Re: Strange kernel build breakage (after r314283?)

On Tue, 7 Mar 2017 09:13:58 -0800 Kevin Oberman <rkoberman_at_gmail.com> wrote

> On Mon, Mar 6, 2017 at 9:18 AM, Lev Serebryakov <lev_at_freebsd.org> wrote:
> 
> > On 06.03.2017 20:10, Lev Serebryakov wrote:
> >
> > >  I've got this error when tried to update my -CURRENT VM to r314772:
> > >
> > > /data/src/sys/cam/cam_xpt.c:84:1: error: static_assert failed
> > > "XPT_PRINT_LEN is too large"
> > > _Static_assert(XPT_PRINT_LEN <= XPT_PRINT_MAXLEN, "XPT_PRINT_LEN is too
> > > large");
> > > ^              ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > >
> > >  I didn't define any XPT_xxxx macro by hands, but I have
> > >
> > > options         PRINTF_BUFR_SIZE=1024
> > >
> > >  in my kernel config.
> >  Yep, removing this option helps, but it is surprising and not obvious
> > at all!
> >
> > --
> > // Lev Serebryakov
> >
> 
> If my memory is good (and it may not be), this option was recommended to
> prevent garbled syslog and console entries, but that was back in v8 days,
> long, long ago. I have not had his problem for a long time and I think that
> the option is no longer required and even they, 1024 was a LOT bigger than
> was recommended at the time. 128 or 256 seems tike the value recommended.

Relax. You're memory is still in good order. :-)
It was in fact the reason. I had to add the then suggested amount:
PRINTF_BUFR_SIZE=128
to my KERNCONF even into early 9. But haven't required it since
~mid-9.

OTOH I'm now seeing something similar on CURRENT. Only somewhat
in reverse.
The last message I receive on the console following halt(8) is
the message telling me the NIC has been brought down. It then
sits there until I hit the enter key to reboot(8).

But that's another topic for another thread. :-)

--Chris
> --
> Kevin Oberman, Part time kid herder and retired Network Engineer
> E-mail: rkoberman_at_gmail.com
> PGP Fingerprint: D03FB98AFA78E3B78C1694B318AB39EF1B055683