March 13, 201610 yr Hardware failure has forced me to upgrade, and this is where I have found Haswell and unRAID don't play nice together when it comes to SpeedStep. root@unRAID:~# cat /proc/cpuinfo | grep MHz cpu MHz : 3500.000 cpu MHz : 3500.273 cpu MHz : 3500.410 cpu MHz : 3500.546 cpu MHz : 3500.820 cpu MHz : 3461.992 cpu MHz : 3500.000 cpu MHz : 3501.367 top - 10:29:01 up 20:50, 1 user, load average: 0.22, 0.31, 0.35 Tasks: 309 total, 1 running, 308 sleeping, 0 stopped, 0 zombie Cpu(s): 0.5%us, 0.1%sy, 0.0%ni, 99.0%id, 0.4%wa, 0.0%hi, 0.0%si, 0.0%st Mem: 32895660k total, 29897840k used, 2997820k free, 136k buffers Swap: 0k total, 0k used, 0k free, 27557820k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 5788 nobody 20 0 4023m 312m 26m S 1 1.0 29:48.25 mono 3778 root 20 0 1021m 19m 812 S 1 0.1 114:18.91 shfs 4045 avahi 20 0 34448 3120 2724 S 0 0.0 0:00.24 avahi-daemon 5719 nobody 20 0 3596m 131m 9.8m S 0 0.4 49:10.41 python2 5790 nobody 20 0 2333m 142m 5880 S 0 0.4 1:42.57 python2 6835 nobody 20 0 420m 85m 13m S 0 0.3 4:43.20 deluged 7225 nobody 20 0 5699m 221m 9180 S 0 0.7 2:22.72 java 8161 nobody 20 0 610m 33m 0 S 0 0.1 4:04.27 mongod 29204 root 20 0 13416 2348 1888 R 0 0.0 0:00.04 top 1 root 20 0 4368 1560 1464 S 0 0.0 0:04.82 init 2 root 20 0 0 0 0 S 0 0.0 0:00.00 kthreadd 3 root 20 0 0 0 0 S 0 0.0 0:11.52 ksoftirqd/0 5 root 0 -20 0 0 0 S 0 0.0 0:00.00 kworker/0:0H 7 root 20 0 0 0 0 S 0 0.0 1:00.10 rcu_preempt 8 root 20 0 0 0 0 S 0 0.0 0:00.00 rcu_sched 9 root 20 0 0 0 0 S 0 0.0 0:00.00 rcu_bh 10 root RT 0 0 0 0 S 0 0.0 0:00.32 migration/0 With my Haswell CPU doing practically nothing it still wants to run at full speed, where my old Bloomfield was very happy to clock down. The state show above is a SSH Session,Docker doing its thing and my kids watching a show. No WebUI. From what I have read it can sort of be adjusted with "intel_pstate=disable" Info about that was found HERE Has 6.2 address this issue or is the work around still required? Thanks
March 17, 201610 yr I've been looking for ages for this fix! Thanks (Sorry, I can't confirm if it's working on the 6.2 beta, not managed to get the beta running on my server yet haha)
March 18, 201610 yr Author Still trying to get an answer from the 6.2 thread. I would try it, but they seem to be having some real issues right now, so 6.1.9 is the much safer place to be
April 19, 201610 yr 6.2 doesn't fix the issue on Haswell. However, I can say that Skylake doesn't have the issue at all - everything works fine on both 6.1.9 and 6.2b21. Well, 6.1.9 doesn't work with the i219V NIC that's run by the Z170 chipset, at least on my Gigabyte Z170N-wifi board. I was able to use the i211AT no problems.
April 19, 201610 yr Is this just Haswell or Haswell-E as well? How did you get the information in the "code" section?
April 20, 201610 yr Is this just Haswell or Haswell-E as well? How did you get the information in the "code" section? From a Linux terminal he did cat /proc/cpuinfo | grep MHz for the first section, and top for the second.
April 20, 201610 yr Is this just Haswell or Haswell-E as well? How did you get the information in the "code" section? This does not effect Haswell-e, my 5930k does not have this issue. My previous 4790s very much had this issue, and the intel_pstate=disable was the best way to "fix" it. It is well known (from people with the issue, as others will tell you differently) that this is not a reporting issue, but a real issue with the CPU not properly throttling. I confirmed this with both temp and wattage at the plug.
April 21, 201610 yr Author good to see this has not been forgotten, maybe with a little luck 6.2 buy the time it hits RC will have it resolved.
August 16, 20169 yr Currently running 6.2 RC3... Speedstep still doesn't work correctly I will have to re-add the pstate back into my config Edit: On a Haswell Refresh Xeon 1231v3
Archived
This topic is now archived and is closed to further replies.