Date: Mon, 1 Nov 2010 20:52:13 +0200 From: Joerg Traeger <jt@xoasis.de> To: Andriy Gapon <avg@icyb.net.ua> Cc: freebsd-acpi@freebsd.org Subject: Re: est CPU support Message-ID: <201011011952.14024.jt@xoasis.de> In-Reply-To: <4CCF0A2A.9060506@icyb.net.ua> References: <201011011052.11084.jt@xoasis.de> <201011011936.55934.jt@xoasis.de> <4CCF0A2A.9060506@icyb.net.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
On Monday 01 November 2010, Andriy Gapon wrote: > on 01/11/2010 20:36 Joerg Traeger said the following: > > On Monday 01 November 2010, Andriy Gapon wrote: > >> It seems that your BIOS makes it a condition that OS supports the > >> following feature: ACPI_CAP_C1_IO_HALT. > >> > >> FreeBSD doesn't really support it, but you can try adding it to > >> 'features' variable in acpi_cpu_attach() in function in > >> sys/dev/acpica/acpi_cpu.c; look for the following line: > >> sc->cpu_features = ACPI_CAP_SMP_SAME | ACPI_CAP_SMP_SAME_C3; > >> > >> I don't think that should break anything for you, but may improve a > >> thing or two. I'd interested in seeing acpidump -d -t produced after the > >> patching. > > > > Hey, est seems to be happy now! > > > > coretemp0: <CPU On-Die Thermal Sensors> on cpu0 > > est0: <Enhanced SpeedStep Frequency Control> on cpu0 > > p4tcc0: <CPU Frequency Thermal Control> on cpu0 > > coretemp1: <CPU On-Die Thermal Sensors> on cpu1 > > est1: <Enhanced SpeedStep Frequency Control> on cpu1 > > p4tcc1: <CPU Frequency Thermal Control> on cpu1 > > > > Even C2 and C3 are anounced. > > > > dev.cpu.0.cx_supported: C1/20 C2/40 C3/60 > > dev.cpu.0.cx_lowest: C3 > > dev.cpu.0.cx_usage: 0.09% 2.48% 97.41% last 207us > > > > But the system behaves strange. The fan comes up 10 times a minute and > > for example "sh /etc/rc autoboot" runs 5 minutes now. Load is too high > > without any processes running. And rebooting takes a long time syncing > > buffers. Are these side effects known? > > Try to not use C3. > > > acpidump output did not change. > > Are you 100% sure? Yes. # diff acpidump.txt acpidump_acpi_patched.txt 109c109 < * Disassembly of /tmp/acpidump.pdLAtj, Mon Nov 1 13:38:24 2010 --- > * Disassembly of /tmp/acpidump.DvGKfH, Mon Nov 1 19:12:53 2010 > If yes, then could you please do the following? > > $ dd if=/dev/mem of=/tmp/ssdt.dump bs=1 skip=0xCBE61C18 count=0x02CC This works. > $ acpidump -d -f /tmp/ssdt.dump > /tmp/ssdt.asl But: # acpidump -d -f /tmp/ssdt.dump > /tmp/ssdt.asl Segmentation fault: 11 > Send me /tmp/ssdt.asl :) The binary file look like this: less /tmp/ssdt.dump "/tmp/ssdt.dump" may be a binary file. See it anyway? SSDT<CC>^B^@^@^A)AMI^@^@^@IST^@^@^@^@^@^A^@^@^@MSFT^A^@^@^C^PD^E\._PR_CPU1PCT^R,^B^Q^T ^Q<82>^L^@^?@^@^@<99>^A^@^@^@^@^@^@y^@^Q^T ^Q<82>^L^@^?^P^@^@<98>^A^@^@^@^@^@^@y^@^T^K_PSS^@<A4>APSS^T _PPC^@<A4> ^@^PD^E\._PR_CPU2PCT^R,^B^Q^T ^Q<82>^L^@^?@^@^@<99>^A^@^@^@^@^@^@y^@^Q^T ^Q<82>^L^@^?^P^@^@<98>^A^@^@^@^@^@^@y^@^T^K_PSS^@<A4>APSS^T _PPC^@<A4> ^@^PD^E\._PR_CPU3PCT^R,^B^Q^T ^Q<82>^L^@^?@^@^@<99>^A^@^@^@^@^@^@y^@^Q^T ^Q<82>^L^@^?^P^@^@<98>^A^@^@^@^@^@^@y^@^T^K_PSS^@<A4>APSS^T _PPC^@<A4> ^@^PD^E\._PR_CPU4PCT^R,^B^Q^T ^Q<82>^L^@^?@^@^@<99>^A^@^@^@^@^@^@y^@^Q^T ^Q<82>^L^@^?^P^@^@<98>^A^@^@^@^@^@^@y^@^T^K_PSS^@<A4>APSS^T _PPC^@<A4> ^@^PD^E\._PR_CPU5PCT^R,^B^Q^T ^Q<82>^L^@^?@^@^@<99>^A^@^@^@^@^@^@y^@^Q^T ^Q<82>^L^@^?^P^@^@<98>^A^@^@^@^@^@^@y^@^T^K_PSS^@<A4>APSS^T _PPC^@<A4> ^@^PD^E\._PR_CPU6PCT^R,^B^Q^T ^Q<82>^L^@^?@^@^@<99>^A^@^@^@^@^@^@y^@^Q^T ^Q<82>^L^@^?^P^@^@<98>^A^@^@^@^@^@^@y^@^T^K_PSS^@<A4>APSS^T _PPC^@<A4> ^@^PD^E\._PR_CPU7PCT^R,^B^Q^T ^Q<82>^L^@^?@^@^@<99>^A^@^@^@^@^@^@y^@^Q^T ^Q<82>^L^@^?^P^@^@<98>^A^@^@^@^@^@^@y^@^T^K_PSS^@<A4>APSS^T _PPC^@<A4> ^@^PD^E\._PR_CPU8PCT^R,^B^Q^T ^Q<82>^L^@^?@^@^@<99>^A^@^@^@^@^@^@y^@^Q^T ^Q<82>^L^@^?^P^@^@<98>^A^@^@^@^@^@^@y^@^T^K_PSS^@<A4>APSS^T _PPC^@<A4> ^@ /tmp/ssdt.dump (END) -- If you say that you can't, then I shall reply, Parsley, sage, rosemary and thyme
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201011011952.14024.jt>