summaryrefslogtreecommitdiffstats
path: root/iptables
Commit message (Collapse)AuthorAgeFilesLines
* nft: Fix for add and delete of same rule in single batchPhil Sutter2019-09-302-5/+16
| | | | | | | | | | | | | | | Another corner-case found when extending restore ordering test: If a delete command in a dump referenced a rule added earlier within the same dump, kernel would reject the resulting NFT_MSG_DELRULE command. Catch this by assigning the rule to delete a RULE_ID value if it doesn't have a handle yet. Since __nft_rule_del() does not duplicate the nftnl_rule object when creating the NFT_COMPAT_RULE_DELETE command, this RULE_ID value is added to both NEWRULE and DELRULE commands - exactly what is needed to establish the reference. Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Florian Westphal <fw@strlen.de>
* tests: shell: Support running for legacy/nft onlyPhil Sutter2019-09-301-8/+20
| | | | | | | | | After some changes, one might want to test a single variant only. Allow this by supporting -n/--nft and -l/--legacy parameters, each disabling the other variant. Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Florian Westphal <fw@strlen.de>
* tests/shell: Speed up ipt-restore/0004-restore-race_0Phil Sutter2019-09-261-2/+2
| | | | | | | | | | | This test tended to cause quite excessive load on my system, sometimes taking longer than all other tests combined. Even with the reduced numbers, it still fails reliably after reverting commit 58d7de0181f61 ("xtables: handle concurrent ruleset modifications"). Fixes: 4000b4cf2ea38 ("tests: add test script for race-free restore") Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Florian Westphal <fw@strlen.de>
* nft: Get rid of NFT_COMPAT_EXPR_MAX definePhil Sutter2019-09-261-4/+2
| | | | | | | | Instead simply use ARRAY_SIZE() macro to not overstep supported_exprs array. Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: Fix typo in nft_parse_limit() error messagePhil Sutter2019-09-261-1/+1
| | | | | | | Seems like a trivial copy'n'paste bug. Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables_error() does not returnPhil Sutter2019-09-255-43/+19
| | | | | | | | | It's a define which resolves into a callback which in turn is declared with noreturn attribute. It will never return, therefore drop all explicit exit() calls or other dead code immediately following it. Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Florian Westphal <fw@strlen.de>
* nft: Fix add_bitwise_u16() on Big EndianPhil Sutter2019-09-232-2/+2
| | | | | | | | | | | | | | | | | | | | | Type used for 'mask' and 'xor' parameters was wrong, 'int' is four bytes on 32 or 64 bit architectures. After casting a uint16_t to int, on Big Endian the first two bytes of data are (the leading) zero which libnftnl then copies instead of the actual value. This problem was noticed when using '--fragment' option: | # iptables-nft -A FORWARD --fragment -j ACCEPT | # nft list ruleset | grep frag-off | ip frag-off & 0 != 0 counter packets 0 bytes 0 accept With this fix in place, the resulting nft rule is correct: | ip frag-off & 8191 != 0 counter packets 0 bytes 0 accept Fixes: 2f1fbab671576 ("iptables: nft: add -f support") Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft Increase mnl_talk() receive buffer sizePhil Sutter2019-09-231-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | This improves cache population quite a bit and therefore helps when dealing with large rulesets. A simple hard to improve use-case is listing the last rule in a large chain. These are the average program run times depending on number of rules: rule count | legacy | nft old | nft new --------------------------------------------------------- 50,000 | .052s | .611s | .406s 100,000 | .115s | 2.12s | 1.24s 150,000 | .265s | 7.63s | 4.14s 200,000 | .411s | 21.0s | 10.6s So while legacy iptables is still magnitudes faster, this simple change doubles iptables-nft performance in ideal cases. Note that using a larger buffer than 32KB doesn't further improve performance since linux kernel won't transmit more data at once. This limit was set (actually extended from 16KB) in kernel commit d35c99ff77ecb ("netlink: do not enter direct reclaim from netlink_dump()"). Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: Introduce nft_bridge_commit()Phil Sutter2019-09-234-4/+9
| | | | | | | | No need to check family value from nft_commit() if we can have a dedicated callback for bridge family. Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: Use nftnl_*_set_str() functionsPhil Sutter2019-09-231-14/+14
| | | | | | | | | | Although it doesn't make a difference in practice, they are the correct API functions to use when assigning string attributes. While doing so, also drop the needless casts to non-const. Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
* DEBUG: Print to stderr to not disturb iptables-savePhil Sutter2019-09-231-1/+1
| | | | | | | | This way there's at least a chance to get meaningful results from testsuite with debugging being turned on. Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
* tests/shell: Make ebtables-basic test more verbosePhil Sutter2019-09-231-12/+16
| | | | | | | Print expected entries count if it doesn't match. Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
* ipables: xtables-restore: output filename option in help textFlorian Westphal2019-09-162-2/+2
| | | | | Closes: https://bugzilla.netfilter.org/show_bug.cgi?id=1341 Signed-off-by: Florian Westphal <fw@strlen.de>
* ebtables: fix over-eager -o checks on custom chainsFlorian Westphal2019-09-122-2/+24
| | | | | | | | | | | | | | | | Arturo reports ebtables-nft reports an error when -o is used in custom chains: -A MYCHAIN -o someif makes ebtables-nft exit with an error: "Use -o only in OUTPUT, FORWARD and POSTROUTING chains." Problem is that all the "-o" checks expect <= NF_BR_POST_ROUTING to mean "builtin", so -1 mistakenly leads to the checks being active. Reported-by: Arturo Borrero Gonzalez <arturo@netfilter.org> Closes: https://bugzilla.netfilter.org/show_bug.cgi?id=1347 Signed-off-by: Florian Westphal <fw@strlen.de>
* nft: Drop stale include directivePhil Sutter2019-08-011-1/+0
| | | | | | | | This is a leftover, the file does not exist in fresh clones. Fixes: 06fd5e46d46f7 ("xtables: Drop support for /etc/xtables.conf") Signed-off-by: Phil Sutter <phil@nwl.cc> Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
* doc: Install ip{6,}tables-restore-translate.8 man pagesPhil Sutter2019-07-292-1/+4
| | | | | | | | | Just like in b738ca3677785 ("doc: Install ip{6,}tables-translate.8 manpages"), create man pages for *-restore-translate tools as semantic links to xtables-translate.8. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* doc: Install nft-variant man pages only if enabledPhil Sutter2019-07-291-6/+6
| | | | | | | | Man pages relevant for nftables backend only (xtables-*, *-translate.8) were installed even if --disable-nftables was given at configure time. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables: Drop support for /etc/xtables.confPhil Sutter2019-07-296-482/+9
| | | | | | | | As decided upon at NFWS2019, drop support for configurable nftables base chains to use with iptables-nft. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: Set errno in nft_rule_flush()Phil Sutter2019-07-291-1/+3
| | | | | | | | | | | When trying to flush a non-existent chain, errno gets set in nft_xtables_config_load(). That is an unintended side-effect and when support for xtables.conf is later removed, iptables-nft will emit the generic "Incompatible with this kernel." error message instead of "No chain/target/match by that name." as it should. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* restore legacy behaviour of iptables-restore when rules start with -4/-6Adel Belhouane2019-07-293-0/+61
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | v2: moved examples to testcase files Legacy implementation of iptables-restore / ip6tables-restore allowed to insert a -4 or -6 option at start of a rule line to ignore it if not matching the command's protocol. This allowed to mix specific ipv4 and ipv6 rules in a single file, as still described in iptables 1.8.3's man page in options -4 and -6. The implementation over nftables doesn't behave correctly in this case: iptables-nft-restore accepts both -4 or -6 lines and ip6tables-nft-restore throws an error on -4. There's a distribution bug report mentioning this problem: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=925343 Restore the legacy behaviour: - let do_parse() return and thus not add a command in those restore special cases - let do_commandx() ignore CMD_NONE instead of bailing out I didn't attempt to fix all minor anomalies, but just to fix the regression. For example in the line below, iptables should throw an error instead of accepting -6 and then adding it as ipv4: % iptables-nft -6 -A INPUT -p tcp -j ACCEPT Signed-off-by: Adel Belhouane <bugs.a.b@free.fr> Signed-off-by: Florian Westphal <fw@strlen.de>
* ebtables-save: Merge into xtables_save_main()Phil Sutter2019-07-231-110/+13
| | | | | | | | The only thing missing was handling of EBTABLES_SAVE_COUNTER env var, but that can be done after parsing parameters in bridge-specific code. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* arptables-save: Merge into xtables_save_main()Phil Sutter2019-07-231-60/+3
| | | | | | | | With all preparations in place, xtables_save_main() can replace it with not further changes. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables-save: Pass format flags to do_output()Phil Sutter2019-07-231-4/+6
| | | | | | | | Let callers define the flags to pass to nft_rule_save() instead of just setting the counters boolean. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables-save: Make COMMIT line optionalPhil Sutter2019-07-231-1/+4
| | | | | | | | | | Explicit commits are not used by either arp- nor ebtables-save. In order to share code between all the different *-save tools without inducing changes to ruleset dump contents, allow for callers of do_output() to turn COMMIT lines on or off. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables-save: Pass optstring/longopts to xtables_save_main()Phil Sutter2019-07-231-7/+13
| | | | | | | | | | | Introduce variables for the different optstrings so short and long options live side-by-side. In order to make xtables_save_main() more versatile, pass optstring and longopts via parameter. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables-save: Avoid mixed code and declarationsPhil Sutter2019-07-231-4/+4
| | | | | | | Also move time() calls to where they are used. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: Make nft_for_each_table() more versatilePhil Sutter2019-07-233-14/+21
| | | | | | | Support passing arbitrary data (via void pointer) to the callback. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables-save: Fix table compatibility checkPhil Sutter2019-07-231-3/+4
| | | | | | | | | | | | | | The builtin table check guarding the 'is incompatible' warning was wrong: The idea was to print the warning only for incompatible tables which are builtin, not for others. Yet the code would print the warning only for non-builtin ones. Also reorder the checks: nft_table_builtin_find() is fast and therefore a quick way to bail for uninteresting tables. The compatibility check is needed for the remaining tables, only. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables-save: Unify *-save header/footer commentsPhil Sutter2019-07-237-33/+23
| | | | | | | | | Make eb- and arptables-save print both header and footer comments, too. Also print them for each table separately - the timing information is worth the extra lines in output. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* ebtables-save: Fix counter formattingPhil Sutter2019-07-234-30/+81
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The initial problem was 'ebtables-save -c' printing iptables-style counters but at the same time not disabling ebtables-style counter output (which was even printed in wrong format for ebtables-save). The code around counter output was complicated enough to motivate a larger rework: * Make FMT_C_COUNTS indicate the appended counter style for ebtables. * Use FMT_EBT_SAVE to distinguish between '-c' style counters and the legacy pcnt/bcnt ones. Consequently, ebtables-save sets format to: FMT_NOCOUNTS - for no counters FMT_EBT_SAVE - for iptables-style counters FMT_EBT_SAVE | FMT_C_COUNTS - for '-c' style counters For regular ebtables, list_rules() always sets FMT_C_COUNTS (iptables-style counters are never used there) and FMT_NOCOUNTS if no counters are requested. The big plus is if neither FMT_NOCOUNTS nor FMT_C_COUNTS is set, iptables-style counters are to be printed - both in iptables and ebtables. This allows to drop the ebtables-specific 'save_counters' callback. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* ebtables: Fix error message for invalid parametersPhil Sutter2019-07-231-1/+1
| | | | | | | | | | | | | | | | | | With empty ruleset, ebtables-nft would report the wrong argv: | % sudo ./install/sbin/ebtables-nft -vnL | ebtables v1.8.3 (nf_tables): Unknown argument: './install/sbin/ebtables-nft' | Try `ebtables -h' or 'ebtables --help' for more information. After a (successful) call to 'ebtables-nft -L', this would even segfault: | % sudo ./install/sbin/ebtables-nft -vnL | zsh: segmentation fault sudo ./install/sbin/ebtables-nft -vnL Fixes: acde6be32036f ("ebtables-translate: Fix segfault while parsing extension options") Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables-save: Use argv[0] as program namePhil Sutter2019-07-181-10/+11
| | | | | | | | Don't hard-code program names. This also fixes for bogus 'xtables-save' name which is no longer used. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: exit in case we can't fetch current genidFlorian Westphal2019-07-151-2/+8
| | | | | | | | | | | When running iptables -nL as non-root user, iptables would loop indefinitely. With this change, it will fail with iptables v1.8.3 (nf_tables): Could not fetch rule set generation id: Permission denied (you must be root) Reported-by: Amish <anon.amish@gmail.com> Signed-off-by: Florian Westphal <fw@strlen.de> Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: Move send/receive buffer sizes into nft_handlePhil Sutter2019-07-032-10/+9
| | | | | | | | | Store them next to the mnl_socket pointer. While being at it, add a comment to mnl_set_rcvbuffer() explaining why the buffer size is changed. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: Pass nft_handle down to mnl_batch_talk()Phil Sutter2019-07-031-23/+18
| | | | | | | | | >From there, pass it along to mnl_nft_socket_sendmsg() and further down to mnl_set_{snd,rcv}buffer(). This prepares the code path for keeping stored socket buffer sizes in struct nft_handle. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: Set socket receive bufferPhil Sutter2019-07-031-4/+23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When trying to delete user-defined chains in a large ruleset, iptables-nft aborts with "No buffer space available". This can be reproduced using the following script: | #! /bin/bash | iptables-nft-restore <( | | echo "*filter" | for i in $(seq 0 200000);do | printf ":chain_%06x - [0:0]\n" $i | done | for i in $(seq 0 200000);do | printf -- "-A INPUT -j chain_%06x\n" $i | printf -- "-A INPUT -j chain_%06x\n" $i | done | echo COMMIT | | ) | iptables-nft -X The problem seems to be the sheer amount of netlink error messages sent back to user space (one EBUSY for each chain). To solve this, set receive buffer size depending on number of commands sent to kernel. Suggested-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables-restore: Fix program names in help textsPhil Sutter2019-06-091-6/+7
| | | | | | | | | | | | | Avoid referring to wrong or even non-existent commands: * When calling xtables_restore_main(), pass the actual program name taken from argv[0]. * Use 'prog_name' in unknown parameter and help output instead of 'xtables-restore' which probably doesn't exist. * While being at it, fix false whitespace in help text. Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* src: replace IPTABLES_VERSION by PACKAGE_VERSIONJan Engelhardt2019-05-3012-27/+30
| | | | | | | | | | | | | The IPTABLES_VERSION C macro replicates the PACKAGE_VERSION C macro (both have the same definition, "@PACKAGE_VERSION@"). Since IPTABLES_VERSION, being located in internal.h, is not exposed to downstream users in any way, it can just be replaced by PACKAGE_VERSION, which saves a configure-time file substitution. This goes towards eliminating unnecessary rebuilds after rerunning ./configure. Signed-off-by: Jan Engelhardt <jengelh@inai.de> Signed-off-by: Florian Westphal <fw@strlen.de>
* nft: reset netlink sender buffer size of socket restartPablo Neira Ayuso2019-05-201-0/+1
| | | | | | | | Otherwise, mnl_set_sndbuffer() skips the buffer update after socket restart. Then, sendmsg() fails with EMSGSIZE later on when sending the batch to the kernel. Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: do not retry on EINTRPablo Neira Ayuso2019-05-201-21/+5
| | | | | | | Patch ab1cd3b510fa ("nft: ensure cache consistency") already handles consistency via generation ID. Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: don't care about previous state in ERESTARTPablo Neira Ayuso2019-05-201-7/+10
| | | | | | | We need to re-evalute based on the existing cache generation. Fixes: 58d7de0181f6 ("xtables: handle concurrent ruleset modifications") Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: don't skip table addition from ERESTARTPablo Neira Ayuso2019-05-201-9/+1
| | | | | | | I don't find a scenario that trigger this case. Fixes: 58d7de0181f6 ("xtables: handle concurrent ruleset modifications") Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* xtables: Fix for explicit rule flushesPhil Sutter2019-05-201-1/+1
| | | | | | | | | | | The commit this fixes added a new parameter to __nft_rule_flush() to mark a rule flush job as implicit or not. Yet the code added to that function ignores the parameter and instead always sets batch job's 'implicit' flag to 1. Fixes: 77e6a93d5c9dc ("xtables: add and set "implict" flag on transaction objects") Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: keep original cache in case of ERESTARTPablo Neira Ayuso2019-05-202-4/+22
| | | | | | | | | | | | | | | Phil Sutter says: "The problem is that data in h->obj_list potentially sits in cache, too. At least rules have to be there so insert with index works correctly. If the cache is flushed before regenerating the batch, use-after-free occurs which crashes the program." This patch keeps around the original cache until we have refreshed the batch. Fixes: 862818ac3a0de ("xtables: add and use nft_build_cache") Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: ensure cache consistencyPablo Neira Ayuso2019-05-201-6/+16
| | | | | | | Check for generation ID before and after fetching the cache to ensure consistency. Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: cache table listPablo Neira Ayuso2019-05-202-45/+58
| | | | | | | nft_table_find() uses the table list cache to look up for existing tables. Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: add flush_cache()Pablo Neira Ayuso2019-05-191-11/+19
| | | | | | | | | This new function takes a struct nft_cache as parameter. This patch also introduces __nft_table_builtin_find() which is required to look up for built-in tables without the nft_handle structure. Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: add __nft_table_builtin_find()Pablo Neira Ayuso2019-05-191-6/+12
| | | | Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: statify nft_rebuild_cache()Pablo Neira Ayuso2019-05-192-2/+1
| | | | Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
* nft: add struct nft_cachePablo Neira Ayuso2019-05-192-19/+25
| | | | | | Add new structure that encloses the cache and update the code to use it. Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>