summaryrefslogtreecommitdiff
path: root/src/ra.c
Commit message (Collapse)AuthorAgeFilesLines
* Sync style with the existing _packed & _unused macrosVladislav Grishenko2016-01-311-1/+1
|
* Use more conventional way for the alignmentVladislav Grishenko2016-01-311-2/+5
|
* Fix alignment of buffers in ra_process and dhcpv6_requestBen Hutchings2016-01-281-1/+2
| | | | | | | | | | The packet buffer needs to be 32-bit aligned to ensure that the various 32-bit fields we pick out are naturally aligned. The control message buffers needs to be naturally aligned for struct cmsghdr. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
* Add padding between odhcp6c_entry structures to ensure 32-bit alignmentBen Hutchings2016-01-281-2/+3
| | | | | | | struct odhcp6c_entry is not declared as __packed, so the compiler may assume it is naturally aligned. Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
* ra: avoid saving an empty search domain entrySteven Barth2015-04-151-7/+10
| | | | Signed-off-by: Steven Barth <steven@midlink.org>
* Improve handling of DNS search domainsSteven Barth2015-04-131-50/+78
| | | | Signed-off-by: Steven Barth <steven@midlink.org>
* Remove obsolete prefix class supportSteven Barth2015-04-121-1/+1
| | | | Signed-off-by: Steven Barth <steven@midlink.org>
* Sanitize RA values lessSteven Barth2015-04-101-1/+1
|
* Pass ICMP configuration to user-space instead of directly applying itSteven Barth2015-04-101-25/+36
|
* properly handle return codesJohn Crispin2015-03-301-2/+4
| | | | Signed-off-by: John Crispin <blogic@openwrt.org>
* ra: use proper struct msghdr initializer to fix build error on muslFelix Fietkau2015-03-291-2/+9
| | | | Signed-off-by: Felix Fietkau <nbd@openwrt.org>
* make route metric more compatible with linux defaultsSteven Barth2015-03-251-2/+2
|
* Get rid of getifaddrs for interface ID detectionSteven Barth2015-03-111-26/+9
| | | | Signed-off-by: Steven Barth <steven@midlink.org>
* Don't apply excess filter to DHCPv6 and unify odhcp6c_update_entrySteven Barth2014-12-101-5/+5
|
* Ensure signal-safety of signal handlersSteven Barth2014-10-081-16/+23
|
* Send RS alternatingly with and without source MACSteven Barth2014-09-271-1/+1
|
* Send router solicitations with source link-layer addressbfdSteven Barth2014-07-211-2/+20
|
* ra: don't set nd_ra_{reachable,retransmit} to 0Steven Barth2014-03-051-4/+6
|
* Revert "RA: fix on-link/off-link handling"Steven Barth2014-02-031-6/+4
| | | | | | New behavior breaks source-restrictions of on-link routes. This reverts commit 9817e4983c7d4e28f53b78d4fc6ea1eabed8affe.
* RA: fix on-link/off-link handlingSteven Barth2014-02-021-4/+6
|
* Bump copyright headersSteven Barth2014-01-291-1/+1
|
* add support for multiple prefixes with distinct IAIDsKaspar Schleiser2014-01-271-1/+1
| | | | | | | | | | | Changes from v1: - removed some unneeded changes - use *_add_state instead of (semantically identical and so unnecessary) *_append_state This is missing IAID validation for prefixes. Contributed by T-Labs, Deutsche Telekom Innovation Laboratories
* Use getifaddrs to get the interface-id when not specifiedMatthias Schiffer2014-01-101-8/+26
| | | | | | | | | | | | | Reimplementing this section also fixes two potential bugs: 1. Only link-local addresses should be used for automatic interface-id selection, as it should always be assigned by the kernel and is always based on the EUI-64 unless an administrator is doing strange things (while other addresses might be chosen according to the privacy extensions or by an administrator) 2. When no address is available (because of a race condition or because an administrator removed it), RA packets should just be ignored; otherwise an address with a zero interface-id might be assigned
* Whitespace fixesMatthias Schiffer2014-01-101-1/+1
| | | | Mostly trailing spaces or whitespace in empty lines.
* Add ICMPv6 validity checkHans Dedecker2013-12-241-4/+28
|
* Revert "Fix parsing of Router Advertisement messsages"Hans Dedecker2013-12-241-59/+16
| | | | This reverts commit de89d2888d38d1ca6676070584ea6ff9c7dd6f23.
* Parse Router Advertisement Hop Limit parameterHans Dedecker2013-12-191-0/+4
|
* Fix parsing of Router Advertisement messsagesHans Dedecker2013-12-121-16/+59
| | | | | Router Advertisement message checks according to RFC4861 paragraph 6.1.2 Fix and further improve router advertisement option parsing
* Basic error checking when initializing raHans Dedecker2013-12-011-1/+11
|
* React to carrier down events as wellSteven Barth2013-10-301-2/+5
|
* Fix compilation on 32-bit platformsSteven Barth2013-10-301-1/+2
|
* Detect "carrier up"-events and restart transactionSteven Barth2013-10-301-4/+68
|
* odhpc6c: status code support in replyHans Dedecker2013-10-231-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | The patch implements support for status code handling in reply messages as described in RFC3313 paragraph 18.1.8. The client will *log the status codes returned by the client *send a request if no binding status code is returned for a given IA *send further renew/rebind if the IA was not present in the reply *terminate message exchange when no prefix/no address status code is returned in reponse to a request *terminate message exchange when unspec fail status code is returned *calculate t1/t2/t3 when all IA's have been processed and based on recorded t1/t2/valid timer values per IA Without this patch I have seen issues with request messages send without any IA_PD/IA_NA and t1/t2/t3 timer values which were not correct. These changes have been tested intensive against an ISC DHCPv6 server Signed-off-by: Hans Dedecker <hans.dedecker@gmail.com>
* Avoid updating too often if ISP spams RAsSteven Barth2013-10-231-6/+7
|
* Check for local addr more oftenSteven Barth2013-10-041-15/+15
|
* ra: validate hop-limitSteven Barth2013-10-021-3/+20
|
* Merge branch 'hnet'Steven Barth2013-08-311-1/+1
|\ | | | | | | | | Conflicts: src/ra.c
| * Merge branch 'master' into hnetMarkus Stenberg2013-06-241-20/+59
| |\ | | | | | | | | | | | | | | | | | | Conflicts: README src/dhcpv6.c src/odhcp6c.h
| * | First hnet-based version.Markus Stenberg2013-03-261-2/+2
| | |
* | | Remove DAD for now due to kernel issuesSteven Barth2013-08-301-70/+1
| |/ |/|
* | Allow overriding the interface-identifier for public addressesSteven Barth2013-06-181-10/+13
| |
* | Add sanity checks for NDP and MTU parameters from RAsSteven Barth2013-06-131-3/+4
| |
* | Work around false-positive DAD-hitsSteven Barth2013-05-311-3/+11
| |
* | Fix handling of multiple RAs in a row.Steven Barth2013-05-311-0/+2
| | | | | | | | | | The default route-entry was not correctly reinitialized. Thanks to pmarks-net for the detailed bugreport.
* | Add more sanity checks and logging to DADSteven Barth2013-05-281-2/+10
| |
* | Fix compilationSteven Barth2013-04-081-3/+6
| |
* | Update license headersSteven Barth2013-04-041-0/+14
|/
* Fix priority for on-link routes for prefixesSteven Barth2013-03-201-1/+1
|
* Avoid redundant deduplicationSteven Barth2013-02-011-1/+1
|
* Fix netlink attribute handlingSteven Barth2013-02-011-1/+1
|
View Lorry Controller Status