /* expanded entry */ #define PF_OSFP_GENERIC 0x002 /* generic signature */ #define PF_OSFP_NODETAIL 0x004 /* no p0f details */ #define PF_OSFP_LEN 32 char fp_class_nm[PF_OSFP_LEN]; char fp_version_nm[PF_OSFP_LEN]; char fp_subtype_nm[PF_OSFP_LEN]; }; #define PF_OSFP_ENTRY_EQ(a, b) \ ((a)->fp_os == (b)->fp_os && \ memcmp((a)->fp_class_nm, (b)->fp_class_nm, PF_OSFP_LEN) == 0 && \ memcmp((a)->fp_version_nm, (b)->fp_version_nm, PF_OSFP_LEN) == 0 && \ memcmp((a)->fp_subtype_nm, (b)->fp_subtype_nm, PF_OSFP_LEN) == 0) /* handle pf_osfp_t packing */ #define _FP_RESERVED_BIT 1 /* For the special negative #defines */ #define _FP_UNUSED_BITS 1 #define _FP_CLASS_BITS 10 /* OS Class (Windows, Linux) */ #define _FP_VERSION_BITS 10 /* OS version (95, 98, NT, 2.4.54, 3.2) */ #define _FP_SUBTYPE_BITS 10 /* patch level (NT SP4, SP3, ECN patch) */ #define PF_OSFP_UNPACK(osfp, class, version, subtype) do { \ (class) = ((osfp) >> (_FP_VERSION_BITS+_FP_SUBTYPE_BITS)) & \ ((1 << _FP_CLASS_BITS) - 1); \ (version) = ((osfp) >> _FP_SUBTYPE_BITS) & \ ((1 << _FP_VERSION_BITS) - 1);\ (subtype) = (osfp) & ((1 << _FP_SUBTYPE_BITS) - 1); \ } while(0) #define PF_OSFP_PACK(osfp, class, version, subtype) do { \ (osfp) = ((class) & ((1 << _FP_CLASS_BITS) - 1)) << (_FP_VERSION_BITS \ + _FP_SUBTYPE_BITS); \ (osfp) |= ((version) & ((1 << _FP_VERSION_BITS) - 1)) << \ _FP_SUBTYPE_BITS; \ (osfp) |= (subtype) & ((1 << _FP_SUBTYPE_BITS) - 1); \ } while(0) /* the fingerprint of an OSes TCP SYN packet */ typedef u_int64_t pf_tcpopts_t; struct pf_os_fingerprint { SLIST_HEAD(pf_osfp_enlist, pf_osfp_entry) fp_oses; /* list of matches */ pf_tcpopts_t fp_tcpopts; /* packed TCP options */ u_int16_t fp_wsize; /* TCP window size */ u_int16_t fp_psize; /* ip->ip_len */ u_int16_t fp_mss; /* TCP MSS */ u_int16_t fp_flags; #define PF_OSFP_WSIZE_MOD 0x0001 /* Window modulus */ #define PF_OSFP_WSIZE_DC 0x0002 /* Window don't care */ #define PF_OSFP_WSIZE_MSS 0x0004 /* Window multiple of MSS */ #define PF_OSFP_WSIZE_MTU 0x0008 /* Window multiple of MTU */ #define PF_OSFP_PSIZE_MOD 0x0010 /* packet size modulus */ #define PF_OSFP_PSIZE_DC 0x0020 /* packet size don't care */ #define PF_OSFP_WSCALE 0x0040 /* TCP window scaling */ #define PF_OSFP_WSCALE_MOD 0x0080 /* TCP window scale modulus */ #define PF_OSFP_WSCALE_DC 0x0100 /* TCP window scale dont-care */ #define PF_OSFP_MSS 0x0200 /* TCP MSS */ #define PF_OSFP_MSS_MOD 0x0400 /* TCP MSS modulus */ #define PF_OSFP_MSS_DC 0x0800 /* TCP MSS dont-care */ #define PF_OSFP_DF 0x1000 /* IPv4 don't fragment bit */ #define PF_OSFP_TS0 0x2000 /* Zero timestamp */ #define PF_OSFP_INET6 0x4000 /* IPv6 */ u_int8_t fp_optcnt; /* TCP option count */ u_int8_t fp_wscale; /* TCP window scaling */ u_int8_t fp_ttl; /* IPv4 TTL */ #define PF_OSFP_MAXTTL_OFFSET 40 /* TCP options packing */ #define PF_OSFP_TCPOPT_NOP 0x0 /* TCP NOP option */ #define PF_OSFP_TCPOPT_WSCALE 0x1 /* TCP window scaling option */ #define PF_OSFP_TCPOPT_MSS 0x2 /* TCP max segment size opt */ #define PF_OSFP_TCPOPT_SACK 0x3 /* TCP SACK OK option */ #define PF_OSFP_TCPOPT_TS 0x4 /* TCP timestamp option */ #define PF_OSFP_TCPOPT_BITS 3 /* bits used by each option */ #define PF_OSFP_MAX_OPTS \ (sizeof(((struct pf_os_fingerprint *)0)->fp_tcpopts) * 8) \ / PF_OSFP_TCPOPT_BITS SLIST_ENTRY(pf_os_fingerprint) fp_next; }; struct pf_osfp_ioctl { struct pf_osfp_entry fp_os; pf_tcpopts_t fp_tcpopts; /* packed TCP options */ u_int16_t fp_wsize; /* TCP window size */ u_int16_t fp_psize; /* ip->ip_len */ u_int16_t fp_mss; /* TCP MSS */ u_int16_t fp_flags; u_int8_t fp_optcnt; /* TCP option count */ u_int8_t fp_wscale; /* TCP window scaling */ u_int8_t fp_ttl; /* IPv4 TTL */ int fp_getnum; /* DIOCOSFPGET number */ }; union pf_rule_ptr { struct pf_rule *ptr; u_int32_t nr; }; #define PF_ANCHOR_NAME_SIZE 64 struct pf_rule { struct pf_rule_addr src; struct pf_rule_addr dst; #define PF_SKIP_IFP 0 #define PF_SKIP_DIR 1 #define PF_SKIP_AF 2 #define PF_SKIP_PROTO 3 #define PF_SKIP_SRC_ADDR 4 #define PF_SKIP_SRC_PORT 5 #define PF_SKIP_DST_ADDR 6 #define PF_SKIP_DST_PORT 7 #define PF_SKIP_COUNT 8 union pf_rule_ptr skip[PF_SKIP_COUNT]; #define PF_RULE_LABEL_SIZE 64 char label[PF_RULE_LABEL_SIZE]; char ifname[IFNAMSIZ]; char qname[PF_QNAME_SIZE]; char pqname[PF_QNAME_SIZE]; #define PF_TAG_NAME_SIZE 64 char tagname[PF_TAG_NAME_SIZE]; char match_tagname[PF_TAG_NAME_SIZE]; char overload_tblname[PF_TABLE_NAME_SIZE]; TAILQ_ENTRY(pf_rule) entries; struct pf_pool rpool; u_int64_t evaluations; u_int64_t packets[2]; u_int64_t bytes[2]; struct pfi_kif *kif; struct pf_anchor *anchor; struct pfr_ktable *overload_tbl; pf_osfp_t os_fingerprint; int rtableid; u_int32_t timeout[PFTM_MAX]; u_int32_t max_states; u_int32_t max_src_nodes; u_int32_t max_src_states; u_int32_t max_src_conn; struct { u_int32_t limit; u_int32_t seconds; } max_src_conn_rate; u_int32_t qid; u_int32_t pqid; u_int32_t rt_listid; u_int32_t nr; u_int32_t prob; uid_t cuid; pid_t cpid; counter_u64_t states_cur; counter_u64_t states_tot; counter_u64_t src_nodes; u_int16_t return_icmp; u_int16_t return_icmp6; u_int16_t max_mss; u_int16_t tag; u_int16_t match_tag; u_int16_t spare2; /* netgraph */ struct pf_rule_uid uid; struct pf_rule_gid gid; u_int32_t rule_flag; u_int8_t action; u_int8_t direction; u_int8_t log; u_int8_t logif; u_int8_t quick; u_int8_t ifnot; u_int8_t match_tag_not; u_int8_t natpass; #define PF_STATE_NORMAL 0x1 #define PF_STATE_MODULATE 0x2 #define PF_STATE_SYNPROXY 0x3 u_int8_t keep_state; sa_family_t af; u_int8_t proto; u_int8_t type; u_int8_t code; u_int8_t flags; u_int8_t flagset; u_int8_t min_ttl; u_int8_t allow_opts; u_int8_t rt; u_int8_t return_ttl; u_int8_t tos; u_int8_t set_tos; u_int8_t anchor_relative; u_int8_t anchor_wildcard; #define PF_FLUSH 0x01 #define PF_FLUSH_GLOBAL 0x02 u_int8_t flush; struct { struct pf_addr addr; u_int16_t port; } divert; uint64_t u_states_cur; uint64_t u_states_tot; uint64_t u_src_nodes; }; /* rule flags */ #define PFRULE_DROP 0x0000 #define PFRULE_RETURNRST 0x0001 #define PFRULE_FRAGMENT 0x0002 #define PFRULE_RETURNICMP 0x0004 #define PFRULE_RETURN 0x0008 #define PFRULE_NOSYNC 0x0010 #define PFRULE_SRCTRACK 0x0020 /* track source states */ #define PFRULE_RULESRCTRACK 0x0040 /* per rule */ #define PFRULE_REFS 0x0080 /* rule has references */ /* scrub flags */ #define PFRULE_NODF 0x0100 -#define PFRULE_FRAGCROP 0x0200 /* non-buffering frag cache */ -#define PFRULE_FRAGDROP 0x0400 /* drop funny fragments */ #define PFRULE_RANDOMID 0x0800 #define PFRULE_REASSEMBLE_TCP 0x1000 #define PFRULE_SET_TOS 0x2000 /* rule flags again */ #define PFRULE_IFBOUND 0x00010000 /* if-bound */ #define PFRULE_STATESLOPPY 0x00020000 /* sloppy state tracking */ #define PFSTATE_HIWAT 10000 /* default state table size */ #define PFSTATE_ADAPT_START 6000 /* default adaptive timeout start */ #define PFSTATE_ADAPT_END 12000 /* default adaptive timeout end */ struct pf_threshold { u_int32_t limit; #define PF_THRESHOLD_MULT 1000 #define PF_THRESHOLD_MAX 0xffffffff / PF_THRESHOLD_MULT u_int32_t seconds; u_int32_t count; u_int32_t last; }; struct pf_src_node { LIST_ENTRY(pf_src_node) entry; struct pf_addr addr; struct pf_addr raddr; union pf_rule_ptr rule; struct pfi_kif *kif; u_int64_t bytes[2]; u_int64_t packets[2]; u_int32_t states; u_int32_t conn; struct pf_threshold conn_rate; u_int32_t creation; u_int32_t expire; sa_family_t af; u_int8_t ruletype; }; #define PFSNODE_HIWAT 10000 /* default source node table size */ struct pf_state_scrub { struct timeval pfss_last; /* time received last packet */ u_int32_t pfss_tsecr; /* last echoed timestamp */ u_int32_t pfss_tsval; /* largest timestamp */ u_int32_t pfss_tsval0; /* original timestamp */ u_int16_t pfss_flags; #define PFSS_TIMESTAMP 0x0001 /* modulate timestamp */ #define PFSS_PAWS 0x0010 /* stricter PAWS checks */ #define PFSS_PAWS_IDLED 0x0020 /* was idle too long. no PAWS */ #define PFSS_DATA_TS 0x0040 /* timestamp on data packets */ #define PFSS_DATA_NOTS 0x0080 /* no timestamp on data packets */ u_int8_t pfss_ttl; /* stashed TTL */ u_int8_t pad; u_int32_t pfss_ts_mod; /* timestamp modulation */ }; struct pf_state_host { struct pf_addr addr; u_int16_t port; u_int16_t pad; }; struct pf_state_peer { struct pf_state_scrub *scrub; /* state is scrubbed */ u_int32_t seqlo; /* Max sequence number sent */ u_int32_t seqhi; /* Max the other end ACKd + win */ u_int32_t seqdiff; /* Sequence number modulator */ u_int16_t max_win; /* largest window (pre scaling) */ u_int16_t mss; /* Maximum segment size option */ u_int8_t state; /* active state level */ u_int8_t wscale; /* window scaling factor */ u_int8_t tcp_est; /* Did we reach TCPS_ESTABLISHED */ u_int8_t pad[1]; }; /* Keep synced with struct pf_state_key. */ struct pf_state_key_cmp { struct pf_addr addr[2]; u_int16_t port[2]; sa_family_t af; u_int8_t proto; u_int8_t pad[2]; }; struct pf_state_key { struct pf_addr addr[2]; u_int16_t port[2]; sa_family_t af; u_int8_t proto; u_int8_t pad[2]; LIST_ENTRY(pf_state_key) entry; TAILQ_HEAD(, pf_state) states[2]; }; /* Keep synced with struct pf_state. */ struct pf_state_cmp { u_int64_t id; u_int32_t creatorid; u_int8_t direction; u_int8_t pad[3]; }; struct pf_state { u_int64_t id; u_int32_t creatorid; u_int8_t direction; u_int8_t pad[3]; u_int refs; TAILQ_ENTRY(pf_state) sync_list; TAILQ_ENTRY(pf_state) key_list[2]; LIST_ENTRY(pf_state) entry; struct pf_state_peer src; struct pf_state_peer dst; union pf_rule_ptr rule; union pf_rule_ptr anchor; union pf_rule_ptr nat_rule; struct pf_addr rt_addr; struct pf_state_key *key[2]; /* addresses stack and wire */ struct pfi_kif *kif; struct pfi_kif *rt_kif; struct pf_src_node *src_node; struct pf_src_node *nat_src_node; u_int64_t packets[2]; u_int64_t bytes[2]; u_int32_t creation; u_int32_t expire; u_int32_t pfsync_time; u_int16_t tag; u_int8_t log; u_int8_t state_flags; #define PFSTATE_ALLOWOPTS 0x01 #define PFSTATE_SLOPPY 0x02 /* was PFSTATE_PFLOW 0x04 */ #define PFSTATE_NOSYNC 0x08 #define PFSTATE_ACK 0x10 u_int8_t timeout; u_int8_t sync_state; /* PFSYNC_S_x */ /* XXX */ u_int8_t sync_updates; u_int8_t _tail[3]; }; /* * Unified state structures for pulling states out of the kernel * used by pfsync(4) and the pf(4) ioctl. */ struct pfsync_state_scrub { u_int16_t pfss_flags; u_int8_t pfss_ttl; /* stashed TTL */ #define PFSYNC_SCRUB_FLAG_VALID 0x01 u_int8_t scrub_flag; u_int32_t pfss_ts_mod; /* timestamp modulation */ } __packed; struct pfsync_state_peer { struct pfsync_state_scrub scrub; /* state is scrubbed */ u_int32_t seqlo; /* Max sequence number sent */ u_int32_t seqhi; /* Max the other end ACKd + win */ u_int32_t seqdiff; /* Sequence number modulator */ u_int16_t max_win; /* largest window (pre scaling) */ u_int16_t mss; /* Maximum segment size option */ u_int8_t state; /* active state level */ u_int8_t wscale; /* window scaling factor */ u_int8_t pad[6]; } __packed; struct pfsync_state_key { struct pf_addr addr[2]; u_int16_t port[2]; }; struct pfsync_state { u_int64_t id; char ifname[IFNAMSIZ]; struct pfsync_state_key key[2]; struct pfsync_state_peer src; struct pfsync_state_peer dst; struct pf_addr rt_addr; u_int32_t rule; u_int32_t anchor; u_int32_t nat_rule; u_int32_t creation; u_int32_t expire; u_int32_t packets[2][2]; u_int32_t bytes[2][2]; u_int32_t creatorid; sa_family_t af; u_int8_t proto; u_int8_t direction; u_int8_t __spare[2]; u_int8_t log; u_int8_t state_flags; u_int8_t timeout; u_int8_t sync_flags; u_int8_t updates; } __packed; #ifdef _KERNEL /* pfsync */ typedef int pfsync_state_import_t(struct pfsync_state *, u_int8_t); typedef void pfsync_insert_state_t(struct pf_state *); typedef void pfsync_update_state_t(struct pf_state *); typedef void pfsync_delete_state_t(struct pf_state *); typedef void pfsync_clear_states_t(u_int32_t, const char *); typedef int pfsync_defer_t(struct pf_state *, struct mbuf *); extern pfsync_state_import_t *pfsync_state_import_ptr; extern pfsync_insert_state_t *pfsync_insert_state_ptr; extern pfsync_update_state_t *pfsync_update_state_ptr; extern pfsync_delete_state_t *pfsync_delete_state_ptr; extern pfsync_clear_states_t *pfsync_clear_states_ptr; extern pfsync_defer_t *pfsync_defer_ptr; void pfsync_state_export(struct pfsync_state *, struct pf_state *); /* pflog */ struct pf_ruleset; struct pf_pdesc; typedef int pflog_packet_t(struct pfi_kif *, struct mbuf *, sa_family_t, u_int8_t, u_int8_t, struct pf_rule *, struct pf_rule *, struct pf_ruleset *, struct pf_pdesc *, int); extern pflog_packet_t *pflog_packet_ptr; #define V_pf_end_threads VNET(pf_end_threads) #endif /* _KERNEL */ #define PFSYNC_FLAG_SRCNODE 0x04 #define PFSYNC_FLAG_NATSRCNODE 0x08 /* for copies to/from network byte order */ /* ioctl interface also uses network byte order */ #define pf_state_peer_hton(s,d) do { \ (d)->seqlo = htonl((s)->seqlo); \ (d)->seqhi = htonl((s)->seqhi); \ (d)->seqdiff = htonl((s)->seqdiff); \ (d)->max_win = htons((s)->max_win); \ (d)->mss = htons((s)->mss); \ (d)->state = (s)->state; \ (d)->wscale = (s)->wscale; \ if ((s)->scrub) { \ (d)->scrub.pfss_flags = \ htons((s)->scrub->pfss_flags & PFSS_TIMESTAMP); \ (d)->scrub.pfss_ttl = (s)->scrub->pfss_ttl; \ (d)->scrub.pfss_ts_mod = htonl((s)->scrub->pfss_ts_mod);\ (d)->scrub.scrub_flag = PFSYNC_SCRUB_FLAG_VALID; \ } \ } while (0) #define pf_state_peer_ntoh(s,d) do { \ (d)->seqlo = ntohl((s)->seqlo); \ (d)->seqhi = ntohl((s)->seqhi); \ (d)->seqdiff = ntohl((s)->seqdiff); \ (d)->max_win = ntohs((s)->max_win); \ (d)->mss = ntohs((s)->mss); \ (d)->state = (s)->state; \ (d)->wscale = (s)->wscale; \ if ((s)->scrub.scrub_flag == PFSYNC_SCRUB_FLAG_VALID && \ (d)->scrub != NULL) { \ (d)->scrub->pfss_flags = \ ntohs((s)->scrub.pfss_flags) & PFSS_TIMESTAMP; \ (d)->scrub->pfss_ttl = (s)->scrub.pfss_ttl; \ (d)->scrub->pfss_ts_mod = ntohl((s)->scrub.pfss_ts_mod);\ } \ } while (0) #define pf_state_counter_hton(s,d) do { \ d[0] = htonl((s>>32)&0xffffffff); \ d[1] = htonl(s&0xffffffff); \ } while (0) #define pf_state_counter_from_pfsync(s) \ (((u_int64_t)(s[0])<<32) | (u_int64_t)(s[1])) #define pf_state_counter_ntoh(s,d) do { \ d = ntohl(s[0]); \ d = d<<32; \ d += ntohl(s[1]); \ } while (0) TAILQ_HEAD(pf_rulequeue, pf_rule); struct pf_anchor; struct pf_ruleset { struct { struct pf_rulequeue queues[2]; struct { struct pf_rulequeue *ptr; struct pf_rule **ptr_array; u_int32_t rcount; u_int32_t ticket; int open; } active, inactive; } rules[PF_RULESET_MAX]; struct pf_anchor *anchor; u_int32_t tticket; int tables; int topen; }; RB_HEAD(pf_anchor_global, pf_anchor); RB_HEAD(pf_anchor_node, pf_anchor); struct pf_anchor { RB_ENTRY(pf_anchor) entry_global; RB_ENTRY(pf_anchor) entry_node; struct pf_anchor *parent; struct pf_anchor_node children; char name[PF_ANCHOR_NAME_SIZE]; char path[MAXPATHLEN]; struct pf_ruleset ruleset; int refcnt; /* anchor rules */ int match; /* XXX: used for pfctl black magic */ }; RB_PROTOTYPE(pf_anchor_global, pf_anchor, entry_global, pf_anchor_compare); RB_PROTOTYPE(pf_anchor_node, pf_anchor, entry_node, pf_anchor_compare); #define PF_RESERVED_ANCHOR "_pf" #define PFR_TFLAG_PERSIST 0x00000001 #define PFR_TFLAG_CONST 0x00000002 #define PFR_TFLAG_ACTIVE 0x00000004 #define PFR_TFLAG_INACTIVE 0x00000008 #define PFR_TFLAG_REFERENCED 0x00000010 #define PFR_TFLAG_REFDANCHOR 0x00000020 #define PFR_TFLAG_COUNTERS 0x00000040 /* Adjust masks below when adding flags. */ #define PFR_TFLAG_USRMASK (PFR_TFLAG_PERSIST | \ PFR_TFLAG_CONST | \ PFR_TFLAG_COUNTERS) #define PFR_TFLAG_SETMASK (PFR_TFLAG_ACTIVE | \ PFR_TFLAG_INACTIVE | \ PFR_TFLAG_REFERENCED | \ PFR_TFLAG_REFDANCHOR) #define PFR_TFLAG_ALLMASK (PFR_TFLAG_PERSIST | \ PFR_TFLAG_CONST | \ PFR_TFLAG_ACTIVE | \ PFR_TFLAG_INACTIVE | \ PFR_TFLAG_REFERENCED | \ PFR_TFLAG_REFDANCHOR | \ PFR_TFLAG_COUNTERS) struct pf_anchor_stackframe; struct pfr_table { char pfrt_anchor[MAXPATHLEN]; char pfrt_name[PF_TABLE_NAME_SIZE]; u_int32_t pfrt_flags; u_int8_t pfrt_fback; }; enum { PFR_FB_NONE, PFR_FB_MATCH, PFR_FB_ADDED, PFR_FB_DELETED, PFR_FB_CHANGED, PFR_FB_CLEARED, PFR_FB_DUPLICATE, PFR_FB_NOTMATCH, PFR_FB_CONFLICT, PFR_FB_NOCOUNT, PFR_FB_MAX }; struct pfr_addr { union { struct in_addr _pfra_ip4addr; struct in6_addr _pfra_ip6addr; } pfra_u; u_int8_t pfra_af; u_int8_t pfra_net; u_int8_t pfra_not; u_int8_t pfra_fback; }; #define pfra_ip4addr pfra_u._pfra_ip4addr #define pfra_ip6addr pfra_u._pfra_ip6addr enum { PFR_DIR_IN, PFR_DIR_OUT, PFR_DIR_MAX }; enum { PFR_OP_BLOCK, PFR_OP_PASS, PFR_OP_ADDR_MAX, PFR_OP_TABLE_MAX }; #define PFR_OP_XPASS PFR_OP_ADDR_MAX struct pfr_astats { struct pfr_addr pfras_a; u_int64_t pfras_packets[PFR_DIR_MAX][PFR_OP_ADDR_MAX]; u_int64_t pfras_bytes[PFR_DIR_MAX][PFR_OP_ADDR_MAX]; long pfras_tzero; }; enum { PFR_REFCNT_RULE, PFR_REFCNT_ANCHOR, PFR_REFCNT_MAX }; struct pfr_tstats { struct pfr_table pfrts_t; u_int64_t pfrts_packets[PFR_DIR_MAX][PFR_OP_TABLE_MAX]; u_int64_t pfrts_bytes[PFR_DIR_MAX][PFR_OP_TABLE_MAX]; u_int64_t pfrts_match; u_int64_t pfrts_nomatch; long pfrts_tzero; int pfrts_cnt; int pfrts_refcnt[PFR_REFCNT_MAX]; }; #define pfrts_name pfrts_t.pfrt_name #define pfrts_flags pfrts_t.pfrt_flags #ifndef _SOCKADDR_UNION_DEFINED #define _SOCKADDR_UNION_DEFINED union sockaddr_union { struct sockaddr sa; struct sockaddr_in sin; struct sockaddr_in6 sin6; }; #endif /* _SOCKADDR_UNION_DEFINED */ struct pfr_kcounters { u_int64_t pfrkc_packets[PFR_DIR_MAX][PFR_OP_ADDR_MAX]; u_int64_t pfrkc_bytes[PFR_DIR_MAX][PFR_OP_ADDR_MAX]; }; SLIST_HEAD(pfr_kentryworkq, pfr_kentry); struct pfr_kentry { struct radix_node pfrke_node[2]; union sockaddr_union pfrke_sa; SLIST_ENTRY(pfr_kentry) pfrke_workq; struct pfr_kcounters *pfrke_counters; long pfrke_tzero; u_int8_t pfrke_af; u_int8_t pfrke_net; u_int8_t pfrke_not; u_int8_t pfrke_mark; }; SLIST_HEAD(pfr_ktableworkq, pfr_ktable); RB_HEAD(pfr_ktablehead, pfr_ktable); struct pfr_ktable { struct pfr_tstats pfrkt_ts; RB_ENTRY(pfr_ktable) pfrkt_tree; SLIST_ENTRY(pfr_ktable) pfrkt_workq; struct radix_node_head *pfrkt_ip4; struct radix_node_head *pfrkt_ip6; struct pfr_ktable *pfrkt_shadow; struct pfr_ktable *pfrkt_root; struct pf_ruleset *pfrkt_rs; long pfrkt_larg; int pfrkt_nflags; }; #define pfrkt_t pfrkt_ts.pfrts_t #define pfrkt_name pfrkt_t.pfrt_name #define pfrkt_anchor pfrkt_t.pfrt_anchor #define pfrkt_ruleset pfrkt_t.pfrt_ruleset #define pfrkt_flags pfrkt_t.pfrt_flags #define pfrkt_cnt pfrkt_ts.pfrts_cnt #define pfrkt_refcnt pfrkt_ts.pfrts_refcnt #define pfrkt_packets pfrkt_ts.pfrts_packets #define pfrkt_bytes pfrkt_ts.pfrts_bytes #define pfrkt_match pfrkt_ts.pfrts_match #define pfrkt_nomatch pfrkt_ts.pfrts_nomatch #define pfrkt_tzero pfrkt_ts.pfrts_tzero /* keep synced with pfi_kif, used in RB_FIND */ struct pfi_kif_cmp { char pfik_name[IFNAMSIZ]; }; struct pfi_kif { char pfik_name[IFNAMSIZ]; union { RB_ENTRY(pfi_kif) _pfik_tree; LIST_ENTRY(pfi_kif) _pfik_list; } _pfik_glue; #define pfik_tree _pfik_glue._pfik_tree #define pfik_list _pfik_glue._pfik_list u_int64_t pfik_packets[2][2][2]; u_int64_t pfik_bytes[2][2][2]; u_int32_t pfik_tzero; u_int pfik_flags; struct ifnet *pfik_ifp; struct ifg_group *pfik_group; u_int pfik_rulerefs; TAILQ_HEAD(, pfi_dynaddr) pfik_dynaddrs; }; #define PFI_IFLAG_REFS 0x0001 /* has state references */ #define PFI_IFLAG_SKIP 0x0100 /* skip filtering on interface */ struct pf_pdesc { struct { int done; uid_t uid; gid_t gid; } lookup; u_int64_t tot_len; /* Make Mickey money */ union { struct tcphdr *tcp; struct udphdr *udp; struct icmp *icmp; #ifdef INET6 struct icmp6_hdr *icmp6; #endif /* INET6 */ void *any; } hdr; struct pf_rule *nat_rule; /* nat/rdr rule applied to packet */ struct pf_addr *src; /* src address */ struct pf_addr *dst; /* dst address */ u_int16_t *sport; u_int16_t *dport; struct pf_mtag *pf_mtag; u_int32_t p_len; /* total length of payload */ u_int16_t *ip_sum; u_int16_t *proto_sum; u_int16_t flags; /* Let SCRUB trigger behavior in * state code. Easier than tags */ #define PFDESC_TCP_NORM 0x0001 /* TCP shall be statefully scrubbed */ #define PFDESC_IP_REAS 0x0002 /* IP frags would've been reassembled */ sa_family_t af; u_int8_t proto; u_int8_t tos; u_int8_t dir; /* direction */ u_int8_t sidx; /* key index for source */ u_int8_t didx; /* key index for destination */ }; /* flags for RDR options */ #define PF_DPORT_RANGE 0x01 /* Dest port uses range */ #define PF_RPORT_RANGE 0x02 /* RDR'ed port uses range */ /* UDP state enumeration */ #define PFUDPS_NO_TRAFFIC 0 #define PFUDPS_SINGLE 1 #define PFUDPS_MULTIPLE 2 #define PFUDPS_NSTATES 3 /* number of state levels */ #define PFUDPS_NAMES { \ "NO_TRAFFIC", \ "SINGLE", \ "MULTIPLE", \ NULL \ } /* Other protocol state enumeration */ #define PFOTHERS_NO_TRAFFIC 0 #define PFOTHERS_SINGLE 1 #define PFOTHERS_MULTIPLE 2 #define PFOTHERS_NSTATES 3 /* number of state levels */ #define PFOTHERS_NAMES { \ "NO_TRAFFIC", \ "SINGLE", \ "MULTIPLE", \ NULL \ } #define ACTION_SET(a, x) \ do { \ if ((a) != NULL) \ *(a) = (x); \ } while (0) #define REASON_SET(a, x) \ do { \ if ((a) != NULL) \ *(a) = (x); \ if (x < PFRES_MAX) \ counter_u64_add(V_pf_status.counters[x], 1); \ } while (0) struct pf_kstatus { counter_u64_t counters[PFRES_MAX]; /* reason for passing/dropping */ counter_u64_t lcounters[LCNT_MAX]; /* limit counters */ counter_u64_t fcounters[FCNT_MAX]; /* state operation counters */ counter_u64_t scounters[SCNT_MAX]; /* src_node operation counters */ uint32_t states; uint32_t src_nodes; uint32_t running; uint32_t since; uint32_t debug; uint32_t hostid; char ifname[IFNAMSIZ]; uint8_t pf_chksum[PF_MD5_DIGEST_LENGTH]; }; struct pf_divert { union { struct in_addr ipv4; struct in6_addr ipv6; } addr; u_int16_t port; }; #define PFFRAG_FRENT_HIWAT 5000 /* Number of fragment entries */ #define PFR_KENTRY_HIWAT 200000 /* Number of table entries */ /* * ioctl parameter structures */ struct pfioc_pooladdr { u_int32_t action; u_int32_t ticket; u_int32_t nr; u_int32_t r_num; u_int8_t r_action; u_int8_t r_last; u_int8_t af; char anchor[MAXPATHLEN]; struct pf_pooladdr addr; }; struct pfioc_rule { u_int32_t action; u_int32_t ticket; u_int32_t pool_ticket; u_int32_t nr; char anchor[MAXPATHLEN]; char anchor_call[MAXPATHLEN]; struct pf_rule rule; }; struct pfioc_natlook { struct pf_addr saddr; struct pf_addr daddr; struct pf_addr rsaddr; struct pf_addr rdaddr; u_int16_t sport; u_int16_t dport; u_int16_t rsport; u_int16_t rdport; sa_family_t af; u_int8_t proto; u_int8_t direction; }; struct pfioc_state { struct pfsync_state state; }; struct pfioc_src_node_kill { sa_family_t psnk_af; struct pf_rule_addr psnk_src; struct pf_rule_addr psnk_dst; u_int psnk_killed; }; struct pfioc_state_kill { struct pf_state_cmp psk_pfcmp; sa_family_t psk_af; int psk_proto; struct pf_rule_addr psk_src; struct pf_rule_addr psk_dst; char psk_ifname[IFNAMSIZ]; char psk_label[PF_RULE_LABEL_SIZE]; u_int psk_killed; }; struct pfioc_states { int ps_len; union { caddr_t psu_buf; struct pfsync_state *psu_states; } ps_u; #define ps_buf ps_u.psu_buf #define ps_states ps_u.psu_states }; struct pfioc_src_nodes { int psn_len; union { caddr_t psu_buf; struct pf_src_node *psu_src_nodes; } psn_u; #define psn_buf psn_u.psu_buf #define psn_src_nodes psn_u.psu_src_nodes }; struct pfioc_if { char ifname[IFNAMSIZ]; }; struct pfioc_tm { int timeout; int seconds; }; struct pfioc_limit { int index; unsigned limit; }; struct pfioc_altq { u_int32_t action; u_int32_t ticket; u_int32_t nr; struct pf_altq altq; }; struct pfioc_qstats { u_int32_t ticket; u_int32_t nr; void *buf; int nbytes; u_int8_t scheduler; }; struct pfioc_ruleset { u_int32_t nr; char path[MAXPATHLEN]; char name[PF_ANCHOR_NAME_SIZE]; }; #define PF_RULESET_ALTQ (PF_RULESET_MAX) #define PF_RULESET_TABLE (PF_RULESET_MAX+1) struct pfioc_trans { int size; /* number of elements */ int esize; /* size of each element in bytes */ struct pfioc_trans_e { int rs_num; char anchor[MAXPATHLEN]; u_int32_t ticket; } *array; }; #define PFR_FLAG_ATOMIC 0x00000001 /* unused */ #define PFR_FLAG_DUMMY 0x00000002 #define PFR_FLAG_FEEDBACK 0x00000004 #define PFR_FLAG_CLSTATS 0x00000008 #define PFR_FLAG_ADDRSTOO 0x00000010 #define PFR_FLAG_REPLACE 0x00000020 #define PFR_FLAG_ALLRSETS 0x00000040 #define PFR_FLAG_ALLMASK 0x0000007F #ifdef _KERNEL #define PFR_FLAG_USERIOCTL 0x10000000 #endif struct pfioc_table { struct pfr_table pfrio_table; void *pfrio_buffer; int pfrio_esize; int pfrio_size; int pfrio_size2; int pfrio_nadd; int pfrio_ndel; int pfrio_nchange; int pfrio_flags; u_int32_t pfrio_ticket; }; #define pfrio_exists pfrio_nadd #define pfrio_nzero pfrio_nadd #define pfrio_nmatch pfrio_nadd #define pfrio_naddr pfrio_size2 #define pfrio_setflag pfrio_size2 #define pfrio_clrflag pfrio_nadd struct pfioc_iface { char pfiio_name[IFNAMSIZ]; void *pfiio_buffer; int pfiio_esize; int pfiio_size; int pfiio_nzero; int pfiio_flags; }; /* * ioctl operations */ #define DIOCSTART _IO ('D', 1) #define DIOCSTOP _IO ('D', 2) #define DIOCADDRULE _IOWR('D', 4, struct pfioc_rule) #define DIOCGETRULES _IOWR('D', 6, struct pfioc_rule) #define DIOCGETRULE _IOWR('D', 7, struct pfioc_rule) /* XXX cut 8 - 17 */ #define DIOCCLRSTATES _IOWR('D', 18, struct pfioc_state_kill) #define DIOCGETSTATE _IOWR('D', 19, struct pfioc_state) #define DIOCSETSTATUSIF _IOWR('D', 20, struct pfioc_if) #define DIOCGETSTATUS _IOWR('D', 21, struct pf_status) #define DIOCCLRSTATUS _IO ('D', 22) #define DIOCNATLOOK _IOWR('D', 23, struct pfioc_natlook) #define DIOCSETDEBUG _IOWR('D', 24, u_int32_t) #define DIOCGETSTATES _IOWR('D', 25, struct pfioc_states) #define DIOCCHANGERULE _IOWR('D', 26, struct pfioc_rule) /* XXX cut 26 - 28 */ #define DIOCSETTIMEOUT _IOWR('D', 29, struct pfioc_tm) #define DIOCGETTIMEOUT _IOWR('D', 30, struct pfioc_tm) #define DIOCADDSTATE _IOWR('D', 37, struct pfioc_state) #define DIOCCLRRULECTRS _IO ('D', 38) #define DIOCGETLIMIT _IOWR('D', 39, struct pfioc_limit) #define DIOCSETLIMIT _IOWR('D', 40, struct pfioc_limit) #define DIOCKILLSTATES _IOWR('D', 41, struct pfioc_state_kill) #define DIOCSTARTALTQ _IO ('D', 42) #define DIOCSTOPALTQ _IO ('D', 43) #define DIOCADDALTQ _IOWR('D', 45, struct pfioc_altq) #define DIOCGETALTQS _IOWR('D', 47, struct pfioc_altq) #define DIOCGETALTQ _IOWR('D', 48, struct pfioc_altq) #define DIOCCHANGEALTQ _IOWR('D', 49, struct pfioc_altq) #define DIOCGETQSTATS _IOWR('D', 50, struct pfioc_qstats) #define DIOCBEGINADDRS _IOWR('D', 51, struct pfioc_pooladdr) #define DIOCADDADDR _IOWR('D', 52, struct pfioc_pooladdr) #define DIOCGETADDRS _IOWR('D', 53, struct pfioc_pooladdr) #define DIOCGETADDR _IOWR('D', 54, struct pfioc_pooladdr) #define DIOCCHANGEADDR _IOWR('D', 55, struct pfioc_pooladdr) /* XXX cut 55 - 57 */ #define DIOCGETRULESETS _IOWR('D', 58, struct pfioc_ruleset) #define DIOCGETRULESET _IOWR('D', 59, struct pfioc_ruleset) #define DIOCRCLRTABLES _IOWR('D', 60, struct pfioc_table) #define DIOCRADDTABLES _IOWR('D', 61, struct pfioc_table) #define DIOCRDELTABLES _IOWR('D', 62, struct pfioc_table) #define DIOCRGETTABLES _IOWR('D', 63, struct pfioc_table) #define DIOCRGETTSTATS _IOWR('D', 64, struct pfioc_table) #define DIOCRCLRTSTATS _IOWR('D', 65, struct pfioc_table) #define DIOCRCLRADDRS _IOWR('D', 66, struct pfioc_table) #define DIOCRADDADDRS _IOWR('D', 67, struct pfioc_table) #define DIOCRDELADDRS _IOWR('D', 68, struct pfioc_table) #define DIOCRSETADDRS _IOWR('D', 69, struct pfioc_table) #define DIOCRGETADDRS _IOWR('D', 70, struct pfioc_table) #define DIOCRGETASTATS _IOWR('D', 71, struct pfioc_table) #define DIOCRCLRASTATS _IOWR('D', 72, struct pfioc_table) #define DIOCRTSTADDRS _IOWR('D', 73, struct pfioc_table) #define DIOCRSETTFLAGS _IOWR('D', 74, struct pfioc_table) #define DIOCRINADEFINE _IOWR('D', 77, struct pfioc_table) #define DIOCOSFPFLUSH _IO('D', 78) #define DIOCOSFPADD _IOWR('D', 79, struct pf_osfp_ioctl) #define DIOCOSFPGET _IOWR('D', 80, struct pf_osfp_ioctl) #define DIOCXBEGIN _IOWR('D', 81, struct pfioc_trans) #define DIOCXCOMMIT _IOWR('D', 82, struct pfioc_trans) #define DIOCXROLLBACK _IOWR('D', 83, struct pfioc_trans) #define DIOCGETSRCNODES _IOWR('D', 84, struct pfioc_src_nodes) #define DIOCCLRSRCNODES _IO('D', 85) #define DIOCSETHOSTID _IOWR('D', 86, u_int32_t) #define DIOCIGETIFACES _IOWR('D', 87, struct pfioc_iface) #define DIOCSETIFFLAG _IOWR('D', 89, struct pfioc_iface) #define DIOCCLRIFFLAG _IOWR('D', 90, struct pfioc_iface) #define DIOCKILLSRCNODES _IOWR('D', 91, struct pfioc_src_node_kill) struct pf_ifspeed { char ifname[IFNAMSIZ]; u_int32_t baudrate; }; #define DIOCGIFSPEED _IOWR('D', 92, struct pf_ifspeed) #ifdef _KERNEL LIST_HEAD(pf_src_node_list, pf_src_node); struct pf_srchash { struct pf_src_node_list nodes; struct mtx lock; }; struct pf_keyhash { LIST_HEAD(, pf_state_key) keys; struct mtx lock; }; struct pf_idhash { LIST_HEAD(, pf_state) states; struct mtx lock; }; extern u_long pf_hashmask; extern u_long pf_srchashmask; #define PF_HASHSIZ (32768) VNET_DECLARE(struct pf_keyhash *, pf_keyhash); VNET_DECLARE(struct pf_idhash *, pf_idhash); #define V_pf_keyhash VNET(pf_keyhash) #define V_pf_idhash VNET(pf_idhash) VNET_DECLARE(struct pf_srchash *, pf_srchash); #define V_pf_srchash VNET(pf_srchash) #define PF_IDHASH(s) (be64toh((s)->id) % (pf_hashmask + 1)) VNET_DECLARE(void *, pf_swi_cookie); #define V_pf_swi_cookie VNET(pf_swi_cookie) VNET_DECLARE(uint64_t, pf_stateid[MAXCPU]); #define V_pf_stateid VNET(pf_stateid) TAILQ_HEAD(pf_altqqueue, pf_altq); VNET_DECLARE(struct pf_altqqueue, pf_altqs[2]); #define V_pf_altqs VNET(pf_altqs) VNET_DECLARE(struct pf_palist, pf_pabuf); #define V_pf_pabuf VNET(pf_pabuf) VNET_DECLARE(u_int32_t, ticket_altqs_active); #define V_ticket_altqs_active VNET(ticket_altqs_active) VNET_DECLARE(u_int32_t, ticket_altqs_inactive); #define V_ticket_altqs_inactive VNET(ticket_altqs_inactive) VNET_DECLARE(int, altqs_inactive_open); #define V_altqs_inactive_open VNET(altqs_inactive_open) VNET_DECLARE(u_int32_t, ticket_pabuf); #define V_ticket_pabuf VNET(ticket_pabuf) VNET_DECLARE(struct pf_altqqueue *, pf_altqs_active); #define V_pf_altqs_active VNET(pf_altqs_active) VNET_DECLARE(struct pf_altqqueue *, pf_altqs_inactive); #define V_pf_altqs_inactive VNET(pf_altqs_inactive) VNET_DECLARE(struct pf_rulequeue, pf_unlinked_rules); #define V_pf_unlinked_rules VNET(pf_unlinked_rules) void pf_initialize(void); void pf_mtag_initialize(void); void pf_mtag_cleanup(void); void pf_cleanup(void); struct pf_mtag *pf_get_mtag(struct mbuf *); extern void pf_calc_skip_steps(struct pf_rulequeue *); #ifdef ALTQ extern void pf_altq_ifnet_event(struct ifnet *, int); #endif VNET_DECLARE(uma_zone_t, pf_state_z); #define V_pf_state_z VNET(pf_state_z) VNET_DECLARE(uma_zone_t, pf_state_key_z); #define V_pf_state_key_z VNET(pf_state_key_z) VNET_DECLARE(uma_zone_t, pf_state_scrub_z); #define V_pf_state_scrub_z VNET(pf_state_scrub_z) extern void pf_purge_thread(void *); extern void pf_intr(void *); extern void pf_purge_expired_src_nodes(void); extern int pf_unlink_state(struct pf_state *, u_int); #define PF_ENTER_LOCKED 0x00000001 #define PF_RETURN_LOCKED 0x00000002 extern int pf_state_insert(struct pfi_kif *, struct pf_state_key *, struct pf_state_key *, struct pf_state *); extern void pf_free_state(struct pf_state *); static __inline void pf_ref_state(struct pf_state *s) { refcount_acquire(&s->refs); } static __inline int pf_release_state(struct pf_state *s) { if (refcount_release(&s->refs)) { pf_free_state(s); return (1); } else return (0); } extern struct pf_state *pf_find_state_byid(uint64_t, uint32_t); extern struct pf_state *pf_find_state_all(struct pf_state_key_cmp *, u_int, int *); extern struct pf_src_node *pf_find_src_node(struct pf_addr *, struct pf_rule *, sa_family_t, int); extern void pf_unlink_src_node(struct pf_src_node *); extern u_int pf_free_src_nodes(struct pf_src_node_list *); extern void pf_print_state(struct pf_state *); extern void pf_print_flags(u_int8_t); extern u_int16_t pf_cksum_fixup(u_int16_t, u_int16_t, u_int16_t, u_int8_t); VNET_DECLARE(struct ifnet *, sync_ifp); #define V_sync_ifp VNET(sync_ifp); VNET_DECLARE(struct pf_rule, pf_default_rule); #define V_pf_default_rule VNET(pf_default_rule) extern void pf_addrcpy(struct pf_addr *, struct pf_addr *, u_int8_t); void pf_free_rule(struct pf_rule *); #ifdef INET int pf_test(int, struct ifnet *, struct mbuf **, struct inpcb *); int pf_normalize_ip(struct mbuf **, int, struct pfi_kif *, u_short *, struct pf_pdesc *); #endif /* INET */ #ifdef INET6 int pf_test6(int, struct ifnet *, struct mbuf **, struct inpcb *); int pf_normalize_ip6(struct mbuf **, int, struct pfi_kif *, u_short *, struct pf_pdesc *); void pf_poolmask(struct pf_addr *, struct pf_addr*, struct pf_addr *, struct pf_addr *, u_int8_t); void pf_addr_inc(struct pf_addr *, sa_family_t); int pf_refragment6(struct ifnet *, struct mbuf **, struct m_tag *); #endif /* INET6 */ u_int32_t pf_new_isn(struct pf_state *); void *pf_pull_hdr(struct mbuf *, int, void *, int, u_short *, u_short *, sa_family_t); void pf_change_a(void *, u_int16_t *, u_int32_t, u_int8_t); void pf_send_deferred_syn(struct pf_state *); int pf_match_addr(u_int8_t, struct pf_addr *, struct pf_addr *, struct pf_addr *, sa_family_t); int pf_match_addr_range(struct pf_addr *, struct pf_addr *, struct pf_addr *, sa_family_t); int pf_match_port(u_int8_t, u_int16_t, u_int16_t, u_int16_t); void pf_normalize_init(void); void pf_normalize_cleanup(void); int pf_normalize_tcp(int, struct pfi_kif *, struct mbuf *, int, int, void *, struct pf_pdesc *); void pf_normalize_tcp_cleanup(struct pf_state *); int pf_normalize_tcp_init(struct mbuf *, int, struct pf_pdesc *, struct tcphdr *, struct pf_state_peer *, struct pf_state_peer *); int pf_normalize_tcp_stateful(struct mbuf *, int, struct pf_pdesc *, u_short *, struct tcphdr *, struct pf_state *, struct pf_state_peer *, struct pf_state_peer *, int *); u_int32_t pf_state_expires(const struct pf_state *); void pf_purge_expired_fragments(void); int pf_routable(struct pf_addr *addr, sa_family_t af, struct pfi_kif *, int); int pf_socket_lookup(int, struct pf_pdesc *, struct mbuf *); struct pf_state_key *pf_alloc_state_key(int); void pfr_initialize(void); void pfr_cleanup(void); int pfr_match_addr(struct pfr_ktable *, struct pf_addr *, sa_family_t); void pfr_update_stats(struct pfr_ktable *, struct pf_addr *, sa_family_t, u_int64_t, int, int, int); int pfr_pool_get(struct pfr_ktable *, int *, struct pf_addr *, sa_family_t); void pfr_dynaddr_update(struct pfr_ktable *, struct pfi_dynaddr *); struct pfr_ktable * pfr_attach_table(struct pf_ruleset *, char *); void pfr_detach_table(struct pfr_ktable *); int pfr_clr_tables(struct pfr_table *, int *, int); int pfr_add_tables(struct pfr_table *, int, int *, int); int pfr_del_tables(struct pfr_table *, int, int *, int); int pfr_get_tables(struct pfr_table *, struct pfr_table *, int *, int); int pfr_get_tstats(struct pfr_table *, struct pfr_tstats *, int *, int); int pfr_clr_tstats(struct pfr_table *, int, int *, int); int pfr_set_tflags(struct pfr_table *, int, int, int, int *, int *, int); int pfr_clr_addrs(struct pfr_table *, int *, int); int pfr_insert_kentry(struct pfr_ktable *, struct pfr_addr *, long); int pfr_add_addrs(struct pfr_table *, struct pfr_addr *, int, int *, int); int pfr_del_addrs(struct pfr_table *, struct pfr_addr *, int, int *, int); int pfr_set_addrs(struct pfr_table *, struct pfr_addr *, int, int *, int *, int *, int *, int, u_int32_t); int pfr_get_addrs(struct pfr_table *, struct pfr_addr *, int *, int); int pfr_get_astats(struct pfr_table *, struct pfr_astats *, int *, int); int pfr_clr_astats(struct pfr_table *, struct pfr_addr *, int, int *, int); int pfr_tst_addrs(struct pfr_table *, struct pfr_addr *, int, int *, int); int pfr_ina_begin(struct pfr_table *, u_int32_t *, int *, int); int pfr_ina_rollback(struct pfr_table *, u_int32_t, int *, int); int pfr_ina_commit(struct pfr_table *, u_int32_t, int *, int *, int); int pfr_ina_define(struct pfr_table *, struct pfr_addr *, int, int *, int *, u_int32_t, int); MALLOC_DECLARE(PFI_MTYPE); VNET_DECLARE(struct pfi_kif *, pfi_all); #define V_pfi_all VNET(pfi_all) void pfi_initialize(void); void pfi_cleanup(void); void pfi_kif_ref(struct pfi_kif *); void pfi_kif_unref(struct pfi_kif *); struct pfi_kif *pfi_kif_find(const char *); struct pfi_kif *pfi_kif_attach(struct pfi_kif *, const char *); int pfi_kif_match(struct pfi_kif *, struct pfi_kif *); void pfi_kif_purge(void); int pfi_match_addr(struct pfi_dynaddr *, struct pf_addr *, sa_family_t); int pfi_dynaddr_setup(struct pf_addr_wrap *, sa_family_t); void pfi_dynaddr_remove(struct pfi_dynaddr *); void pfi_dynaddr_copyout(struct pf_addr_wrap *); void pfi_update_status(const char *, struct pf_status *); void pfi_get_ifaces(const char *, struct pfi_kif *, int *); int pfi_set_flags(const char *, int); int pfi_clear_flags(const char *, int); int pf_match_tag(struct mbuf *, struct pf_rule *, int *, int); int pf_tag_packet(struct mbuf *, struct pf_pdesc *, int); int pf_addr_cmp(struct pf_addr *, struct pf_addr *, sa_family_t); void pf_qid2qname(u_int32_t, char *); VNET_DECLARE(struct pf_kstatus, pf_status); #define V_pf_status VNET(pf_status) struct pf_limit { uma_zone_t zone; u_int limit; }; VNET_DECLARE(struct pf_limit, pf_limits[PF_LIMIT_MAX]); #define V_pf_limits VNET(pf_limits) #endif /* _KERNEL */ #ifdef _KERNEL VNET_DECLARE(struct pf_anchor_global, pf_anchors); #define V_pf_anchors VNET(pf_anchors) VNET_DECLARE(struct pf_anchor, pf_main_anchor); #define V_pf_main_anchor VNET(pf_main_anchor) #define pf_main_ruleset V_pf_main_anchor.ruleset #endif /* these ruleset functions can be linked into userland programs (pfctl) */ int pf_get_ruleset_number(u_int8_t); void pf_init_ruleset(struct pf_ruleset *); int pf_anchor_setup(struct pf_rule *, const struct pf_ruleset *, const char *); int pf_anchor_copyout(const struct pf_ruleset *, const struct pf_rule *, struct pfioc_rule *); void pf_anchor_remove(struct pf_rule *); void pf_remove_if_empty_ruleset(struct pf_ruleset *); struct pf_ruleset *pf_find_ruleset(const char *); struct pf_ruleset *pf_find_or_create_ruleset(const char *); void pf_rs_initialize(void); /* The fingerprint functions can be linked into userland programs (tcpdump) */ int pf_osfp_add(struct pf_osfp_ioctl *); #ifdef _KERNEL struct pf_osfp_enlist * pf_osfp_fingerprint(struct pf_pdesc *, struct mbuf *, int, const struct tcphdr *); #endif /* _KERNEL */ void pf_osfp_flush(void); int pf_osfp_get(struct pf_osfp_ioctl *); int pf_osfp_match(struct pf_osfp_enlist *, pf_osfp_t); #ifdef _KERNEL void pf_print_host(struct pf_addr *, u_int16_t, u_int8_t); void pf_step_into_anchor(struct pf_anchor_stackframe *, int *, struct pf_ruleset **, int, struct pf_rule **, struct pf_rule **, int *); int pf_step_out_of_anchor(struct pf_anchor_stackframe *, int *, struct pf_ruleset **, int, struct pf_rule **, struct pf_rule **, int *); int pf_map_addr(u_int8_t, struct pf_rule *, struct pf_addr *, struct pf_addr *, struct pf_addr *, struct pf_src_node **); struct pf_rule *pf_get_translation(struct pf_pdesc *, struct mbuf *, int, int, struct pfi_kif *, struct pf_src_node **, struct pf_state_key **, struct pf_state_key **, struct pf_addr *, struct pf_addr *, uint16_t, uint16_t, struct pf_anchor_stackframe *); struct pf_state_key *pf_state_key_setup(struct pf_pdesc *, struct pf_addr *, struct pf_addr *, u_int16_t, u_int16_t); struct pf_state_key *pf_state_key_clone(struct pf_state_key *); #endif /* _KERNEL */ #endif /* _NET_PFVAR_H_ */ Index: head/sys/netpfil/pf/pf_norm.c =================================================================== --- head/sys/netpfil/pf/pf_norm.c (revision 287221) +++ head/sys/netpfil/pf/pf_norm.c (revision 287222) @@ -1,2288 +1,1840 @@ /*- * Copyright 2001 Niels Provos * Copyright 2011 Alexander Bluhm * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. * * $OpenBSD: pf_norm.c,v 1.114 2009/01/29 14:11:45 henning Exp $ */ #include __FBSDID("$FreeBSD$"); #include "opt_inet.h" #include "opt_inet6.h" #include "opt_pf.h" #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #ifdef INET6 #include #endif /* INET6 */ struct pf_frent { TAILQ_ENTRY(pf_frent) fr_next; struct mbuf *fe_m; uint16_t fe_hdrlen; /* ipv4 header lenght with ip options ipv6, extension, fragment header */ uint16_t fe_extoff; /* last extension header offset or 0 */ uint16_t fe_len; /* fragment length */ uint16_t fe_off; /* fragment offset */ uint16_t fe_mff; /* more fragment flag */ }; struct pf_fragment_cmp { struct pf_addr frc_src; struct pf_addr frc_dst; uint32_t frc_id; sa_family_t frc_af; uint8_t frc_proto; }; struct pf_fragment { struct pf_fragment_cmp fr_key; #define fr_src fr_key.frc_src #define fr_dst fr_key.frc_dst #define fr_id fr_key.frc_id #define fr_af fr_key.frc_af #define fr_proto fr_key.frc_proto RB_ENTRY(pf_fragment) fr_entry; TAILQ_ENTRY(pf_fragment) frag_next; - uint8_t fr_flags; /* status flags */ -#define PFFRAG_SEENLAST 0x0001 /* Seen the last fragment for this */ -#define PFFRAG_NOBUFFER 0x0002 /* Non-buffering fragment cache */ -#define PFFRAG_DROP 0x0004 /* Drop all fragments */ -#define BUFFER_FRAGMENTS(fr) (!((fr)->fr_flags & PFFRAG_NOBUFFER)) - uint16_t fr_max; /* fragment data max */ uint32_t fr_timeout; uint16_t fr_maxlen; /* maximum length of single fragment */ TAILQ_HEAD(pf_fragq, pf_frent) fr_queue; }; struct pf_fragment_tag { uint16_t ft_hdrlen; /* header length of reassembled pkt */ uint16_t ft_extoff; /* last extension header offset or 0 */ uint16_t ft_maxlen; /* maximum fragment payload length */ uint32_t ft_id; /* fragment id */ }; static struct mtx pf_frag_mtx; MTX_SYSINIT(pf_frag_mtx, &pf_frag_mtx, "pf fragments", MTX_DEF); #define PF_FRAG_LOCK() mtx_lock(&pf_frag_mtx) #define PF_FRAG_UNLOCK() mtx_unlock(&pf_frag_mtx) #define PF_FRAG_ASSERT() mtx_assert(&pf_frag_mtx, MA_OWNED) VNET_DEFINE(uma_zone_t, pf_state_scrub_z); /* XXX: shared with pfsync */ static VNET_DEFINE(uma_zone_t, pf_frent_z); #define V_pf_frent_z VNET(pf_frent_z) static VNET_DEFINE(uma_zone_t, pf_frag_z); #define V_pf_frag_z VNET(pf_frag_z) TAILQ_HEAD(pf_fragqueue, pf_fragment); TAILQ_HEAD(pf_cachequeue, pf_fragment); static VNET_DEFINE(struct pf_fragqueue, pf_fragqueue); #define V_pf_fragqueue VNET(pf_fragqueue) -static VNET_DEFINE(struct pf_cachequeue, pf_cachequeue); -#define V_pf_cachequeue VNET(pf_cachequeue) RB_HEAD(pf_frag_tree, pf_fragment); static VNET_DEFINE(struct pf_frag_tree, pf_frag_tree); #define V_pf_frag_tree VNET(pf_frag_tree) -static VNET_DEFINE(struct pf_frag_tree, pf_cache_tree); -#define V_pf_cache_tree VNET(pf_cache_tree) static int pf_frag_compare(struct pf_fragment *, struct pf_fragment *); static RB_PROTOTYPE(pf_frag_tree, pf_fragment, fr_entry, pf_frag_compare); static RB_GENERATE(pf_frag_tree, pf_fragment, fr_entry, pf_frag_compare); static void pf_flush_fragments(void); static void pf_free_fragment(struct pf_fragment *); static void pf_remove_fragment(struct pf_fragment *); static int pf_normalize_tcpopt(struct pf_rule *, struct mbuf *, struct tcphdr *, int, sa_family_t); static struct pf_frent *pf_create_fragment(u_short *); static struct pf_fragment *pf_find_fragment(struct pf_fragment_cmp *key, struct pf_frag_tree *tree); static struct pf_fragment *pf_fillup_fragment(struct pf_fragment_cmp *, struct pf_frent *, u_short *); static int pf_isfull_fragment(struct pf_fragment *); static struct mbuf *pf_join_fragment(struct pf_fragment *); #ifdef INET static void pf_scrub_ip(struct mbuf **, uint32_t, uint8_t, uint8_t); static int pf_reassemble(struct mbuf **, struct ip *, int, u_short *); -static struct mbuf *pf_fragcache(struct mbuf **, struct ip*, - struct pf_fragment **, int, int, int *); #endif /* INET */ #ifdef INET6 static int pf_reassemble6(struct mbuf **, struct ip6_hdr *, struct ip6_frag *, uint16_t, uint16_t, u_short *); static void pf_scrub_ip6(struct mbuf **, uint8_t); #endif /* INET6 */ #define DPFPRINTF(x) do { \ if (V_pf_status.debug >= PF_DEBUG_MISC) { \ printf("%s: ", __func__); \ printf x ; \ } \ } while(0) #ifdef INET static void pf_ip2key(struct ip *ip, int dir, struct pf_fragment_cmp *key) { key->frc_src.v4 = ip->ip_src; key->frc_dst.v4 = ip->ip_dst; key->frc_af = AF_INET; key->frc_proto = ip->ip_p; key->frc_id = ip->ip_id; } #endif /* INET */ void pf_normalize_init(void) { V_pf_frag_z = uma_zcreate("pf frags", sizeof(struct pf_fragment), NULL, NULL, NULL, NULL, UMA_ALIGN_PTR, 0); V_pf_frent_z = uma_zcreate("pf frag entries", sizeof(struct pf_frent), NULL, NULL, NULL, NULL, UMA_ALIGN_PTR, 0); V_pf_state_scrub_z = uma_zcreate("pf state scrubs", sizeof(struct pf_state_scrub), NULL, NULL, NULL, NULL, UMA_ALIGN_PTR, 0); V_pf_limits[PF_LIMIT_FRAGS].zone = V_pf_frent_z; V_pf_limits[PF_LIMIT_FRAGS].limit = PFFRAG_FRENT_HIWAT; uma_zone_set_max(V_pf_frent_z, PFFRAG_FRENT_HIWAT); uma_zone_set_warning(V_pf_frent_z, "PF frag entries limit reached"); TAILQ_INIT(&V_pf_fragqueue); - TAILQ_INIT(&V_pf_cachequeue); } void pf_normalize_cleanup(void) { uma_zdestroy(V_pf_state_scrub_z); uma_zdestroy(V_pf_frent_z); uma_zdestroy(V_pf_frag_z); } static int pf_frag_compare(struct pf_fragment *a, struct pf_fragment *b) { int diff; if ((diff = a->fr_id - b->fr_id) != 0) return (diff); if ((diff = a->fr_proto - b->fr_proto) != 0) return (diff); if ((diff = a->fr_af - b->fr_af) != 0) return (diff); if ((diff = pf_addr_cmp(&a->fr_src, &b->fr_src, a->fr_af)) != 0) return (diff); if ((diff = pf_addr_cmp(&a->fr_dst, &b->fr_dst, a->fr_af)) != 0) return (diff); return (0); } void pf_purge_expired_fragments(void) { struct pf_fragment *frag; u_int32_t expire = time_uptime - V_pf_default_rule.timeout[PFTM_FRAG]; PF_FRAG_LOCK(); while ((frag = TAILQ_LAST(&V_pf_fragqueue, pf_fragqueue)) != NULL) { - KASSERT((BUFFER_FRAGMENTS(frag)), - ("BUFFER_FRAGMENTS(frag) == 0: %s", __FUNCTION__)); if (frag->fr_timeout > expire) break; DPFPRINTF(("expiring %d(%p)\n", frag->fr_id, frag)); pf_free_fragment(frag); } - while ((frag = TAILQ_LAST(&V_pf_cachequeue, pf_cachequeue)) != NULL) { - KASSERT((!BUFFER_FRAGMENTS(frag)), - ("BUFFER_FRAGMENTS(frag) != 0: %s", __FUNCTION__)); - if (frag->fr_timeout > expire) - break; - - DPFPRINTF(("expiring %d(%p)\n", frag->fr_id, frag)); - pf_free_fragment(frag); - KASSERT((TAILQ_EMPTY(&V_pf_cachequeue) || - TAILQ_LAST(&V_pf_cachequeue, pf_cachequeue) != frag), - ("!(TAILQ_EMPTY() || TAILQ_LAST() == farg): %s", - __FUNCTION__)); - } PF_FRAG_UNLOCK(); } /* * Try to flush old fragments to make space for new ones */ static void pf_flush_fragments(void) { - struct pf_fragment *frag, *cache; + struct pf_fragment *frag; int goal; PF_FRAG_ASSERT(); goal = uma_zone_get_cur(V_pf_frent_z) * 9 / 10; DPFPRINTF(("trying to free %d frag entriess\n", goal)); while (goal < uma_zone_get_cur(V_pf_frent_z)) { frag = TAILQ_LAST(&V_pf_fragqueue, pf_fragqueue); if (frag) pf_free_fragment(frag); - cache = TAILQ_LAST(&V_pf_cachequeue, pf_cachequeue); - if (cache) - pf_free_fragment(cache); - if (frag == NULL && cache == NULL) + else break; } } /* Frees the fragments and all associated entries */ static void pf_free_fragment(struct pf_fragment *frag) { struct pf_frent *frent; PF_FRAG_ASSERT(); /* Free all fragments */ - if (BUFFER_FRAGMENTS(frag)) { - for (frent = TAILQ_FIRST(&frag->fr_queue); frent; - frent = TAILQ_FIRST(&frag->fr_queue)) { - TAILQ_REMOVE(&frag->fr_queue, frent, fr_next); + for (frent = TAILQ_FIRST(&frag->fr_queue); frent; + frent = TAILQ_FIRST(&frag->fr_queue)) { + TAILQ_REMOVE(&frag->fr_queue, frent, fr_next); - m_freem(frent->fe_m); - uma_zfree(V_pf_frent_z, frent); - } - } else { - for (frent = TAILQ_FIRST(&frag->fr_queue); frent; - frent = TAILQ_FIRST(&frag->fr_queue)) { - TAILQ_REMOVE(&frag->fr_queue, frent, fr_next); - - KASSERT((TAILQ_EMPTY(&frag->fr_queue) || - TAILQ_FIRST(&frag->fr_queue)->fe_off > - frent->fe_len), - ("! (TAILQ_EMPTY() || TAILQ_FIRST()->fe_off >" - " frent->fe_len): %s", __func__)); - - uma_zfree(V_pf_frent_z, frent); - } + m_freem(frent->fe_m); + uma_zfree(V_pf_frent_z, frent); } pf_remove_fragment(frag); } static struct pf_fragment * pf_find_fragment(struct pf_fragment_cmp *key, struct pf_frag_tree *tree) { struct pf_fragment *frag; PF_FRAG_ASSERT(); frag = RB_FIND(pf_frag_tree, tree, (struct pf_fragment *)key); if (frag != NULL) { /* XXX Are we sure we want to update the timeout? */ frag->fr_timeout = time_uptime; - if (BUFFER_FRAGMENTS(frag)) { - TAILQ_REMOVE(&V_pf_fragqueue, frag, frag_next); - TAILQ_INSERT_HEAD(&V_pf_fragqueue, frag, frag_next); - } else { - TAILQ_REMOVE(&V_pf_cachequeue, frag, frag_next); - TAILQ_INSERT_HEAD(&V_pf_cachequeue, frag, frag_next); - } + TAILQ_REMOVE(&V_pf_fragqueue, frag, frag_next); + TAILQ_INSERT_HEAD(&V_pf_fragqueue, frag, frag_next); } return (frag); } /* Removes a fragment from the fragment queue and frees the fragment */ static void pf_remove_fragment(struct pf_fragment *frag) { PF_FRAG_ASSERT(); - if (BUFFER_FRAGMENTS(frag)) { - RB_REMOVE(pf_frag_tree, &V_pf_frag_tree, frag); - TAILQ_REMOVE(&V_pf_fragqueue, frag, frag_next); - uma_zfree(V_pf_frag_z, frag); - } else { - RB_REMOVE(pf_frag_tree, &V_pf_cache_tree, frag); - TAILQ_REMOVE(&V_pf_cachequeue, frag, frag_next); - uma_zfree(V_pf_frag_z, frag); - } + RB_REMOVE(pf_frag_tree, &V_pf_frag_tree, frag); + TAILQ_REMOVE(&V_pf_fragqueue, frag, frag_next); + uma_zfree(V_pf_frag_z, frag); } static struct pf_frent * pf_create_fragment(u_short *reason) { struct pf_frent *frent; PF_FRAG_ASSERT(); frent = uma_zalloc(V_pf_frent_z, M_NOWAIT); if (frent == NULL) { pf_flush_fragments(); frent = uma_zalloc(V_pf_frent_z, M_NOWAIT); if (frent == NULL) { REASON_SET(reason, PFRES_MEMORY); return (NULL); } } return (frent); } static struct pf_fragment * pf_fillup_fragment(struct pf_fragment_cmp *key, struct pf_frent *frent, u_short *reason) { struct pf_frent *after, *next, *prev; struct pf_fragment *frag; uint16_t total; PF_FRAG_ASSERT(); /* No empty fragments. */ if (frent->fe_len == 0) { DPFPRINTF(("bad fragment: len 0")); goto bad_fragment; } /* All fragments are 8 byte aligned. */ if (frent->fe_mff && (frent->fe_len & 0x7)) { DPFPRINTF(("bad fragment: mff and len %d", frent->fe_len)); goto bad_fragment; } /* Respect maximum length, IP_MAXPACKET == IPV6_MAXPACKET. */ if (frent->fe_off + frent->fe_len > IP_MAXPACKET) { DPFPRINTF(("bad fragment: max packet %d", frent->fe_off + frent->fe_len)); goto bad_fragment; } DPFPRINTF((key->frc_af == AF_INET ? "reass frag %d @ %d-%d" : "reass frag %#08x @ %d-%d", key->frc_id, frent->fe_off, frent->fe_off + frent->fe_len)); /* Fully buffer all of the fragments in this fragment queue. */ frag = pf_find_fragment(key, &V_pf_frag_tree); /* Create a new reassembly queue for this packet. */ if (frag == NULL) { frag = uma_zalloc(V_pf_frag_z, M_NOWAIT); if (frag == NULL) { pf_flush_fragments(); frag = uma_zalloc(V_pf_frag_z, M_NOWAIT); if (frag == NULL) { REASON_SET(reason, PFRES_MEMORY); goto drop_fragment; } } *(struct pf_fragment_cmp *)frag = *key; - frag->fr_flags = 0; frag->fr_timeout = time_second; frag->fr_maxlen = frent->fe_len; TAILQ_INIT(&frag->fr_queue); RB_INSERT(pf_frag_tree, &V_pf_frag_tree, frag); TAILQ_INSERT_HEAD(&V_pf_fragqueue, frag, frag_next); /* We do not have a previous fragment. */ TAILQ_INSERT_HEAD(&frag->fr_queue, frent, fr_next); return (frag); } KASSERT(!TAILQ_EMPTY(&frag->fr_queue), ("!TAILQ_EMPTY()->fr_queue")); /* Remember maximum fragment len for refragmentation. */ if (frent->fe_len > frag->fr_maxlen) frag->fr_maxlen = frent->fe_len; /* Maximum data we have seen already. */ total = TAILQ_LAST(&frag->fr_queue, pf_fragq)->fe_off + TAILQ_LAST(&frag->fr_queue, pf_fragq)->fe_len; /* Non terminal fragments must have more fragments flag. */ if (frent->fe_off + frent->fe_len < total && !frent->fe_mff) goto bad_fragment; /* Check if we saw the last fragment already. */ if (!TAILQ_LAST(&frag->fr_queue, pf_fragq)->fe_mff) { if (frent->fe_off + frent->fe_len > total || (frent->fe_off + frent->fe_len == total && frent->fe_mff)) goto bad_fragment; } else { if (frent->fe_off + frent->fe_len == total && !frent->fe_mff) goto bad_fragment; } /* Find a fragment after the current one. */ prev = NULL; TAILQ_FOREACH(after, &frag->fr_queue, fr_next) { if (after->fe_off > frent->fe_off) break; prev = after; } KASSERT(prev != NULL || after != NULL, ("prev != NULL || after != NULL")); if (prev != NULL && prev->fe_off + prev->fe_len > frent->fe_off) { uint16_t precut; precut = prev->fe_off + prev->fe_len - frent->fe_off; if (precut >= frent->fe_len) goto bad_fragment; DPFPRINTF(("overlap -%d", precut)); m_adj(frent->fe_m, precut); frent->fe_off += precut; frent->fe_len -= precut; } for (; after != NULL && frent->fe_off + frent->fe_len > after->fe_off; after = next) { uint16_t aftercut; aftercut = frent->fe_off + frent->fe_len - after->fe_off; DPFPRINTF(("adjust overlap %d", aftercut)); if (aftercut < after->fe_len) { m_adj(after->fe_m, aftercut); after->fe_off += aftercut; after->fe_len -= aftercut; break; } /* This fragment is completely overlapped, lose it. */ next = TAILQ_NEXT(after, fr_next); m_freem(after->fe_m); TAILQ_REMOVE(&frag->fr_queue, after, fr_next); uma_zfree(V_pf_frent_z, after); } if (prev == NULL) TAILQ_INSERT_HEAD(&frag->fr_queue, frent, fr_next); else TAILQ_INSERT_AFTER(&frag->fr_queue, prev, frent, fr_next); return (frag); bad_fragment: REASON_SET(reason, PFRES_FRAG); drop_fragment: uma_zfree(V_pf_frent_z, frent); return (NULL); } static int pf_isfull_fragment(struct pf_fragment *frag) { struct pf_frent *frent, *next; uint16_t off, total; /* Check if we are completely reassembled */ if (TAILQ_LAST(&frag->fr_queue, pf_fragq)->fe_mff) return (0); /* Maximum data we have seen already */ total = TAILQ_LAST(&frag->fr_queue, pf_fragq)->fe_off + TAILQ_LAST(&frag->fr_queue, pf_fragq)->fe_len; /* Check if we have all the data */ off = 0; for (frent = TAILQ_FIRST(&frag->fr_queue); frent; frent = next) { next = TAILQ_NEXT(frent, fr_next); off += frent->fe_len; if (off < total && (next == NULL || next->fe_off != off)) { DPFPRINTF(("missing fragment at %d, next %d, total %d", off, next == NULL ? -1 : next->fe_off, total)); return (0); } } DPFPRINTF(("%d < %d?", off, total)); if (off < total) return (0); KASSERT(off == total, ("off == total")); return (1); } static struct mbuf * pf_join_fragment(struct pf_fragment *frag) { struct mbuf *m, *m2; struct pf_frent *frent, *next; frent = TAILQ_FIRST(&frag->fr_queue); next = TAILQ_NEXT(frent, fr_next); m = frent->fe_m; m_adj(m, (frent->fe_hdrlen + frent->fe_len) - m->m_pkthdr.len); uma_zfree(V_pf_frent_z, frent); for (frent = next; frent != NULL; frent = next) { next = TAILQ_NEXT(frent, fr_next); m2 = frent->fe_m; /* Strip off ip header. */ m_adj(m2, frent->fe_hdrlen); /* Strip off any trailing bytes. */ m_adj(m2, frent->fe_len - m2->m_pkthdr.len); uma_zfree(V_pf_frent_z, frent); m_cat(m, m2); } /* Remove from fragment queue. */ pf_remove_fragment(frag); return (m); } #ifdef INET static int pf_reassemble(struct mbuf **m0, struct ip *ip, int dir, u_short *reason) { struct mbuf *m = *m0; struct pf_frent *frent; struct pf_fragment *frag; struct pf_fragment_cmp key; uint16_t total, hdrlen; /* Get an entry for the fragment queue */ if ((frent = pf_create_fragment(reason)) == NULL) return (PF_DROP); frent->fe_m = m; frent->fe_hdrlen = ip->ip_hl << 2; frent->fe_extoff = 0; frent->fe_len = ntohs(ip->ip_len) - (ip->ip_hl << 2); frent->fe_off = (ntohs(ip->ip_off) & IP_OFFMASK) << 3; frent->fe_mff = ntohs(ip->ip_off) & IP_MF; pf_ip2key(ip, dir, &key); if ((frag = pf_fillup_fragment(&key, frent, reason)) == NULL) return (PF_DROP); /* The mbuf is part of the fragment entry, no direct free or access */ m = *m0 = NULL; if (!pf_isfull_fragment(frag)) return (PF_PASS); /* drop because *m0 is NULL, no error */ /* We have all the data */ frent = TAILQ_FIRST(&frag->fr_queue); KASSERT(frent != NULL, ("frent != NULL")); total = TAILQ_LAST(&frag->fr_queue, pf_fragq)->fe_off + TAILQ_LAST(&frag->fr_queue, pf_fragq)->fe_len; hdrlen = frent->fe_hdrlen; m = *m0 = pf_join_fragment(frag); frag = NULL; if (m->m_flags & M_PKTHDR) { int plen = 0; for (m = *m0; m; m = m->m_next) plen += m->m_len; m = *m0; m->m_pkthdr.len = plen; } ip = mtod(m, struct ip *); ip->ip_len = htons(hdrlen + total); ip->ip_off &= ~(IP_MF|IP_OFFMASK); if (hdrlen + total > IP_MAXPACKET) { DPFPRINTF(("drop: too big: %d", total)); ip->ip_len = 0; REASON_SET(reason, PFRES_SHORT); /* PF_DROP requires a valid mbuf *m0 in pf_test() */ return (PF_DROP); } DPFPRINTF(("complete: %p(%d)\n", m, ntohs(ip->ip_len))); return (PF_PASS); } #endif /* INET */ #ifdef INET6 static int pf_reassemble6(struct mbuf **m0, struct ip6_hdr *ip6, struct ip6_frag *fraghdr, uint16_t hdrlen, uint16_t extoff, u_short *reason) { struct mbuf *m = *m0; struct pf_frent *frent; struct pf_fragment *frag; struct pf_fragment_cmp key; struct m_tag *mtag; struct pf_fragment_tag *ftag; int off; uint32_t frag_id; uint16_t total, maxlen; uint8_t proto; PF_FRAG_LOCK(); /* Get an entry for the fragment queue. */ if ((frent = pf_create_fragment(reason)) == NULL) { PF_FRAG_UNLOCK(); return (PF_DROP); } frent->fe_m = m; frent->fe_hdrlen = hdrlen; frent->fe_extoff = extoff; frent->fe_len = sizeof(struct ip6_hdr) + ntohs(ip6->ip6_plen) - hdrlen; frent->fe_off = ntohs(fraghdr->ip6f_offlg & IP6F_OFF_MASK); frent->fe_mff = fraghdr->ip6f_offlg & IP6F_MORE_FRAG; key.frc_src.v6 = ip6->ip6_src; key.frc_dst.v6 = ip6->ip6_dst; key.frc_af = AF_INET6; /* Only the first fragment's protocol is relevant. */ key.frc_proto = 0; key.frc_id = fraghdr->ip6f_ident; if ((frag = pf_fillup_fragment(&key, frent, reason)) == NULL) { PF_FRAG_UNLOCK(); return (PF_DROP); } /* The mbuf is part of the fragment entry, no direct free or access. */ m = *m0 = NULL; if (!pf_isfull_fragment(frag)) { PF_FRAG_UNLOCK(); return (PF_PASS); /* Drop because *m0 is NULL, no error. */ } /* We have all the data. */ extoff = frent->fe_extoff; maxlen = frag->fr_maxlen; frag_id = frag->fr_id; frent = TAILQ_FIRST(&frag->fr_queue); KASSERT(frent != NULL, ("frent != NULL")); total = TAILQ_LAST(&frag->fr_queue, pf_fragq)->fe_off + TAILQ_LAST(&frag->fr_queue, pf_fragq)->fe_len; hdrlen = frent->fe_hdrlen - sizeof(struct ip6_frag); m = *m0 = pf_join_fragment(frag); frag = NULL; PF_FRAG_UNLOCK(); /* Take protocol from first fragment header. */ m = m_getptr(m, hdrlen + offsetof(struct ip6_frag, ip6f_nxt), &off); KASSERT(m, ("%s: short mbuf chain", __func__)); proto = *(mtod(m, caddr_t) + off); m = *m0; /* Delete frag6 header */ if (ip6_deletefraghdr(m, hdrlen, M_NOWAIT) != 0) goto fail; if (m->m_flags & M_PKTHDR) { int plen = 0; for (m = *m0; m; m = m->m_next) plen += m->m_len; m = *m0; m->m_pkthdr.len = plen; } if ((mtag = m_tag_get(PF_REASSEMBLED, sizeof(struct pf_fragment_tag), M_NOWAIT)) == NULL) goto fail; ftag = (struct pf_fragment_tag *)(mtag + 1); ftag->ft_hdrlen = hdrlen; ftag->ft_extoff = extoff; ftag->ft_maxlen = maxlen; ftag->ft_id = frag_id; m_tag_prepend(m, mtag); ip6 = mtod(m, struct ip6_hdr *); ip6->ip6_plen = htons(hdrlen - sizeof(struct ip6_hdr) + total); if (extoff) { /* Write protocol into next field of last extension header. */ m = m_getptr(m, extoff + offsetof(struct ip6_ext, ip6e_nxt), &off); KASSERT(m, ("%s: short mbuf chain", __func__)); *(mtod(m, char *) + off) = proto; m = *m0; } else ip6->ip6_nxt = proto; if (hdrlen - sizeof(struct ip6_hdr) + total > IPV6_MAXPACKET) { DPFPRINTF(("drop: too big: %d", total)); ip6->ip6_plen = 0; REASON_SET(reason, PFRES_SHORT); /* PF_DROP requires a valid mbuf *m0 in pf_test6(). */ return (PF_DROP); } DPFPRINTF(("complete: %p(%d)", m, ntohs(ip6->ip6_plen))); return (PF_PASS); fail: REASON_SET(reason, PFRES_MEMORY); /* PF_DROP requires a valid mbuf *m0 in pf_test6(), will free later. */ return (PF_DROP); } #endif /* INET6 */ -#ifdef INET -static struct mbuf * -pf_fragcache(struct mbuf **m0, struct ip *h, struct pf_fragment **frag, int mff, - int drop, int *nomem) -{ - struct mbuf *m = *m0; - struct pf_frent *frp, *fra, *cur = NULL; - int ip_len = ntohs(h->ip_len) - (h->ip_hl << 2); - u_int16_t off = ntohs(h->ip_off) << 3; - u_int16_t max = ip_len + off; - int hosed = 0; - - PF_FRAG_ASSERT(); - KASSERT((*frag == NULL || !BUFFER_FRAGMENTS(*frag)), - ("!(*frag == NULL || !BUFFER_FRAGMENTS(*frag)): %s", __FUNCTION__)); - - /* Create a new range queue for this packet */ - if (*frag == NULL) { - *frag = uma_zalloc(V_pf_frag_z, M_NOWAIT); - if (*frag == NULL) { - pf_flush_fragments(); - *frag = uma_zalloc(V_pf_frag_z, M_NOWAIT); - if (*frag == NULL) - goto no_mem; - } - - /* Get an entry for the queue */ - cur = uma_zalloc(V_pf_frent_z, M_NOWAIT); - if (cur == NULL) { - uma_zfree(V_pf_frag_z, *frag); - *frag = NULL; - goto no_mem; - } - - (*frag)->fr_flags = PFFRAG_NOBUFFER; - (*frag)->fr_max = 0; - (*frag)->fr_src.v4 = h->ip_src; - (*frag)->fr_dst.v4 = h->ip_dst; - (*frag)->fr_af = AF_INET; - (*frag)->fr_proto = h->ip_p; - (*frag)->fr_id = h->ip_id; - (*frag)->fr_timeout = time_uptime; - - cur->fe_off = off; - cur->fe_len = max; /* TODO: fe_len = max - off ? */ - TAILQ_INIT(&(*frag)->fr_queue); - TAILQ_INSERT_HEAD(&(*frag)->fr_queue, cur, fr_next); - - RB_INSERT(pf_frag_tree, &V_pf_cache_tree, *frag); - TAILQ_INSERT_HEAD(&V_pf_cachequeue, *frag, frag_next); - - DPFPRINTF(("fragcache[%d]: new %d-%d\n", h->ip_id, off, max)); - - goto pass; - } - - /* - * Find a fragment after the current one: - * - off contains the real shifted offset. - */ - frp = NULL; - TAILQ_FOREACH(fra, &(*frag)->fr_queue, fr_next) { - if (fra->fe_off > off) - break; - frp = fra; - } - - KASSERT((frp != NULL || fra != NULL), - ("!(frp != NULL || fra != NULL): %s", __FUNCTION__)); - - if (frp != NULL) { - int precut; - - precut = frp->fe_len - off; - if (precut >= ip_len) { - /* Fragment is entirely a duplicate */ - DPFPRINTF(("fragcache[%d]: dead (%d-%d) %d-%d\n", - h->ip_id, frp->fe_off, frp->fe_len, off, max)); - goto drop_fragment; - } - if (precut == 0) { - /* They are adjacent. Fixup cache entry */ - DPFPRINTF(("fragcache[%d]: adjacent (%d-%d) %d-%d\n", - h->ip_id, frp->fe_off, frp->fe_len, off, max)); - frp->fe_len = max; - } else if (precut > 0) { - /* The first part of this payload overlaps with a - * fragment that has already been passed. - * Need to trim off the first part of the payload. - * But to do so easily, we need to create another - * mbuf to throw the original header into. - */ - - DPFPRINTF(("fragcache[%d]: chop %d (%d-%d) %d-%d\n", - h->ip_id, precut, frp->fe_off, frp->fe_len, off, - max)); - - off += precut; - max -= precut; - /* Update the previous frag to encompass this one */ - frp->fe_len = max; - - if (!drop) { - /* XXX Optimization opportunity - * This is a very heavy way to trim the payload. - * we could do it much faster by diddling mbuf - * internals but that would be even less legible - * than this mbuf magic. For my next trick, - * I'll pull a rabbit out of my laptop. - */ - *m0 = m_dup(m, M_NOWAIT); - if (*m0 == NULL) - goto no_mem; - /* From KAME Project : We have missed this! */ - m_adj(*m0, (h->ip_hl << 2) - - (*m0)->m_pkthdr.len); - - KASSERT(((*m0)->m_next == NULL), - ("(*m0)->m_next != NULL: %s", - __FUNCTION__)); - m_adj(m, precut + (h->ip_hl << 2)); - m_cat(*m0, m); - m = *m0; - if (m->m_flags & M_PKTHDR) { - int plen = 0; - struct mbuf *t; - for (t = m; t; t = t->m_next) - plen += t->m_len; - m->m_pkthdr.len = plen; - } - - - h = mtod(m, struct ip *); - - KASSERT(((int)m->m_len == - ntohs(h->ip_len) - precut), - ("m->m_len != ntohs(h->ip_len) - precut: %s", - __FUNCTION__)); - h->ip_off = htons(ntohs(h->ip_off) + - (precut >> 3)); - h->ip_len = htons(ntohs(h->ip_len) - precut); - } else { - hosed++; - } - } else { - /* There is a gap between fragments */ - - DPFPRINTF(("fragcache[%d]: gap %d (%d-%d) %d-%d\n", - h->ip_id, -precut, frp->fe_off, frp->fe_len, off, - max)); - - cur = uma_zalloc(V_pf_frent_z, M_NOWAIT); - if (cur == NULL) - goto no_mem; - - cur->fe_off = off; - cur->fe_len = max; - TAILQ_INSERT_AFTER(&(*frag)->fr_queue, frp, cur, fr_next); - } - } - - if (fra != NULL) { - int aftercut; - int merge = 0; - - aftercut = max - fra->fe_off; - if (aftercut == 0) { - /* Adjacent fragments */ - DPFPRINTF(("fragcache[%d]: adjacent %d-%d (%d-%d)\n", - h->ip_id, off, max, fra->fe_off, fra->fe_len)); - fra->fe_off = off; - merge = 1; - } else if (aftercut > 0) { - /* Need to chop off the tail of this fragment */ - DPFPRINTF(("fragcache[%d]: chop %d %d-%d (%d-%d)\n", - h->ip_id, aftercut, off, max, fra->fe_off, - fra->fe_len)); - fra->fe_off = off; - max -= aftercut; - - merge = 1; - - if (!drop) { - m_adj(m, -aftercut); - if (m->m_flags & M_PKTHDR) { - int plen = 0; - struct mbuf *t; - for (t = m; t; t = t->m_next) - plen += t->m_len; - m->m_pkthdr.len = plen; - } - h = mtod(m, struct ip *); - KASSERT(((int)m->m_len == ntohs(h->ip_len) - aftercut), - ("m->m_len != ntohs(h->ip_len) - aftercut: %s", - __FUNCTION__)); - h->ip_len = htons(ntohs(h->ip_len) - aftercut); - } else { - hosed++; - } - } else if (frp == NULL) { - /* There is a gap between fragments */ - DPFPRINTF(("fragcache[%d]: gap %d %d-%d (%d-%d)\n", - h->ip_id, -aftercut, off, max, fra->fe_off, - fra->fe_len)); - - cur = uma_zalloc(V_pf_frent_z, M_NOWAIT); - if (cur == NULL) - goto no_mem; - - cur->fe_off = off; - cur->fe_len = max; - TAILQ_INSERT_HEAD(&(*frag)->fr_queue, cur, fr_next); - } - - - /* Need to glue together two separate fragment descriptors */ - if (merge) { - if (cur && fra->fe_off <= cur->fe_len) { - /* Need to merge in a previous 'cur' */ - DPFPRINTF(("fragcache[%d]: adjacent(merge " - "%d-%d) %d-%d (%d-%d)\n", - h->ip_id, cur->fe_off, cur->fe_len, off, - max, fra->fe_off, fra->fe_len)); - fra->fe_off = cur->fe_off; - TAILQ_REMOVE(&(*frag)->fr_queue, cur, fr_next); - uma_zfree(V_pf_frent_z, cur); - cur = NULL; - - } else if (frp && fra->fe_off <= frp->fe_len) { - /* Need to merge in a modified 'frp' */ - KASSERT((cur == NULL), ("cur != NULL: %s", - __FUNCTION__)); - DPFPRINTF(("fragcache[%d]: adjacent(merge " - "%d-%d) %d-%d (%d-%d)\n", - h->ip_id, frp->fe_off, frp->fe_len, off, - max, fra->fe_off, fra->fe_len)); - fra->fe_off = frp->fe_off; - TAILQ_REMOVE(&(*frag)->fr_queue, frp, fr_next); - uma_zfree(V_pf_frent_z, frp); - frp = NULL; - - } - } - } - - if (hosed) { - /* - * We must keep tracking the overall fragment even when - * we're going to drop it anyway so that we know when to - * free the overall descriptor. Thus we drop the frag late. - */ - goto drop_fragment; - } - - - pass: - /* Update maximum data size */ - if ((*frag)->fr_max < max) - (*frag)->fr_max = max; - - /* This is the last segment */ - if (!mff) - (*frag)->fr_flags |= PFFRAG_SEENLAST; - - /* Check if we are completely reassembled */ - if (((*frag)->fr_flags & PFFRAG_SEENLAST) && - TAILQ_FIRST(&(*frag)->fr_queue)->fe_off == 0 && - TAILQ_FIRST(&(*frag)->fr_queue)->fe_len == (*frag)->fr_max) { - /* Remove from fragment queue */ - DPFPRINTF(("fragcache[%d]: done 0-%d\n", h->ip_id, - (*frag)->fr_max)); - pf_free_fragment(*frag); - *frag = NULL; - } - - return (m); - - no_mem: - *nomem = 1; - - /* Still need to pay attention to !IP_MF */ - if (!mff && *frag != NULL) - (*frag)->fr_flags |= PFFRAG_SEENLAST; - - m_freem(m); - return (NULL); - - drop_fragment: - - /* Still need to pay attention to !IP_MF */ - if (!mff && *frag != NULL) - (*frag)->fr_flags |= PFFRAG_SEENLAST; - - if (drop) { - /* This fragment has been deemed bad. Don't reass */ - if (((*frag)->fr_flags & PFFRAG_DROP) == 0) - DPFPRINTF(("fragcache[%d]: dropping overall fragment\n", - h->ip_id)); - (*frag)->fr_flags |= PFFRAG_DROP; - } - - m_freem(m); - return (NULL); -} -#endif /* INET */ - #ifdef INET6 int pf_refragment6(struct ifnet *ifp, struct mbuf **m0, struct m_tag *mtag) { struct mbuf *m = *m0, *t; struct pf_fragment_tag *ftag = (struct pf_fragment_tag *)(mtag + 1); struct pf_pdesc pd; uint32_t frag_id; uint16_t hdrlen, extoff, maxlen; uint8_t proto; int error, action; hdrlen = ftag->ft_hdrlen; extoff = ftag->ft_extoff; maxlen = ftag->ft_maxlen; frag_id = ftag->ft_id; m_tag_delete(m, mtag); mtag = NULL; ftag = NULL; if (extoff) { int off; /* Use protocol from next field of last extension header */ m = m_getptr(m, extoff + offsetof(struct ip6_ext, ip6e_nxt), &off); KASSERT((m != NULL), ("pf_refragment6: short mbuf chain")); proto = *(mtod(m, caddr_t) + off); *(mtod(m, char *) + off) = IPPROTO_FRAGMENT; m = *m0; } else { struct ip6_hdr *hdr; hdr = mtod(m, struct ip6_hdr *); proto = hdr->ip6_nxt; hdr->ip6_nxt = IPPROTO_FRAGMENT; } /* * Maxlen may be less than 8 if there was only a single * fragment. As it was fragmented before, add a fragment * header also for a single fragment. If total or maxlen * is less than 8, ip6_fragment() will return EMSGSIZE and * we drop the packet. */ error = ip6_fragment(ifp, m, hdrlen, proto, maxlen, frag_id); m = (*m0)->m_nextpkt; (*m0)->m_nextpkt = NULL; if (error == 0) { /* The first mbuf contains the unfragmented packet. */ m_freem(*m0); *m0 = NULL; action = PF_PASS; } else { /* Drop expects an mbuf to free. */ DPFPRINTF(("refragment error %d", error)); action = PF_DROP; } for (t = m; m; m = t) { t = m->m_nextpkt; m->m_nextpkt = NULL; m->m_flags |= M_SKIP_FIREWALL; memset(&pd, 0, sizeof(pd)); pd.pf_mtag = pf_find_mtag(m); if (error == 0) ip6_forward(m, 0); else m_freem(m); } return (action); } #endif /* INET6 */ #ifdef INET int pf_normalize_ip(struct mbuf **m0, int dir, struct pfi_kif *kif, u_short *reason, struct pf_pdesc *pd) { struct mbuf *m = *m0; struct pf_rule *r; - struct pf_fragment *frag = NULL; - struct pf_fragment_cmp key; struct ip *h = mtod(m, struct ip *); int mff = (ntohs(h->ip_off) & IP_MF); int hlen = h->ip_hl << 2; u_int16_t fragoff = (ntohs(h->ip_off) & IP_OFFMASK) << 3; u_int16_t max; int ip_len; int ip_off; int tag = -1; int verdict; PF_RULES_RASSERT(); r = TAILQ_FIRST(pf_main_ruleset.rules[PF_RULESET_SCRUB].active.ptr); while (r != NULL) { r->evaluations++; if (pfi_kif_match(r->kif, kif) == r->ifnot) r = r->skip[PF_SKIP_IFP].ptr; else if (r->direction && r->direction != dir) r = r->skip[PF_SKIP_DIR].ptr; else if (r->af && r->af != AF_INET) r = r->skip[PF_SKIP_AF].ptr; else if (r->proto && r->proto != h->ip_p) r = r->skip[PF_SKIP_PROTO].ptr; else if (PF_MISMATCHAW(&r->src.addr, (struct pf_addr *)&h->ip_src.s_addr, AF_INET, r->src.neg, kif, M_GETFIB(m))) r = r->skip[PF_SKIP_SRC_ADDR].ptr; else if (PF_MISMATCHAW(&r->dst.addr, (struct pf_addr *)&h->ip_dst.s_addr, AF_INET, r->dst.neg, NULL, M_GETFIB(m))) r = r->skip[PF_SKIP_DST_ADDR].ptr; else if (r->match_tag && !pf_match_tag(m, r, &tag, pd->pf_mtag ? pd->pf_mtag->tag : 0)) r = TAILQ_NEXT(r, entries); else break; } if (r == NULL || r->action == PF_NOSCRUB) return (PF_PASS); else { r->packets[dir == PF_OUT]++; r->bytes[dir == PF_OUT] += pd->tot_len; } /* Check for illegal packets */ - if (hlen < (int)sizeof(struct ip)) + if (hlen < (int)sizeof(struct ip)) { + REASON_SET(reason, PFRES_NORM); goto drop; + } - if (hlen > ntohs(h->ip_len)) + if (hlen > ntohs(h->ip_len)) { + REASON_SET(reason, PFRES_NORM); goto drop; + } /* Clear IP_DF if the rule uses the no-df option */ if (r->rule_flag & PFRULE_NODF && h->ip_off & htons(IP_DF)) { u_int16_t ip_off = h->ip_off; h->ip_off &= htons(~IP_DF); h->ip_sum = pf_cksum_fixup(h->ip_sum, ip_off, h->ip_off, 0); } /* We will need other tests here */ if (!fragoff && !mff) goto no_fragment; /* We're dealing with a fragment now. Don't allow fragments * with IP_DF to enter the cache. If the flag was cleared by * no-df above, fine. Otherwise drop it. */ if (h->ip_off & htons(IP_DF)) { DPFPRINTF(("IP_DF\n")); goto bad; } ip_len = ntohs(h->ip_len) - hlen; ip_off = (ntohs(h->ip_off) & IP_OFFMASK) << 3; /* All fragments are 8 byte aligned */ if (mff && (ip_len & 0x7)) { DPFPRINTF(("mff and %d\n", ip_len)); goto bad; } /* Respect maximum length */ if (fragoff + ip_len > IP_MAXPACKET) { DPFPRINTF(("max packet %d\n", fragoff + ip_len)); goto bad; } max = fragoff + ip_len; - if ((r->rule_flag & (PFRULE_FRAGCROP|PFRULE_FRAGDROP)) == 0) { + /* Fully buffer all of the fragments + * Might return a completely reassembled mbuf, or NULL */ + PF_FRAG_LOCK(); + DPFPRINTF(("reass frag %d @ %d-%d\n", h->ip_id, fragoff, max)); + verdict = pf_reassemble(m0, h, dir, reason); + PF_FRAG_UNLOCK(); - /* Fully buffer all of the fragments */ - PF_FRAG_LOCK(); + if (verdict != PF_PASS) + return (PF_DROP); - pf_ip2key(h, dir, &key); - frag = pf_find_fragment(&key, &V_pf_frag_tree); + m = *m0; + if (m == NULL) + return (PF_DROP); - /* Check if we saw the last fragment already */ - if (frag != NULL && (frag->fr_flags & PFFRAG_SEENLAST) && - max > frag->fr_max) - goto bad; + h = mtod(m, struct ip *); - /* Might return a completely reassembled mbuf, or NULL */ - DPFPRINTF(("reass frag %d @ %d-%d\n", h->ip_id, fragoff, max)); - verdict = pf_reassemble(m0, h, dir, reason); - PF_FRAG_UNLOCK(); - - if (verdict != PF_PASS) - return (PF_DROP); - - m = *m0; - if (m == NULL) - return (PF_DROP); - - h = mtod(m, struct ip *); - } else { - /* non-buffering fragment cache (drops or masks overlaps) */ - int nomem = 0; - - if (dir == PF_OUT && pd->pf_mtag && - pd->pf_mtag->flags & PF_TAG_FRAGCACHE) { - /* - * Already passed the fragment cache in the - * input direction. If we continued, it would - * appear to be a dup and would be dropped. - */ - goto fragment_pass; - } - - PF_FRAG_LOCK(); - pf_ip2key(h, dir, &key); - frag = pf_find_fragment(&key, &V_pf_cache_tree); - - /* Check if we saw the last fragment already */ - if (frag != NULL && (frag->fr_flags & PFFRAG_SEENLAST) && - max > frag->fr_max) { - if (r->rule_flag & PFRULE_FRAGDROP) - frag->fr_flags |= PFFRAG_DROP; - goto bad; - } - - *m0 = m = pf_fragcache(m0, h, &frag, mff, - (r->rule_flag & PFRULE_FRAGDROP) ? 1 : 0, &nomem); - PF_FRAG_UNLOCK(); - if (m == NULL) { - if (nomem) - goto no_mem; - goto drop; - } - - if (dir == PF_IN) { - /* Use mtag from copied and trimmed mbuf chain. */ - pd->pf_mtag = pf_get_mtag(m); - if (pd->pf_mtag == NULL) { - m_freem(m); - *m0 = NULL; - goto no_mem; - } - pd->pf_mtag->flags |= PF_TAG_FRAGCACHE; - } - - if (frag != NULL && (frag->fr_flags & PFFRAG_DROP)) - goto drop; - goto fragment_pass; - } - no_fragment: /* At this point, only IP_DF is allowed in ip_off */ if (h->ip_off & ~htons(IP_DF)) { u_int16_t ip_off = h->ip_off; h->ip_off &= htons(IP_DF); h->ip_sum = pf_cksum_fixup(h->ip_sum, ip_off, h->ip_off, 0); } - /* not missing a return here */ - - fragment_pass: pf_scrub_ip(&m, r->rule_flag, r->min_ttl, r->set_tos); - if ((r->rule_flag & (PFRULE_FRAGCROP|PFRULE_FRAGDROP)) == 0) - pd->flags |= PFDESC_IP_REAS; return (PF_PASS); - no_mem: - REASON_SET(reason, PFRES_MEMORY); - if (r != NULL && r->log) - PFLOG_PACKET(kif, m, AF_INET, dir, *reason, r, NULL, NULL, pd, - 1); - return (PF_DROP); - - drop: - REASON_SET(reason, PFRES_NORM); - if (r != NULL && r->log) - PFLOG_PACKET(kif, m, AF_INET, dir, *reason, r, NULL, NULL, pd, - 1); - return (PF_DROP); - bad: DPFPRINTF(("dropping bad fragment\n")); - - /* Free associated fragments */ - if (frag != NULL) { - pf_free_fragment(frag); - PF_FRAG_UNLOCK(); - } - REASON_SET(reason, PFRES_FRAG); + drop: if (r != NULL && r->log) PFLOG_PACKET(kif, m, AF_INET, dir, *reason, r, NULL, NULL, pd, 1); return (PF_DROP); } #endif #ifdef INET6 int pf_normalize_ip6(struct mbuf **m0, int dir, struct pfi_kif *kif, u_short *reason, struct pf_pdesc *pd) { struct mbuf *m = *m0; struct pf_rule *r; struct ip6_hdr *h = mtod(m, struct ip6_hdr *); int extoff; int off; struct ip6_ext ext; struct ip6_opt opt; struct ip6_opt_jumbo jumbo; struct ip6_frag frag; u_int32_t jumbolen = 0, plen; int optend; int ooff; u_int8_t proto; int terminal; PF_RULES_RASSERT(); r = TAILQ_FIRST(pf_main_ruleset.rules[PF_RULESET_SCRUB].active.ptr); while (r != NULL) { r->evaluations++; if (pfi_kif_match(r->kif, kif) == r->ifnot) r = r->skip[PF_SKIP_IFP].ptr; else if (r->direction && r->direction != dir) r = r->skip[PF_SKIP_DIR].ptr; else if (r->af && r->af != AF_INET6) r = r->skip[PF_SKIP_AF].ptr; #if 0 /* header chain! */ else if (r->proto && r->proto != h->ip6_nxt) r = r->skip[PF_SKIP_PROTO].ptr; #endif else if (PF_MISMATCHAW(&r->src.addr, (struct pf_addr *)&h->ip6_src, AF_INET6, r->src.neg, kif, M_GETFIB(m))) r = r->skip[PF_SKIP_SRC_ADDR].ptr; else if (PF_MISMATCHAW(&r->dst.addr, (struct pf_addr *)&h->ip6_dst, AF_INET6, r->dst.neg, NULL, M_GETFIB(m))) r = r->skip[PF_SKIP_DST_ADDR].ptr; else break; } if (r == NULL || r->action == PF_NOSCRUB) return (PF_PASS); else { r->packets[dir == PF_OUT]++; r->bytes[dir == PF_OUT] += pd->tot_len; } /* Check for illegal packets */ if (sizeof(struct ip6_hdr) + IPV6_MAXPACKET < m->m_pkthdr.len) goto drop; extoff = 0; off = sizeof(struct ip6_hdr); proto = h->ip6_nxt; terminal = 0; do { switch (proto) { case IPPROTO_FRAGMENT: goto fragment; break; case IPPROTO_AH: case IPPROTO_ROUTING: case IPPROTO_DSTOPTS: if (!pf_pull_hdr(m, off, &ext, sizeof(ext), NULL, NULL, AF_INET6)) goto shortpkt; extoff = off; if (proto == IPPROTO_AH) off += (ext.ip6e_len + 2) * 4; else off += (ext.ip6e_len + 1) * 8; proto = ext.ip6e_nxt; break; case IPPROTO_HOPOPTS: if (!pf_pull_hdr(m, off, &ext, sizeof(ext), NULL, NULL, AF_INET6)) goto shortpkt; extoff = off; optend = off + (ext.ip6e_len + 1) * 8; ooff = off + sizeof(ext); do { if (!pf_pull_hdr(m, ooff, &opt.ip6o_type, sizeof(opt.ip6o_type), NULL, NULL, AF_INET6)) goto shortpkt; if (opt.ip6o_type == IP6OPT_PAD1) { ooff++; continue; } if (!pf_pull_hdr(m, ooff, &opt, sizeof(opt), NULL, NULL, AF_INET6)) goto shortpkt; if (ooff + sizeof(opt) + opt.ip6o_len > optend) goto drop; switch (opt.ip6o_type) { case IP6OPT_JUMBO: if (h->ip6_plen != 0) goto drop; if (!pf_pull_hdr(m, ooff, &jumbo, sizeof(jumbo), NULL, NULL, AF_INET6)) goto shortpkt; memcpy(&jumbolen, jumbo.ip6oj_jumbo_len, sizeof(jumbolen)); jumbolen = ntohl(jumbolen); if (jumbolen <= IPV6_MAXPACKET) goto drop; if (sizeof(struct ip6_hdr) + jumbolen != m->m_pkthdr.len) goto drop; break; default: break; } ooff += sizeof(opt) + opt.ip6o_len; } while (ooff < optend); off = optend; proto = ext.ip6e_nxt; break; default: terminal = 1; break; } } while (!terminal); /* jumbo payload option must be present, or plen > 0 */ if (ntohs(h->ip6_plen) == 0) plen = jumbolen; else plen = ntohs(h->ip6_plen); if (plen == 0) goto drop; if (sizeof(struct ip6_hdr) + plen > m->m_pkthdr.len) goto shortpkt; pf_scrub_ip6(&m, r->min_ttl); return (PF_PASS); fragment: /* Jumbo payload packets cannot be fragmented. */ plen = ntohs(h->ip6_plen); if (plen == 0 || jumbolen) goto drop; if (sizeof(struct ip6_hdr) + plen > m->m_pkthdr.len) goto shortpkt; if (!pf_pull_hdr(m, off, &frag, sizeof(frag), NULL, NULL, AF_INET6)) goto shortpkt; /* Offset now points to data portion. */ off += sizeof(frag); /* Returns PF_DROP or *m0 is NULL or completely reassembled mbuf. */ if (pf_reassemble6(m0, h, &frag, off, extoff, reason) != PF_PASS) return (PF_DROP); m = *m0; if (m == NULL) return (PF_DROP); pd->flags |= PFDESC_IP_REAS; return (PF_PASS); shortpkt: REASON_SET(reason, PFRES_SHORT); if (r != NULL && r->log) PFLOG_PACKET(kif, m, AF_INET6, dir, *reason, r, NULL, NULL, pd, 1); return (PF_DROP); drop: REASON_SET(reason, PFRES_NORM); if (r != NULL && r->log) PFLOG_PACKET(kif, m, AF_INET6, dir, *reason, r, NULL, NULL, pd, 1); return (PF_DROP); } #endif /* INET6 */ int pf_normalize_tcp(int dir, struct pfi_kif *kif, struct mbuf *m, int ipoff, int off, void *h, struct pf_pdesc *pd) { struct pf_rule *r, *rm = NULL; struct tcphdr *th = pd->hdr.tcp; int rewrite = 0; u_short reason; u_int8_t flags; sa_family_t af = pd->af; PF_RULES_RASSERT(); r = TAILQ_FIRST(pf_main_ruleset.rules[PF_RULESET_SCRUB].active.ptr); while (r != NULL) { r->evaluations++; if (pfi_kif_match(r->kif, kif) == r->ifnot) r = r->skip[PF_SKIP_IFP].ptr; else if (r->direction && r->direction != dir) r = r->skip[PF_SKIP_DIR].ptr; else if (r->af && r->af != af) r = r->skip[PF_SKIP_AF].ptr; else if (r->proto && r->proto != pd->proto) r = r->skip[PF_SKIP_PROTO].ptr; else if (PF_MISMATCHAW(&r->src.addr, pd->src, af, r->src.neg, kif, M_GETFIB(m))) r = r->skip[PF_SKIP_SRC_ADDR].ptr; else if (r->src.port_op && !pf_match_port(r->src.port_op, r->src.port[0], r->src.port[1], th->th_sport)) r = r->skip[PF_SKIP_SRC_PORT].ptr; else if (PF_MISMATCHAW(&r->dst.addr, pd->dst, af, r->dst.neg, NULL, M_GETFIB(m))) r = r->skip[PF_SKIP_DST_ADDR].ptr; else if (r->dst.port_op && !pf_match_port(r->dst.port_op, r->dst.port[0], r->dst.port[1], th->th_dport)) r = r->skip[PF_SKIP_DST_PORT].ptr; else if (r->os_fingerprint != PF_OSFP_ANY && !pf_osfp_match( pf_osfp_fingerprint(pd, m, off, th), r->os_fingerprint)) r = TAILQ_NEXT(r, entries); else { rm = r; break; } } if (rm == NULL || rm->action == PF_NOSCRUB) return (PF_PASS); else { r->packets[dir == PF_OUT]++; r->bytes[dir == PF_OUT] += pd->tot_len; } if (rm->rule_flag & PFRULE_REASSEMBLE_TCP) pd->flags |= PFDESC_TCP_NORM; flags = th->th_flags; if (flags & TH_SYN) { /* Illegal packet */ if (flags & TH_RST) goto tcp_drop; if (flags & TH_FIN) goto tcp_drop; } else { /* Illegal packet */ if (!(flags & (TH_ACK|TH_RST))) goto tcp_drop; } if (!(flags & TH_ACK)) { /* These flags are only valid if ACK is set */ if ((flags & TH_FIN) || (flags & TH_PUSH) || (flags & TH_URG)) goto tcp_drop; } /* Check for illegal header length */ if (th->th_off < (sizeof(struct tcphdr) >> 2)) goto tcp_drop; /* If flags changed, or reserved data set, then adjust */ if (flags != th->th_flags || th->th_x2 != 0) { u_int16_t ov, nv; ov = *(u_int16_t *)(&th->th_ack + 1); th->th_flags = flags; th->th_x2 = 0; nv = *(u_int16_t *)(&th->th_ack + 1); th->th_sum = pf_cksum_fixup(th->th_sum, ov, nv, 0); rewrite = 1; } /* Remove urgent pointer, if TH_URG is not set */ if (!(flags & TH_URG) && th->th_urp) { th->th_sum = pf_cksum_fixup(th->th_sum, th->th_urp, 0, 0); th->th_urp = 0; rewrite = 1; } /* Process options */ if (r->max_mss && pf_normalize_tcpopt(r, m, th, off, pd->af)) rewrite = 1; /* copy back packet headers if we sanitized */ if (rewrite) m_copyback(m, off, sizeof(*th), (caddr_t)th); return (PF_PASS); tcp_drop: REASON_SET(&reason, PFRES_NORM); if (rm != NULL && r->log) PFLOG_PACKET(kif, m, AF_INET, dir, reason, r, NULL, NULL, pd, 1); return (PF_DROP); } int pf_normalize_tcp_init(struct mbuf *m, int off, struct pf_pdesc *pd, struct tcphdr *th, struct pf_state_peer *src, struct pf_state_peer *dst) { u_int32_t tsval, tsecr; u_int8_t hdr[60]; u_int8_t *opt; KASSERT((src->scrub == NULL), ("pf_normalize_tcp_init: src->scrub != NULL")); src->scrub = uma_zalloc(V_pf_state_scrub_z, M_ZERO | M_NOWAIT); if (src->scrub == NULL) return (1); switch (pd->af) { #ifdef INET case AF_INET: { struct ip *h = mtod(m, struct ip *); src->scrub->pfss_ttl = h->ip_ttl; break; } #endif /* INET */ #ifdef INET6 case AF_INET6: { struct ip6_hdr *h = mtod(m, struct ip6_hdr *); src->scrub->pfss_ttl = h->ip6_hlim; break; } #endif /* INET6 */ } /* * All normalizations below are only begun if we see the start of * the connections. They must all set an enabled bit in pfss_flags */ if ((th->th_flags & TH_SYN) == 0) return (0); if (th->th_off > (sizeof(struct tcphdr) >> 2) && src->scrub && pf_pull_hdr(m, off, hdr, th->th_off << 2, NULL, NULL, pd->af)) { /* Diddle with TCP options */ int hlen; opt = hdr + sizeof(struct tcphdr); hlen = (th->th_off << 2) - sizeof(struct tcphdr); while (hlen >= TCPOLEN_TIMESTAMP) { switch (*opt) { case TCPOPT_EOL: /* FALLTHROUGH */ case TCPOPT_NOP: opt++; hlen--; break; case TCPOPT_TIMESTAMP: if (opt[1] >= TCPOLEN_TIMESTAMP) { src->scrub->pfss_flags |= PFSS_TIMESTAMP; src->scrub->pfss_ts_mod = htonl(arc4random()); /* note PFSS_PAWS not set yet */ memcpy(&tsval, &opt[2], sizeof(u_int32_t)); memcpy(&tsecr, &opt[6], sizeof(u_int32_t)); src->scrub->pfss_tsval0 = ntohl(tsval); src->scrub->pfss_tsval = ntohl(tsval); src->scrub->pfss_tsecr = ntohl(tsecr); getmicrouptime(&src->scrub->pfss_last); } /* FALLTHROUGH */ default: hlen -= MAX(opt[1], 2); opt += MAX(opt[1], 2); break; } } } return (0); } void pf_normalize_tcp_cleanup(struct pf_state *state) { if (state->src.scrub) uma_zfree(V_pf_state_scrub_z, state->src.scrub); if (state->dst.scrub) uma_zfree(V_pf_state_scrub_z, state->dst.scrub); /* Someday... flush the TCP segment reassembly descriptors. */ } int pf_normalize_tcp_stateful(struct mbuf *m, int off, struct pf_pdesc *pd, u_short *reason, struct tcphdr *th, struct pf_state *state, struct pf_state_peer *src, struct pf_state_peer *dst, int *writeback) { struct timeval uptime; u_int32_t tsval, tsecr; u_int tsval_from_last; u_int8_t hdr[60]; u_int8_t *opt; int copyback = 0; int got_ts = 0; KASSERT((src->scrub || dst->scrub), ("%s: src->scrub && dst->scrub!", __func__)); /* * Enforce the minimum TTL seen for this connection. Negate a common * technique to evade an intrusion detection system and confuse * firewall state code. */ switch (pd->af) { #ifdef INET case AF_INET: { if (src->scrub) { struct ip *h = mtod(m, struct ip *); if (h->ip_ttl > src->scrub->pfss_ttl) src->scrub->pfss_ttl = h->ip_ttl; h->ip_ttl = src->scrub->pfss_ttl; } break; } #endif /* INET */ #ifdef INET6 case AF_INET6: { if (src->scrub) { struct ip6_hdr *h = mtod(m, struct ip6_hdr *); if (h->ip6_hlim > src->scrub->pfss_ttl) src->scrub->pfss_ttl = h->ip6_hlim; h->ip6_hlim = src->scrub->pfss_ttl; } break; } #endif /* INET6 */ } if (th->th_off > (sizeof(struct tcphdr) >> 2) && ((src->scrub && (src->scrub->pfss_flags & PFSS_TIMESTAMP)) || (dst->scrub && (dst->scrub->pfss_flags & PFSS_TIMESTAMP))) && pf_pull_hdr(m, off, hdr, th->th_off << 2, NULL, NULL, pd->af)) { /* Diddle with TCP options */ int hlen; opt = hdr + sizeof(struct tcphdr); hlen = (th->th_off << 2) - sizeof(struct tcphdr); while (hlen >= TCPOLEN_TIMESTAMP) { switch (*opt) { case TCPOPT_EOL: /* FALLTHROUGH */ case TCPOPT_NOP: opt++; hlen--; break; case TCPOPT_TIMESTAMP: /* Modulate the timestamps. Can be used for * NAT detection, OS uptime determination or * reboot detection. */ if (got_ts) { /* Huh? Multiple timestamps!? */ if (V_pf_status.debug >= PF_DEBUG_MISC) { DPFPRINTF(("multiple TS??")); pf_print_state(state); printf("\n"); } REASON_SET(reason, PFRES_TS); return (PF_DROP); } if (opt[1] >= TCPOLEN_TIMESTAMP) { memcpy(&tsval, &opt[2], sizeof(u_int32_t)); if (tsval && src->scrub && (src->scrub->pfss_flags & PFSS_TIMESTAMP)) { tsval = ntohl(tsval); pf_change_a(&opt[2], &th->th_sum, htonl(tsval + src->scrub->pfss_ts_mod), 0); copyback = 1; } /* Modulate TS reply iff valid (!0) */ memcpy(&tsecr, &opt[6], sizeof(u_int32_t)); if (tsecr && dst->scrub && (dst->scrub->pfss_flags & PFSS_TIMESTAMP)) { tsecr = ntohl(tsecr) - dst->scrub->pfss_ts_mod; pf_change_a(&opt[6], &th->th_sum, htonl(tsecr), 0); copyback = 1; } got_ts = 1; } /* FALLTHROUGH */ default: hlen -= MAX(opt[1], 2); opt += MAX(opt[1], 2); break; } } if (copyback) { /* Copyback the options, caller copys back header */ *writeback = 1; m_copyback(m, off + sizeof(struct tcphdr), (th->th_off << 2) - sizeof(struct tcphdr), hdr + sizeof(struct tcphdr)); } } /* * Must invalidate PAWS checks on connections idle for too long. * The fastest allowed timestamp clock is 1ms. That turns out to * be about 24 days before it wraps. XXX Right now our lowerbound * TS echo check only works for the first 12 days of a connection * when the TS has exhausted half its 32bit space */ #define TS_MAX_IDLE (24*24*60*60) #define TS_MAX_CONN (12*24*60*60) /* XXX remove when better tsecr check */ getmicrouptime(&uptime); if (src->scrub && (src->scrub->pfss_flags & PFSS_PAWS) && (uptime.tv_sec - src->scrub->pfss_last.tv_sec > TS_MAX_IDLE || time_uptime - state->creation > TS_MAX_CONN)) { if (V_pf_status.debug >= PF_DEBUG_MISC) { DPFPRINTF(("src idled out of PAWS\n")); pf_print_state(state); printf("\n"); } src->scrub->pfss_flags = (src->scrub->pfss_flags & ~PFSS_PAWS) | PFSS_PAWS_IDLED; } if (dst->scrub && (dst->scrub->pfss_flags & PFSS_PAWS) && uptime.tv_sec - dst->scrub->pfss_last.tv_sec > TS_MAX_IDLE) { if (V_pf_status.debug >= PF_DEBUG_MISC) { DPFPRINTF(("dst idled out of PAWS\n")); pf_print_state(state); printf("\n"); } dst->scrub->pfss_flags = (dst->scrub->pfss_flags & ~PFSS_PAWS) | PFSS_PAWS_IDLED; } if (got_ts && src->scrub && dst->scrub && (src->scrub->pfss_flags & PFSS_PAWS) && (dst->scrub->pfss_flags & PFSS_PAWS)) { /* Validate that the timestamps are "in-window". * RFC1323 describes TCP Timestamp options that allow * measurement of RTT (round trip time) and PAWS * (protection against wrapped sequence numbers). PAWS * gives us a set of rules for rejecting packets on * long fat pipes (packets that were somehow delayed * in transit longer than the time it took to send the * full TCP sequence space of 4Gb). We can use these * rules and infer a few others that will let us treat * the 32bit timestamp and the 32bit echoed timestamp * as sequence numbers to prevent a blind attacker from * inserting packets into a connection. * * RFC1323 tells us: * - The timestamp on this packet must be greater than * or equal to the last value echoed by the other * endpoint. The RFC says those will be discarded * since it is a dup that has already been acked. * This gives us a lowerbound on the timestamp. * timestamp >= other last echoed timestamp * - The timestamp will be less than or equal to * the last timestamp plus the time between the * last packet and now. The RFC defines the max * clock rate as 1ms. We will allow clocks to be * up to 10% fast and will allow a total difference * or 30 seconds due to a route change. And this * gives us an upperbound on the timestamp. * timestamp <= last timestamp + max ticks * We have to be careful here. Windows will send an * initial timestamp of zero and then initialize it * to a random value after the 3whs; presumably to * avoid a DoS by having to call an expensive RNG * during a SYN flood. Proof MS has at least one * good security geek. * * - The TCP timestamp option must also echo the other * endpoints timestamp. The timestamp echoed is the * one carried on the earliest unacknowledged segment * on the left edge of the sequence window. The RFC * states that the host will reject any echoed * timestamps that were larger than any ever sent. * This gives us an upperbound on the TS echo. * tescr <= largest_tsval * - The lowerbound on the TS echo is a little more * tricky to determine. The other endpoint's echoed * values will not decrease. But there may be * network conditions that re-order packets and * cause our view of them to decrease. For now the * only lowerbound we can safely determine is that * the TS echo will never be less than the original * TS. XXX There is probably a better lowerbound. * Remove TS_MAX_CONN with better lowerbound check. * tescr >= other original TS * * It is also important to note that the fastest * timestamp clock of 1ms will wrap its 32bit space in * 24 days. So we just disable TS checking after 24 * days of idle time. We actually must use a 12d * connection limit until we can come up with a better * lowerbound to the TS echo check. */ struct timeval delta_ts; int ts_fudge; /* * PFTM_TS_DIFF is how many seconds of leeway to allow * a host's timestamp. This can happen if the previous * packet got delayed in transit for much longer than * this packet. */ if ((ts_fudge = state->rule.ptr->timeout[PFTM_TS_DIFF]) == 0) ts_fudge = V_pf_default_rule.timeout[PFTM_TS_DIFF]; /* Calculate max ticks since the last timestamp */ #define TS_MAXFREQ 1100 /* RFC max TS freq of 1Khz + 10% skew */ #define TS_MICROSECS 1000000 /* microseconds per second */ delta_ts = uptime; timevalsub(&delta_ts, &src->scrub->pfss_last); tsval_from_last = (delta_ts.tv_sec + ts_fudge) * TS_MAXFREQ; tsval_from_last += delta_ts.tv_usec / (TS_MICROSECS/TS_MAXFREQ); if ((src->state >= TCPS_ESTABLISHED && dst->state >= TCPS_ESTABLISHED) && (SEQ_LT(tsval, dst->scrub->pfss_tsecr) || SEQ_GT(tsval, src->scrub->pfss_tsval + tsval_from_last) || (tsecr && (SEQ_GT(tsecr, dst->scrub->pfss_tsval) || SEQ_LT(tsecr, dst->scrub->pfss_tsval0))))) { /* Bad RFC1323 implementation or an insertion attack. * * - Solaris 2.6 and 2.7 are known to send another ACK * after the FIN,FIN|ACK,ACK closing that carries * an old timestamp. */ DPFPRINTF(("Timestamp failed %c%c%c%c\n", SEQ_LT(tsval, dst->scrub->pfss_tsecr) ? '0' : ' ', SEQ_GT(tsval, src->scrub->pfss_tsval + tsval_from_last) ? '1' : ' ', SEQ_GT(tsecr, dst->scrub->pfss_tsval) ? '2' : ' ', SEQ_LT(tsecr, dst->scrub->pfss_tsval0)? '3' : ' ')); DPFPRINTF((" tsval: %u tsecr: %u +ticks: %u " "idle: %jus %lums\n", tsval, tsecr, tsval_from_last, (uintmax_t)delta_ts.tv_sec, delta_ts.tv_usec / 1000)); DPFPRINTF((" src->tsval: %u tsecr: %u\n", src->scrub->pfss_tsval, src->scrub->pfss_tsecr)); DPFPRINTF((" dst->tsval: %u tsecr: %u tsval0: %u" "\n", dst->scrub->pfss_tsval, dst->scrub->pfss_tsecr, dst->scrub->pfss_tsval0)); if (V_pf_status.debug >= PF_DEBUG_MISC) { pf_print_state(state); pf_print_flags(th->th_flags); printf("\n"); } REASON_SET(reason, PFRES_TS); return (PF_DROP); } /* XXX I'd really like to require tsecr but it's optional */ } else if (!got_ts && (th->th_flags & TH_RST) == 0 && ((src->state == TCPS_ESTABLISHED && dst->state == TCPS_ESTABLISHED) || pd->p_len > 0 || (th->th_flags & TH_SYN)) && src->scrub && dst->scrub && (src->scrub->pfss_flags & PFSS_PAWS) && (dst->scrub->pfss_flags & PFSS_PAWS)) { /* Didn't send a timestamp. Timestamps aren't really useful * when: * - connection opening or closing (often not even sent). * but we must not let an attacker to put a FIN on a * data packet to sneak it through our ESTABLISHED check. * - on a TCP reset. RFC suggests not even looking at TS. * - on an empty ACK. The TS will not be echoed so it will * probably not help keep the RTT calculation in sync and * there isn't as much danger when the sequence numbers * got wrapped. So some stacks don't include TS on empty * ACKs :-( * * To minimize the disruption to mostly RFC1323 conformant * stacks, we will only require timestamps on data packets. * * And what do ya know, we cannot require timestamps on data * packets. There appear to be devices that do legitimate * TCP connection hijacking. There are HTTP devices that allow * a 3whs (with timestamps) and then buffer the HTTP request. * If the intermediate device has the HTTP response cache, it * will spoof the response but not bother timestamping its * packets. So we can look for the presence of a timestamp in * the first data packet and if there, require it in all future * packets. */ if (pd->p_len > 0 && (src->scrub->pfss_flags & PFSS_DATA_TS)) { /* * Hey! Someone tried to sneak a packet in. Or the * stack changed its RFC1323 behavior?!?! */ if (V_pf_status.debug >= PF_DEBUG_MISC) { DPFPRINTF(("Did not receive expected RFC1323 " "timestamp\n")); pf_print_state(state); pf_print_flags(th->th_flags); printf("\n"); } REASON_SET(reason, PFRES_TS); return (PF_DROP); } } /* * We will note if a host sends his data packets with or without * timestamps. And require all data packets to contain a timestamp * if the first does. PAWS implicitly requires that all data packets be * timestamped. But I think there are middle-man devices that hijack * TCP streams immediately after the 3whs and don't timestamp their * packets (seen in a WWW accelerator or cache). */ if (pd->p_len > 0 && src->scrub && (src->scrub->pfss_flags & (PFSS_TIMESTAMP|PFSS_DATA_TS|PFSS_DATA_NOTS)) == PFSS_TIMESTAMP) { if (got_ts) src->scrub->pfss_flags |= PFSS_DATA_TS; else { src->scrub->pfss_flags |= PFSS_DATA_NOTS; if (V_pf_status.debug >= PF_DEBUG_MISC && dst->scrub && (dst->scrub->pfss_flags & PFSS_TIMESTAMP)) { /* Don't warn if other host rejected RFC1323 */ DPFPRINTF(("Broken RFC1323 stack did not " "timestamp data packet. Disabled PAWS " "security.\n")); pf_print_state(state); pf_print_flags(th->th_flags); printf("\n"); } } } /* * Update PAWS values */ if (got_ts && src->scrub && PFSS_TIMESTAMP == (src->scrub->pfss_flags & (PFSS_PAWS_IDLED|PFSS_TIMESTAMP))) { getmicrouptime(&src->scrub->pfss_last); if (SEQ_GEQ(tsval, src->scrub->pfss_tsval) || (src->scrub->pfss_flags & PFSS_PAWS) == 0) src->scrub->pfss_tsval = tsval; if (tsecr) { if (SEQ_GEQ(tsecr, src->scrub->pfss_tsecr) || (src->scrub->pfss_flags & PFSS_PAWS) == 0) src->scrub->pfss_tsecr = tsecr; if ((src->scrub->pfss_flags & PFSS_PAWS) == 0 && (SEQ_LT(tsval, src->scrub->pfss_tsval0) || src->scrub->pfss_tsval0 == 0)) { /* tsval0 MUST be the lowest timestamp */ src->scrub->pfss_tsval0 = tsval; } /* Only fully initialized after a TS gets echoed */ if ((src->scrub->pfss_flags & PFSS_PAWS) == 0) src->scrub->pfss_flags |= PFSS_PAWS; } } /* I have a dream.... TCP segment reassembly.... */ return (0); } static int pf_normalize_tcpopt(struct pf_rule *r, struct mbuf *m, struct tcphdr *th, int off, sa_family_t af) { u_int16_t *mss; int thoff; int opt, cnt, optlen = 0; int rewrite = 0; u_char opts[TCP_MAXOLEN]; u_char *optp = opts; thoff = th->th_off << 2; cnt = thoff - sizeof(struct tcphdr); if (cnt > 0 && !pf_pull_hdr(m, off + sizeof(*th), opts, cnt, NULL, NULL, af)) return (rewrite); for (; cnt > 0; cnt -= optlen, optp += optlen) { opt = optp[0]; if (opt == TCPOPT_EOL) break; if (opt == TCPOPT_NOP) optlen = 1; else { if (cnt < 2) break; optlen = optp[1]; if (optlen < 2 || optlen > cnt) break; } switch (opt) { case TCPOPT_MAXSEG: mss = (u_int16_t *)(optp + 2); if ((ntohs(*mss)) > r->max_mss) { th->th_sum = pf_cksum_fixup(th->th_sum, *mss, htons(r->max_mss), 0); *mss = htons(r->max_mss); rewrite = 1; } break; default: break; } } if (rewrite) m_copyback(m, off + sizeof(*th), thoff - sizeof(*th), opts); return (rewrite); } #ifdef INET static void pf_scrub_ip(struct mbuf **m0, u_int32_t flags, u_int8_t min_ttl, u_int8_t tos) { struct mbuf *m = *m0; struct ip *h = mtod(m, struct ip *); /* Clear IP_DF if no-df was requested */ if (flags & PFRULE_NODF && h->ip_off & htons(IP_DF)) { u_int16_t ip_off = h->ip_off; h->ip_off &= htons(~IP_DF); h->ip_sum = pf_cksum_fixup(h->ip_sum, ip_off, h->ip_off, 0); } /* Enforce a minimum ttl, may cause endless packet loops */ if (min_ttl && h->ip_ttl < min_ttl) { u_int16_t ip_ttl = h->ip_ttl; h->ip_ttl = min_ttl; h->ip_sum = pf_cksum_fixup(h->ip_sum, ip_ttl, h->ip_ttl, 0); } /* Enforce tos */ if (flags & PFRULE_SET_TOS) { u_int16_t ov, nv; ov = *(u_int16_t *)h; h->ip_tos = tos; nv = *(u_int16_t *)h; h->ip_sum = pf_cksum_fixup(h->ip_sum, ov, nv, 0); } /* random-id, but not for fragments */ if (flags & PFRULE_RANDOMID && !(h->ip_off & ~htons(IP_DF))) { uint16_t ip_id = h->ip_id; ip_fillid(h); h->ip_sum = pf_cksum_fixup(h->ip_sum, ip_id, h->ip_id, 0); } } #endif /* INET */ #ifdef INET6 static void pf_scrub_ip6(struct mbuf **m0, u_int8_t min_ttl) { struct mbuf *m = *m0; struct ip6_hdr *h = mtod(m, struct ip6_hdr *); /* Enforce a minimum ttl, may cause endless packet loops */ if (min_ttl && h->ip6_hlim < min_ttl) h->ip6_hlim = min_ttl; } #endif