aboutsummaryrefslogtreecommitdiffstatshomepage
path: root/src/queueing.c (unfollow)
Commit message (Collapse)AuthorFilesLines
2017-10-09crypto/x86_64: satisfy stack validation 2.0Jason A. Donenfeld3-31/+29
We change this to look like the code gcc generates, so as to keep the objtool checker somewhat happy.
2017-10-09qemu: bump stable kernelJason A. Donenfeld1-1/+1
2017-10-09Makefile: quiet recursive makeJason A. Donenfeld1-10/+10
2017-10-08tools: try again if dump is interruptedJason A. Donenfeld1-1/+4
2017-10-06socket: gcc inlining makes this fasterJason A. Donenfeld3-11/+10
2017-10-06socket: don't bother recomparing afterwardsJason A. Donenfeld1-6/+4
It doesn't actually matter if this races, so there's no point in making the hot path slower with the stack copy. Suggested-by: Willy Tarreau <w@1wt.eu>
2017-10-06socket: compare while unlocked firstJason A. Donenfeld2-27/+26
This should be a bit faster. Suggested-by: Willy Tarreau <w@1wt.eu>
2017-10-06receive: do not consider 0 jiffies as being setJason A. Donenfeld1-4/+4
This causes tests to fail if run within the first 5 minutes. We also move to jiffies 64, so that there's low chance of wrapping in case handshakes are spread far apart. Reported-by: René van Dorst <opensource@vdorst.com>
2017-10-06version: bump snapshot0.0.20171005Jason A. Donenfeld2-2/+2
2017-10-05queueing: cleanup skb_paddingJason A. Donenfeld2-14/+14
2017-10-05compat: macro rewrite netlink instead of clutteringJason A. Donenfeld2-19/+23
2017-10-05Makefile: clang now builds the kernel, so use scan-buildJason A. Donenfeld4-10/+9
Also add little stub for coccinelle and clean up semicolon issue it found.
2017-10-05queueing: move from ctx to cbJason A. Donenfeld5-151/+119
2017-10-05receive: do not store endpoint in ctxJason A. Donenfeld2-6/+21
2017-10-05compat: RHEL backported netlink changesJason A. Donenfeld1-1/+1
2017-10-05compat: conditionally redefine GENL_UNS_ADMIN_PERMJason A. Donenfeld1-0/+3
2017-10-05compat: support ptr_ring for old kernelsJason A. Donenfeld2-0/+644
2017-10-05queueing: use ptr_ring instead of linked listsJason A. Donenfeld7-76/+67
2017-10-05receive: we're not planning on turning that into a while loop nowJason A. Donenfeld1-6/+5
2017-10-05send: put keypair referenceJason A. Donenfeld2-0/+20
2017-10-03receive: use local keypair, not ctx keypair in error pathJason A. Donenfeld1-1/+1
2017-10-03Makefile: add non-verbose mode to toolsJason A. Donenfeld2-2/+23
2017-10-03global: add space around variable declarationsJason A. Donenfeld13-1/+59
2017-10-03global: use _WG prefix for include guardsJason A. Donenfeld20-60/+60
Suggested-by: Sultan Alsawaf <sultanxda@gmail.com>
2017-10-03global: satisfy bitshift pedantryJason A. Donenfeld8-24/+24
Suggested-by: Sultan Alsawaf <sultanxda@gmail.com>
2017-10-03receive: simplify message type validationJason A. Donenfeld2-42/+33
2017-10-02curve25519-neon-arm: force ARM encoding, since this is unrepresentable in ThumbJason A. Donenfeld1-0/+1
2017-10-02kernel-tree: remember UAPI in patch creationJason A. Donenfeld1-1/+1
2017-10-02contrib: remove worthless build artifactJason A. Donenfeld1-0/+0
2017-10-02tools: compile on non-LinuxJason A. Donenfeld1-16/+22
2017-10-02tools: simmer down silly compilersJason A. Donenfeld1-1/+1
2017-10-02version: bump snapshot0.0.20171001Jason A. Donenfeld2-2/+2
2017-10-02tools: do not warn on unrecognized itemsJason A. Donenfeld2-26/+0
Upstream advice is to simply be silent.
2017-10-02ratelimiter: wait for destruction, not for read_unlockJason A. Donenfeld1-1/+1
2017-10-02qemu: add watchdog for not hanging on oopsJason A. Donenfeld3-1/+24
2017-10-02messages: reduce maximum staged packets per peerJason A. Donenfeld1-1/+1
2017-10-02noise: use spinlock for rotating keysJason A. Donenfeld3-21/+29
This should only really be contended in extremely exceptional cases, so changing from a mutex to a spinlock is likely fine.
2017-10-02wg-quick: check permissions of parent directoryJason A. Donenfeld1-1/+1
Also prefix octal 0, in case these files are actually of modes that don't start with 0 by accident (such as SUID or sticky bit).
2017-10-02wg-quick: verify wireguard interface in more clever wayJason A. Donenfeld1-1/+1
This helps with old Debian which has ancient iproute2, as well as paving the path toward this script supporting userspace implementations.
2017-10-02wg-quick: anchor sysctl regex to start and endJason A. Donenfeld1-1/+1
This doesn't actually fix a real problem, but it is more correct than not having it. Suggested-by: Aaron Sigel <aaron@vtty.com>
2017-10-02debug: add better insert targetJason A. Donenfeld1-4/+4
2017-10-02peer: remove from RCU lists when the kref is zeroJason A. Donenfeld1-4/+5
2017-10-02timers: ensure safe timer removalJason A. Donenfeld1-12/+17
2017-10-02receive: do not consider netfilter drop a real dropJason A. Donenfeld1-5/+3
2017-10-02peer: ensure that lookup tables are added lastJason A. Donenfeld1-3/+3
2017-10-02netlink: switch from ioctl to netlink for configurationJason A. Donenfeld28-1008/+1934
2017-09-26tools: uapi: only make sure socket file is socketJason A. Donenfeld1-4/+9
2017-09-26receive: mark function staticJason A. Donenfeld1-1/+1
2017-09-25queueing: more standard init/uninit namesJason A. Donenfeld3-7/+7
2017-09-24tools: use key_is_zero for comparing to zerosJason A. Donenfeld7-24/+27
Maybe an attacker on the system could use the infoleak in /proc to gauge how long a wg(8) process takes to complete and determine the number of leading zeros. This is somewhat ridiculous, but it's possible somebody somewhere might at somepoint care in the future, so alright.
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.