| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 1998-04-25 | Aparently the socketpair bug is fixed in current Linux kernels so | Todd C. Miller | |
| back out my last change. | |||
| 1998-04-25 | Linux has SHUT_RD and SHUT_WR reversed in the ``how'' argument. Presumably | Todd C. Miller | |
| they will fix this sometime, after which this change should be backed out. Pointed out by tchrist@perl.com | |||
| 1998-02-10 | recover from lint changes elsewhere, not completely tested | Theo de Raadt | |
| 1998-02-07 | emulate linux connect() braindamage even better -- only do so on async | Theo de Raadt | |
| sockets; pb@fasterix.freenix.org | |||
| 1997-12-14 | support IP_HDRINCL | Theo de Raadt | |
| 1997-12-10 | add GIFMETRIC and GIFMTU | Niels Provos | |
| 1997-12-10 | emulate linux SIOCGIFHWADDR; stolen from freebsd; not tested yet | Theo de Raadt | |
| 1997-11-27 | oh, connect() semantics are EVEN WORSE | Theo de Raadt | |
| 1997-11-26 | If using async connect(), after select for writeable, if getpeername() | Theo de Raadt | |
| indicates connection failure, you can use getsockopt(SOL_SOCKET, SO_ERROR, ...) to find the error. But the linux people I guess didn't know about this, and changed connect() so that a 2nd connect() call on a failed socket returns the error directly. Fake this demented non-standard behaviour which Linux users are now starting to code to. Talk about dilution of the API... | |||
| 1997-11-26 | indent | Theo de Raadt | |
| 1996-04-17 | Good fixes and changes from latest Net. | Michael Shalayeff | |
| 1996-03-08 | From NetBSD: | Michael Shalayeff | |
| linux compatibility changes. | |||
| 1995-10-18 | initial import of NetBSD tree | Theo de Raadt | |
 |
index : src | |
| OpenBSD base system |
| summaryrefslogtreecommitdiff |