Commit message (Collapse) | Author | Files | Lines | ||
---|---|---|---|---|---|
2019-07-11 | noise: immediately rekey all peers after changing device private key | Jason A. Donenfeld | 6 | -7/+47 | |
Reported-by: Derrick Pallas <derrick@pallas.us> | |||||
2019-07-08 | netlink: enforce that unused bits of flags are zero | Jason A. Donenfeld | 2 | -5/+16 | |
Reported-by: Toke Høiland-Jørgensen <toke@toke.dk> | |||||
2019-07-08 | wg-quick: android: refactor and add incoming allow rules | Jason A. Donenfeld | 1 | -65/+73 | |
Suggested-by: Yağmur Oymak <yagmur.oymak@gmail.com> | |||||
2019-07-04 | compat: define conversion constants for ancient kernels | Jason A. Donenfeld | 2 | -1/+5 | |
2019-07-02 | version: bump snapshot0.0.20190702 | Jason A. Donenfeld | 2 | -2/+2 | |
2019-07-02 | receive: queue dead packets to napi queue instead of empty rx_queue | Jason A. Donenfeld | 2 | -5/+3 | |
2019-06-28 | peer: use LIST_HEAD macro | Jason A. Donenfeld | 1 | -1/+1 | |
Suggested-by: Sultan Alsawaf <sultan@kerneltoast.com> | |||||
2019-06-28 | netlink: cast struct over cb->args for type safety | Jason A. Donenfeld | 1 | -29/+34 | |
This eliminates the headache of managing cb->args[??]. Suggested-by: Johannes Berg <johannes@sipsolutions.net> | |||||
2019-06-28 | compat: support RHEL8's skb_mark_not_on_list backport | Jason A. Donenfeld | 1 | -1/+3 | |
2019-06-25 | compat: rhel backported list modifications | Jason A. Donenfeld | 1 | -1/+1 | |
2019-06-25 | global: switch to coarse ktime | Jason A. Donenfeld | 11 | -36/+43 | |
Coarse ktime is broken until [1] in 5.2 and kernels without the backport, so we use fallback code there. The fallback code has also been improved significantly. It now only uses slower clocks on kernels < 3.17, at the expense of some accuracy we're not overly concerned about. [1] https://lore.kernel.org/lkml/tip-e3ff9c3678b4d80e22d2557b68726174578eaf52@git.kernel.org/ Suggested-by: Arnd Bergmann <arnd@arndb.de> | |||||
2019-06-24 | compat: unify custom function prefix/suffix | Jason A. Donenfeld | 3 | -48/+48 | |
2019-06-24 | compat: some kernels weirdly backport prandom_u32_max | Jason A. Donenfeld | 1 | -2/+3 | |
2019-06-24 | wg-quick: darwin: support being called from launchd | Jason A. Donenfeld | 3 | -1/+53 | |
This causes wg-quick up to wait for the monitor to exit before it exits, so that launchd can correctly wait on it. Reported-by: Cameron Palmer <cameron@promon.no> | |||||
2019-06-18 | qemu: show signal when failing | Jason A. Donenfeld | 1 | -3/+14 | |
2019-06-03 | blake2s: spacing | Jason A. Donenfeld | 2 | -123/+123 | |
2019-06-02 | curve25519: not all linkers support bmi2 and adx | Jason A. Donenfeld | 3 | -6/+58 | |
2019-06-01 | version: bump snapshot0.0.20190601 | Jason A. Donenfeld | 2 | -2/+2 | |
2019-05-31 | compat: don't call xgetbv on cpus with no XSAVE | Jason A. Donenfeld | 1 | -1/+1 | |
2019-05-31 | blake2s: add ssse3 to nobs | Jason A. Donenfeld | 1 | -1/+2 | |
2019-05-31 | blake2s: do not use xgetbv for ssse3 detection | Jason A. Donenfeld | 1 | -3/+1 | |
2019-05-31 | tools: pass WG_ENDPOINT_RESOLUTION_RETRIES=infinity to systemd unit | Jason A. Donenfeld | 1 | -0/+1 | |
2019-05-31 | version: bump snapshot0.0.20190531 | Jason A. Donenfeld | 2 | -2/+2 | |
2019-05-31 | tools: add wincompat layer to wg(8) | Jason A. Donenfeld | 15 | -0/+321 | |
2019-05-29 | tools: allow setting WG_ENDPOINT_RESOLUTION_RETRIES | Jason A. Donenfeld | 2 | -4/+25 | |
2019-05-29 | zinc: update copyright | Jason A. Donenfeld | 2 | -2/+2 | |
2019-05-29 | blake2s: shorten ssse3 loop | Samuel Neves | 1 | -857/+66 | |
This (mostly) preserves the performance (as measured on Haswell and *lake) of last commit, but it drastically reduces code size. Signed-off-by: Samuel Neves <sneves@dei.uc.pt> | |||||
2019-05-29 | blake2s,chacha: latency tweak | Samuel Neves | 5 | -618/+982 | |
In every odd-numbered round, instead of operating over the state x00 x01 x02 x03 x05 x06 x07 x04 x10 x11 x08 x09 x15 x12 x13 x14 we operate over the rotated state x03 x00 x01 x02 x04 x05 x06 x07 x09 x10 x11 x08 x14 x15 x12 x13 The advantage here is that this requires no changes to the 'x04 x05 x06 x07' row, which is in the critical path. This results in a noticeable latency improvement of roughly R cycles, for R diagonal rounds in the primitive. In the case of BLAKE2s, which I also moved from requiring AVX to only requiring SSSE3, we save approximately 30 cycles per compression function call on Haswell and Skylake. In other words, this is an improvement of ~0.6 cpb. This idea was pointed out to me by Shunsuke Shimizu, though it appears to have been around for longer. Signed-off-by: Samuel Neves <sneves@dei.uc.pt> | |||||
2019-05-29 | qemu: do not check for alignment with ubsan | Jason A. Donenfeld | 1 | -0/+1 | |
It insta-crashes on x86. | |||||
2019-05-29 | zinc: arm64: use cpu_get_elf_hwcap accessor for 5.2 | Jason A. Donenfeld | 3 | -2/+6 | |
2019-05-29 | kbuild: account for recent upstream changes | Jason A. Donenfeld | 2 | -11/+11 | |
Apparently cdd750bfb1f76fe9be8cfb53cbe77b2e811081ab changed things, so we fall back onto this hack. Reported-by: Alex Xu <alex@alxu.ca> | |||||
2019-05-29 | timers: add jitter on ack failure reinitiation | Jason A. Donenfeld | 1 | -2/+4 | |
2019-05-29 | qemu: build iproute2 with libmnl support | Jason A. Donenfeld | 1 | -1/+1 | |
2019-05-29 | wg-quick: specify protocol to ip(8), because of inconsistencies | Jason A. Donenfeld | 1 | -3/+7 | |
2019-05-29 | netlink: use new strict length types in policy for 5.2 | Jason A. Donenfeld | 3 | -14/+29 | |
Reported-by: Bruno Wolff III <bruno@wolff.to> | |||||
2019-05-29 | wg-quick: look up existing routes properly | Jason A. Donenfeld | 1 | -1/+1 | |
This was never really correct, and then 5.1 broke it entirely. Reported-by: piraty1@inbox.ru | |||||
2019-05-29 | qemu: use newer iproute2 and kernel | Jason A. Donenfeld | 1 | -9/+4 | |
2019-05-29 | socket: set ignore_df=1 on xmit | Joe Holden | 2 | -0/+6 | |
This allows the kernel to generate ipv6 fragments. Apply the same to ipv4 for consistency. Signed-off-by: Joe Holden <jwh@zorins.us> | |||||
2019-04-23 | wg-quick: make darwin and freebsd path search strict like linux | Jason A. Donenfeld | 2 | -2/+4 | |
2019-04-23 | wg-quick: freebsd: workaround SIOCGIFSTATUS race in FreeBSD kernel | Jason A. Donenfeld | 1 | -9/+22 | |
2019-04-14 | compat: udp_tunnel: force cast sk_data_ready | Jason A. Donenfeld | 1 | -1/+3 | |
Reference: https://lists.zx2c4.com/pipermail/wireguard/2019-April/004081.html | |||||
2019-04-06 | version: bump snapshot0.0.20190406 | Jason A. Donenfeld | 2 | -2/+2 | |
2019-04-06 | allowedips: initialize list head when removing intermediate nodes | Jason A. Donenfeld | 2 | -1/+5 | |
Otherwise if this list item is later reused, we'll crash on list poison or worse. Also, add a version of Mimka's reproducer to netns.sh to catch these types of bugs in the future. Reported-by: Mimka <mikma.wg@lists.m7n.se> | |||||
2019-03-29 | compat: backport skb_mark_not_on_list | Jason A. Donenfeld | 1 | -0/+7 | |
2019-03-27 | blake2s: remove outlen parameter from final | Jason A. Donenfeld | 4 | -12/+11 | |
2019-03-27 | blake2s: simplify | Samuel Neves | 2 | -40/+12 | |
Signed-off-by: Samuel Neves <sneves@dei.uc.pt> | |||||
2019-03-25 | qemu: set framewarn 1280 for 64bit and 1024 for 32bit | Jason A. Donenfeld | 14 | -4/+14 | |
2019-03-25 | device: use skb accessor functions where possible | Jason A. Donenfeld | 1 | -2/+2 | |
Suggested-by: David Miller <davem@davemloft.net> | |||||
2019-03-25 | allowedips: do not use __always_inline | Jason A. Donenfeld | 1 | -9/+9 | |
DaveM doth forbid. Suggested-by: David Miller <davem@davemloft.net> | |||||
2019-03-25 | peerlookup: rename from hashtables | Jason A. Donenfeld | 7 | -9/+9 | |