Pablo reports set test fails with HZ=250, as it lists "324ms" instead
of "321". This is because of rounding errors that occur when converting
from user-side millisecond scale to kernel-internal jiffies one.
use 100ms for now to avoid this error.
Alternatives would be to store use-provided value in kernel or to avoid
the conversions; this would require a change to make timeout independent from
jiffies on kernel side.
Reported-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Florian Westphal <fw@strlen.de>
Florian Westphal [Wed, 23 May 2018 21:30:24 +0000 (23:30 +0200)]
src: netlink_delinarize: don't stop when encountering unsupported expression
The error message is still shown, but try to make sense of further
expressions (if any).
I tried to replace the expression by a textual representation.
Two variants I tested are:
1. append as comment:
ip saddr 127.0.0.2 drop comment "unknown expression 'foo'"
This allows nft -f, but it adds/alters a comment.
2. substitute in-place (i.e., add a constant expression
with the error message instead of 'unkown' message:
unknown expression "foo" ip saddr 127.0.0.2 drop
This won't allow 'nft -f' however.
I'm not sure this is a problem, in any case we lose information
if we can't find an expression (e.g. added by newer version).
Florian Westphal [Tue, 29 May 2018 14:16:44 +0000 (16:16 +0200)]
tests: shell: add crash reproducer
Two reports point to a crash in nft when 'flush' is provided
on existing ruleset. In that case, nft will crash with a null-ptr
dereference.
"evaluate: do not inconditionally update cache from flush command"
causes the commit to fail due to a cache inconsistency, we then trip
over NULL location->indesc. Cause of 2nd bug not known yet, not sure
how to fix cache issue either, so only adding reproducer so this can be
fixed later.
Without erec bug, the (errnoeous) error message would be
Could not process rule: File exists
If 'flush ruleset' command is done, release the cache but still keep the
generation ID around. Hence, follow up calls to cache_update() will
assume that cache is updated and will not perform a netlink dump.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Wed, 6 Jun 2018 11:21:49 +0000 (13:21 +0200)]
evaluate: explicitly deny concatenated types in interval sets
Previously, this triggered a program abort:
| # nft add table ip t
| # nft add set ip t my_set '{ type ipv4_addr . inet_service ; flags interval ; }'
| # nft add element ip t my_set '{10.0.0.1 . tcp }'
| BUG: invalid range expression type concat
| nft: expression.c:1085: range_expr_value_low: Assertion `0' failed.
With this patch in place, the 'add set' command above gives an error
message:
| # nft add set ip t my_set3 '{ type ipv4_addr . inet_service ; flags interval ; }'
| Error: concatenated types not supported in interval sets
| add set ip t my_set3 { type ipv4_addr . inet_service ; flags interval ; }
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Florian Westphal <fw@strlen.de>
Phil Sutter [Wed, 6 Jun 2018 08:56:26 +0000 (10:56 +0200)]
nft.8: Fix reject statement documentation
First of all, 'with icmp6' is invalid, expected is 'with icmpv6'. In
addition to that, parameter 'type' expects an icmp*_code type, not
icmp*_type. The respective table column was already correct, but in
synopsis it was wrong.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Florian Westphal <fw@strlen.de>
Phil Sutter [Fri, 1 Jun 2018 15:15:07 +0000 (17:15 +0200)]
log: Add support for audit logging
This is implemented via a pseudo log level. The kernel ignores any other
parameter, so reject those at evaluation stage. Audit logging is
therefore simply a matter of:
| log level audit
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Fri, 1 Jun 2018 15:32:11 +0000 (17:32 +0200)]
JSON: Review large number parsing/printing
When parsing large (uint64_t) values, capital 'I' has to be used in
format string. While being at it, make sure JSON output code handles
those variables correctly, too.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Fri, 1 Jun 2018 15:32:09 +0000 (17:32 +0200)]
JSON: Fix add rule with index
This was missed by commit fb557b5546084 ("JSON: Sort out rule position
and handles in general"): When adding a rule, the parser would still
expect "pos" property instead of "index".
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Mon, 28 May 2018 16:51:08 +0000 (18:51 +0200)]
tests/py: Add missing JSON equivalent for inet/sets.t
This adds the missing JSON variant for the two rules which are supposed
to work.
Fixes: 2efbdf7b8fcf7 ("tests: py: allow to specify sets with a timeout") Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Mon, 28 May 2018 16:51:07 +0000 (18:51 +0200)]
tests/py: Adjust JSON for changes in any/ct.t
Commit 71624f25f22b1 ("tests: py: add expires tests with different time
bases") removed two testcases and added five other ones, adjust JSON
equivalent and recorded output to those changes.
Fixes: 71624f25f22b1 ("tests: py: add expires tests with different time bases") Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Mon, 28 May 2018 16:51:06 +0000 (18:51 +0200)]
JSON: Simplify immediate value parsing
Since an explicit "immediate" expression doesn't exist in JSON (the
values are represented as plain JSON string/integer/boolean types),
there is no need for json_parse_immediate_expr() to comply to the common
expression parser parameter format.
Apart from that:
* Drop CTX_F_RHS checks - caller assures sane context already.
* Improve error message a bit for denied JSON types.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Mon, 28 May 2018 16:51:05 +0000 (18:51 +0200)]
JSON: Fix parsing of meter statement key
The key must be a set elem expression, but if a "regular" expression was
provided (which should be commonly accepted in case no set elem specific
properties are required), the resulting object tree crashed libnftables.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Mon, 28 May 2018 16:51:01 +0000 (18:51 +0200)]
JSON: Review set elem expressions
* There is no need to prefix element-specific properties with 'elem_',
they can't conflict.
* In json_parse_set_stmt(), searching for above properties is pointless
since that's already done by called function.
* Fix potential NULL-pointer deref in json_parse_set_elem_expr_stmt():
json_parse_flagged_expr() may return NULL.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Mon, 28 May 2018 16:51:00 +0000 (18:51 +0200)]
JSON: Review meter statement support
Meter name being optional seems to come from old flow statement, so
don't support this. Also add size support as was recently added to
standard syntax.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Mon, 28 May 2018 16:50:59 +0000 (18:50 +0200)]
JSON: Sort out rule position and handles in general
First of all, don't print the position property when listing rules. This
was there only because libnftnl JSON output has it too, but since the
preferred way to *add* a rule at some location is via 'handle' keyword,
keeping "position" in output would be non-intuitive. Changing "position"
property name to "handle" instead is also a no-go since that would clash
with the real rule handle.
Secondly, turn all handle output on regardless of octx->handle setting.
For a programmatic API like JSON, this should be fine.
Thirdly, fix rule locations when parsing JSON: Respect "handle" property
for CMD_INSERT and CMD_ADD and ignore "pos" at all (actually even a
typo, should have read "position"). Also support "index" property
recently added to standard syntax.
Finally, adjust nft-test.py for the above changes: There is no
"position" property to drop from rule output, and "handle" property will
always be present.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Mon, 28 May 2018 16:50:56 +0000 (18:50 +0200)]
objref: Use "ct helper" for NFT_OBJECT_CT_HELPER
Change name of NFT_OBJECT_CT_HELPER in objref_type table to "ct helper"
for consistency. Note that this is not used in regular nft output since
objref_stmt_print() treats this object type separately.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
statement: old kernels are allowing burst zero, don't print it
Don't print limit burst zero which was the default value in old kernels,
this is not allowed in more recent kernels that now operate like
iptables xt_limit which is what users are expecting.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
segtree: incorrect handling of comments and timeouts with mapping
Check if expression is a mapping to do the right handling.
Fixes: 35fedcf540bf ("segtree: missing comments in range and prefix expressions in sets") Fixes: be90e03dd1fa ("segtree: add timeout for range and prefix expressions in sets") Reported-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Sabrina Dubroca [Fri, 25 May 2018 13:23:16 +0000 (15:23 +0200)]
fix printing of "tcp flags syn" and "tcp flags == syn" expressions
Commit 6979625686ec ("relational: Eliminate meta OPs") introduced some
bugs when printing bitmask types.
First, during the post-processing phase of delinearization, the
expression for "tcp flags syn" (PAYLOAD & flag != 0) gets converted to
PAYLOAD == flag, which is not equivalent. This should be
PAYLOAD (IMPL) flag.
Then, during output, the "==" sign from "tcp flags == syn" is dropped,
because the bitmask condition in must_print_eq_op() was removed. Let's
restore it, so that "tcp flags == syn" doesn't get printed as
"tcp flags syn". An extra check for value types is added, so that we
don't start printing "==" for sets such as "tcp flags {syn,ack}"
Finally, add a regression test for this particular case.
segtree: add timeout for range and prefix expressions in sets
# nft add table x
# nft add set x y { type ipv4_addr\; flags timeout,interval\; }
# nft add element x y { 7.4.4.5-8.8.8.8 comment "good guy" timeout 30s}
# nft list ruleset
table ip x {
set y {
type ipv4_addr
flags interval,timeout
elements = { 7.4.4.5-8.8.8.8 timeout 30s expires 27s956ms comment "good guy" }
}
}
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Máté Eckl [Thu, 17 May 2018 07:36:59 +0000 (09:36 +0200)]
tests: py: print_msg refactor
The errstr attribute was hard-coded to "ERROR:"
errstr has been moved in the parameter list. As print_msg is only
used from the other print_* this is not an issue, and as there is a
print_error function, I don't think that strerr should default to
"ERROR:".
Also this kind of messages now get written to stderr. This can be
beneficial if someone wants to redirect output to a file.
Máté Eckl [Thu, 17 May 2018 07:36:58 +0000 (09:36 +0200)]
tests: py: updated test file structure descripion in README
The order of the table and chain definitions have changed in test files.
Now the name of the chain has to be specified in the definition of the
table, so their order is reverted.
Phil Sutter [Tue, 15 May 2018 09:37:56 +0000 (11:37 +0200)]
libnftables: Move scanner object into struct nft_ctx
The initial approach of keeping as much of lex/yacc-specific data
local to the relevant parsing routines was flawed in that input
descriptors which parsed commands' location information points at were
freed after parsing (in scanner_destroy()) although they were required
later for error reporting in case a command was rejected by the kernel.
To overcome this, keep the scanner pointer in struct nft_ctx so that it
can be kept in place until kernel communication has finished.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Fri, 11 May 2018 10:33:46 +0000 (12:33 +0200)]
evaluate: Return ENOENT if rule index is too large
Since EINVAL usually indicates errors from kernel, avoid using it here
too. Instead return ENOENT to indicate there's no entry to append or
prepend the rule to.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Tue, 8 May 2018 11:08:45 +0000 (13:08 +0200)]
tests/py: Support testing JSON input and output as well
This extends nft-test.py by optional JSON testing capabilities,
activated via '-j'/'--enable-json' parameter).
JSON testing happens for all rules which are supposed to work: After a
rule has been added and the existing tests (payload, ruleset listing
output) have been performed, basically the same test is done again using
a recorded JSON equivalent and (if necessary) a recorded listing output.
The code tries to ease new test case creation overhead by
auto-generating JSON equivalent input via listing the (non-JSON) rule in
JSON format. Also, differing netlink debug and listing output are stored
in *.got files to assist in analyzing/fixing failing test cases.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Tue, 8 May 2018 11:08:43 +0000 (13:08 +0200)]
tests/py: Don't read expected payload for each table
When testing rule adding to different table families, expected payload
was read for each tested family again. Instead, read it just once and
just try to read a family-specific payload for each tested family.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Tue, 8 May 2018 11:08:38 +0000 (13:08 +0200)]
libnftables: Implement JSON parser
If JSON output setting is active in current context, try parsing any
input as JSON. If the initial loading of the buffer or filename by
libjansson fails, fall back to regular syntax parser.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Tue, 8 May 2018 11:08:37 +0000 (13:08 +0200)]
libnftables: Implement JSON output support
Although technically there already is support for JSON output via 'nft
export json' command, it is hardly useable since it exports all the gory
details of nftables VM. Also, libnftables has no control over what is
exported since the content comes directly from libnftnl.
Instead, implement JSON format support for regular 'nft list' commands.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Tue, 8 May 2018 11:08:34 +0000 (13:08 +0200)]
libnftables: Make some arrays globally accessible
This removes static flag and adds declarations in headers for the
following arrays:
* ct_templates from src/ct.c
* mark_tbl from src/datatype.c
* meta_templates and devgroup_tbl from src/meta.c
* table_flags_name from src/rule.c
* set_stmt_op_names from src/statement.c
* tcpopthdr_protocols from src/tcpopt.c
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Combine it with 0002insertposition_0 due to the many similarities,
extend it to test 'handle' and 'index' parameters as well and rename the
testcase accordingly.
Also add a new 0002addinsertlocation_1 which tests that wrong argument
to all of the location parameters fails.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Wed, 9 May 2018 14:03:42 +0000 (16:03 +0200)]
Support 'add/insert rule index <IDX>'
Allow to specify an absolute rule position in add/insert commands like
with iptables. The translation to rule handle takes place in userspace,
so no kernel support for this is needed. Possible undesired effects are
pointed out in man page to make users aware that this way of specifying
a rule location might not be ideal.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Wed, 9 May 2018 14:03:40 +0000 (16:03 +0200)]
Deprecate add/insert rule 'position' argument
Instead, use 'handle' keyword for the same effect since that is more
consistent with respect to replace/delete commands. The old keyword is
still supported for backwards compatibility and also listed in man page
along with a hint that it shouldn't be used anymore.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Sat, 5 May 2018 12:44:53 +0000 (14:44 +0200)]
scanner: Support rfc4291 IPv4-compatible addresses
These are defined in section 2.5.5.1. Although it is stated that they
are deprecated and new implementations are not required to support them,
they occur in ruleset output if an address in the form '::feed:babe' was
given in input. In order to support reinsertion of that rule, we have to
support those deprecated addresses as well.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Phil Sutter [Sat, 5 May 2018 12:44:26 +0000 (14:44 +0200)]
proto: Fix wrong token in proto_icmp6
'token' value of ICMP6HDR_MTU field must be 'mtu', not 'packet-too-big'.
This went unnoticed because rule delinearization for icmp/icmpv6 payload
expressions is problematic anyway in that different fields point to the
same offset and therefore are indistinguishable. In this case, an
expression like e.g. 'icmpv6 mtu 1500' will be printed later as 'icmpv6
parameter-problem 1500'.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
projects / thirdparty / nftables.git / log
