Re: writev() to /dev/bpf from different sockets block

[Jason Curl <jcurl%arcor.de@localhost>]

On 1/25/25 21:04, Mouse wrote:
> I mostly don't have anything to say on this thread.  But...
>
> > [...]
> >       if (errno == ENOBUFS) {
> >         delay(750000);
> > [...]
>
> > Secondary problem: I couldn't find an effective way to handle
> > ENOBUFS, other than with a 750us delay
>
> That looks to me like 750ms, not 750us.  At least, on 9.1 delay() is
> documented as taking us, not ns.

It's pseudo code. The actual implementation uses clock_nanosleep() in 
C++ with the literal 750us given to the function. But when testing on 
NetBSD, it never gets here anyway (but does on FreeBSD quite regularly).

> /~\ The ASCII				  Mouse
> \ / Ribbon Campaign
>   X  Against HTML		mouse%rodents-montreal.org@localhost
> / \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B