| Age | Commit message (Expand) | Author |
|---|
| 2015-02-10 | Groundwork for better route support over multiple interfaces by | Kenneth R Westerback |
| 2015-02-08 | Get rid of a bunch of memset()'s where struct variables and arrays | Kenneth R Westerback |
| 2015-02-07 | Add support for interface-mtu (option 26). | Kenneth R Westerback |
| 2015-02-07 | GC unused parameters now that ifname and rdomain are not used in | Kenneth R Westerback |
| 2015-02-06 | The write_file() privsep interface was too permissive and | Reyk Floeter |
| 2015-02-06 | matthew@ points out that it's a bad idea to rely on the non-privileged, | Kenneth R Westerback |
| 2015-02-06 | Don't let errors leak the memory returned by getifaddrs(). | Kenneth R Westerback |
| 2015-02-05 | Use sizeof(struct X) rather than sizeof(*p) in calloc() invocations. | Kenneth R Westerback |
| 2015-02-01 | free(NULL) works, so stop checking for non-zero length allocation | Kenneth R Westerback |
| 2015-01-31 | free(NULL) works, so stop checking for non-zero length allocation | Kenneth R Westerback |
| 2015-01-31 | Don't leak the "access protected, zero sized object" returned by | Kenneth R Westerback |
| 2015-01-31 | Set is_static to 0 before trying to free a lease. Otherwise it | Kenneth R Westerback |
| 2015-01-31 | Always free(buf) if errmsg is set while trying to flush routes. | Kenneth R Westerback |
| 2015-01-30 | Don't leak incompletely cloned static lease. Should never happen | Kenneth R Westerback |
| 2015-01-30 | Tweak error message to say 'realloc' since that's what's failing, | Kenneth R Westerback |
| 2015-01-16 | Replace <sys/param.h> with <limits.h> and other less dirty headers where | Theo de Raadt |
| 2014-12-10 | Reusing the same field for different things in different contexts may | Kenneth R Westerback |
| 2014-12-10 | The consensus is that RTM_IFINFO messages do not always carry the | Kenneth R Westerback |
| 2014-12-09 | Move the initial link status check and possible forceup back before | Kenneth R Westerback |
| 2014-12-08 | Try harder to avoid displaying "no link ... got link" when initial | Kenneth R Westerback |
| 2014-12-07 | Do not move from S_PREBOOT to S_REBOOT just because link_timeout | Kenneth R Westerback |
| 2014-12-07 | Ensure we grab the interface LLADDR as soon as the link state is | Kenneth R Westerback |
| 2014-12-07 | Retract check for IFF_RUNNING when checking for link state. At least | Kenneth R Westerback |
| 2014-12-05 | Move initial wait for network link into the state machine by adding | Kenneth R Westerback |
| 2014-12-03 | Correct comment to state that the BPF program is patched in | Kenneth R Westerback |
| 2014-11-30 | Don't check for hardware address change unless link was up. Delay | Kenneth R Westerback |
| 2014-11-29 | Explicitly initialize client->state to S_REBOOTING rather than | Kenneth R Westerback |
| 2014-11-27 | Use if_nametoindex() to get the interface index, and thereby check | Kenneth R Westerback |
| 2014-11-26 | Reset ifi->errors after a successful receive_packet(), rather than | Kenneth R Westerback |
| 2014-11-26 | 'ifi' is successfully calloc'd during initialization or dhclient | Kenneth R Westerback |
| 2014-11-26 | Rename 'got_one()' to 'packethandler()'. | Kenneth R Westerback |
| 2014-11-25 | r1.118 starting using LINK_STATE_IS_UP() to check the link status | Kenneth R Westerback |
| 2014-11-23 | Rename discover_interface() to more meaningful get_hw_address(). | Kenneth R Westerback |
| 2014-11-23 | Cleanup some struct interface_info fields. Make 'nomedia' a flag. Replace | Kenneth R Westerback |
| 2014-11-20 | A little too agressive reordering. Must initialize the client struct | Kenneth R Westerback |
| 2014-11-17 | Further juggle startup code, putting config file reading after forking | Kenneth R Westerback |
| 2014-11-16 | Move some memory allocations/initialization and file reading to | Kenneth R Westerback |
| 2014-11-16 | Read your compiler output even when it succeeds. Add missing variable to | Kenneth R Westerback |
| 2014-11-15 | Make -q (quiet) and -d (don't daemonize) options mutually exclusive. | Kenneth R Westerback |
| 2014-11-13 | Tweak some comments and error messages. Put all "I can't handle | Kenneth R Westerback |
| 2014-11-08 | Use open(O_CLOEXEC) rather than open() + fcntl(FD_CLOEXEC). Inspired | Kenneth R Westerback |
| 2014-11-08 | Exhort user to recompile 'dhclient', not 'dhcpd', when bpf versions are | Kenneth R Westerback |
| 2014-11-04 | Of the 14 occurances of free_client_lease() only 2 deal with leases | Kenneth R Westerback |
| 2014-11-04 | Instead of correcting things after the fact, use SOCK_NONBLOCK and SOCK_CLOEXEC | Philip Guenther |
| 2014-11-03 | Don't leak static leases when the 'lease {}' parsing fails or one static | Kenneth R Westerback |
| 2014-11-03 | No need to immediately remove an expired lease from the list of | Kenneth R Westerback |
| 2014-11-02 | Pesky whitespace and spurious parenthesis. | Kenneth R Westerback |
| 2014-11-01 | And finally move convert.c to the attic. | Kenneth R Westerback |
| 2014-11-01 | Unhook convert.c. It's empty but for the license text. | Kenneth R Westerback |
| 2014-11-01 | Try to ensure that the various lease timeouts are sane. i.e., renew | Kenneth R Westerback |
