ich777 Posted July 26, 2022 Share Posted July 26, 2022 3 minutes ago, Kulisch said: Großartig. Müssen nach einspielen des Updates Anpassungen am Bios gemacht werden? (C-States z.B.) Das kann ich dir leider nicht sagen, hab leider kein AMD System, hier können aber sicher @Pillendreher & @kdwg & @mgutt weiter helfen. Habe nur dafür gesorgt dass das modul in dem release enthalten ist. Hab mich auch nur oberflächlich eingelesen weil ich wie gesagt kein AMD System hab. Du musst den Treiber aber aktivieren, im BIOS sollten keine Einstellungen nötig sein soweit ich gelesen hab. 1 Quote Link to comment
mgutt Posted July 26, 2022 Author Share Posted July 26, 2022 34 minutes ago, Kulisch said: Müssen nach einspielen des Updates Anpassungen am Bios gemacht werden? Du solltest diesen Thread mal lesen, weil das selbst in Windows nicht trivial ist: https://www.computerbase.de/forum/threads/renoir-und-b550-die-idle-kuenstler.1967755/ Hier auch mein Beitrag: https://www.computerbase.de/forum/threads/renoir-und-b550-die-idle-kuenstler.1967755/post-24812476 Es sollte in jedem Fall ASPM und ALPM forciert werden. Je nach Board hat die Deaktivierung von Komponenten wie zb LED Controller keinen oder einen großen Einfluss auf den Verbrauch. Da hilft nur durchprobieren und jede Änderung messen. 1 Quote Link to comment
kdwg Posted July 26, 2022 Share Posted July 26, 2022 14 hours ago, Kulisch said: Großartig. Müssen nach einspielen des Updates Anpassungen am Bios gemacht werden? (C-States z.B.) du solltest die Optionen für CPPC im BIOS aktivieren. Konnte ich bisher aber noch nicht verifizieren. Danke für den Einsatz @ich777 ! Ich hab mein System gerade erst wieder stabil bekommen, will es eigentlich ungern auf ein Prerelease loslassen.... aber neugierig was man messen kann bin ich schon ... vielleicht nächste Woche. 1 Quote Link to comment
ich777 Posted July 27, 2022 Share Posted July 27, 2022 5 hours ago, kdwg said: Ich hab mein System gerade erst wieder stabil bekommen, will es eigentlich ungern auf ein Prerelease loslassen.... aber neugierig was man messen kann bin ich schon ... vielleicht nächste Woche. Glaub mir prereleases auf Unraid sind nicht wie sonst, die werden exzessiv getestet. Quote Link to comment
Pillendreher Posted July 30, 2022 Share Posted July 30, 2022 Bin jetzt auf 6.11.0 rc2. root@Tower:~# cat /sys/devices/system/cpu//cpu0/cpufreq/scaling_available_governors conservative ondemand userspace powersave performance schedutil root@Tower:~# grep amd-pstate /proc/kallsyms root@Tower:~# grep PSTATE /usr/src/linux-*/.config CONFIG_X86_INTEL_PSTATE=y CONFIG_X86_AMD_PSTATE=m Woran müsste ich die Verbesserungen nun erkennen? Laut corefreq bleibt weiterhin alles in C1, auch wenn die Frequenz der einzelnen Kerne im Leerlauf natürlich sehr niedrig ist: Quote Link to comment
mgutt Posted July 30, 2022 Author Share Posted July 30, 2022 Erstmal müsstest du ermitteln ob der neue Treiber überhaupt genutzt wird. Führe mal das aus: cpufreq-info Quote Link to comment
Pillendreher Posted July 30, 2022 Share Posted July 30, 2022 Das wird bei mir ausgegeben: Linux 5.18.14-Unraid. root@Tower:~# cpufreq-info cpufrequtils 008: cpufreq-info (C) Dominik Brodowski 2004-2009 Report errors and bugs to [email protected], please. analyzing CPU 0: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 0 CPUs which need to have their frequency coordinated by software: 0 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.40 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:1.77%, 1.70 GHz:8.97%, 1.40 GHz:89.26% (49898) analyzing CPU 1: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 1 CPUs which need to have their frequency coordinated by software: 1 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.40 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:2.35%, 1.70 GHz:9.57%, 1.40 GHz:88.09% (46195) analyzing CPU 2: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 2 CPUs which need to have their frequency coordinated by software: 2 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.40 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:2.58%, 1.70 GHz:8.94%, 1.40 GHz:88.49% (43083) analyzing CPU 3: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 3 CPUs which need to have their frequency coordinated by software: 3 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.40 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:2.56%, 1.70 GHz:8.94%, 1.40 GHz:88.50% (41017) analyzing CPU 4: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 4 CPUs which need to have their frequency coordinated by software: 4 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.40 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:2.55%, 1.70 GHz:9.09%, 1.40 GHz:88.35% (41321) analyzing CPU 5: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 5 CPUs which need to have their frequency coordinated by software: 5 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.40 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:2.41%, 1.70 GHz:9.16%, 1.40 GHz:88.43% (41958) analyzing CPU 6: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 6 CPUs which need to have their frequency coordinated by software: 6 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.40 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:2.26%, 1.70 GHz:8.28%, 1.40 GHz:89.46% (38701) analyzing CPU 7: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 7 CPUs which need to have their frequency coordinated by software: 7 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.40 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:2.18%, 1.70 GHz:8.10%, 1.40 GHz:89.71% (38502) analyzing CPU 8: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 8 CPUs which need to have their frequency coordinated by software: 8 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.40 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:2.44%, 1.70 GHz:8.03%, 1.40 GHz:89.53% (36377) analyzing CPU 9: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 9 CPUs which need to have their frequency coordinated by software: 9 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.70 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:1.93%, 1.70 GHz:8.34%, 1.40 GHz:89.73% (38459) analyzing CPU 10: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 10 CPUs which need to have their frequency coordinated by software: 10 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.70 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:2.49%, 1.70 GHz:12.26%, 1.40 GHz:85.25% (42878) analyzing CPU 11: driver: acpi-cpufreq CPUs which run at the same hardware frequency: 11 CPUs which need to have their frequency coordinated by software: 11 maximum transition latency: 4294.55 ms. hardware limits: 1.40 GHz - 4.46 GHz available frequency steps: 3.90 GHz, 1.70 GHz, 1.40 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, per formance, schedutil current policy: frequency should be within 1.40 GHz and 3.90 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.70 GHz (asserted by call to hardware). cpufreq stats: 3.90 GHz:2.78%, 1.70 GHz:10.33%, 1.40 GHz:86.89% (42064) root@Tower:~# Quote Link to comment
mgutt Posted July 30, 2022 Author Share Posted July 30, 2022 Ist nicht aktiv. Bitte das ausführen: lscpu | grep "cppc" Wenn vorhanden, dann folgende Kernel Option setzen und Server neu starten: amd_pstate.replace=1 Von hier: https://wiki.archlinux.org/title/CPU_frequency_scaling#Scaling_drivers Quote Link to comment
Pillendreher Posted July 31, 2022 Share Posted July 31, 2022 (edited) root@Tower:~# lscpu | grep "cppc" Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd cppc arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif v_spec_ctrl umip pku ospke vaes vpclmulqdq rdpid overflow_recov succor smca fsrm Ich sehe jetzt hier nichts, was auf amd_pstate hindeuten würde? EDIT: Hab mal testweise mein go-file angepasst (hoffe, dort gehört der Befehl hin): Unraid OS kernel /bzimage append initrd=/bzroot amd_pstate.replace=1 Aber der pstate-Treiber ist wohl trotzdem nicht aktiviert: root@Tower:~# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver acpi-cpufreq EDIT2: Das Blockieren des ACPI-Treibers nach dieser Anleitung funktioniert zwar, aber es wird dann nicht der neue, sondern gar kein Treiber mehr geladen: root@Tower:~# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver cat: /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver: No such file or directory root@Tower:~# cpufreq-info cpufrequtils 008: cpufreq-info (C) Dominik Brodowski 2004-2009 Report errors and bugs to [email protected], please. analyzing CPU 0: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 1: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 2: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 3: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 4: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 5: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 6: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 7: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 8: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 9: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 10: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. analyzing CPU 11: no or unknown cpufreq driver is active on this CPU maximum transition latency: 4294.55 ms. Edited July 31, 2022 by Pillendreher Quote Link to comment
jj1987 Posted July 31, 2022 Share Posted July 31, 2022 24 minutes ago, Pillendreher said: Hab mal testweise mein go-file angepasst (hoffe, dort gehört der Befehl hin): Kernel Befehle müssen in die syslinux.cfg. dafür unter "Main" auf "Flash" klicken, siehe auch hier (anderer Kontext aber als Beispiel wohl gut genug): https://forums.unraid.net/topic/108966-strom-sparen-mit-powertop-stromverbrauch-von-unraid-verbessern/?do=findComment&comment=1151917 Quote Link to comment
Pillendreher Posted July 31, 2022 Share Posted July 31, 2022 Just now, jj1987 said: Kernel Befehle müssen in die syslinux.cfg. dafür unter "Main" auf "Flash" klicken, siehe auch hier (anderer Kontext aber als Beispiel wohl gut genug): https://forums.unraid.net/topic/108966-strom-sparen-mit-powertop-stromverbrauch-von-unraid-verbessern/?do=findComment&comment=1151917 Dann hab ich das richtig gemacht, aber gedacht ich wäre im go-file unterwegs. Keine Ahnung wie ich darauf kam, wahrscheinlich Schlafmangel Quote Link to comment
Pillendreher Posted July 31, 2022 Share Posted July 31, 2022 So, jetzt scheint zumindest der Treiber zu laufen: root@Tower:~# modprobe amd-pstate root@Tower:~# cpufreq-info cpufrequtils 008: cpufreq-info (C) Dominik Brodowski 2004-2009 Report errors and bugs to [email protected], please. analyzing CPU 0: driver: amd-pstate CPUs which run at the same hardware frequency: 0 CPUs which need to have their frequency coordinated by software: 0 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.27 GHz. analyzing CPU 1: driver: amd-pstate CPUs which run at the same hardware frequency: 1 CPUs which need to have their frequency coordinated by software: 1 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.84 GHz. analyzing CPU 2: driver: amd-pstate CPUs which run at the same hardware frequency: 2 CPUs which need to have their frequency coordinated by software: 2 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.27 GHz. analyzing CPU 3: driver: amd-pstate CPUs which run at the same hardware frequency: 3 CPUs which need to have their frequency coordinated by software: 3 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.28 GHz. analyzing CPU 4: driver: amd-pstate CPUs which run at the same hardware frequency: 4 CPUs which need to have their frequency coordinated by software: 4 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.28 GHz. analyzing CPU 5: driver: amd-pstate CPUs which run at the same hardware frequency: 5 CPUs which need to have their frequency coordinated by software: 5 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.34 GHz. analyzing CPU 6: driver: amd-pstate CPUs which run at the same hardware frequency: 6 CPUs which need to have their frequency coordinated by software: 6 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.34 GHz. analyzing CPU 7: driver: amd-pstate CPUs which run at the same hardware frequency: 7 CPUs which need to have their frequency coordinated by software: 7 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.30 GHz. analyzing CPU 8: driver: amd-pstate CPUs which run at the same hardware frequency: 8 CPUs which need to have their frequency coordinated by software: 8 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.43 GHz. analyzing CPU 9: driver: amd-pstate CPUs which run at the same hardware frequency: 9 CPUs which need to have their frequency coordinated by software: 9 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 3.05 GHz. analyzing CPU 10: driver: amd-pstate CPUs which run at the same hardware frequency: 10 CPUs which need to have their frequency coordinated by software: 10 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.43 GHz. analyzing CPU 11: driver: amd-pstate CPUs which run at the same hardware frequency: 11 CPUs which need to have their frequency coordinated by software: 11 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "schedutil" may decide which speed to use within this range. current CPU frequency is 2.28 GHz. root@Tower:~# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver amd-pstate Und so sieht corefreq-cli aus: Quote Link to comment
mgutt Posted July 31, 2022 Author Share Posted July 31, 2022 45 minutes ago, Pillendreher said: So, jetzt scheint zumindest der Treiber zu laufen: Gut. Nächster Schritt wäre es zu prüfen ob ASPM funktioniert: lspci -vv | awk '/ASPM/{print $0}' RS= | grep --color -P '(^[a-z0-9:.]+|ASPM )' Ziel wäre, dass möglichst viele PCI Geräte ausgegeben werden und möglichst alle auf ASPM Enabled stehen. Beispiel: https://forums.unraid.net/topic/111921-why-is-aspm-disabled-in-unraid-while-its-enabled-in-ubuntu/?do=findComment&comment=1053964 Quote Link to comment
mgutt Posted July 31, 2022 Author Share Posted July 31, 2022 1 hour ago, Pillendreher said: Ich sehe jetzt hier nichts, was auf amd_pstate hindeuten würde? Nur, wenn die CPU cppc beherrscht, kann sie vom amd_pstate Treiber angesprochen werden (und deine CPU kann das): https://www.kernel.org/doc/html/latest/admin-guide/pm/amd-pstate.html Quote Link to comment
Pillendreher Posted July 31, 2022 Share Posted July 31, 2022 3 hours ago, mgutt said: Gut. Nächster Schritt wäre es zu prüfen ob ASPM funktioniert: lspci -vv | awk '/ASPM/{print $0}' RS= | grep --color -P '(^[a-z0-9:.]+|ASPM )' Ziel wäre, dass möglichst viele PCI Geräte ausgegeben werden und möglichst alle auf ASPM Enabled stehen. Beispiel: https://forums.unraid.net/topic/111921-why-is-aspm-disabled-in-unraid-while-its-enabled-in-ubuntu/?do=findComment&comment=1053964 Bei 6 von 11 Geräten wird "ASPM L1 Enabled" angezeigt. ASPM mit dem von Dir verlinkten Befehl zu forcieren ändert daran nichts. 3 hours ago, mgutt said: Nur, wenn die CPU cppc beherrscht, kann sie vom amd_pstate Treiber angesprochen werden (und deine CPU kann das): https://www.kernel.org/doc/html/latest/admin-guide/pm/amd-pstate.html Ach, um CPPC gings! Ja, das habe ich im UEFI schon aktiviert! Quote Link to comment
mgutt Posted July 31, 2022 Author Share Posted July 31, 2022 OK auch super. Hast du powertop --auto-tune ausgeführt, damit ungenutze SATA und USB Ports schlagen gehen können? Quote Link to comment
Pillendreher Posted July 31, 2022 Share Posted July 31, 2022 Ja, hab das alles in nem UserScript drinnen, welches bei jedem Start ausgeführt wird. Auch deine 2.14 Version zeigt mir bei den "tunables" keine offenen Punkte mehr an. Komme im Moment so auf 20 W im Leerlauf mit schlafenden Festplatten. Laute corefreq-cli aber immer noch nur C1. Interessanterweise habe ich den modprobe Befehl jetzt auch in Userscripts und habe wohl jetzt nen anderen CPU governor?! root@Tower:~# cpufreq-info cpufrequtils 008: cpufreq-info (C) Dominik Brodowski 2004-2009 Report errors and bugs to [email protected], please. analyzing CPU 0: driver: amd-pstate CPUs which run at the same hardware frequency: 0 CPUs which need to have their frequency coordinated by software: 0 maximum transition latency: 131 us. hardware limits: 400 MHz - 4.46 GHz available cpufreq governors: conservative, ondemand, userspace, powersave, performance, schedutil current policy: frequency should be within 400 MHz and 4.46 GHz. The governor "ondemand" may decide which speed to use within this range. current CPU frequency is 1.60 GHz. Quote Link to comment
mgutt Posted July 31, 2022 Author Share Posted July 31, 2022 Stell mal auf powersave: /etc/rc.d/rc.cpufreq powersave Ansonsten wäre die Frage ob es was bringt, wenn man LAN zieht oder sogar der Unraid USB Stick die C-States verhindern könnte. Das habe ich ja bei einem Intel Board gehabt. Welches Board hast du? Quote Link to comment
Pillendreher Posted August 1, 2022 Share Posted August 1, 2022 Hab auf Powersave umgestellt. Hier corefreq und powertop zur selben Zeit: Als Mainboard kommt ein ASRock B550M Pro4 zum Einsatz. Quote Link to comment
jj1987 Posted August 1, 2022 Share Posted August 1, 2022 4 minutes ago, Pillendreher said: Als Mainboard kommt ein ASRock B550M Pro4 zum Einsatz. Interessant, bei dir ist beim Realtek NIC L1 aktiviert - bei meinem B550M Pro4 wollte der das nicht. Leicht OT: Auf welchem Bios bist du gerade? Muss ich vielleicht mal upgraden und nochmal neu austesten. Was mir noch auffällt, sind die beiden USB Ports die ASPM disabled haben. Ich weiß nicht was du so alles an USB Gedöns womöglich angeschlossen hast, da es aber 3.1 Ports sind sollte da ja eigentlich NICHT der Unraid Stick dran sein. Du könntest also entweder die Ports im Bios komplett abschalten oder mal testweise auf USB2 drosseln - einfach um mal zu schauen ob das nicht womöglich irgendwelche Auswirkungen hat. Quote Link to comment
Pillendreher Posted August 1, 2022 Share Posted August 1, 2022 (edited) 2.30, also das aktuelle. Ich schaue gleich mal, wo der Stick steckt. Ich hab jedenfalls noch nen Empfänger für die Funktastatur dran, um im BIOS rumspielen zu können, aber der muss in der Tat nicht dranhängen, den entferne ich gleich mal. EDIT: Stick steckt in einem USB 2.0 Anschluss (zumindest gehe ich davon aus, denn der verwendete Steckplatz ist nicht "blau"); hab den Funkempfänger entfernt, jedoch hat sich am ASPM Status nichts geändert. Auch der Leerlaufverbrauch liegt bei abgezogenem HDMI Kabel und entferntem Funkempfänger weiterhin bei ca. 20 W. Edited August 1, 2022 by Pillendreher Quote Link to comment
kdwg Posted August 3, 2022 Share Posted August 3, 2022 Ich hab den Sprung auf 6.11rc2 ebenfalls gewagt Musste ebenfalls acpi_cpufreq deaktivieren. So sind also die Settigns: amd_pstate.replace=1 amd_pstate.shared_mem=1 initcall_blacklist=acpi_cpufreq_init Außerdem "modprobe amd_pstate" im go-file Ich beobachte das ganze mal und vergleiche die Ergebnisse Quote Link to comment
Pillendreher Posted August 4, 2022 Share Posted August 4, 2022 (edited) Also so ganz rund scheint das Ganze noch nicht zu laufen. Ich lasse gerade im Hintergrund den MKVToolnix-Docker eine fast 40 GB große MKV-Datei aufteilen und der Prozessor dümpelt so vor sich hin (selbst "ondemand" ändert nix): Auf "performance" ein wenig besser, aber selbst da schießt er nicht Richtung Turbo hoch... Das ist wahrscheinlich so nicht richtig, oder? EDIT: Vielleicht lastet auch der Docker selber nicht stark genug aus? MKVToolnix + Bilderimport per Lightroom in einer W10 VM: Aber so richtig voll über alle Kerne wird da immer noch nichts ausgelastet: Mal Prime95 ausprobieren 😄 Edited August 4, 2022 by Pillendreher Quote Link to comment
kdwg Posted August 7, 2022 Share Posted August 7, 2022 On 8/4/2022 at 8:21 PM, Pillendreher said: Also so ganz rund scheint das Ganze noch nicht zu laufen. Ich lasse gerade im Hintergrund den MKVToolnix-Docker eine fast 40 GB große MKV-Datei aufteilen und der Prozessor dümpelt so vor sich hin (selbst "ondemand" ändert nix): Auf "performance" ein wenig besser, aber selbst da schießt er nicht Richtung Turbo hoch... Das ist wahrscheinlich so nicht richtig, oder? EDIT: Vielleicht lastet auch der Docker selber nicht stark genug aus? MKVToolnix + Bilderimport per Lightroom in einer W10 VM: Aber so richtig voll über alle Kerne wird da immer noch nichts ausgelastet: Mal Prime95 ausprobieren 😄 ich hatte das Gefühl der Turbo kommt nicht richtig. Wie sieht es bei dir aus ? Quote Link to comment
Pillendreher Posted August 7, 2022 Share Posted August 7, 2022 Er dreht z.T. schon voll auf: Ich habe eher den Eindruck, dass z.T. die Last nicht wirklich "ankommt", muss aber zugeben, dass ich das bisher nie beobachtet habe. Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.