Commit message (Collapse) | Author | Files | Lines | ||
---|---|---|---|---|---|
2019-05-03 | setupapi: safer aliasing of slice types | 5 | -23/+40 | ||
2019-05-03 | wintun: work around GetInterface staleness bug | 2 | -18/+30 | ||
2019-05-02 | wintun: don't retry when not creating | 2 | -9/+26 | ||
The only time we're trying to counteract the race condition is when we're creating a driver. When we're simply looking up all drivers, it doesn't make sense to retry. | |||||
2019-04-29 | wintun: try harder to open registry key | 1 | -2/+2 | ||
This sucks. Can we please find a deterministic way of doing this instead? | |||||
2019-04-29 | go.mod: use vendored winio | 2 | -1/+4 | ||
2019-04-23 | tun: freebsd: work around numerous kernel panics on shutdown | 1 | -23/+82 | ||
There are numerous race conditions. But even this will crash it: while true; do ifconfig tun0 create; ifconfig tun0 destroy; done It seems like LLv6 is related, which we're not using anyway, so explicitly disable it on the interface. | |||||
2019-04-19 | setupapi: Fix struct size mismatches | 4 | -8/+27 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-04-19 | setupapi: actually fix padding by rounding up to sizeof(void*) | 1 | -1/+1 | ||
2019-04-19 | tun: darwin: write routeSocket variable in helper | 1 | -2/+4 | ||
Otherwise the race detector "complains". | |||||
2019-04-19 | main: revise warnings | 2 | -25/+17 | ||
2019-04-18 | device: send: check packet length before freeing element | 1 | -3/+3 | ||
2019-04-18 | setupapi: Revise DrvInfoDetailData struct size calculation | 2 | -3/+7 | ||
Go adds trailing padding to DrvInfoDetailData struct in GOARCH=386 which confuses SetupAPI expecting exactly sizeof(SP_DRVINFO_DETAIL_DATA). Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-04-13 | windows: use proper constants from updated x/sys | 5 | -25/+28 | ||
2019-04-13 | conn: linux: RTA_MARK has moved to x/sys | 1 | -1/+1 | ||
2019-04-11 | tun: windows: Adopt new error codes returned by Wintun | 1 | -3/+2 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-04-11 | tun: windows: do not sleep after OPERATION_ABORTED on write | 1 | -2/+1 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-04-09 | main_windows: use proper version constant | 1 | -1/+1 | ||
2019-04-09 | version: put version in right place0.0.20190409 | 2 | -4/+1 | ||
2019-04-09 | version: bump snapshot | 1 | -1/+1 | ||
2019-04-09 | tun: windows: do not sleep after OPERATION_ABORTED | 1 | -2/+1 | ||
2019-04-04 | tun: windows: Retry R/W on ERROR_OPERATION_ABORTED | 1 | -18/+44 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-04-03 | tun: windows: Attempt to reopen handle on all errors | 1 | -15/+9 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-04-01 | wintun: rename device using undocumented API that netsh.exe uses | 2 | -1/+41 | ||
2019-04-01 | wintun: add more retry loops | 3 | -30/+56 | ||
2019-03-26 | tun: windows: cancel ongoing reads on closing and delete after close | 1 | -4/+11 | ||
This reverts commit 52ec440d7977fad966002c3710ed9df957943407 and adds some spice. | |||||
2019-03-22 | wintun: query for NetCfgInstanceId several times | 1 | -5/+16 | ||
2019-03-22 | tun: windows: delete interface before deleting file handles | 1 | -9/+4 | ||
2019-03-22 | tun: windows: Make adapter rename asynchronous | 2 | -60/+13 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-03-22 | tun: windows: Adapter devices renamed to WINTUN<LUID Index> | 4 | -52/+63 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-03-22 | tun: windows: Increase unavailable adapter timeout to 30sec | 1 | -1/+1 | ||
5 seconds was too short when debugging. Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-03-22 | tun: windows: Make writing persistent too | 1 | -15/+17 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-03-22 | tun: windows: Fix paused adapter test | 1 | -1/+1 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-03-21 | tun: windows: just open two file handles | 3 | -162/+83 | ||
2019-03-21 | receive: implement flush semantics | 12 | -98/+147 | ||
2019-03-21 | tun: windows: add dummy overlapped events back | 1 | -2/+21 | ||
These seem basically wrong to me, but we get crashes without them. | |||||
2019-03-20 | tun: windows: use new constants in sys | 3 | -13/+15 | ||
2019-03-20 | uapi: remove unhelpful log messages | 1 | -7/+0 | ||
2019-03-21 | wintun: Use native Win32 API for I/O | 5 | -38/+173 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-03-19 | uapi: report endpoint error | 1 | -1/+1 | ||
2019-03-18 | tun: windows: wintun does iocp | 2 | -142/+49 | ||
2019-03-13 | tun: windows: temporary hack for forcing MTU | 1 | -1/+8 | ||
2019-03-10 | wintun: Poll more often | 1 | -1/+1 | ||
2019-03-10 | uapi: make ipcerror conform to interface | 1 | -2/+2 | ||
2019-03-08 | wintun: Make errors more descriptive | 1 | -15/+20 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-03-08 | wintun: Return correct reboot-req flag on CreateInterface() error too | 1 | -1/+1 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-03-08 | wintun: Fix double-quoted strings escaping on output | 2 | -4/+4 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-03-08 | wintun: Introduce SetupAPI enumerator and machineName consts | 1 | -4/+7 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> | |||||
2019-03-08 | Windows: Apply strict security descriptor on pipe server | 1 | -1/+12 | ||
Signed-off-by: Odd Stranne <odd@mullvad.net> | |||||
2019-03-08 | uapi: windows: work out pipe semantics | 1 | -2/+2 | ||
Pipes can be arranged like this, so that's fine. We also apply a strict SDDL that can't be inherited and only gives access to local system. Developed-with: Odd Stranne <odd@mullvad.net> | |||||
2019-03-07 | wintun: Cleanup | 1 | -1/+1 | ||
Signed-off-by: Simon Rozman <simon@rozman.si> |