wireguard-openbsd/sys/kern/syscalls.master, branch jd/queueboosts WireGuard implementation for the OpenBSD kernel https://git.zx2c4.com/wireguard-openbsd/atom/sys/kern/syscalls.master?h=jd%2Fqueueboosts 2020-03-18T19:33:36Z Unlock flock(2). 2020-03-18T19:33:36Z anton anton@openbsd.org 2020-03-18T19:33:36Z urn:sha1:5d5740f67c444be4c74dd146b171e2f87ee3b937 ok mpi@ visa@ Unlock fcntl(2). 2020-03-10T19:20:14Z anton anton@openbsd.org 2020-03-10T19:20:14Z urn:sha1:c015c199e245ddc716bcb4a532be68f318f6f5bf ok visa@ Unlock ioctl(2). 2020-02-22T12:00:24Z anton anton@openbsd.org 2020-02-22T12:00:24Z urn:sha1:7570a38724322c11cb288407b5d959e114aef9c8 ok kettenis@ mpi@ visa@ Unlock close*(2) and dup*(2). 2020-01-30T15:24:06Z visa visa@openbsd.org 2020-01-30T15:24:06Z urn:sha1:3b8ae0fda4ba36bce11528de388ac5140fbe8ee7 Tested by krw@, anton@ OK mpi@, anton@ add a new __tmpfd system call that creates a new unnamed file in /tmp. 2020-01-26T04:28:18Z tedu tedu@openbsd.org 2020-01-26T04:28:18Z urn:sha1:e1764fa663c6cc4e53b337926f7b315843c43dad intended for shm/fd passing, but for programs that may otherwise like filesystem access. ok beck deraadt kettenis Make __thrsleep(2) and __thrwakeup(2) MP-safe 2020-01-21T15:20:47Z visa visa@openbsd.org 2020-01-21T15:20:47Z urn:sha1:1021dd3d027b31effd3975d80598b0ea0412dae3 Threads in __thrsleep(2) are tracked using queues, one queue per each process for synchronization between threads of a process, and one system-wide queue for the special ident -1 handling. Each of these queues has an associated rwlock that serializes access. The queue lock is released when calling copyin() and copyout() in thrsleep(). This preserves the existing behaviour where a blocked copy operation does not prevent other threads from making progress. Tested by anton@, claudio@ OK anton@, claudio@, tedu@, mpi@ Unlock a tiny part of the kqueue subsystem: kqueue(2). 2020-01-13T14:01:18Z visa visa@openbsd.org 2020-01-13T14:01:18Z urn:sha1:c979e23800b8081b9f866d81e27689d19dd2dd92 Tested by and OK anton@ OK mpi@, kettenis@ Unlock pipe(2) and pipe2(2). The kernel lock does not provide any 2020-01-02T19:03:20Z anton anton@openbsd.org 2020-01-02T19:03:20Z urn:sha1:b0d39d8ac2d57848aed66f1b67a85ff088f9500f serialization at this point since everything is protected by the pipe_lock. Thanks to tb@ and visa@ for testing; ok mpi@ visa@ Unlock nanosleep(2). 2019-12-24T04:23:43Z visa visa@openbsd.org 2019-12-24T04:23:43Z urn:sha1:4227f8354c875534e4b2e3153d17da3fa427d28c OK cheloha@, anton@, mpi@ Add dummy msyscall(2) system call which is currently a noop. This will 2019-11-27T01:04:13Z deraadt deraadt@openbsd.org 2019-11-27T01:04:13Z urn:sha1:43981c5226a00cb8b9ea74db0c340b73dbde2350 be used by kernel and ld.so in the near future. Adding the system call earlier will reduce the number of people who try to build through and encounter agony. ok kettenis guenther