Age | Commit message (Collapse) | Author |
|
multiple packets off the iface.
The actual packet data is preceded by a "struct bpf_hdr" header we need to skip. Because of that, passing a raw buffer is not going to be very useful, so bpf-read-packet! can be removed from the API.
|
|
|
|
useless for nonblocking I/O...
|
|
always the same in a macro. This should also inline it
|
|
|
|
|
|
|
|
For now, we just work with the underlying integer values. Maybe later we'll translate these to symbols?
|
|
|
|
to more closely match tcpdump-generated output
|
|
the conditional ones, too), make uint? and label? also recognise zero as valid and fix error expression of disassembler.
Implement back-and-forth transcoding of tcpdump's "decimal encoding" output of compiled packet matching instructions (using -ddd).
Fixed a few bugs I ran into with bitstring, so depend on the version to be released next (this code triggers error situations in the current version)
|
|
|
|
procedures as well as the procedure to retrieve the buffer length.
Switch to c-pointer and use LOCATION to get pointers from scheme objects.
|
|
Add an initial opaque record type holding a connection to the BPF
device. At least this now allows us to open a BPF device onto a network
interface and request which network device we're connected to.
|
|
to use this anyway in order to be able to use its IOCTLs
|
|
internal consistency tests
|