Re: CPU frequency doesn't drop below 1200MHz (like it used to)
Frequency control may not be relevant on that platform. Try installing the intel-pcm package; then # kldload cpuctl # pcm.x 1 Then paste some of that in here. Let's see if the CPU is idling some other way. -adrian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Sat, May 23, 2015 at 10:41 PM, Adrian Chadd adr...@freebsd.org wrote: Frequency control may not be relevant on that platform. Try installing the intel-pcm package; then # kldload cpuctl # pcm.x 1 Then paste some of that in here. Let's see if the CPU is idling some other way. -adrian Five iterations one every second: Script started on Sun May 24 00:07:18 2015 command: sudo pcm.x 1 -i=5 Intel(r) Performance Counter Monitor V2.8 (2014-12-18 12:52:39 +0100 ID=ba39a89) Copyright (c) 2009-2014 Intel Corporation Number of physical cores: 1 Number of logical cores: 4 Number of online logical cores: 4 Threads (logical cores) per physical core: 4 Num sockets: 1 Physical cores per socket: 1 Core PMU (perfmon) version: 3 Number of core PMU generic (programmable) counters: 2 Width of generic (programmable) counters: 40 bits Number of core PMU fixed counters: 3 Width of fixed counters: 40 bits Nominal core frequency: 166000 Hz Delay: 1 Detected Intel(R) Atom(TM) CPU D510 @ 1.66GHz Intel(r) microarchitecture codename Atom(tm) EXEC : instructions per nominal CPU cycle IPC : instructions per CPU cycle FREQ : relation to nominal CPU frequency='unhalted clock ticks'/'invariant timer ticks' (includes Intel Turbo Boost) L2MISS: L2 cache misses L2HIT : L2 cache hit ratio (0.00-1.00) TEMP : Temperature reading in 1 degree Celsius relative to the TjMax temperature (thermal headroom): 0 corresponds to the max temperature Core (SKT) | EXEC | IPC | FREQ | L2MISS | L2HIT | TEMP 00 0.00 0.19 0.00 5513 0.85 89 10 0.00 0.37 0.00 2676 0.84 89 20 0.00 0.39 0.01 21 K0.83 N/A 30 0.00 0.28 0.00 4731 0.64 N/A - TOTAL * 0.00 0.33 0.00 34 K0.82 N/A Instructions retired: K ; Active cycles: 20 M ; Time (TSC): 1765 Mticks ; C0 (active,non-halted) core residency: 0.28 % C1 core residency: 99.72 %; C2 package residency: 0.00 %; C4 package residency: 0.00 %; C6 package residency: 0.00 %; PHYSICAL CORE IPC : 1.33 = corresponds to 66.42 % utilization for cores in active state Instructions per nominal CPU cycle: 0.00 = corresponds to 0.19 % core utilization over time interval -- EXEC : instructions per nominal CPU cycle IPC : instructions per CPU cycle FREQ : relation to nominal CPU frequency='unhalted clock ticks'/'invariant timer ticks' (includes Intel Turbo Boost) L2MISS: L2 cache misses L2HIT : L2 cache hit ratio (0.00-1.00) TEMP : Temperature reading in 1 degree Celsius relative to the TjMax temperature (thermal headroom): 0 corresponds to the max temperature Core (SKT) | EXEC | IPC | FREQ | L2MISS | L2HIT | TEMP 00 0.00 0.19 0.00 6296 0.82 89 10 0.00 0.35 0.00 12 K0.81 89 20 0.00 0.44 0.00 6378 0.84 N/A 30 0.00 0.24 0.00 3846 0.86 N/A - TOTAL * 0.00 0.34 0.00 29 K0.83 N/A Instructions retired: 6646 K ; Active cycles: 19 M ; Time (TSC): 1766 Mticks ; C0 (active,non-halted) core residency: 0.28 % C1 core residency: 99.72 %; C2 package residency: 0.00 %; C4 package residency: 0.00 %; C6 package residency: 0.00 %; PHYSICAL CORE IPC : 1.34 = corresponds to 67.19 % utilization for cores in active state Instructions per nominal CPU cycle: 0.00 = corresponds to 0.19 % core utilization over time interval -- EXEC : instructions per nominal CPU cycle IPC : instructions per CPU cycle FREQ : relation to nominal CPU frequency='unhalted clock ticks'/'invariant timer ticks' (includes Intel Turbo Boost) L2MISS: L2 cache misses L2HIT : L2 cache hit ratio (0.00-1.00) TEMP : Temperature reading in 1 degree Celsius relative to the TjMax temperature (thermal headroom): 0 corresponds to the max temperature Core (SKT) | EXEC | IPC | FREQ | L2MISS | L2HIT | TEMP 00 0.00 0.25 0.00 12 K0.74 89 10 0.00 0.42 0.00 3166 0.94 89 20 0.00 0.19 0.00 4869 0.68 94 30 0.00 0.36 0.00 13 K0.81 94 - TOTAL * 0.00 0.34 0.00 33 K0.82 N/A Instructions retired: 8041 K ; Active cycles: 23 M ; Time (TSC): 1755 Mticks ; C0 (active,non-halted) core residency: 0.34 % C1 core residency: 99.66 %; C2 package residency: 0.00 %; C4 package
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Fri, 22 May 2015 20:26:40 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 8:19 PM, Ian Smith smi...@nimnet.asn.au wrote: On Fri, 22 May 2015 16:28:49 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 10:42 AM, Ivan Klymenko fi...@ukr.net wrote: [..] Try changing the options in /boot/device.hints hint.acpi_throttle.0.disabled=0 hint.p4tcc.0.disabled=0 Thanks, those also fixed powerd(8) for me that stopped working after upgrading to stable/10 from releng/10.1. Why are those setting suddenly needed now? -Kimmo [..] Can you say exactly in what way powerd stopped working then? Powerd(8) complained (excerpt from dmesg -a): Starting powerd. powerd: no cpufreq(4) support -- aborting: No such file or directory /etc/rc: WARNING: failed to start powerd Putting those two settings in loader.conf and rebooting fixed the problem and powerd started working again apparently because cpufreq(4) device was available again. Ok, if anabling acpi_throttle and/or p4tcc made cpufreq - and thus powerd - work for you, then it seems likely that you do not have EST enabled in your BIOS. Or at least, we've seen another instance where that was the case, which was fixed by enabling EST (or however your particular BIOS refers to it .. AMD for example use different terms). What CPU is this? In what machine? If EST (ono) IS enabled in your BIOS, this needs further investigation. As is, powerd may be running, but it's doing so highly inefficiently; refer to Stefan, Adrian and Kevin's responses for details. cheers, Ian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Sat, 23 May 2015 14:01:16 +0300, Kimmo Paasiala wrote: On Sat, May 23, 2015 at 10:00 AM, Ian Smith smi...@nimnet.asn.au wrote: On Fri, 22 May 2015 20:26:40 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 8:19 PM, Ian Smith smi...@nimnet.asn.au wrote: On Fri, 22 May 2015 16:28:49 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 10:42 AM, Ivan Klymenko fi...@ukr.net wrote: [..] Try changing the options in /boot/device.hints hint.acpi_throttle.0.disabled=0 hint.p4tcc.0.disabled=0 Thanks, those also fixed powerd(8) for me that stopped working after upgrading to stable/10 from releng/10.1. Why are those setting suddenly needed now? [..] Can you say exactly in what way powerd stopped working then? Powerd(8) complained (excerpt from dmesg -a): Starting powerd. powerd: no cpufreq(4) support -- aborting: No such file or directory /etc/rc: WARNING: failed to start powerd Putting those two settings in loader.conf and rebooting fixed the problem and powerd started working again apparently because cpufreq(4) device was available again. Ok, if anabling acpi_throttle and/or p4tcc made cpufreq - and thus powerd - work for you, then it seems likely that you do not have EST enabled in your BIOS. Or at least, we've seen another instance where that was the case, which was fixed by enabling EST (or however your particular BIOS refers to it .. AMD for example use different terms). What CPU is this? In what machine? If EST (ono) IS enabled in your BIOS, this needs further investigation. As is, powerd may be running, but it's doing so highly inefficiently; refer to Stefan, Adrian and Kevin's responses for details. It's an Intel Atom running amd64 version of FreeBSD stable/10: FreeBSD firewall.rdnzl.info 10.1-STABLE FreeBSD 10.1-STABLE #1 r283292: Sat May 23 01:08:03 EEST 2015 r...@firewall.rdnzl.info:/usr/obj/usr/src/sys/GENERIC amd64 CPU: Intel(R) Atom(TM) CPU D510 @ 1.66GHz (1666.68-MHz K8-class CPU) Origin=GenuineIntel Id=0x106ca Family=0x6 Model=0x1c Stepping=10 Features=0xbfebfbffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE Features2=0x40e31dSSE3,DTES64,MON,DS_CPL,TM2,SSSE3,CX16,xTPR,PDCM,MOVBE AMD Features=0x20100800SYSCALL,NX,LM AMD Features2=0x1LAHF TSC: P-state invariant, performance statistics Powerd was working on 10.1-RELEASE but stopped working after upgrade to 10-STABLE and nothing was changed in BIOS settings. Which would be consistent with EST not being enabled in your BIOS; with no EST, cpufreq(4) still checks for 'relative' drivers such as p4tcc or acpi_throttle and uses that, as a last resort really; with those also disabled, no cpufreq, so no powerd. Have you checked BIOS settings to confirm that you do have SpeedStep (however termed) properly enabled? Please show `sysctl dev.cpu dev.est` and `sysctl -a | grep freq_levels` However, reading the other replies to this thread I get the impression that powerd(8) doesn't actually save energy on this platform and I'm better off without it? No, I don't think that's correct; using deeper C-states is most likely a bigger win, but higher than needed CPU freq will still use extra power, so run hotter. `sysctl dev.cpu` will also reveal your C-state usage. Reason I'm pursuing this is that this change shouldn't hurt, but it will flush out those cases where people were only getting cpufreq due to use of a 'relative' cpufreq driver like p4tcc, unless EST's enabled in BIOS; I suspect yours may be one such case :) If not, there's a bug to fix. cheers, Ian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Sat, May 23, 2015 at 10:00 AM, Ian Smith smi...@nimnet.asn.au wrote: On Fri, 22 May 2015 20:26:40 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 8:19 PM, Ian Smith smi...@nimnet.asn.au wrote: On Fri, 22 May 2015 16:28:49 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 10:42 AM, Ivan Klymenko fi...@ukr.net wrote: [..] Try changing the options in /boot/device.hints hint.acpi_throttle.0.disabled=0 hint.p4tcc.0.disabled=0 Thanks, those also fixed powerd(8) for me that stopped working after upgrading to stable/10 from releng/10.1. Why are those setting suddenly needed now? -Kimmo [..] Can you say exactly in what way powerd stopped working then? Powerd(8) complained (excerpt from dmesg -a): Starting powerd. powerd: no cpufreq(4) support -- aborting: No such file or directory /etc/rc: WARNING: failed to start powerd Putting those two settings in loader.conf and rebooting fixed the problem and powerd started working again apparently because cpufreq(4) device was available again. Ok, if anabling acpi_throttle and/or p4tcc made cpufreq - and thus powerd - work for you, then it seems likely that you do not have EST enabled in your BIOS. Or at least, we've seen another instance where that was the case, which was fixed by enabling EST (or however your particular BIOS refers to it .. AMD for example use different terms). What CPU is this? In what machine? If EST (ono) IS enabled in your BIOS, this needs further investigation. As is, powerd may be running, but it's doing so highly inefficiently; refer to Stefan, Adrian and Kevin's responses for details. cheers, Ian It's an Intel Atom running amd64 version of FreeBSD stable/10: FreeBSD firewall.rdnzl.info 10.1-STABLE FreeBSD 10.1-STABLE #1 r283292: Sat May 23 01:08:03 EEST 2015 r...@firewall.rdnzl.info:/usr/obj/usr/src/sys/GENERIC amd64 CPU: Intel(R) Atom(TM) CPU D510 @ 1.66GHz (1666.68-MHz K8-class CPU) Origin=GenuineIntel Id=0x106ca Family=0x6 Model=0x1c Stepping=10 Features=0xbfebfbffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE Features2=0x40e31dSSE3,DTES64,MON,DS_CPL,TM2,SSSE3,CX16,xTPR,PDCM,MOVBE AMD Features=0x20100800SYSCALL,NX,LM AMD Features2=0x1LAHF TSC: P-state invariant, performance statistics Powerd was working on 10.1-RELEASE but stopped working after upgrade to 10-STABLE and nothing was changed in BIOS settings. However, reading the other replies to this thread I get the impression that powerd(8) doesn't actually save energy on this platform and I'm better off without it? -Kimmo ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Sat, 23 May 2015 17:40:26 +0300, Kimmo Paasiala wrote: On Sat, May 23, 2015 at 5:15 PM, Ian Smith smi...@nimnet.asn.au wrote: [..] It's an Intel Atom running amd64 version of FreeBSD stable/10: FreeBSD firewall.rdnzl.info 10.1-STABLE FreeBSD 10.1-STABLE #1 r283292: Sat May 23 01:08:03 EEST 2015 r...@firewall.rdnzl.info:/usr/obj/usr/src/sys/GENERIC amd64 CPU: Intel(R) Atom(TM) CPU D510 @ 1.66GHz (1666.68-MHz K8-class CPU) Origin=GenuineIntel Id=0x106ca Family=0x6 Model=0x1c Stepping=10 Features=0xbfebfbffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE Features2=0x40e31dSSE3,DTES64,MON,DS_CPL,TM2,SSSE3,CX16,xTPR,PDCM,MOVBE AMD Features=0x20100800SYSCALL,NX,LM AMD Features2=0x1LAHF TSC: P-state invariant, performance statistics Powerd was working on 10.1-RELEASE but stopped working after upgrade to 10-STABLE and nothing was changed in BIOS settings. [..] However, reading the other replies to this thread I get the impression that powerd(8) doesn't actually save energy on this platform and I'm better off without it? No, I don't think that's correct; using deeper C-states is most likely a bigger win, but higher than needed CPU freq will still use extra power, so run hotter. `sysctl dev.cpu` will also reveal your C-state usage. Reason I'm pursuing this is that this change shouldn't hurt, but it will flush out those cases where people were only getting cpufreq due to use of a 'relative' cpufreq driver like p4tcc, unless EST's enabled in BIOS; I suspect yours may be one such case :) If not, there's a bug to fix. Seems _I've_ got a bug to fix; I need to stop assuming all modern Intel CPUs are going to make SpeedStep and/or deeper C-states available :( Looking deeper into this it appears I don't have speedstep (EST) support in the CPU it being a crappy Atom D510: http://ark.intel.com/products/43098 Indeed. It is rated at only 13W TDP, so relatively low power anyway. This the full 'sysctl dev.cpu' output: % sysctl dev.cpu dev.cpu.3.cx_usage: 100.00% last 65712us dev.cpu.3.cx_lowest: C1 dev.cpu.3.cx_supported: C1/1/0 [..] dev.cpu.0.cx_usage: 100.00% last 3132us dev.cpu.0.cx_lowest: C1 dev.cpu.0.cx_supported: C1/1/0 dev.cpu.0.%parent: acpi0 dev.cpu.0.%pnpinfo: _HID=none _UID=0 dev.cpu.0.%location: handle=\_PR_.P001 dev.cpu.0.%driver: cpu dev.cpu.0.%desc: ACPI CPU dev.cpu.%parent: It doesn't even provide dev.cpu.0.freq, and has no deeper C-states ('Idle States' on that page) available, so it looks like you may as well not bother running powerd. Others maybe can offer better suggestions. So I should keep those two hints in loader.conf to use p4tcc I guess? If this is a desktop I'd just let it run flat out, ie disable p4tcc and acpi_throttle, have no cpufreq and forget powerd. If it's a laptop and power consumption on battery matters to you, you could see if p4tcc's lower frequencies actually save any power much, by running 'powerd -v' in a terminal while testing with different loads, or if your 'acpiconf -i0' shows discharging rates in mA or mW, or both. Sorry again for my poor assumption, and thanks for the data point! cheers, Ian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
Hm, no thermal monitoring and no speedstep. Could be dangerous/fun. What's the output of sysctl dev.cpu.0 ? -adrian On 23 May 2015 at 07:40, Kimmo Paasiala kpaas...@gmail.com wrote: On Sat, May 23, 2015 at 5:15 PM, Ian Smith smi...@nimnet.asn.au wrote: On Sat, 23 May 2015 14:01:16 +0300, Kimmo Paasiala wrote: On Sat, May 23, 2015 at 10:00 AM, Ian Smith smi...@nimnet.asn.au wrote: On Fri, 22 May 2015 20:26:40 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 8:19 PM, Ian Smith smi...@nimnet.asn.au wrote: On Fri, 22 May 2015 16:28:49 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 10:42 AM, Ivan Klymenko fi...@ukr.net wrote: [..] Try changing the options in /boot/device.hints hint.acpi_throttle.0.disabled=0 hint.p4tcc.0.disabled=0 Thanks, those also fixed powerd(8) for me that stopped working after upgrading to stable/10 from releng/10.1. Why are those setting suddenly needed now? [..] Can you say exactly in what way powerd stopped working then? Powerd(8) complained (excerpt from dmesg -a): Starting powerd. powerd: no cpufreq(4) support -- aborting: No such file or directory /etc/rc: WARNING: failed to start powerd Putting those two settings in loader.conf and rebooting fixed the problem and powerd started working again apparently because cpufreq(4) device was available again. Ok, if anabling acpi_throttle and/or p4tcc made cpufreq - and thus powerd - work for you, then it seems likely that you do not have EST enabled in your BIOS. Or at least, we've seen another instance where that was the case, which was fixed by enabling EST (or however your particular BIOS refers to it .. AMD for example use different terms). What CPU is this? In what machine? If EST (ono) IS enabled in your BIOS, this needs further investigation. As is, powerd may be running, but it's doing so highly inefficiently; refer to Stefan, Adrian and Kevin's responses for details. It's an Intel Atom running amd64 version of FreeBSD stable/10: FreeBSD firewall.rdnzl.info 10.1-STABLE FreeBSD 10.1-STABLE #1 r283292: Sat May 23 01:08:03 EEST 2015 r...@firewall.rdnzl.info:/usr/obj/usr/src/sys/GENERIC amd64 CPU: Intel(R) Atom(TM) CPU D510 @ 1.66GHz (1666.68-MHz K8-class CPU) Origin=GenuineIntel Id=0x106ca Family=0x6 Model=0x1c Stepping=10 Features=0xbfebfbffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE Features2=0x40e31dSSE3,DTES64,MON,DS_CPL,TM2,SSSE3,CX16,xTPR,PDCM,MOVBE AMD Features=0x20100800SYSCALL,NX,LM AMD Features2=0x1LAHF TSC: P-state invariant, performance statistics Powerd was working on 10.1-RELEASE but stopped working after upgrade to 10-STABLE and nothing was changed in BIOS settings. Which would be consistent with EST not being enabled in your BIOS; with no EST, cpufreq(4) still checks for 'relative' drivers such as p4tcc or acpi_throttle and uses that, as a last resort really; with those also disabled, no cpufreq, so no powerd. Have you checked BIOS settings to confirm that you do have SpeedStep (however termed) properly enabled? Please show `sysctl dev.cpu dev.est` and `sysctl -a | grep freq_levels` However, reading the other replies to this thread I get the impression that powerd(8) doesn't actually save energy on this platform and I'm better off without it? No, I don't think that's correct; using deeper C-states is most likely a bigger win, but higher than needed CPU freq will still use extra power, so run hotter. `sysctl dev.cpu` will also reveal your C-state usage. Reason I'm pursuing this is that this change shouldn't hurt, but it will flush out those cases where people were only getting cpufreq due to use of a 'relative' cpufreq driver like p4tcc, unless EST's enabled in BIOS; I suspect yours may be one such case :) If not, there's a bug to fix. cheers, Ian Looking deeper into this it appears I don't have speedstep (EST) support in the CPU it being a crappy Atom D510: http://ark.intel.com/products/43098 This the full 'sysctl dev.cpu' output: % sysctl dev.cpu dev.cpu.3.cx_usage: 100.00% last 65712us dev.cpu.3.cx_lowest: C1 dev.cpu.3.cx_supported: C1/1/0 dev.cpu.3.%parent: acpi0 dev.cpu.3.%pnpinfo: _HID=none _UID=0 dev.cpu.3.%location: handle=\_PR_.P004 dev.cpu.3.%driver: cpu dev.cpu.3.%desc: ACPI CPU dev.cpu.2.cx_usage: 100.00% last 41518us dev.cpu.2.cx_lowest: C1 dev.cpu.2.cx_supported: C1/1/0 dev.cpu.2.%parent: acpi0 dev.cpu.2.%pnpinfo: _HID=none _UID=0 dev.cpu.2.%location: handle=\_PR_.P003 dev.cpu.2.%driver: cpu dev.cpu.2.%desc: ACPI CPU dev.cpu.1.cx_usage: 100.00% last 12706us dev.cpu.1.cx_lowest: C1 dev.cpu.1.cx_supported: C1/1/0 dev.cpu.1.%parent: acpi0 dev.cpu.1.%pnpinfo: _HID=none
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Sat, May 23, 2015 at 6:57 PM, Ian Smith smi...@nimnet.asn.au wrote: On Sat, 23 May 2015 17:40:26 +0300, Kimmo Paasiala wrote: On Sat, May 23, 2015 at 5:15 PM, Ian Smith smi...@nimnet.asn.au wrote: [..] It's an Intel Atom running amd64 version of FreeBSD stable/10: FreeBSD firewall.rdnzl.info 10.1-STABLE FreeBSD 10.1-STABLE #1 r283292: Sat May 23 01:08:03 EEST 2015 r...@firewall.rdnzl.info:/usr/obj/usr/src/sys/GENERIC amd64 CPU: Intel(R) Atom(TM) CPU D510 @ 1.66GHz (1666.68-MHz K8-class CPU) Origin=GenuineIntel Id=0x106ca Family=0x6 Model=0x1c Stepping=10 Features=0xbfebfbffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE Features2=0x40e31dSSE3,DTES64,MON,DS_CPL,TM2,SSSE3,CX16,xTPR,PDCM,MOVBE AMD Features=0x20100800SYSCALL,NX,LM AMD Features2=0x1LAHF TSC: P-state invariant, performance statistics Powerd was working on 10.1-RELEASE but stopped working after upgrade to 10-STABLE and nothing was changed in BIOS settings. [..] However, reading the other replies to this thread I get the impression that powerd(8) doesn't actually save energy on this platform and I'm better off without it? No, I don't think that's correct; using deeper C-states is most likely a bigger win, but higher than needed CPU freq will still use extra power, so run hotter. `sysctl dev.cpu` will also reveal your C-state usage. Reason I'm pursuing this is that this change shouldn't hurt, but it will flush out those cases where people were only getting cpufreq due to use of a 'relative' cpufreq driver like p4tcc, unless EST's enabled in BIOS; I suspect yours may be one such case :) If not, there's a bug to fix. Seems _I've_ got a bug to fix; I need to stop assuming all modern Intel CPUs are going to make SpeedStep and/or deeper C-states available :( Looking deeper into this it appears I don't have speedstep (EST) support in the CPU it being a crappy Atom D510: http://ark.intel.com/products/43098 Indeed. It is rated at only 13W TDP, so relatively low power anyway. This the full 'sysctl dev.cpu' output: % sysctl dev.cpu dev.cpu.3.cx_usage: 100.00% last 65712us dev.cpu.3.cx_lowest: C1 dev.cpu.3.cx_supported: C1/1/0 [..] dev.cpu.0.cx_usage: 100.00% last 3132us dev.cpu.0.cx_lowest: C1 dev.cpu.0.cx_supported: C1/1/0 dev.cpu.0.%parent: acpi0 dev.cpu.0.%pnpinfo: _HID=none _UID=0 dev.cpu.0.%location: handle=\_PR_.P001 dev.cpu.0.%driver: cpu dev.cpu.0.%desc: ACPI CPU dev.cpu.%parent: It doesn't even provide dev.cpu.0.freq, and has no deeper C-states ('Idle States' on that page) available, so it looks like you may as well not bother running powerd. Others maybe can offer better suggestions. So I should keep those two hints in loader.conf to use p4tcc I guess? If this is a desktop I'd just let it run flat out, ie disable p4tcc and acpi_throttle, have no cpufreq and forget powerd. If it's a laptop and power consumption on battery matters to you, you could see if p4tcc's lower frequencies actually save any power much, by running 'powerd -v' in a terminal while testing with different loads, or if your 'acpiconf -i0' shows discharging rates in mA or mW, or both. Sorry again for my poor assumption, and thanks for the data point! cheers, Ian It's a firewall/router with some minimal services like nginx running. I'll just leave it like it's now without any frequency control. Thanks, -Kimmo ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Sat, May 23, 2015 at 5:15 PM, Ian Smith smi...@nimnet.asn.au wrote: On Sat, 23 May 2015 14:01:16 +0300, Kimmo Paasiala wrote: On Sat, May 23, 2015 at 10:00 AM, Ian Smith smi...@nimnet.asn.au wrote: On Fri, 22 May 2015 20:26:40 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 8:19 PM, Ian Smith smi...@nimnet.asn.au wrote: On Fri, 22 May 2015 16:28:49 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 10:42 AM, Ivan Klymenko fi...@ukr.net wrote: [..] Try changing the options in /boot/device.hints hint.acpi_throttle.0.disabled=0 hint.p4tcc.0.disabled=0 Thanks, those also fixed powerd(8) for me that stopped working after upgrading to stable/10 from releng/10.1. Why are those setting suddenly needed now? [..] Can you say exactly in what way powerd stopped working then? Powerd(8) complained (excerpt from dmesg -a): Starting powerd. powerd: no cpufreq(4) support -- aborting: No such file or directory /etc/rc: WARNING: failed to start powerd Putting those two settings in loader.conf and rebooting fixed the problem and powerd started working again apparently because cpufreq(4) device was available again. Ok, if anabling acpi_throttle and/or p4tcc made cpufreq - and thus powerd - work for you, then it seems likely that you do not have EST enabled in your BIOS. Or at least, we've seen another instance where that was the case, which was fixed by enabling EST (or however your particular BIOS refers to it .. AMD for example use different terms). What CPU is this? In what machine? If EST (ono) IS enabled in your BIOS, this needs further investigation. As is, powerd may be running, but it's doing so highly inefficiently; refer to Stefan, Adrian and Kevin's responses for details. It's an Intel Atom running amd64 version of FreeBSD stable/10: FreeBSD firewall.rdnzl.info 10.1-STABLE FreeBSD 10.1-STABLE #1 r283292: Sat May 23 01:08:03 EEST 2015 r...@firewall.rdnzl.info:/usr/obj/usr/src/sys/GENERIC amd64 CPU: Intel(R) Atom(TM) CPU D510 @ 1.66GHz (1666.68-MHz K8-class CPU) Origin=GenuineIntel Id=0x106ca Family=0x6 Model=0x1c Stepping=10 Features=0xbfebfbffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE Features2=0x40e31dSSE3,DTES64,MON,DS_CPL,TM2,SSSE3,CX16,xTPR,PDCM,MOVBE AMD Features=0x20100800SYSCALL,NX,LM AMD Features2=0x1LAHF TSC: P-state invariant, performance statistics Powerd was working on 10.1-RELEASE but stopped working after upgrade to 10-STABLE and nothing was changed in BIOS settings. Which would be consistent with EST not being enabled in your BIOS; with no EST, cpufreq(4) still checks for 'relative' drivers such as p4tcc or acpi_throttle and uses that, as a last resort really; with those also disabled, no cpufreq, so no powerd. Have you checked BIOS settings to confirm that you do have SpeedStep (however termed) properly enabled? Please show `sysctl dev.cpu dev.est` and `sysctl -a | grep freq_levels` However, reading the other replies to this thread I get the impression that powerd(8) doesn't actually save energy on this platform and I'm better off without it? No, I don't think that's correct; using deeper C-states is most likely a bigger win, but higher than needed CPU freq will still use extra power, so run hotter. `sysctl dev.cpu` will also reveal your C-state usage. Reason I'm pursuing this is that this change shouldn't hurt, but it will flush out those cases where people were only getting cpufreq due to use of a 'relative' cpufreq driver like p4tcc, unless EST's enabled in BIOS; I suspect yours may be one such case :) If not, there's a bug to fix. cheers, Ian Looking deeper into this it appears I don't have speedstep (EST) support in the CPU it being a crappy Atom D510: http://ark.intel.com/products/43098 This the full 'sysctl dev.cpu' output: % sysctl dev.cpu dev.cpu.3.cx_usage: 100.00% last 65712us dev.cpu.3.cx_lowest: C1 dev.cpu.3.cx_supported: C1/1/0 dev.cpu.3.%parent: acpi0 dev.cpu.3.%pnpinfo: _HID=none _UID=0 dev.cpu.3.%location: handle=\_PR_.P004 dev.cpu.3.%driver: cpu dev.cpu.3.%desc: ACPI CPU dev.cpu.2.cx_usage: 100.00% last 41518us dev.cpu.2.cx_lowest: C1 dev.cpu.2.cx_supported: C1/1/0 dev.cpu.2.%parent: acpi0 dev.cpu.2.%pnpinfo: _HID=none _UID=0 dev.cpu.2.%location: handle=\_PR_.P003 dev.cpu.2.%driver: cpu dev.cpu.2.%desc: ACPI CPU dev.cpu.1.cx_usage: 100.00% last 12706us dev.cpu.1.cx_lowest: C1 dev.cpu.1.cx_supported: C1/1/0 dev.cpu.1.%parent: acpi0 dev.cpu.1.%pnpinfo: _HID=none _UID=0 dev.cpu.1.%location: handle=\_PR_.P002 dev.cpu.1.%driver: cpu dev.cpu.1.%desc: ACPI CPU dev.cpu.0.cx_usage: 100.00% last 3132us dev.cpu.0.cx_lowest: C1 dev.cpu.0.cx_supported: C1/1/0 dev.cpu.0.%parent: acpi0
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
Fri, 22 May 2015 09:33:15 +0200 Nikos Vassiliadis nv...@gmx.com написав: Hi, I just noticed that my CPU's frequency doesn't support dropping below 1200MHz. It used to be able to go down to 150MHz, if I am not mistaken. I'd like it to go down to 600MHz via powerd, like it used to go. This is a month's old 10-STABLE. [nik@moby ~]$ sysctl dev.cpu.0.freq_levels dev.cpu.0.freq_levels: 2400/35000 2300/32872 2200/31127 2100/29417 2000/27740 1900/26096 1800/24490 1700/22588 1600/21045 1500/19534 1400/18055 1300/16611 1200/15194 This is the CPU: hw.model: Intel(R) Core(TM) i3-3110M CPU @ 2.40GHz Thanks in advance for any ideas, Nikos Try changing the options in /boot/device.hints hint.acpi_throttle.0.disabled=0 hint.p4tcc.0.disabled=0 ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
CPU frequency doesn't drop below 1200MHz (like it used to)
Hi, I just noticed that my CPU's frequency doesn't support dropping below 1200MHz. It used to be able to go down to 150MHz, if I am not mistaken. I'd like it to go down to 600MHz via powerd, like it used to go. This is a month's old 10-STABLE. [nik@moby ~]$ sysctl dev.cpu.0.freq_levels dev.cpu.0.freq_levels: 2400/35000 2300/32872 2200/31127 2100/29417 2000/27740 1900/26096 1800/24490 1700/22588 1600/21045 1500/19534 1400/18055 1300/16611 1200/15194 This is the CPU: hw.model: Intel(R) Core(TM) i3-3110M CPU @ 2.40GHz Thanks in advance for any ideas, Nikos ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Fri, May 22, 2015 at 11:47 AM, Adrian Chadd adr...@freebsd.org wrote: Hi, The whole point of throttling on modern hardware isn't to get really low clock rates, it's to deal with being out of thermal envelope. But, the modern intel cores will do that for you without OS involvement. So, you don't have to actually use p4tcc and it may actually configure your hardware wrong. Just throttle down to 1200MHz and go into deeper sleep states (C1). I checked this on a variety of older and modern hardware; they all worked better just doing lowest ACPI P state and lowest ACPI C state. -adrian It's actually worse than this. TCC (which was first available on the Pentium 4) was always documented by Intel as a Thermal Control Circuit. Nothing about power management. FreeBSD suborned it for power management a long time go, but it never actually saved power. EST , which adjusts frequency and voltage does save a bit. Cx states save a lot. Using TCC for this was a very bad idea. (I did research on this back when I worked for Berkeley Lab.) There are a couple of very limited corner cases where throttling MAY save an utterly insignificant amount of power, but when C-states came about, Intel never considered the impact of C-states when the OS was playing around with throttling. (Windows never did that.) If you want to save power, set both economy_cx_lowest (battery) and performance_cx_lowest (AC) to Cmax in rc.conf. (This is the default in head.) Do not set a minimum frequency for powerd or it may fail to start if you specify a frequncy that is no longer available. That capability was to prevent system lockups when TCC and Cx collided. With TCC off, there is no need to worry about it. Please don't just turn TCC back on. It really just makes things worse. Read mav's excellent article on the issues on the FreeBSD wiki at https://wiki.freebsd.org/TuningPowerConsumption. His research and mine came to virtually identical conclusions. -- Kevin Oberman, Network Engineer, Retired E-mail: rkober...@gmail.com ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On 05/22/15 09:42, Ivan Klymenko wrote: Try changing the options in /boot/device.hints hint.acpi_throttle.0.disabled=0 hint.p4tcc.0.disabled=0 Thanks Ivan, now it works as it did before! ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Fri, May 22, 2015 at 10:42 AM, Ivan Klymenko fi...@ukr.net wrote: Fri, 22 May 2015 09:33:15 +0200 Nikos Vassiliadis nv...@gmx.com написав: Hi, I just noticed that my CPU's frequency doesn't support dropping below 1200MHz. It used to be able to go down to 150MHz, if I am not mistaken. I'd like it to go down to 600MHz via powerd, like it used to go. This is a month's old 10-STABLE. [nik@moby ~]$ sysctl dev.cpu.0.freq_levels dev.cpu.0.freq_levels: 2400/35000 2300/32872 2200/31127 2100/29417 2000/27740 1900/26096 1800/24490 1700/22588 1600/21045 1500/19534 1400/18055 1300/16611 1200/15194 This is the CPU: hw.model: Intel(R) Core(TM) i3-3110M CPU @ 2.40GHz Thanks in advance for any ideas, Nikos Try changing the options in /boot/device.hints hint.acpi_throttle.0.disabled=0 hint.p4tcc.0.disabled=0 Thanks, those also fixed powerd(8) for me that stopped working after upgrading to stable/10 from releng/10.1. Why are those setting suddenly needed now? -Kimmo ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
Hi, The whole point of throttling on modern hardware isn't to get really low clock rates, it's to deal with being out of thermal envelope. But, the modern intel cores will do that for you without OS involvement. So, you don't have to actually use p4tcc and it may actually configure your hardware wrong. Just throttle down to 1200MHz and go into deeper sleep states (C1). I checked this on a variety of older and modern hardware; they all worked better just doing lowest ACPI P state and lowest ACPI C state. -adrian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Fri, 22 May 2015 16:28:49 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 10:42 AM, Ivan Klymenko fi...@ukr.net wrote: Fri, 22 May 2015 09:33:15 +0200 Nikos Vassiliadis nv...@gmx.com ÿÿ: Hi, I just noticed that my CPU's frequency doesn't support dropping below 1200MHz. It used to be able to go down to 150MHz, if I am not mistaken. I'd like it to go down to 600MHz via powerd, like it used to go. This is a month's old 10-STABLE. [nik@moby ~]$ sysctl dev.cpu.0.freq_levels dev.cpu.0.freq_levels: 2400/35000 2300/32872 2200/31127 2100/29417 2000/27740 1900/26096 1800/24490 1700/22588 1600/21045 1500/19534 1400/18055 1300/16611 1200/15194 This is the CPU: hw.model: Intel(R) Core(TM) i3-3110M CPU @ 2.40GHz Thanks in advance for any ideas, Nikos Try changing the options in /boot/device.hints hint.acpi_throttle.0.disabled=0 hint.p4tcc.0.disabled=0 Thanks, those also fixed powerd(8) for me that stopped working after upgrading to stable/10 from releng/10.1. Why are those setting suddenly needed now? -Kimmo Looks like the changes to these two hints, now defaulting to 1, committed to -head some months ago has been merged to stable/10. Can you say exactly in what way powerd stopped working then? Except that the minimum frequency that may be set with powerd's -m switch will be higher without p4tcc (or acpi_throttle) running, this change shouldn't hurt powerd; if anything it should be more efficient, as the lower p4tcc-generated frequencies don't save much if any power. If you compare dev.cpu.0.freq_levels, as above, both before and after booting with the changed hints, you can see the ones due to p4tcc's use of subfrequencies with factors of 1/8 to 7/8 of some base freq, but the power use in milliWatts provided for these seems largely ficticious. On my Lenovo X200, Core2Duo 2.4GHz, idling on battery at 800MHz (minimum EST freq) or at 100MHz using p4tcc draws almost exactly the same power, about 7.6W measured from the battery - but responsiveness as performance is required is a great deal better using just the base EST freqs; YMMV. This generally gets discussed on the freebsd-mobile and freebsd-acpi lists; not sure if a deeper discussion of issues is warranted here. cheers, Ian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
Am 22.05.2015 um 09:33 schrieb Nikos Vassiliadis: Hi, I just noticed that my CPU's frequency doesn't support dropping below 1200MHz. It used to be able to go down to 150MHz, if I am not mistaken. I'd like it to go down to 600MHz via powerd, like it used to go. This is a month's old 10-STABLE. [nik@moby ~]$ sysctl dev.cpu.0.freq_levels dev.cpu.0.freq_levels: 2400/35000 2300/32872 2200/31127 2100/29417 2000/27740 1900/26096 1800/24490 1700/22588 1600/21045 1500/19534 1400/18055 1300/16611 1200/15194 This is the CPU: hw.model: Intel(R) Core(TM) i3-3110M CPU @ 2.40GHz Well, your CPU does not support clock frequencies below 1200 MHz. Throttling works by injection of wait cycles that reduce the amount of work the CPU can perform per unit of time, but does not really lower the CPU frequency. That means, that with throttling the CPU will need more energy to perform some calculation than it would without. If you select 150 MHz, then your CPU will be clocked at 1200 MHz, but will only perform any operations on each 8th clock cycle. This limits peak energy consumption (and that was the reason this feature was introduced in the power-hungry Pentium-4 processors), but increases the amount of energy needed to perform the computation. The power consumption of your CPU may be (an estimated) 50% to 70% at 150 Mhz compared to 1200 Mhz. But you'll need 8 times as long until the CPU can fall into a deep sleep state. Since RAM and other components see the same clock whether throttling is enabled or not, you'll need 8 times as long full power for your RAM (which will also go into a low power refresh mode, when the CPU is idle). Throttling has been disabled, because there are no longer any CPUs which need it to prevent overheating. (Or rather: there are now better mechanisms than throttling, which are implemented in any modern x68 CPU.) Throttling could also impact system stability. It really serves no purpose anymore and it was never suitable to improve the power efficiency of e.g. a laptop computer. You'll see better battery live if you keep throttling disabled. Regards, STefan ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: CPU frequency doesn't drop below 1200MHz (like it used to)
On Fri, May 22, 2015 at 8:19 PM, Ian Smith smi...@nimnet.asn.au wrote: On Fri, 22 May 2015 16:28:49 +0300, Kimmo Paasiala wrote: On Fri, May 22, 2015 at 10:42 AM, Ivan Klymenko fi...@ukr.net wrote: Fri, 22 May 2015 09:33:15 +0200 Nikos Vassiliadis nv...@gmx.com яя: Hi, I just noticed that my CPU's frequency doesn't support dropping below 1200MHz. It used to be able to go down to 150MHz, if I am not mistaken. I'd like it to go down to 600MHz via powerd, like it used to go. This is a month's old 10-STABLE. [nik@moby ~]$ sysctl dev.cpu.0.freq_levels dev.cpu.0.freq_levels: 2400/35000 2300/32872 2200/31127 2100/29417 2000/27740 1900/26096 1800/24490 1700/22588 1600/21045 1500/19534 1400/18055 1300/16611 1200/15194 This is the CPU: hw.model: Intel(R) Core(TM) i3-3110M CPU @ 2.40GHz Thanks in advance for any ideas, Nikos Try changing the options in /boot/device.hints hint.acpi_throttle.0.disabled=0 hint.p4tcc.0.disabled=0 Thanks, those also fixed powerd(8) for me that stopped working after upgrading to stable/10 from releng/10.1. Why are those setting suddenly needed now? -Kimmo Looks like the changes to these two hints, now defaulting to 1, committed to -head some months ago has been merged to stable/10. Can you say exactly in what way powerd stopped working then? Powerd(8) complained (excerpt from dmesg -a): Starting powerd. powerd: no cpufreq(4) support -- aborting: No such file or directory /etc/rc: WARNING: failed to start powerd Putting those two settings in loader.conf and rebooting fixed the problem and powerd started working again apparently because cpufreq(4) device was available again. -Kimmo Except that the minimum frequency that may be set with powerd's -m switch will be higher without p4tcc (or acpi_throttle) running, this change shouldn't hurt powerd; if anything it should be more efficient, as the lower p4tcc-generated frequencies don't save much if any power. If you compare dev.cpu.0.freq_levels, as above, both before and after booting with the changed hints, you can see the ones due to p4tcc's use of subfrequencies with factors of 1/8 to 7/8 of some base freq, but the power use in milliWatts provided for these seems largely ficticious. On my Lenovo X200, Core2Duo 2.4GHz, idling on battery at 800MHz (minimum EST freq) or at 100MHz using p4tcc draws almost exactly the same power, about 7.6W measured from the battery - but responsiveness as performance is required is a great deal better using just the base EST freqs; YMMV. This generally gets discussed on the freebsd-mobile and freebsd-acpi lists; not sure if a deeper discussion of issues is warranted here. cheers, Ian ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org