On Mon, Nov 06, 2017 at 03:33:43PM +0000, Patrick Welche wrote: > On Thu, Nov 02, 2017 at 03:22:56PM +0000, Patrick Welche wrote: > > On Thu, Nov 02, 2017 at 10:27:24AM +0000, Patrick Welche wrote: > > > -current/amd64 instability spread to a box I updated to -current > > > last night: it panicked and rebooted over night, with no core, so > > > no clues... The only clue so far has been the pool_cache_get > > > returning NULL. > > > > Instability spread to a laptop - just cvs updated its copy of src, built, > > rebooted. Spontaneous reboot, no core. I did see once on start, i.e., while > > the rc scripts were run "pid... (sh) out of swap" (!) > > > > Still no clues left behind... > > Still occur after mlelstv's pool_grow fix (which probably fixed > the one decent core dump I had). 4 different (different cpu / > chipset) amd64 computers are suffering. October 23 netbsd.old > kernels are stable. Not obvious how to cause the panic either, > so still no clues...
As that random reboot was on a laptop, it could just be that LOCKDEBUG + DIAGNOSTIC pushed it over the heat edge. One of the desktops is now stable, so maybe mlelstv's pool_grow fix really did fix all of this. :-) Thanks, Patrick
