summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAge
...
* We need to blacklist IP addresses at the packet levelRoy Marples2009-03-10
| | | | so we can ignore NAKs from rogue servers who don't supply a ServerID, or supply a fake one.
* Release dhcpcd-4.99.14Roy Marples2009-03-05
|
* Improve INFORM and STATIC support.Roy Marples2009-03-05
| | | | | | It's now possible to flip between DHCP, INFORM and STATIC on already running instances.
* 0 is not a valid CIDRRoy Marples2009-03-05
|
* Don't add subnet routes to INADDR_ANY or those without a gateway.Roy Marples2009-03-05
|
* If we reboot an IPv4LL address, start DHCP negotiation right away.Roy Marples2009-03-03
|
* Set the interval to 0 when carrier comes up.Roy Marples2009-03-02
|
* Validate UDP better by ensuring data len is not bigger than our structRoy Marples2009-02-28
| | | | | | and that the claimed length by the header is not greater than our length. Thanks to Michael Olney.
* Blacklist now accepts networks as well as addresses.Roy Marples2009-02-27
| | | | It also matches on the address offered as well as the server address.
* Use state dir for tmp files and try to preserve orig file more.Roy Marples2009-02-26
|
* TypoRoy Marples2009-02-26
|
* Remove compile warnings on Sparc64Roy Marples2009-02-26
|
* Stop enforcing a umask in dhcpcd.Roy Marples2009-02-25
| | | | | | Instead we will ensure that /etc/resolv.conf is 644. Other config files will retain their existing umask or inherit the current one.
* Gah, remove some debugRoy Marples2009-02-25
|
* Don't timeout if we're daemonised.Roy Marples2009-02-25
|
* Optimize allocation a little.Roy Marples2009-02-25
|
* Why bother with a one off array when we can use our eloop?Roy Marples2009-02-24
|
* Close sockets better and just run the script when we're not touching the ↵Roy Marples2009-02-24
| | | | interface.
* Remove debug.Roy Marples2009-02-24
|
* Fix starting all ifaces.Roy Marples2009-02-24
|
* Finally fix compile on vanilla kernel headers.Roy Marples2009-02-24
| | | | | | | | | This is done by moving the wireless calls to a separate file that just includes linux/wireless.h. This is needed because linux/wireless.h includes linux/if.h which may or may not have been fixed to co-exist with net/if.h. Maybe one day this won't be needed ....
* Fix debug working in dhcpcd.confRoy Marples2009-02-23
|
* We need to include linux/wireless.h as wireless.h belongs to wireless-toolsRoy Marples2009-02-23
|
* Try a new tactic to compile with old broken linux kernel headers.Roy Marples2009-02-23
|
* If we fail to send a packet we should still continue our code path.Roy Marples2009-02-20
| | | | | Existing failure timers should change it if needed.
* We should expire instead of stop as we will start again.Roy Marples2009-02-20
|
* If we fail to send a packet start over.Roy Marples2009-02-19
| | | | | | We should have our own array of interfaces to start, as interface order could change or be removed during start_interface.
* Only auto background with no interfaces.Roy Marples2009-02-18
|
* If running as a controlling daemon with link management, background earlier.Roy Marples2009-02-17
|
* Allow socket to report location of the configuration file.Roy Marples2009-02-13
|
* Fix up some more formatting.Roy Marples2009-02-12
|
* More style updates.Roy Marples2009-02-12
|
* Enforce NetBSD KNF style moreRoy Marples2009-02-11
|
* Log our control commands.Roy Marples2009-02-05
|
* Ensure the buffer is NULL terminated.Roy Marples2009-02-05
|
* Use Aq properly.Roy Marples2009-02-05
|
* Release dhcpcd-4.99.13Roy Marples2009-02-04
|
* IPV4LL addresses are not BOOTP.Roy Marples2009-02-01
|
* RFC2131 implies that we should only ARP check after an offer.Roy Marples2009-02-01
|
* We should zero the entire message struct so any options don't overflow into ↵Roy Marples2009-01-30
| | | | unallocated memory when evaluating them.
* CopyrightRoy Marples2009-01-29
|
* Just terminate the DHCP message if not terminated. DHO_PAD could be valid data.Roy Marples2009-01-29
|
* Shorten error msg.Roy Marples2009-01-29
|
* Clean up compiler warnings from overly pedantic checking.Roy Marples2009-01-29
|
* dhcpcd should not send hostname by default.Roy Marples2009-01-29
| | | | | | | | However, the default config file we ship enables the sending of the hostname by default. This makes things more explicit I think, and also allows the FQDN to be sent but not the hostname if someone ever needs this.
* We updated these files this year.Roy Marples2009-01-28
|
* We should state which interface we're removing the route from correctly.Roy Marples2009-01-27
|
* Fix compile on BSD.Roy Marples2009-01-27
|
* Add route removal detection to Linux.Roy Marples2009-01-27
| | | | | | This involved a rejig of the socket code so we can detect if we removed the route or not.
* We need to ensure that manual route deletions are deleted from our internal ↵Roy Marples2009-01-26
| | | | routing table also. Partial fix for #159.
generated by cgit v1.2.3 (git 2.25.1) at 2021-03-13 22:29:29 +0000