Re: startmol >=.56 locks machine hard [was Re: Improved 9.1 support]


Subject: Re: startmol >=.56 locks machine hard [was Re: Improved 9.1 support]
From: Samuel Rydh (samuel@ibrium.se)
Date: Wed Apr 11 2001 - 12:36:37 MDT


On Sat, Apr 07, 2001 at 01:29:29PM -0700, phandel@cise.ufl.edu wrote:
> I rsync'ed linuxppc_2_4 from penguinppc.org::linux-2.4-bk ( with a slight
> patch to work around the missing process_bridge_ranges function), then
> rsync'ed your zorn.theophys.kth.se::mol tree into a new directory, did a
> make libimport, make, make install, and startmol. It locks up the machine
> hard at the GC 80800000 stage.
>

I think I've eliminated this bug now - it was caused
by the large latency of the mftb/mfdec instructions
on 604 (the 7450 was probably also affected).

The fix is in the bitkeeper repository. The rsync
snapshot will be updated within a couple of hours.

/Samuel

----------------------------------------------------------
 E-mail <samuel@ibrium.se> WWW: <http://www.ibrium.se>
  Phone/fax: (home) +46 8 4418431, (work) +46 8 7908470
----------------------------------------------------------



This archive was generated by hypermail 2a24 : Wed Apr 11 2001 - 12:39:57 MDT