| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Contrary to the comment and expected output, nft does _not_ eliminate
the redundant `ip protocol` expression from the second test. Dependency
elimination requires a higher level expression. `ip saddr` cannot lead
to the elimination of `ip protocol` since they are both L3 expressions.
`tcp dport` cannot because although `ip saddr` and `ip protocol` both
imply that the L3 protocol is `ip`, only protocol matches are stored as
dependencies, so the redundancy is not apparent, and in fact,
`payload_may_dependency_kill` explicitly checks for the combination of
inet, bridge or netdev family, L4 expression and L3 ipv4 or ipv6
dependency and returns false.
Correct the expected output and comment.
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
| |
In one of the bridge payloads, the wrong command is given to load the
protocol.
[ fw@strlen.de: remove the duplicated netdev payload ]
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
| |
The netdev payload for one of the inet/sets.t tests was cut-and-pasted
from the inet payload without being properly updated.
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When passing no upper size limit, the dynset expression forces
an internal 64k upperlimit.
In some cases, this can result in 'nft -f' to restore the ruleset.
Avoid this by always setting the EVAL flag on a set definition when
we encounter packet-path update attempt in the batch.
Reported-by: Yi Chen <yiche@redhat.com>
Suggested-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Devices with interface names starting with a digit can not be configured
in flowtables. Trying to do so throws the following error:
Error: syntax error, unexpected number, expecting comma or '}'
devices = { eth0, 6in4-wan6 };
This is however a perfectly valid interface name. Solve the issue by
allowing the use of quoted strings.
Suggested-by: Jo-Philipp Wich <jo@mein.io>
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
| |
automake recommends shipping the output of bison and lex in distribution
tar-balls and runs bison and lex during `make dist` (this has the
advantage that end-users don't need to have bison or lex installed to
compile the software). Accordingly, automake also recommends removing
these files with `make maintainer-clean` and generates rules to do so.
Therefore, remove scanner.c and parser_bison.c from `CLEANFILES`.
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
Move the check for NULL protocol description away from the loop to avoid
too long line.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
| |
Add a few tests for the catchall features and maps.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
autoconf complains about three obsolete macros.
`AC_CONFIG_HEADER` has been superseded by `AC_CONFIG_HEADERS`, so
replace it.
`AM_PROG_LEX` calls `AC_PROG_LEX` with no arguments, but this usage is
deprecated. The only difference between `AM_PROG_LEX` and `AC_PROG_LEX`
is that the former defines `$LEX` as "./build-aux/missing lex" if no lex
is found to ensure a useful error is reported when make is run. How-
ever, the configure script checks that we have a working lex and exits
with an error if none is available, so `$LEX` will never be called and
we can replace `AM_PROG_LEX` with `AC_PROG_LEX`.
`AM_PROG_LIBTOOL` has been superseded by `LT_INIT`, which is already in
configure.ac, so remove it.
We can also replace `AC_DISABLE_STATIC` with an argument to `LT_INIT`.
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
0040mark_shift_0 was passing --debug=eval to nft. Remove it.
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When we are evaluating a `reject` statement in the `inet` family, we may
have `ether` and `ip` or `ip6` as the L2 and L3 protocols in the
evaluation context:
table inet filter {
chain input {
type filter hook input priority filter;
ether saddr aa:bb:cc:dd:ee:ff ip daddr 192.168.0.1 reject
}
}
Since no `reject` option is given, nft attempts to infer one and fails:
BUG: unsupported familynft: evaluate.c:2766:stmt_evaluate_reject_inet_family: Assertion `0' failed.
Aborted
The reason it fails is that the ethernet protocol numbers for IPv4 and
IPv6 (`ETH_P_IP` and `ETH_P_IPV6`) do not match `NFPROTO_IPV4` and
`NFPROTO_IPV6`. Add support for the ethernet protocol numbers.
Replace the current `BUG("unsupported family")` error message with
something more informative that tells the user to provide an explicit
reject option.
Add a Python test case.
Fixes: 5fdd0b6a0600 ("nft: complete reject support")
Link: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1001360
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
There are a couple of mistakes in comments. Fix them.
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
Each `struct proto_desc` contains a fixed-size array of higher layer
protocols. Only the first few are not NULL. Therefore, we can stop
iterating over the array once we reach a NULL member.
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Remove shifts-by-0. These can occur after binop postprocessing
has adjusted the RHS value to account for a mask operation.
Example: frag frag-off @s4
Is internally represented via:
[ exthdr load ipv6 2b @ 44 + 2 => reg 1 ]
[ bitwise reg 1 = ( reg 1 & 0x0000f8ff ) ^ 0x00000000 ]
[ bitwise reg 1 = ( reg 1 >> 0x00000003 ) ]
[ lookup reg 1 set s ]
First binop masks out unwanted parts of the 16-bit field.
Second binop needs to left-shift so that lookups in the set will work.
When decoding, the first binop is removed after the exthdr load
has been adjusted accordingly. Constant propagation adjusts the
shift-value to 0 on removal. This change then gets rid of the
shift-by-0 entirely.
After this change, 'frag frag-off @s4' input is shown as-is.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Before this patch:
in: frag frag-off @s4
in: ip version @s8
out: (@nh,0,8 & 0xf0) >> 4 == @s8
out: (frag unknown & 0xfff8 [invalid type]) >> 3 == @s4
after:
out: frag frag-off >> 0 == @s4
out: ip version >> 0 == @s8
Next patch adds support for zero-shift removal.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
meta templates are only there for ease of use (input/parsing).
When listing, they should be ignored:
set s4 { typeof ip version elements = { 1, } }
chain c4 { ip version @s4 accept }
gets listed as 'ip l4proto ...' which is nonsensical.
after this patch we get:
in: ip version @s4
out: (@nh,0,8 & 0xf0) >> 4 == @s4
.. which is (marginally) better.
Next patch adds support for payload decoding.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
These tests work, but I omitted a few lines that do not:
in: frag frag-off @s4 accept
in: ip version @s8
out: (frag unknown & 0xfff8 [invalid type]) >> 3 == @s4
out: (ip l4proto & pfsync) >> 4 == @s8
Next patches resolve this.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Wider testing has shown that 128 kB stack is too low (e.g. for systems
with 64 kB page size), leading to false failures in some environments.
Based on results from a matrix of RHEL 8 and RHEL 9 systems across
x86_64, aarch64, ppc64le and s390x architectures as well as some
anecdotal testing of other Linux distros on x86_64 machines, 400 kB
seems safe: the normal nft stack (which should stay constant during
this test) on all tested systems doesn't exceed 200 kB (stays around
100 kB on typical systems with 4 kB page size), while always growing
beyond 500 kB in the failing case (nftables before baecd1cf2685) with
the increased set size.
Fixes: d8ccad2a2b73 ("tests: cover baecd1cf2685 ("segtree: Fix segfault when restoring a huge interval set")")
Signed-off-by: Štěpán Němec <snemec@redhat.com>
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
% nft describe ip option rr value
segmentation fault
after this fix, this exits with 'Error: unknown ip option type/field'.
Problem is that 'rr' doesn't have a value template, so the template struct is
all-zeroes, so we crash when trying to use tmpl->dtype (its NULL).
Furthermore, expr_describe tries to print expr->identifier but expr is
exthdr, not symbol: ->identifier contains garbage.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
| |
This did not store the 'op' member and listing always treated this as ipv6
extension header.
Add test cases for this.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
| |
Its always 0, so remove it.
Looks like this was intended to support variable options that have
array-like members, but so far this isn't implemented, better remove
dead code and implement it properly when such support is needed.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
| |
Extend nft_cache_filter to hold a flowtable name so 'list flowtable'
command causes fetching the requested flowtable only.
Dump flowtables just once instead of for each table, merely assign
fetched data to tables inside the loop.
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
| |
Fetch either all tables' sets at once, a specific table's sets or even a
specific set if needed instead of iterating over the list of previously
fetched tables and fetching for each, then ignoring anything returned
that doesn't match the filter.
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When operating on a specific chain, add payload to NFT_MSG_GETCHAIN so
kernel returns only relevant data. Since ENOENT is an expected return
code, do not treat this as error.
While being at it, improve code in chain_cache_cb() a bit:
- Check chain's family first, it is a less expensive check than
comparing table names.
- Do not extract chain name of uninteresting chains.
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
|
|
| |
Instead of fetching all existing rules in kernel's ruleset and filtering
in user space, add payload to the dump request specifying the table and
chain to filter for.
Since list_rule_cb() no longer needs the filter, pass only netlink_ctx
to the callback and drop struct rule_cache_dump_ctx.
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
|
|
| |
Instead of requesting a dump of all tables and filtering the data in
user space, construct a non-dump request if filter contains a table so
kernel returns only that single table.
This should improve nft performance in rulesets with many tables
present.
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
| |
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Unify binop handling for ipv6 extension header, ip option and tcp option
processing.
Pass the real offset and length expected, not the one used in the kernel.
This was already done for extension headers and ip options, but tcp
option parsing did not do this.
This was fine before because no existing tcp option template
had a non-byte sized member.
With mptcp addition this isn't the case anymore, subtype field is
only 4 bits wide, but tcp option delinearization passed 8bits instead.
Pass the offset and mask delta, just like ip option/ipv6 exthdr.
This makes nft show 'tcp option mptcp subtype 1' instead of
'tcp option mptcp unknown & 240 == 16'.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
MPTCP multiplexes the various mptcp signalling data using the
first 4 bits of the mptcp option.
This allows to match on the mptcp subtype via:
tcp option mptcp subtype 1
This misses delinearization support. mptcp subtype is the first tcp
option field that has a length of less than one byte.
Serialization processing will add a binop for this, but netlink
delinearization can't remove them, yet.
Also misses a new datatype/symbol table to allow to use mnemonics like
'mp_join' instead of raw numbers.
For this reason, no tests are added yet.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
| |
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
| |
Allow to use "fastopen", "md5sig" and "mptcp" mnemonics rather than the
raw option numbers.
These new keywords are only recognized while scanner is in tcp state.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
At this time the parser will accept nonsensical input like
tcp option mss left 2
which will be treated as 'tcp option maxseg size 2'.
This is because the enum space overlaps.
Split the rules so that 'tcp option mss' will only
accept field names specific to the mss/maxseg option kind.
Signed-off-by: Florian Westphal <fw@strlen.de>
(cherry picked from commit 46168852c03d73c29b557c93029dc512ca6e233a)
|
|
|
|
|
|
|
|
| |
This moves tcp options not used anywhere else (e.g. in synproxy) to a
distinct scope. This will also allow to avoid exposing new option
keywords in the ruleset context.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
tcp option <foo> kind ... never makes any sense, as "tcp option <foo>"
already tells the kernel to look for the foo <kind>.
"tcp option sack kind 5" matches if the sack option is present; its a
more complicated form of the simpler "tcp option sack exists".
"tcp option sack kind 1" (or any other value than 5) will never match.
So remove this.
Test cases are converted to "exists".
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
| |
This function can be called for different expression types, including
some (EXPR_MAP) where expr->left/right alias to different member
variables.
This makes accesses to those members conditional by checking the
expression type ahead of the access.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
|
| |
relational_binop_postprocess() is called for EXPR_RELATIONAL,
so "expr->right" is safe to use.
But the RHS can be something other than a value.
This has been extended to handle other types, so rename to 'right'.
No code changes intended.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
| |
expr is a map, so this should use expr->map, not expr->left.
These fields are aliased, so this would break if that is ever changed.
Signed-off-by: Florian Westphal <fw@strlen.de>
|
|
|
|
|
|
|
|
|
|
|
| |
Missing call to cli_exit() to save the history when ctrl-d is pressed in
nft -i.
Moreover, remove call to rl_callback_handler_remove() in cli_exit() for
editline cli since it does not call rl_callback_handler_install().
Fixes: bc2d5f79c2ea ("cli: use plain readline() interface with libedit")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
| |
The original nul-char detection was not functional on Big Endian.
Instead, go a simpler route by exporting the string and working on the
exported data to check for a nul-char and escape a trailing asterisk if
present. With the data export already happening in the caller, fold
escaped_string_wildcard_expr_alloc() into it as well.
Fixes: b851ba4731d9f ("src: add interface wildcard matching")
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
| |
Size of array to export the bit value into was eight times too large, so
on Big Endian the data written into the data reg was always zero.
Fixes: 2fcce8b0677b3 ("ct: connlabel matching support")
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
| |
Used by 'ct expiration', time_type is supposed to be 32bits. Passing a
64bits variable to constant_expr_alloc() causes the value to be always
zero on Big Endian.
Fixes: 0974fa84f162a ("datatype: seperate time parsing/printing from time_type")
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
|
|
|
| |
In kernel as well as when parsing, hour_type is assumed to be 32bits.
Having the struct datatype field set to 64bits breaks Big Endian and so
does passing a 64bit value and 32 as length to constant_expr_alloc() as
it makes it import the upper 32bits. Fix this by turning 'result' into a
uint32_t and introduce a temporary uint64_t just for the call to
time_parse() which expects that.
Fixes: f8f32deda31df ("meta: Introduce new conditions 'time', 'day' and 'hour'")
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Using a 64bit variable to temporarily hold the parsed value works only
on Little Endian. uid_t and gid_t (and therefore also pw->pw_uid and
gr->gr_gid) are 32bit.
To fix this, use uid_t/gid_t for the temporary variable but keep the
64bit one for numeric parsing so values exceeding 32bits are still
detected.
Fixes: e0ed4c45d9ad2 ("meta: relax restriction on UID/GID parsing")
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
| |
The mask used to select bits to keep must be exported in the same
byteorder as the payload statement itself, also the length of the
exported data must match the number of bytes extracted earlier.
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
|
|
| |
Commit 0e52cab1e64ab improved error reporting by adding rule's table and
chain names to netlink message directly, prefixed by their location
info. This in turn caused netlink dumps of the rule to not contain table
and chain name anymore. Fix this by inserting the missing info before
dumping and remove it afterwards to not cause duplicated entries in
netlink message.
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
| |
Unknown exthdr type with NFT_EXTHDR_F_PRESENT flag set caused
NULL-pointer deref. Fix this by moving the conditional exthdr.desc deref
atop the function and use the result in all cases.
Fixes: e02bd59c4009b ("exthdr: Implement existence check")
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
|
| |
If payloads don't contain family-specific bits, they may sit in a single
*.payload file for all tested families. In such case, nft-test.py will
consequently write dissenting payloads into a single *.got file. To
avoid the duplicate entries, check if a matching record exists already
before writing it out.
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
| |
This should store the index of the protocol template, but
&x[i] - &x[0] is always i, so remove the divide. Also add test case.
Fixes: 01fbc1574b9e ("exthdr: add parse and build userdata interface")
Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
| |
This header is not required to compile nftables with editline, remove
it, this unbreak compilation in several distros which have no symlink
from history.h to editline.h
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
mnl.c: In function ‘mnl_batch_talk’:
mnl.c:417:17: warning: comparison of integer expressions of different signedness: ‘unsigned in’ and ‘long int’ [-Wsign-compare]
if (rcvbufsiz < NFT_MNL_ECHO_RCVBUFF_DEFAULT)
^
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|