summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* version: bump snapshot0.0.20190805Jason A. Donenfeld2019-08-051-1/+1
|
* device: drop lock before expiring keysJason A. Donenfeld2019-08-051-4/+11
|
* uapi: skip peers with invalid keysJason A. Donenfeld2019-08-052-4/+16
|
* tun: windows: spin for only a millisecond/80Jason A. Donenfeld2019-08-031-1/+1
| | | | Performance stays the same as before.
* wintun: merge opening device registry keySimon Rozman2019-08-021-21/+15
| | | | | | | | | This also introduces waiting for key to appear on initial access. See if this resolves the issue caused by HDD power-up delay resulting in failure to create the adapter. Signed-off-by: Simon Rozman <simon@rozman.si>
* wintun: simplify checking reboot requirementSimon Rozman2019-08-021-18/+7
| | | | | | We never checked checkReboot() reported error anyway. Signed-off-by: Simon Rozman <simon@rozman.si>
* wintun: refactor `err == nil` error checkingSimon Rozman2019-08-022-133/+166
| | | | Signed-off-by: Simon Rozman <simon@rozman.si>
* wintun: handle error for deadgwdetectJason A. Donenfeld2019-08-021-2/+3
|
* tun: darwin: do not attempt to close tun.event twiceJason A. Donenfeld2019-08-021-4/+1
| | | | | Previously it was possible for this to race. It turns out we really don't need to set anything to -1 anyway.
* wintun: get interface path properly with cfgmgrJason A. Donenfeld2019-07-234-30/+93
|
* tun: windows: styleJason A. Donenfeld2019-07-231-1/+1
|
* wintun: simplify resolution of dev nodeJason A. Donenfeld2019-07-231-13/+10
|
* wintun: enable sharing of pnp nodeJason A. Donenfeld2019-07-221-1/+1
|
* tun: windows: close event handle on shutdownJason A. Donenfeld2019-07-221-0/+1
|
* ipc: windows: match SDDL of WDK and make monkeyableJason A. Donenfeld2019-07-191-11/+3
|
* tun: windows: get rid of retry logicJason A. Donenfeld2019-07-192-218/+100
| | | | Things work fine on Windows 8.
* tun: windows: use specific IOCTL codeJason A. Donenfeld2019-07-191-1/+1
|
* device: do not crash on nil'd bind in windows bindingJason A. Donenfeld2019-07-181-0/+5
|
* tun: windows: open file at startup timeJason A. Donenfeld2019-07-181-0/+9
|
* tun: windows: silently drop packet when ring is fullJason A. Donenfeld2019-07-181-2/+2
|
* tun: windows: switch to NDIS device objectJason A. Donenfeld2019-07-182-7/+6
|
* wintun: calculate path of NDIS device object symbolic linkJason A. Donenfeld2019-07-181-0/+20
|
* tun: openbsd: don't change MTU when it's already the expected sizeMichael Zeltner2019-07-181-4/+7
| | | | | | Allows for running wireguard-go as non-root user. Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
* tun: windows: spin for a bit before falling back to event objectJason A. Donenfeld2019-07-181-4/+19
|
* tun: windows: implement ring buffersSimon Rozman2019-07-171-186/+184
| | | | Signed-off-by: Simon Rozman <simon@rozman.si>
* device: immediately rekey all peers after changing device private keyJason A. Donenfeld2019-07-112-0/+25
| | | | Reported-by: Derrick Pallas <derrick@pallas.us>
* README: update windows infoJason A. Donenfeld2019-07-081-1/+1
|
* tun: windows: registration of write buffer no longer requiredJason A. Donenfeld2019-07-052-14/+1
|
* tun: windows: decrease alignment to 4Jason A. Donenfeld2019-07-051-3/+3
|
* tun: windows: delay initial writeJason A. Donenfeld2019-07-041-16/+22
| | | | Otherwise we provoke Wintun 0.3.
* device: receive: uniform message for source address checkJason A. Donenfeld2019-07-011-1/+1
|
* tun: windows: packetNum is unusedJason A. Donenfeld2019-07-011-3/+0
|
* device: receive: simplify flush loopJason A. Donenfeld2019-07-011-35/+12
|
* tun: windows: inform wintun of maximum buffer length for writesJason A. Donenfeld2019-06-261-0/+6
|
* tun: windows: never retry open on Windows 10Jason A. Donenfeld2019-06-184-13/+22
|
* tun: remove TUN prefix from types to reduce stutter elsewhereMatt Layher2019-06-1410-67/+66
| | | | Signed-off-by: Matt Layher <mdlayher@gmail.com>
* device: update transfer counters correctlyJason A. Donenfeld2019-06-113-4/+9
| | | | | The rule is to always update them to the full packet size minus UDP/IP encapsulation for all authenticated packet types.
* wintun: increase registry timeoutJason A. Donenfeld2019-06-111-1/+1
|
* wintun: add helper for cleaning upJason A. Donenfeld2019-06-101-9/+82
|
* wintun: simplify error matching and remove dumb commentsJason A. Donenfeld2019-06-101-19/+4
|
* wintun: fix comments and remove hwnd paramJason A. Donenfeld2019-06-102-124/+55
| | | | This now looks more idiomatic.
* setupapi: add SetDeviceRegistryPropertyString descriptionSimon Rozman2019-06-101-1/+2
| | | | Signed-off-by: Simon Rozman <simon@rozman.si>
* setupapi: unify ERROR_INSUFFICIENT_BUFFER handlingSimon Rozman2019-06-101-58/+32
| | | | Signed-off-by: Simon Rozman <simon@rozman.si>
* wintun: allow controlling GUIDJason A. Donenfeld2019-06-102-2/+29
|
* setupapi: add DeviceInstanceID()Jason A. Donenfeld2019-06-103-0/+41
|
* global: fixup TODO comment spacingJason A. Donenfeld2019-06-067-11/+11
|
* wintun: guid functions are upstreamJason A. Donenfeld2019-06-067-120/+14
|
* wintun: simplify DeleteInterface method signatureSimon Rozman2019-06-062-58/+79
| | | | Signed-off-by: Simon Rozman <simon@rozman.si>
* wintun: don't run HrRenameConnection in separate threadJason A. Donenfeld2019-06-051-3/+2
| | | | | It's very slow, but unfortunately we haven't a choice. NLA needs this to have completed.
* tun: windows: obsolete 256 packets per exchange buffer limitationSimon Rozman2019-06-051-2/+1
| | | | Signed-off-by: Simon Rozman <simon@rozman.si>