Date: Tue, 04 Jul 2017 09:51:14 +0000 From: bugzilla-noreply@freebsd.org To: freebsd-bugs@FreeBSD.org Subject: [Bug 219399] System panics after several hours of 14-threads-compilation orgies using poudriere on AMD Ryzen... Message-ID: <bug-219399-8-nqFtBWEuxn@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-219399-8@https.bugs.freebsd.org/bugzilla/> References: <bug-219399-8@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D219399 --- Comment #41 from Nils Beyer <nbe@renzel.net> --- I think you mean this post here: https://community.amd.com/message/2802591#comment-2802591 Well, maybe the hysteresis temperature of Ryzen is not 75=C2=B0C but rather= 50=C2=B0C, meaning as soon as the CPU reaches (IMHO rather low) 50=C2=B0C it starts be= coming susceptible to data corruption. That would explain the behaviour here on my system. One guy here "tosilva" suspects that there is a signal integrity issue: https://community.amd.com/message/2808194#comment-2808194 that could possibly be fixed by raising voltages; what voltages excactly he doesn't tell - so I assume every voltage. And by what amount - no idea. That's what I'm trying right now, raising every voltage by 100mV: CPU Voltage: 1.18750 -> 1.28750 VPPM: 2.550 -> 2.650 2.50V Voltage: 2.520 -> 2.620 DRAM Voltage: 1.210 -> 1.310 +1.8 Voltage: 1.820 -> 1.920 VDDP: 0.950 -> 1.050 1.05V Voltage: 1.072 -> 1.072 (can only be raised several mV, so I left= it AUTO) --=20 You are receiving this mail because: You are the assignee for the bug.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-219399-8-nqFtBWEuxn>