Current-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: build simply stops



On 25 July 2012 17:33, Patrick Welche <prlw1%cam.ac.uk@localhost> wrote:
> On Tue, Jul 24, 2012 at 10:16:51PM +0100, Chavdar Ivanov wrote:
>> -j24 enough parallel?
>
> That should be!

It is a six core Xeon E5-1660 with HT and 16GB memory, build.sh for
amd64 and i386, all of release, iso-image, live-image and
install-image takes about 90 minutes. BUILD-NetBSD doubles the threads
- that's where the 24 comes from.

>
> So, I just tried again - cvs update at Jul 25 13:57 GMT, new kernel,
> newfs the /usr/obj partition, build -j4 -E -x build, and this time
> I'm as far as
>
> ln -sf libgssapi.so.10.0 libgssapi.so.10.tmp
> mv -f libgssapi.so.10.tmp libgssapi.so.10
> ln -sf libgssapi.so.10.0 libgssapi.so.tmp
> mv -f libgssapi.so.tmp libgssapi.so
>
> and nothing is happening, 100% idle, 9 nbmake processes in S.
> Maybe a clue:  3997 ttyp2 Z     0:00.00 (x86_64--netbsd-o) ?
>
> Still no idea...

Well, there may be something. The build I showed the log above
completed without a problem, I then did 'INSTALL-NetBSD install',
synced and rebooted, losing the connection to the host (I was at home
at the time, did the work over VPN). This morning I found the system
hanging, still with the X display from yesterday, tried to interrupt
and do a blind 'reboot 0x104', to no avail; I then reset the machine
and rebuilt everything from scratch (erasing the old obj directory and
cvs updating). Now it seems fine. I usually am not too bothered when I
have an occasional glitch - it is -current, after all, I wait for a
while, update, rebuild and carry on. In my view -current is actually
very stable for an operating system in constant update.

>
> Cheers,
>
> Patrick

Chavdar


-- 
----


Home | Main Index | Thread Index | Old Index