> As Matthew Jacob wrote ...
> >
> > Q'uelle strange- what I have... well, we'll see what my nightly build
> > brings...
>
> Mine fell over in 'truss'. I already reported it to Marcel, looks
> like something was changed in kdump/mkioctls, a script that 'truss' also
> uses.
Mine's still going (I start mine at 0430 PST8PDT, the update from
freefall runs at 0107 PST8PDT)).... Nope, what a coincidence. It *just*
finished. Went all the way through w/o errors on a buildworld.
>
> >
> > On Fri, 3 Dec 1999, Jordan K. Hubbard wrote:
> >
> > > > What is beast? I was happily able to build -current at about 5 this
> > > > morning on a pc164.
> > >
> > > It's an Aspen systems DEC Durango PC164 motherboard based system.
>
> Doesn't that one have ECC memory (refering to a previous post where
> you theorised a bit on memory problems and sig-11s ;)
It may or may not, but memory problems are typically detected pretty well
on alphas and you'll get a 670 machine check if it was recoverable (which
the PALcode has already done, except for TurboLaser if the memory problem
was caused by one of the I/O boards (PALcode cannot go offboard to read
registers on another board to find out what address was porked, IIRC)), or
a 660 machine check, which is a panic. At any rate, this is a different
entry point than mmu traps.
It's not like I don't get sig-11s... ctags always seems to core dump on
me- somewhere up in libc.
-matt
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
