On Sat, Oct 17, 2015 at 01:33:32PM +0300, Andreas Gustafsson wrote: > The discussion that ensued branched out in lots of different > directions, but as far as I can see, none of them actually > offered a diagnosis of the original problem reported. > > To me, this looks very much like the same problem as PR 50060, which > was recently fixed in -current by riastradh@. A pullup to -7 is still > pending.
Thanks for the confirmation. I'll try out current. Since the machine was critical I reverted to 6.1.5 for now. Mayuresh