Commit message (Collapse) | Author | Files | Lines | ||
---|---|---|---|---|---|
2016-12-23 | wg-config: cleanups | Jason A. Donenfeld | 2 | -7/+8 | |
2016-12-23 | tools: do not use AI_ADDRCONFIG | Jason A. Donenfeld | 1 | -2/+1 | |
Some people run wg(8) using hard coded v6 addresses before interfaces have v6 addresses, causing getaddrinfo to fail. Since AI_ADDRCONFIG doesn't actualy change the sorting, but just the queries made, we don't really need AI_ADDRCONFIG anyway, since we're always only taking the first result. Reported-by: Benedikt Morbach <benedikt.morbach@googlemail.com> | |||||
2016-12-23 | tools: allowed-ips is easier to parse with spaces instead of ", " | Jason A. Donenfeld | 2 | -2/+2 | |
2016-12-23 | routing-table: simplify and mask reparented root | Jason A. Donenfeld | 1 | -16/+9 | |
2016-12-23 | wg-config: cleanup ip parsing | Jason A. Donenfeld | 1 | -2/+2 | |
This also sorts routes by cidr. | |||||
2016-12-23 | config: allow removing multiple peers at once | Jason A. Donenfeld | 1 | -1/+2 | |
2016-12-18 | version: bump snapshot0.0.20161218 | Jason A. Donenfeld | 2 | -2/+2 | |
2016-12-16 | siphash: preserve endian-ness for quick helper | Jason A. Donenfeld | 1 | -5/+9 | |
This fixes errors on big endian machines. | |||||
2016-12-16 | ratelimiter: remember newline in error | Jason A. Donenfeld | 1 | -2/+2 | |
2016-12-16 | version: bump snapshot0.0.20161216 | Jason A. Donenfeld | 2 | -2/+2 | |
2016-12-16 | hashtables: use counter and int to ensure forward progress | Jason A. Donenfeld | 2 | -17/+2 | |
2016-12-16 | messages: increase header by 3 bytes for alignment | Jason A. Donenfeld | 4 | -15/+22 | |
2016-12-16 | siphash: update against upstream submission | Jason A. Donenfeld | 15 | -187/+403 | |
2016-12-16 | tools: fix latest-handshake typo in documentation | Jason A. Donenfeld | 2 | -2/+2 | |
Reported-by: Dan Lüdtke <mail@danrl.com> | |||||
2016-12-16 | ratelimiter: drop family from action for 4.10 | Jason A. Donenfeld | 1 | -2/+0 | |
This was never actually used, and now that member of the struct has been removed all together from the kernel. | |||||
2016-12-16 | tests: avoid non-strict writes via printf | Jason A. Donenfeld | 1 | -11/+8 | |
2016-12-16 | noise: update comments | Jason A. Donenfeld | 1 | -8/+8 | |
2016-12-16 | types: enforce consistency | Jason A. Donenfeld | 4 | -24/+26 | |
2016-12-16 | config: cleanups | Jason A. Donenfeld | 1 | -33/+19 | |
2016-12-13 | peer: don't use sockaddr_storage to reduce memory usage | Jason A. Donenfeld | 9 | -30/+37 | |
2016-12-12 | hashtables: ensure we get 64-bits of randomness | Jason A. Donenfeld | 1 | -1/+7 | |
2016-12-12 | messages: remove unused constants | Jason A. Donenfeld | 1 | -2/+0 | |
2016-12-12 | siphash: useless rice | Jason A. Donenfeld | 1 | -3/+1 | |
2016-12-11 | crypto: use kernel's bitops functions | Jason A. Donenfeld | 2 | -18/+15 | |
2016-12-11 | blake2s: move self tests to correct directory | Jason A. Donenfeld | 2 | -554/+555 | |
2016-12-11 | global: move to consistent use of uN instead of uintN_t for kernel code | Jason A. Donenfeld | 23 | -419/+419 | |
2016-12-11 | receive: simplify ip header checking logic | Jason A. Donenfeld | 1 | -15/+2 | |
2016-12-11 | device: ensure icmp skb length check is done for v6 | Jason A. Donenfeld | 1 | -6/+2 | |
2016-12-11 | device: allocate tstats in newlink | Jason A. Donenfeld | 1 | -27/+20 | |
Every place else does this in ndo_init, but this seems completely redundant, so to simplify things, I'm moving it into newlink. | |||||
2016-12-11 | device: clean up xmit error path | Jason A. Donenfeld | 3 | -14/+18 | |
2016-12-11 | main: consistent lines | Jason A. Donenfeld | 1 | -6/+1 | |
2016-12-11 | device: simplify device_init, since it only returns -EEXIST | Jason A. Donenfeld | 1 | -6/+1 | |
2016-12-11 | device: dellink is already implictly unregister_netdevice_queue | Jason A. Donenfeld | 1 | -10/+1 | |
2016-12-11 | device: disable ipv6 auto address generation | Jason A. Donenfeld | 1 | -2/+8 | |
2016-12-11 | device: rc -> ret | Jason A. Donenfeld | 1 | -3/+3 | |
2016-12-09 | version: bump snapshot0.0.20161209 | Jason A. Donenfeld | 2 | -2/+2 | |
2016-12-09 | cookie: kill redundant forward declaration | Jason A. Donenfeld | 1 | -2/+0 | |
2016-12-09 | compat: build dep errors belong here, since it's out of tree specific | Jason A. Donenfeld | 2 | -8/+10 | |
2016-12-09 | socket: clear src address when retrying handshake | Jason A. Donenfeld | 3 | -0/+15 | |
2016-12-09 | contrib: add wg-config | Jason A. Donenfeld | 4 | -0/+383 | |
2016-12-09 | device: traditional if is cleaner than switch for this small | Jason A. Donenfeld | 1 | -3/+1 | |
This also generates identical assembly. | |||||
2016-12-09 | device: make suspend code conditional on CONFIG_PM_SLEEP | Jason A. Donenfeld | 2 | -0/+10 | |
This isn't actually neccessary, since there are no-op stubs for these functions, but the MIPS people don't want any unneccessary bloat. | |||||
2016-12-09 | main: cleaner error teardown | Jason A. Donenfeld | 1 | -12/+12 | |
2016-12-09 | device: cleaner error teardown | Jason A. Donenfeld | 1 | -43/+33 | |
2016-12-09 | timers: add random jitter to handshake retry | Jason A. Donenfeld | 2 | -1/+2 | |
2016-12-09 | device: clear all peer ephemeral keys on sleep | Jason A. Donenfeld | 2 | -2/+35 | |
2016-12-09 | data: reset tc when resetting skb | Jason A. Donenfeld | 1 | -0/+6 | |
2016-12-09 | qemu: bump kernel version | Jason A. Donenfeld | 1 | -3/+3 | |
2016-12-09 | build system: add dkms installation | Jason A. Donenfeld | 2 | -2/+32 | |
2016-12-04 | tests: directly kill nmap | Jason A. Donenfeld | 1 | -7/+12 | |