aboutsummaryrefslogtreecommitdiffstatshomepage
path: root/src/uapi (unfollow)
Commit message (Collapse)AuthorFilesLines
2017-11-14compat: 3.16.50 got proper rt6_get_cookieJason A. Donenfeld1-1/+1
2017-11-14qemu: update kernelJason A. Donenfeld1-1/+1
2017-11-13allowedips: do not write out of boundsJason A. Donenfeld1-2/+4
2017-11-13selftest: allowedips: randomized test mutex updateJason A. Donenfeld1-5/+6
2017-11-12wg-quick: document localhost exception and v6 ruleJason A. Donenfeld1-3/+4
Reported-by: Hermann Lienstromberg <nurtic-vibe@grmml.net>
2017-11-11version: bump snapshot0.0.20171111Jason A. Donenfeld2-2/+2
2017-11-11tools: allow for NULL keys everywhereJason A. Donenfeld5-29/+50
2017-11-11curve25519: reject deriving from NULL private keysJason A. Donenfeld2-7/+16
These aren't actually valid 25519 points pre-normalization, and doing this is required to make unsetting private keys based on all zeros.
2017-11-11tools: remove ioctl cruftJason A. Donenfeld1-4/+0
2017-11-10allowedips: rename from routingtableJason A. Donenfeld11-149/+149
Makes it more clear that this _not_ a routing table replacement.
2017-11-10wg-quick: allow for tabs in keysJason A. Donenfeld2-20/+20
2017-11-10netlink: make sure we reserve space for NLMSG_DONEJason A. Donenfeld2-1/+25
Otherwise, if messages pack really close together, we'll exceed the size of the sk_buff and return ENOBUFS. I suspect this has been the cause of a lot of weird bugs that people just worked around by increasing the receive buffer size. This actually addresses the root cause.
2017-11-10compat: 4.4.0 has strange ECN functionJason A. Donenfeld1-1/+2
2017-11-10wg-quick: stat the correct enclosing folder of config fileJason A. Donenfeld1-1/+2
2017-11-10qemu: bump kernel versionJason A. Donenfeld1-1/+1
2017-11-10receive: hoist fpu outside of receive loopJason A. Donenfeld3-18/+19
2017-11-10qemu: more debuggingJason A. Donenfeld3-1/+15
2017-11-10device: wait for all peers to be freed before destroyingJason A. Donenfeld1-0/+1
2017-11-10qemu: check for memory leaksJason A. Donenfeld3-6/+28
2017-11-03netlink: plug memory leakJason A. Donenfeld1-1/+1
2017-11-03device: please lockdepJason A. Donenfeld1-1/+1
2017-11-02global: revert checkpatch.pl changesJason A. Donenfeld2-2/+2
These changes were suggested by checkpatch.pl, but actually cause big problems depending on the options. Revert.
2017-11-02Kconfig: remove trailing whitespaceGreg Kroah-Hartman1-1/+1
Git complains when applying this Kconfig file as part of a patch to the kernel tree, so drop the trailing whitespace. Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-01version: bump snapshot0.0.20171101Jason A. Donenfeld2-2/+2
2017-11-01compat: unbreak unloading on kernels 4.6 through 4.9Jason A. Donenfeld1-1/+4
2017-11-01timers: switch to kees' new timer_list functionsJason A. Donenfeld2-17/+22
2017-11-01wg-quick: save all hooks on saveJason A. Donenfeld1-5/+13
2017-10-31version: bump snapshot0.0.20171031Jason A. Donenfeld2-2/+2
2017-10-31wg-quick: fsync the temporary file before renamingJason A. Donenfeld1-0/+1
This ensures that on an unclean shutdown, we either see the old content or the new content, but not empty content. Suggested-by: Ka Ho Ng <ngkaho1234@gmail.com>
2017-10-31wg-quick: allow for saving existing interfaceJason A. Donenfeld2-2/+14
2017-10-31contrib: add reresolve-dnsJason A. Donenfeld2-0/+48
2017-10-31tools: correct type for CTRL_ATTR_FAMILY_IDJason A. Donenfeld1-4/+4
Suggested-by: Jörg Thalheim <joerg@thalheim.io>
2017-10-31wg-quick: allow for the hatchet, but not by defaultJason A. Donenfeld4-0/+55
2017-10-31global: use fewer BUG_ONsJason A. Donenfeld2-3/+5
Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-10-31timers: guard entire setting in blockJason A. Donenfeld1-4/+6
2017-10-31curve25519: only enable int128 if compiler support is soundJason A. Donenfeld1-1/+1
2017-10-31device: expand scope of destruct lockJason A. Donenfeld1-2/+2
2017-10-31global: get rid of useless forward declarationsJason A. Donenfeld3-11/+0
2017-10-31device: only take reference if netns is differentJason A. Donenfeld3-9/+56
If we take two references, the namespace and the device are never freed in the usual manner. We should thus only take a reference to another namespace when it is a different namespace from our own.
2017-10-31wg-quick: remember to rewind DNS settings on failureJason A. Donenfeld1-3/+8
2017-10-31wg-quick: allow specifiying multiple hooksJason A. Donenfeld2-18/+21
2017-10-31qemu: test using four coresJason A. Donenfeld1-1/+1
2017-10-31global: style nitsJason A. Donenfeld27-274/+470
2017-10-31qemu: work around ccache bugsJason A. Donenfeld1-1/+1
2017-10-31global: infuriating kernel iterator styleJason A. Donenfeld13-48/+48
One types: for (i = 0 ... So one should also type: for_each_obj (obj ... But the upstream kernel style guidelines are insane, and so we must instead do: for_each_obj(obj ... Ugly, but one must choose his battles wisely.
2017-10-31peer: store total number of peers instead of iteratingJason A. Donenfeld3-15/+4
This is faster, since it means adding a new peer is O(1) instead of O(n). It's also safe to do because we're holding the device_update_lock on both the ++ and the --.
2017-10-31peer: get rid of peer_for_each magicJason A. Donenfeld3-41/+17
Since the peer list is protected by the device_update_lock, and since items are removed from the peer list before putting their final reference, we don't actually need to take a reference when iterating. This allows us to simplify the macro considerably. Suggested-by: Johannes Berg <johannes@sipsolutions.net>
2017-10-31compat: be sure to include header before testingJason A. Donenfeld1-0/+1
Reported-by: Aaron M. D. Jones <aaronmdjones@gmail.com>
2017-10-31qemu: allow for cross compilationJason A. Donenfeld14-98/+249
2017-10-31crypto/avx: make sure we can actually use ymm registersJason A. Donenfeld4-3/+23
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.