| Age | Commit message (Expand) | Author |
|---|
| 2015-06-16 | Store a unique ID, an interface index, rather than a pointer to the | Martin Pieuchot |
| 2015-05-13 | test mbuf pointers against NULL not 0 | Jonathan Gray |
| 2015-02-10 | First step towards making uiomove() take a size_t size argument: | Miod Vallat |
| 2015-02-10 | make bpf(4) able to filter based on a pf(4) queue ID for tcpdump -Q qname | Martin Pelikan |
| 2015-01-29 | back bpf.c down to 1.113, from before most recent timeout changes. | Ted Unangst |
| 2015-01-28 | when doing a blocking read with a timeout, after the sleep reset | David Gwynne |
| 2015-01-09 | correctly handle no timeouts and make timeout handling in general better. | Ted Unangst |
| 2014-12-16 | primary change: move uvm_vnode out of vnode, keeping only a pointer. | Ted Unangst |
| 2014-12-02 | replace some malloc multiplies with mallocarry. ok deraadt henning | Ted Unangst |
| 2014-11-23 | length argument for some free() calls; ok doug | Theo de Raadt |
| 2014-10-07 | when running bpf on an outgoing vlan interface that doesnt have a | David Gwynne |
| 2014-09-23 | lock around the sysctl code that sets the bpf buffer sizes so if we ever | David Gwynne |
| 2014-09-22 | remove a stupid comment above bpfilterattach about how we dont do anything | David Gwynne |
| 2014-09-22 | stash a pointer to bpf_d in the knotes kn_hook instead of the device id. | David Gwynne |
| 2014-09-22 | it's easy to allow bpfwrites bigger than MCLBYTES now that we have | David Gwynne |
| 2014-09-22 | if you request a read timeout and then use kqueues to wait for them, you | David Gwynne |
| 2014-09-19 | passing M_NOWAIT to m_tag_get means it can fail, which could hit | David Gwynne |
| 2014-07-12 | add a size argument to free. will be used soon, but for now default to 0. | Ted Unangst |
| 2014-07-12 | sizeof(afh), afh being uint32, is cooler than literal "4" | Henning Brauer |
| 2014-07-10 | time to claim copyright | Henning Brauer |
| 2014-07-10 | some say you don't need NULL checks before free(). Not 0 either. | Henning Brauer |
| 2014-07-10 | introduce the revolutionary concept of NULL pointers. ok gcc | Henning Brauer |
| 2014-07-10 | introduce bpf_mcopy_stripvlan, which cuts the 4 extra bytes out of the | Henning Brauer |
| 2014-07-09 | Add support bpfwrite on DLT_LOOP interfaces. | YASUOKA Masahiko |
| 2014-07-09 | Herr Reyk correctly pointed out that we don't need the if_pflog.h include | Henning Brauer |
| 2014-07-09 | tedu bpf_mtap_pflog(). | Henning Brauer |
| 2014-07-09 | bpf code surgery / shuffling / simplification. | Henning Brauer |
| 2014-04-23 | Don't attempt to deal with link types supported by no drivers in the | Jeremie Courreges-Anglas |
| 2014-04-14 | "struct pkthdr" holds a routing table ID, not a routing domain one. | Martin Pieuchot |
| 2014-03-30 | Eliminates struct pcred by moving the real and saved ugids into | Philip Guenther |
| 2013-12-24 | rearrange/correct timeout conditionals to work better. | Ted Unangst |
| 2013-11-29 | panics still being reported. send bpf.c back to 1.85 | Ted Unangst |
| 2013-11-17 | speeling | David Gwynne |
| 2013-11-15 | calculate the line in the sand before comparing it to ticks, which looks | David Gwynne |
| 2013-11-12 | try bpf.c r1.84 again, this time without semantic changes to if statements. | David Gwynne |
| 2013-11-11 | Revert bpf.c 1.84 / bpfdesc.h 1.19 for now, "panic: timeout_add: to_ticks (-1) | Stuart Henderson |
| 2013-11-11 | replace the user of ticks in a condition like "interval + start < ticks" | David Gwynne |
| 2012-12-28 | change the malloc(9) flags from M_DONTWAIT to M_NOWAIT; OK millert@ | Gleydson Soares |
| 2012-12-21 | Rather than calling mircotime in bpf_catchpacket each time it's called | Mike Belopuhov |
| 2012-12-21 | bpf allocates packet buffers in the ioctl path and can sleep | Mike Belopuhov |
| 2012-04-14 | Use DLT_LOOP for all tunneling interfaces. | YASUOKA Masahiko |
| 2012-01-16 | bpf devices behave similar to raw sockets and never block on write, | Philip Guenthe |
| 2011-07-02 | kqueue attach functions should return an errno or 0, not a plain 1. Fix | Nicholas Marriott |
| 2011-01-04 | in bpf_movein(), range-check mbuf size against MCLBYTES before | Theo de Raadt |
| 2010-09-21 | pflog overhaul | Henning Brauer |
| 2009-11-09 | Every selwakeup() should have a matching KNOTE() (even if kqueue isn't | Nicholas Marriott |
| 2009-10-26 | Set the rdomain in bpfwrite() to the interface rdomain so that bpf sender | Claudio Jeker |
| 2009-09-21 | Properly reference count bpf descriptors when using kqueue. | Can Erkin Acar |
| 2009-09-07 | de-inline a function which gains absolutely no benefit at all from it | Theo de Raadt |
| 2008-11-26 | dont have bpf.h expose the kernel ticks variable wherever it is includeing. | David Gwynne |
