| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Elements' data was printed as type DATA_VALUE no matter the actual type.
For verdicts, this meant no printing at all (because reg->len is either
zero or garbage).
To fix this, nftnl_set_elem_snprintf_default() needs type info held in
struct nftnl_set. Pass it via parameter to that function, make it
non-static and call it from nftnl_set_snprintf_default() instead of the
generic nftnl_set_elem_snprintf(). This way no changes have to be done
to exported functions, also the output type is already defined when
nftnl_set_snprintf_default() runs so checking type value again is
pointless.
Signed-off-by: Phil Sutter <phil@nwl.cc>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
NFTNL_SET_EXPR defines the stateful expression type that this set stores
in each element. This provides the set definition in terms of stateful
expressions. The expression that is passed via NFNTL_SET_ELEM_EXPR must
equal to this set stateful expression type, otherwise the kernel bails
out.
This patch adds support for the set expression list, which generalizes
NFTNL_SET_EXPR.
This patch also adds nftnl_set_add_expr() to add new expressions to a set
and nftnl_set_elem_expr_foreach() to iterate over the list of expressions.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
| |
This patch adds support for the NFTA_SET_EXPR netlink attribute.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If NFTNL_SET_DESC_CONCAT data is passed, pass that to the kernel
as NFTA_SET_DESC_CONCAT attributes: it describes the length of
single concatenated fields, in bytes.
Similarly, parse NFTA_SET_DESC_CONCAT attributes if received
from the kernel.
This is the libnftnl counterpart for nftables patch:
src: Add support for NFTNL_SET_DESC_CONCAT
v4:
- move NFTNL_SET_DESC_CONCAT to the end of enum nftnl_set_attr
to avoid ABI breakage (Pablo Neira Ayuso)
v3:
- use NFTNL_SET_DESC_CONCAT and NFTA_SET_DESC_CONCAT instead of a
stand-alone NFTA_SET_SUBKEY attribute (Pablo Neira Ayuso)
- pass field length in bytes instead of bits, fields would get
unnecessarily big otherwise
v2:
- fixed grammar in commit message
- removed copy of array bytes in nftnl_set_nlmsg_build_subkey_payload(),
we're simply passing values to htonl() (Phil Sutter)
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
These functions make assumptions on size of passed data pointer and
therefore tend to hide programming mistakes. Instead either one of the
type-specific setters or the generic *_set_data() setter should be used.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
By calling nftnl_set_set(), any data size checks are effectively
bypassed. Better call nftnl_set_set_data() directly, passing the real
size for validation.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
Rename and optimize internal function nftnl_set_lookup() for external
use. Just like with nftnl_chain_list, use a hash table for fast set name
lookups.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Rolf Eike Beer <eike@sf-mail.de> reported that nft-expr_quota-test fails
with a SIGBUS on SPARC due to unaligned accesses. This patch resolves
that and fixes additional sources of unaligned accesses matching the
same pattern. Both nft-expr_quota-test and nft-expr_objref-test
generated unaligned accesses on DEC Alpha.
Bug: https://bugs.gentoo.org/666448
Signed-off-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
| |
We have better json support in libnftables these days.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
Correct one typo for parsing set handles.
Signed-off-by: Harsha Sharma <harshasharmaiitr@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
This patch adds code to allocate set handles and delete sets via set
handle.
Signed-off-by: Harsha Sharma <harshasharmaiitr@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When EXPORT_SYMBOL is located after function definition, clang won't
properly export the function, resulting in a library with no symbols
when built with clang.
Based on libmnl commit dcdb47373a37 ("Move declaration of visibility
attributes before definition.")
Closes: https://bugzilla.netfilter.org/show_bug.cgi?id=1205
Signed-off-by: Armin K <krejzi@email.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
Public API already does this for us, no need to do this again from
internal helper functions.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
This function bails out with -1 if we cannot print, and this buffer is
internally allocated. No need for this overly deffensive initialization.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
In _snprintf() functions definition the buffer is null terminated.
Signed-off-by: Varsha Rao <rvarsha016@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
SNPRINTF_BUFFER_SIZE() macro declaration and definition is changed so
that it accepts three arguments ret, remain and offset. Parameters size
and len are not required instead parameter remain keeps track of
available space in the buffer.
Signed-off-by: Varsha Rao <rvarsha016@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
This was causing a memory leak when using set.
Catched by an ASAN run:
==21004==ERROR: LeakSanitizer: detected memory leaks
Direct leak of 12 byte(s) in 2 object(s) allocated from:
#0 0x4cde58 in malloc (/usr/local/sbin/nft+0x4cde58)
#1 0x7ffff79b8c19 in nftnl_set_set_data /home/eric/git/netfilter/libnftnl/src/set.c:179
|
|
|
|
|
|
|
|
|
|
|
| |
This machinery was introduced to avoid sudden compilation breakage of
old nftables releases. With the upcoming release of 0.7 (and 0.6 which
is now 6 months old) this is not required anymore. Moreover, users gain
nothing from older releases since they are half-boiled and buggy.
So let's get rid of aliases now. Bump LIBVERSION and update map file.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
| |
This new attribute specifies the stateful object type this set stores.
Similar to data type, but specific to store objects. You must set the
NFT_SET_OBJECT flag to use this.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch removes the libmxml integration in libnftnl, since we have
JSON in place and there is no need to support two at the same time.
The JSON support is much better, for example libjansson has a better
parsing error reporting.
Moreover, libmxml 2.10 breaks the integration with libnftnl somehow,
as reported in Debian bug #83870 [0].
Also, the XML support inside libnftnl has never been in good shape, with
several tiny inconsitencies.
[0] https://bugs.debian.org/838370
Signed-off-by: Arturo Borrero Gonzalez <arturo.borrero.glez@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
Maximum set description attribute is NFTA_SET_DESC_MAX, instead of
NFTA_SET_MAX.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
Although the 'err' pointer should be interesting for users only if the
parser returned non-zero, having it point to uninitialized data is
generally a bad thing.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
During list populating, in error case the function returns without
freeing the newly allocated 'elem' object, thereby losing any references
to it.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
| |
Iterators do not modify objects which they iterate, so input pointer must
be const.
Signed-off-by: Carlos Falgueras García <carlosfg@riseup.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
| |
All getters must set the memory size of the attributes, ie. this
includes the nul-termination in strings.
For references to opaque objects hidden behind the curtain, report
a zero size.
Signed-off-by: Carlos Falgueras García <carlosfg@riseup.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
| |
The string length must be one character longer to include the
nul-termination.
Signed-off-by: Carlos Falgueras García <carlosfg@riseup.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
| |
The new structure 'user' holds a pointer to user data and its length. The
kernel must have the flag NFTA_SET_USERDATA to support this feature.
Signed-off-by: Carlos Falgueras García <carlosfg@riseup.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bail out on errors in several nftnl_*_nlmsg_parse() functions. We can
overwrite the previous error value, and may execute code which should
not.
Bad way:
int f() {
int ret;
ret = g();
ret = h();
return ret;
}
Good way:
int f() {
int ret;
ret = g();
if (ret < 0)
return ret;
ret = h();
if (ret < 0)
return ret;
return 0;
}
Signed-off-by: Carlos Falgueras García <carlosfg@riseup.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
Now that unsetters don't set pointers to NULL, check if the attribute is
set before trying to release it.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
If the attribute is set as we already check at the beginning of this
function, then we can release the object.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
| |
And pass up an error to the caller.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
| |
So the client can bail out of memory allocation errors. Or in case of
daemon, make sure things are left in consistent state before bailing
out.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
| |
If this attribute is not supported by the library, we should rise an
assertion so the client knows something is wrong, instead of silently
going through.
The only case I can think may hit this problem is version mismatch
between library and tools. This should not ever really happen, so better
bail out from the library itself in this case.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
These functions are passed as parameter, so we basically get nothing
with this.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
flow table support needs constant object arguments to printing functions
to avoid ugly casts. While at it, also constify object arguments to message
construction, destructor and a few helper functions.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
Future symbols don't need backwards-compat aliases.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
The function names are already large, trim off the _ATTR_ infix in the
attribute definitions.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
The constant names are already large, trim off the _ATTR_ infix in the
attribute definitions.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
| |
Use a shorter name for this, morever this can be used from sets so the _rule_
is misleading.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
So we can use the nft_* prefix anytime soon for our upcoming higher level
library.
After this patch, the nft_* symbols become an alias of the nftnl_* symbols.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
| |
This patch introduces the nftnl_ symbols as aliases for the existing nft_
symbols through the EXPORT_SYMBOL(...) macro.
We would like to use the nft_* prefix from our upcoming higher level library,
meanwhile with this move we avoid that old binaries break because of missing
symbol dependencies.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
| |
Free object attributes before overwrite it. Fix 'nlmsg_parse' methods of
following objects: 'table', 'chain', 'rule', 'set' and 'set_element'.
Signed-off-by: Carlos Falgueras García <carlosfg@riseup.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
| |
Signed-off-by: Patrick McHardy <kaber@trash.net>
|
|
|
|
|
| |
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When we parse a ruleset which has a rule using a set. First step is to
parse the set, set up an ID and add it to a set list. Later, we use this
set list to find the set associated to the rule and we set up the set ID
to the expression (lookup expression) of the rule.
The problem is that if we return this set to the callback function
nft_ruleset_parse_file_cb() and we free this set, we have a crash when
we try to iterate in the set list.
This patch solves it, cloning the set and adding the new set to the set
list.
Signed-off-by: Alvaro Neira Ayuso <alvaroneay@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
The internal.h file started being a small file with private definitions.
Its size has been increasing over time more and more, so let's split
this in small header files that map to the corresponding class where the
functions belong to.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch adds support to parse the new command tag in XML/JSON. This
patch adds two new functions:
* nft_ruleset_parse_file_cb
* nft_ruleset_parse_buffer_cb
The idea is to invoke the callback function that is passed as parameter is
called for each object that is parsed from the corresponding input. Each
callback has access to the nft_parse_ctx structure that provides the necessary
context such as the command, the object type and the object itself.
This change also adds support to update the content of a set incrementally.
{"nftables":[{"add":[{"element":{"name":"blackhole","table":"filter",
"family":"ip","key_type":7,"key_len":4,"set_elem":[{"key":{
"reg":{"type":"value","len":4,"data0":"0x0403a8c0"}}}]}}]}]}
This also patch consolidates the xml/json ruleset import path.
Signed-off-by: Alvaro Neira Ayuso <alvaroneay@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently, we can't do incremental updates via JSON/XML.
This patch enriches the existing output to indicate the kind of
update that you want to perform.
So, if we have a ruleset like:
table ip filter {
chain input {
type filter hook input priority 0;
}
}
The new output looks like:
{"nftables":[{"add":[{"table":{"name":"filter",...}}]}]}
^^^^^
Where we explicitly indicate that we want to add a table.
We support all the actions that we can do with nft, they are:
- Add, delete and flush tables and chains.
- Add, delete, replace and insert rules.
- Add and delete sets.
- Add and delete set elements.
- Flush ruleset.
You only need to add the command tag:
{"nftables":[{"delete":[{...}, {...},...}]}]}
^^^^^^^^
The possible command tags that you can use are "add", "delete", "insert",
"replace" and "flush".
- Flush table or chain, eg.:
{"nftables":[{"flush":[{"table":{"name":...}}]}]}
- Delete table, chain, set or rule:
{"nftables":[{"delete":[{"chain":{"name":...}]}]}
- Replace a rule (you have to specify the handle):
{"nftables":[{"replace":[{"rule":{...}}]}]}
- Insert a rule:
{"nftables":[{"insert":[{"rule":{...}}]}]}
Signed-off-by: Alvaro Neira Ayuso <alvaroneay@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
| |
This patch refactors code to parse the set in two functions
nft_jansson_parse_set_info and nft_jansson_parse_set. These changes are
used in follow up patches.
Signed-off-by: Alvaro Neira Ayuso <alvaroneay@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
|
|
|
|
|
|
|
|
| |
Currently, we create iterator without test if the list is empty. If the
list is empty, we have a crash when we set up the current element. With
this patch, we test if the list is empty before to create the iterator.
If the list is empty the iterator return NULL.
Signed-off-by: Alvaro Neira Ayuso <alvaroneay@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|