Re: Core Duo - only one cpu being used
Hello, On 05/05/06 06:44, Scott Long wrote: CPU1 is being treated as a hyperthreading core instead of a real core, and is being disabled per our policy on Intel hyperthreading. By 'disabled' I mean that it is started, but it is being excluded from scheduling decisions, and thus is only running its idle proc. It's also handling any interrupts that come to it, such as timer and IPI interrupts, so it's at 99% instead of 100% for the idle proc. There is nothing broken about the number you are seeing, your system is just running under a scheduling policy that it should not be. This should have been fixed a week or so ago by a commit to HEAD, RELENG_6, and RELENG_6_1 by Colin Percival. How old is kernel? Isn't this fixed by jkim's commit? I don't have any Core Duos here, but on a Xeon LV (Sossaman) with CURRENT, everything is OK: http://www.freebsd.org/cgi/query-pr.cgi?pr=96203 BTW, just set machdep.hyperthreading_allowed=1 until the fix arrives. -- Attila Nagy e-mail: [EMAIL PROTECTED] Free Software Network (FSN.HU) phone: +3630 306 6758 http://www.fsn.hu/ ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Core Duo - only one cpu being used
Attila Nagy wrote: Hello, On 05/05/06 06:44, Scott Long wrote: CPU1 is being treated as a hyperthreading core instead of a real core, and is being disabled per our policy on Intel hyperthreading. By 'disabled' I mean that it is started, but it is being excluded from scheduling decisions, and thus is only running its idle proc. It's also handling any interrupts that come to it, such as timer and IPI interrupts, so it's at 99% instead of 100% for the idle proc. There is nothing broken about the number you are seeing, your system is just running under a scheduling policy that it should not be. This should have been fixed a week or so ago by a commit to HEAD, RELENG_6, and RELENG_6_1 by Colin Percival. How old is kernel? Isn't this fixed by jkim's commit? I don't have any Core Duos here, but on a Xeon LV (Sossaman) with CURRENT, everything is OK: http://www.freebsd.org/cgi/query-pr.cgi?pr=96203 BTW, just set machdep.hyperthreading_allowed=1 until the fix arrives. I can confirm now that it is fixed in todays 6-STABLE. Thanks everyone! Eric -- Eric AndersonSr. Systems AdministratorCentaur Technology Anything that works is better than anything that doesn't. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Core Duo - only one cpu being used
Forgive me if I've missed this on a list somewhere, but My new laptop with a Core Duo doesn't seem to use both CPU's. It sees both, but I never see anything on cpu 1. Here's a top snippet: last pid: 20852; load averages: 1.31, 1.27, 1.00 up 0+01:20:55 21:05:47 100 processes: 2 running, 98 sleeping CPU states: 47.4% user, 0.0% nice, 1.3% system, 1.3% interrupt, 50.0% idle Mem: 383M Active, 391M Inact, 154M Wired, 34M Cache, 110M Buf, 25M Free Swap: 2048M Total, 2048M Free PID USERNAMETHR PRI NICE SIZERES STATE C TIME WCPU COMMAND 20851 root 1 1270 23024K 22400K RUN0 0:03 88.66% cc1plus 861 anderson 1 960 108M 85384K select 0 1:22 0.98% Xorg 958 anderson 5 200 109M 97228K kserel 0 2:14 0.00% firefox-bin 922 anderson 1 960 19556K 14944K select 0 0:54 0.00% xfce4-panel 592 root 1 80 1348K 872K nanslp 0 0:43 0.00% powerd 932 anderson 6 200 75360K 60860K kserel 0 0:20 0.00% thunderbird-bin 28378 root 1 80 31388K 30920K wait 0 0:17 0.00% ruby18 909 anderson 1 960 29416K 15816K select 0 0:12 0.00% kdeinit 920 anderson 1 960 16412K 11228K select 0 0:03 0.00% xfdesktop 559 root 1 960 1448K 872K select 0 0:02 0.00% bthidd All processes are on cpu 0. Output of sysctl dev.cpu shows: dev.cpu.0.%desc: ACPI CPU dev.cpu.0.%driver: cpu dev.cpu.0.%location: handle=\_PR_.CPU0 dev.cpu.0.%pnpinfo: _HID=none _UID=0 dev.cpu.0.%parent: acpi0 dev.cpu.0.freq: 2005 dev.cpu.0.freq_levels: 2005/-1 1754/-1 1503/-1 1253/-1 1002/-1 751/-1 501/-1 250/-1 dev.cpu.1.%desc: ACPI CPU dev.cpu.1.%driver: cpu dev.cpu.1.%location: handle=\_PR_.CPU1 dev.cpu.1.%pnpinfo: _HID=none _UID=0 dev.cpu.1.%parent: acpi0 More information (dmesg, sysctl -a, devinfo, pciconf, etc) can be found here: http://www.googlebit.com/freebsd/200605041945/ Thanks.. Eric -- Eric AndersonSr. Systems AdministratorCentaur Technology Anything that works is better than anything that doesn't. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Core Duo - only one cpu being used
On Thu, 04 May 2006 21:16:15 -0500, Eric Anderson wrote: Forgive me if I've missed this on a list somewhere, but My new laptop with a Core Duo doesn't seem to use both CPU's. It sees both, but I never see anything on cpu 1. Here's a top snippet: My new desktop (M/B is ASUS N4L-VM DH) with a Core Duo is fine to use both CPU's in 6.1-RC like this: last pid: 546; load averages: 0.01, 0.11, 0.07up 0+00:06:20 11:23:32 23 processes: 1 running, 22 sleeping CPU states: 0.0% user, 0.0% nice, 0.0% system, 0.0% interrupt, 100% idle Mem: 9312K Active, 84M Inact, 81M Wired, 112M Buf, 1822M Free Swap: 4096M Total, 4096M Free PID USERNAME THR PRI NICE SIZERES STATE C TIME WCPU COMMAND 533 root1 760 2272K 1532K CPU0 0 0:00 0.00% top 509 root1 40 6104K 3064K sbwait 1 0:00 0.00% sshd 401 root1 760 2840K 1700K select 0 0:00 0.00% ntpd 520 root1 200 5388K 3064K pause 1 0:00 0.00% tcsh 512 yosimoto1 760 6080K 3076K select 0 0:00 0.00% sshd 513 yosimoto1 200 5060K 2848K pause 1 0:00 0.00% tcsh 296 root1 760 1300K 948K select 0 0:00 0.00% syslogd 519 yosimoto1 80 1604K 1292K wait 1 0:00 0.00% su 432 root1 760 3400K 2728K select 0 0:00 0.00% sendmail 442 root1 80 1312K 1044K nanslp 0 0:00 0.00% cron 502 root1 50 1268K 904K ttyin 0 0:00 0.00% getty 504 root1 50 1268K 904K ttyin 1 0:00 0.00% getty 507 root1 50 1268K 904K ttyin 1 0:00 0.00% getty 508 root1 50 1268K 904K ttyin 1 0:00 0.00% getty 505 root1 50 1268K 904K ttyin 1 0:00 0.00% getty 506 root1 50 1268K 904K ttyin 1 0:00 0.00% getty 501 root1 50 1268K 904K ttyin 1 0:00 0.00% getty 503 root1 50 1268K 904K ttyin 1 0:00 0.00% getty 426 root1 760 3356K 2540K select 0 0:00 0.00% sshd 436 smmsp 1 200 3300K 2704K pause 0 0:00 0.00% sendmail 262 root1 820 500K 360K select 1 0:00 0.00% devd 145 root1 200 1176K 680K pause 0 0:00 0.00% adjkerntz 463 root1 1390 1260K 768K select 0 0:00 0.00% moused All processes are on cpu 0. Output of sysctl dev.cpu shows: In my case: # sysctl -a | grep dev.cpu dev.cpu.0.%desc: ACPI CPU dev.cpu.0.%driver: cpu dev.cpu.0.%location: handle=\_PR_.CPU1 dev.cpu.0.%pnpinfo: _HID=none _UID=0 dev.cpu.0.%parent: acpi0 dev.cpu.1.%desc: ACPI CPU dev.cpu.1.%driver: cpu dev.cpu.1.%location: handle=\_PR_.CPU2 dev.cpu.1.%pnpinfo: _HID=none _UID=0 dev.cpu.1.%parent: acpi0 --- Shin-ichi YOSHIMOTO [EMAIL PROTECTED] http://darwin.waishi.jp/diary/ http://www.ugon.jp/ ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Core Duo - only one cpu being used
Eric Anderson wrote: Forgive me if I've missed this on a list somewhere, but My new laptop with a Core Duo doesn't seem to use both CPU's. It sees both, but I never see anything on cpu 1. Here's a top snippet: Your top output shows a single process eating the CPU. A single process can't span CPUs, so you're only going to see one CPU in use. You need to do something in parallel, like make -j N where N 1. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Core Duo - only one cpu being used
Darren Pilgrim wrote: Eric Anderson wrote: Forgive me if I've missed this on a list somewhere, but My new laptop with a Core Duo doesn't seem to use both CPU's. It sees both, but I never see anything on cpu 1. Here's a top snippet: Your top output shows a single process eating the CPU. A single process can't span CPUs, so you're only going to see one CPU in use. You need to do something in parallel, like make -j N where N 1. I understand that a single process won't span cpu's, but there isn't a single process on the second cpu, and running multiple: cat /dev/random | md5 Still slams one cpu. Eric -- Eric AndersonSr. Systems AdministratorCentaur Technology Anything that works is better than anything that doesn't. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Core Duo - only one cpu being used
Hi, Eric Anderson wrote: All processes are on cpu 0. Output of sysctl dev.cpu shows: can you show all tasks running? Erich ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Core Duo - only one cpu being used
Erich Dollansky wrote: Hi, Eric Anderson wrote: All processes are on cpu 0. Output of sysctl dev.cpu shows: can you show all tasks running? Erich last pid: 33015; load averages: 1.20, 1.08, 1.08 up 0+02:35:03 22:19:55 98 processes: 4 running, 94 sleeping CPU states: 19.7% user, 0.0% nice, 28.9% system, 1.3% interrupt, 50.1% idle Mem: 388M Active, 251M Inact, 132M Wired, 11M Cache, 110M Buf, 206M Free Swap: 2048M Total, 2048M Free PID USERNAMETHR PRI NICE SIZERES STATE C TIME WCPU COMMAND 33010 root 1 1020 4984K 4364K RUN0 0:06 27.21% bzip2 33012 anderson 1 1000 1296K 636K RUN0 0:04 21.09% cat 33014 anderson 1 1000 1296K 636K RUN0 0:03 19.97% cat 33013 anderson 1 -80 1356K 708K piperd 0 0:01 3.72% md5 33015 anderson 1 -80 1356K 708K piperd 0 0:01 3.14% md5 33011 root 1 -80 1736K 1224K piperd 0 0:01 2.36% bsdtar 861 anderson 1 960 115M 89532K select 0 2:24 1.12% Xorg 592 root 1 80 1348K 872K nanslp 0 1:24 0.05% powerd 958 anderson 5 200 109M 96828K kserel 0 3:04 0.00% firefox-bin 922 anderson 1 960 19604K 14992K select 0 1:48 0.00% xfce4-panel 932 anderson 6 200 79688K 65352K kserel 0 0:45 0.00% thunderbird-bin 909 anderson 1 960 29416K 16044K select 0 0:24 0.00% kdeinit 28378 root 1 80 31388K 30912K wait 0 0:19 0.00% ruby18 920 anderson 1 960 16520K 11356K select 0 0:06 0.00% xfdesktop 1039 anderson 1 960 6388K 4752K select 0 0:04 0.00% xterm 559 root 1 960 1448K 872K select 0 0:03 0.00% bthidd 870 anderson 1 960 27000K 17472K select 0 0:03 0.00% gaim 1037 anderson 1 960 7916K 5824K select 0 0:02 0.00% nedit 888 anderson 1 960 18600K 13752K select 0 0:02 0.00% xfce4-session 916 anderson 1 960 14592K 10344K select 0 0:02 0.00% xfwm4 897 anderson 1 960 14404K 8460K select 0 0:02 0.00% xfce-mcs-manager 918 anderson 1 960 13116K 8968K select 0 0:01 0.00% xftaskbar4 869 anderson 1 960 6820K 5080K select 0 0:01 0.00% xterm 25663 anderson 1 960 5540K 3904K select 0 0:01 0.00% xterm 1192 anderson 1 960 6128K 4492K select 0 0:01 0.00% xterm 235 root 1 960 3056K 1684K select 0 0:00 0.00% wpa_supplicant 630 root 1 960 2620K 1952K select 0 0:00 0.00% httpd 68119 anderson 1 960 6244K 4548K select 0 0:00 0.00% xterm 92204 anderson 1 960 5536K 3848K select 0 0:00 0.00% xterm 901 anderson 1 970 25284K 11308K select 0 0:00 0.00% kdeinit 683 root 1 960 3712K 2404K select 0 0:00 0.00% sendmail 68120 anderson 1 960 3900K 2416K select 0 0:00 0.00% ssh 934 anderson 1 960 4360K 2880K select 0 0:00 0.00% gconfd-2 1193 anderson 1 50 3360K 1924K ttyin 0 0:00 0.00% bash 32933 root 1 80 1624K 1476K wait 0 0:00 0.00% make 25666 root 1 50 3276K 1824K ttyin 0 0:00 0.00% bash 30424 anderson 1 960 5536K 4192K select 0 0:00 0.00% xterm 889 anderson 1 80 3348K 1908K wait 0 0:00 0.00% bash 29892 anderson 1 960 2600K 1696K CPU0 0 0:00 0.00% top 480 root 1 960 1412K 924K select 0 0:00 0.00% syslogd 904 anderson 1 960 24004K 10172K select 0 0:00 0.00% kdeinit 907 anderson 1 960 24600K 10772K select 0 0:00 0.00% kdeinit 92205 anderson 1 50 3308K 1876K ttyin 0 0:00 0.00% bash 694 root 1 80 1428K 1028K nanslp 0 0:00 0.00% cron 924 anderson 1 80 1768K 1136K wait 0 0:00 0.00% sh 785 root 1 80 1748K 1256K wait 0 0:00 0.00% login 30453 anderson 1 50 3308K 1920K ttyin 0 0:00 0.00% bash 830 anderson 1 80 3320K 1864K wait 0 0:00 0.00% bash 1040 anderson 1 80 3344K 1908K wait 0 0:00 0.00% bash 25664 anderson 1 80 3308K 1872K wait 0 0:00 0.00% bash 1043 root 1 80 3304K 1852K wait 0 0:00 0.00% bash 936 anderson 1 80 1768K 1136K wait 0 0:00 0.00% sh 616 root 1 960 1320K 768K select 0 0:00 0.00% usbd 25665 anderson 1 80 1728K 1244K wait 0 0:00 0.00% su 1042 anderson 1 80 1728K 1236K wait 0 0:00 0.00% su 871 anderson 1 80 1764K 1060K wait 0 0:00 0.00% sh 32932 root 1 1050 1376K 784K select 0 0:00 0.00% script 242 root 1 960 1500K 988K select 0 0:00 0.00% dhclient 841 anderson 1 80 1792K 1108K wait 0 0:00 0.00% sh 248 _dhcp
Re: Core Duo - only one cpu being used
Hi, this is hard to believe. Just add -S to the command line or enter S while top is running. Top should show you then on top the two idle threads. Erich Eric Anderson wrote: Erich Dollansky wrote: Hi, Eric Anderson wrote: All processes are on cpu 0. Output of sysctl dev.cpu shows: can you show all tasks running? Erich last pid: 33015; load averages: 1.20, 1.08, 1.08 up 0+02:35:03 22:19:55 98 processes: 4 running, 94 sleeping CPU states: 19.7% user, 0.0% nice, 28.9% system, 1.3% interrupt, 50.1% idle Mem: 388M Active, 251M Inact, 132M Wired, 11M Cache, 110M Buf, 206M Free Swap: 2048M Total, 2048M Free PID USERNAMETHR PRI NICE SIZERES STATE C TIME WCPU COMMAND 33010 root 1 1020 4984K 4364K RUN0 0:06 27.21% bzip2 33012 anderson 1 1000 1296K 636K RUN0 0:04 21.09% cat 33014 anderson 1 1000 1296K 636K RUN0 0:03 19.97% cat 33013 anderson 1 -80 1356K 708K piperd 0 0:01 3.72% md5 33015 anderson 1 -80 1356K 708K piperd 0 0:01 3.14% md5 33011 root 1 -80 1736K 1224K piperd 0 0:01 2.36% bsdtar 861 anderson 1 960 115M 89532K select 0 2:24 1.12% Xorg 592 root 1 80 1348K 872K nanslp 0 1:24 0.05% powerd 958 anderson 5 200 109M 96828K kserel 0 3:04 0.00% firefox-bin 922 anderson 1 960 19604K 14992K select 0 1:48 0.00% xfce4-panel 932 anderson 6 200 79688K 65352K kserel 0 0:45 0.00% thunderbird-bin 909 anderson 1 960 29416K 16044K select 0 0:24 0.00% kdeinit 28378 root 1 80 31388K 30912K wait 0 0:19 0.00% ruby18 920 anderson 1 960 16520K 11356K select 0 0:06 0.00% xfdesktop 1039 anderson 1 960 6388K 4752K select 0 0:04 0.00% xterm 559 root 1 960 1448K 872K select 0 0:03 0.00% bthidd 870 anderson 1 960 27000K 17472K select 0 0:03 0.00% gaim 1037 anderson 1 960 7916K 5824K select 0 0:02 0.00% nedit 888 anderson 1 960 18600K 13752K select 0 0:02 0.00% xfce4-session 916 anderson 1 960 14592K 10344K select 0 0:02 0.00% xfwm4 897 anderson 1 960 14404K 8460K select 0 0:02 0.00% xfce-mcs-manager 918 anderson 1 960 13116K 8968K select 0 0:01 0.00% xftaskbar4 869 anderson 1 960 6820K 5080K select 0 0:01 0.00% xterm 25663 anderson 1 960 5540K 3904K select 0 0:01 0.00% xterm 1192 anderson 1 960 6128K 4492K select 0 0:01 0.00% xterm 235 root 1 960 3056K 1684K select 0 0:00 0.00% wpa_supplicant 630 root 1 960 2620K 1952K select 0 0:00 0.00% httpd 68119 anderson 1 960 6244K 4548K select 0 0:00 0.00% xterm 92204 anderson 1 960 5536K 3848K select 0 0:00 0.00% xterm 901 anderson 1 970 25284K 11308K select 0 0:00 0.00% kdeinit 683 root 1 960 3712K 2404K select 0 0:00 0.00% sendmail 68120 anderson 1 960 3900K 2416K select 0 0:00 0.00% ssh 934 anderson 1 960 4360K 2880K select 0 0:00 0.00% gconfd-2 1193 anderson 1 50 3360K 1924K ttyin 0 0:00 0.00% bash 32933 root 1 80 1624K 1476K wait 0 0:00 0.00% make 25666 root 1 50 3276K 1824K ttyin 0 0:00 0.00% bash 30424 anderson 1 960 5536K 4192K select 0 0:00 0.00% xterm 889 anderson 1 80 3348K 1908K wait 0 0:00 0.00% bash 29892 anderson 1 960 2600K 1696K CPU0 0 0:00 0.00% top 480 root 1 960 1412K 924K select 0 0:00 0.00% syslogd 904 anderson 1 960 24004K 10172K select 0 0:00 0.00% kdeinit 907 anderson 1 960 24600K 10772K select 0 0:00 0.00% kdeinit 92205 anderson 1 50 3308K 1876K ttyin 0 0:00 0.00% bash 694 root 1 80 1428K 1028K nanslp 0 0:00 0.00% cron 924 anderson 1 80 1768K 1136K wait 0 0:00 0.00% sh 785 root 1 80 1748K 1256K wait 0 0:00 0.00% login 30453 anderson 1 50 3308K 1920K ttyin 0 0:00 0.00% bash 830 anderson 1 80 3320K 1864K wait 0 0:00 0.00% bash 1040 anderson 1 80 3344K 1908K wait 0 0:00 0.00% bash 25664 anderson 1 80 3308K 1872K wait 0 0:00 0.00% bash 1043 root 1 80 3304K 1852K wait 0 0:00 0.00% bash 936 anderson 1 80 1768K 1136K wait 0 0:00 0.00% sh 616 root 1 960 1320K 768K select 0 0:00 0.00% usbd 25665 anderson 1 80 1728K 1244K wait 0 0:00 0.00% su 1042 anderson 1 80 1728K 1236K wait 0 0:00 0.00% su 871 anderson 1 80 1764K 1060K wait 0 0:00 0.00% sh 32932 root 1 1050 1376K 784K select 0 0:00
Re: Core Duo - only one cpu being used
Erich Dollansky wrote: Hi, this is hard to believe. Just add -S to the command line or enter S while top is running. Top should show you then on top the two idle threads. Here's the output: last pid: 2654; load averages: 1.10, 1.16, 1.12 up 0+03:30:41 23:15:33 159 processes: 5 running, 132 sleeping, 22 waiting CPU states: 46.5% user, 0.0% nice, 3.2% system, 0.4% interrupt, 49.9% idle Mem: 408M Active, 398M Inact, 134M Wired, 32M Cache, 110M Buf, 15M Free Swap: 2048M Total, 2048M Free PID USERNAMETHR PRI NICE SIZERES STATE C TIME WCPU COMMAND 11 root 1 171 52 0K 8K CPU1 0 0:00 99.02% idle: cpu1 2653 root 1 1280 18564K 17560K RUN0 0:01 34.00% cc1plus 24 root 1 80 0K 8K - 0 4:34 1.32% ath0 taskq 23 root 1 -68 -187 0K 8K WAIT 0 2:37 0.88% irq17: ath0 861 anderson 1 960 112M 90744K select 0 3:27 0.10% Xorg 922 anderson 1 960 19672K 15060K select 0 2:54 0.05% xfce4-panel 12 root 1 171 52 0K 8K RUN0 51:04 0.00% idle: cpu0 958 anderson 5 200 109M 96876K kserel 0 3:48 0.00% firefox-bin 592 root 1 80 1348K 872K nanslp 0 2:09 0.00% powerd 932 anderson 6 200 87196K 72904K kserel 0 1:02 0.00% thunderbird-bin 909 anderson 1 960 29416K 15964K select 0 0:39 0.00% kdeinit 46 root 1 171 52 0K 8K RUN0 0:34 0.00% pagezero 14 root 1 -32 -151 0K 8K WAIT 0 0:33 0.00% swi4: clock sio 4 root 1 -80 0K 8K - 0 0:23 0.00% g_down 22 root 1 -80 -199 0K 8K WAIT 0 0:21 0.00% irq16: nvidia0 3 root 1 -80 0K 8K - 0 0:19 0.00% g_up 28378 root 1 80 31388K 30912K wait 0 0:19 0.00% ruby18 49 root 1 200 0K 8K syncer 0 0:09 0.00% syncer 38 root 1 -64 -183 0K 8K WAIT 0 0:08 0.00% irq14: ata0 920 anderson 1 960 16536K 11372K select 0 0:07 0.00% xfdesktop 40 root 1 00 0K 8K tzpoll 0 0:06 0.00% acpi_thermal 559 root 1 960 1448K 872K select 0 0:05 0.00% bthidd 870 anderson 1 960 26996K 17060K select 0 0:05 0.00% gaim 1039 anderson 1 960 6388K 4752K select 0 0:05 0.00% xterm 1037 anderson 1 960 7916K 5824K select 0 0:03 0.00% nedit 916 anderson 1 960 14592K 10340K select 0 0:02 0.00% xfwm4 13 root 1 -44 -163 0K 8K WAIT 0 0:02 0.00% swi1: net 888 anderson 1 960 18600K 13752K select 0 0:02 0.00% xfce4-session 897 anderson 1 960 1K 8500K select 0 0:02 0.00% xfce-mcs-manager 26 root 1 -64 -183 0K 8K WAIT 0 0:02 0.00% irq20: uhci0 ehci0 918 anderson 1 960 13116K 8976K select 0 0:02 0.00% xftaskbar4 16 root 1 -160 0K 8K - 0 0:02 0.00% yarrow 869 anderson 1 960 6280K 4540K select 0 0:01 0.00% xterm 2 root 1 -80 0K 8K - 0 0:01 0.00% g_event 55 root 1 120 0K 8K - 0 0:01 0.00% schedcpu 1192 anderson 1 960 6908K 5272K select 0 0:01 0.00% xterm 18 root 1 -24 -143 0K 8K WAIT 0 0:01 0.00% swi6: + 25663 anderson 1 960 5540K 3904K select 0 0:01 0.00% xterm 235 root 1 960 3056K 1684K select 0 0:01 0.00% wpa_supplicant 19 root 1 -24 -143 0K 8K WAIT 0 0:01 0.00% swi6: task queue 41 root 1 -60 -179 0K 8K WAIT 0 0:01 0.00% irq1: atkbd0 [..snip..] Eric Anderson wrote: Erich Dollansky wrote: Hi, Eric Anderson wrote: All processes are on cpu 0. Output of sysctl dev.cpu shows: can you show all tasks running? Erich last pid: 33015; load averages: 1.20, 1.08, 1.08 up 0+02:35:03 22:19:55 98 processes: 4 running, 94 sleeping CPU states: 19.7% user, 0.0% nice, 28.9% system, 1.3% interrupt, 50.1% idle Mem: 388M Active, 251M Inact, 132M Wired, 11M Cache, 110M Buf, 206M Free Swap: 2048M Total, 2048M Free PID USERNAMETHR PRI NICE SIZERES STATE C TIME WCPU COMMAND 33010 root 1 1020 4984K 4364K RUN0 0:06 27.21% bzip2 33012 anderson 1 1000 1296K 636K RUN0 0:04 21.09% cat 33014 anderson 1 1000 1296K 636K RUN0 0:03 19.97% cat 33013 anderson 1 -80 1356K 708K piperd 0 0:01 3.72% md5 33015 anderson 1 -80 1356K 708K piperd 0 0:01 3.14% md5 33011 root 1 -80 1736K 1224K piperd 0 0:01 2.36% bsdtar 861 anderson 1 960 115M 89532K select 0 2:24 1.12% Xorg 592 root 1 80 1348K 872K nanslp 0
Re: Core Duo - only one cpu being used
Hi, Eric Anderson wrote: PID USERNAMETHR PRI NICE SIZERES STATE C TIME WCPU COMMAND 11 root 1 171 52 0K 8K CPU1 0 0:00 99.02% idle: cpu1 2653 root 1 1280 18564K 17560K RUN0 0:01 34.00% cc1plus could it be that it is just a problem with top itself? It cannot be that CPU1 uses 99% for the idle process and 34% for the compiler. Play with the other sort options. You might find the the idle process for CPU0. Erich ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Core Duo - only one cpu being used
Erich Dollansky wrote: Hi, Eric Anderson wrote: PID USERNAMETHR PRI NICE SIZERES STATE C TIME WCPU COMMAND 11 root 1 171 52 0K 8K CPU1 0 0:00 99.02% idle: cpu1 2653 root 1 1280 18564K 17560K RUN0 0:01 34.00% cc1plus could it be that it is just a problem with top itself? It cannot be that CPU1 uses 99% for the idle process and 34% for the compiler. Play with the other sort options. You might find the the idle process for CPU0. Is this what you want: $ ps -auxw | grep idle root11 99.0 0.0 0 8 ?? RL 7:45PM 0:00.00 [idle: cpu1] root12 0.0 0.0 0 8 ?? RL 7:45PM 51:04.57 [idle: cpu0] I'm sure it could be a top issue, but other tools that don't use top report only 50% total cpu usage.. Eric -- Eric AndersonSr. Systems AdministratorCentaur Technology Anything that works is better than anything that doesn't. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Core Duo - only one cpu being used
Hi, Eric Anderson wrote: Erich Dollansky wrote: Hi, Eric Anderson wrote: PID USERNAMETHR PRI NICE SIZERES STATE C TIME WCPU COMMAND 11 root 1 171 52 0K 8K CPU1 0 0:00 99.02% idle: cpu1 2653 root 1 1280 18564K 17560K RUN0 0:01 34.00% cc1plus could it be that it is just a problem with top itself? It cannot be that CPU1 uses 99% for the idle process and 34% for the compiler. Play with the other sort options. You might find the the idle process for CPU0. Is this what you want: $ ps -auxw | grep idle root11 99.0 0.0 0 8 ?? RL 7:45PM 0:00.00 [idle: cpu1] root12 0.0 0.0 0 8 ?? RL 7:45PM 51:04.57 [idle: cpu0] something is really wrong here. CPU1 gets 99% of the time but uses then only 0 seconds while CPU0 gets 0% of the time but uses 51 hours? This is, what I get: root 11 98.2 0.0 0 8 ?? RL 21Apr06 19013:02.26 [idle: cpu1] root 12 94.9 0.0 0 8 ?? RL 21Apr06 18456:39.99 [idle: cpu0] This is a dual-Athlon system running 6.1 RC1. Erich ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Core Duo - only one cpu being used
Erich Dollansky wrote: Hi, Eric Anderson wrote: Erich Dollansky wrote: Hi, Eric Anderson wrote: PID USERNAMETHR PRI NICE SIZERES STATE C TIME WCPU COMMAND 11 root 1 171 52 0K 8K CPU1 0 0:00 99.02% idle: cpu1 2653 root 1 1280 18564K 17560K RUN0 0:01 34.00% cc1plus could it be that it is just a problem with top itself? It cannot be that CPU1 uses 99% for the idle process and 34% for the compiler. Play with the other sort options. You might find the the idle process for CPU0. Is this what you want: $ ps -auxw | grep idle root11 99.0 0.0 0 8 ?? RL 7:45PM 0:00.00 [idle: cpu1] root12 0.0 0.0 0 8 ?? RL 7:45PM 51:04.57 [idle: cpu0] something is really wrong here. CPU1 gets 99% of the time but uses then only 0 seconds while CPU0 gets 0% of the time but uses 51 hours? CPU1 is being treated as a hyperthreading core instead of a real core, and is being disabled per our policy on Intel hyperthreading. By 'disabled' I mean that it is started, but it is being excluded from scheduling decisions, and thus is only running its idle proc. It's also handling any interrupts that come to it, such as timer and IPI interrupts, so it's at 99% instead of 100% for the idle proc. There is nothing broken about the number you are seeing, your system is just running under a scheduling policy that it should not be. This should have been fixed a week or so ago by a commit to HEAD, RELENG_6, and RELENG_6_1 by Colin Percival. How old is kernel? Scott ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Core Duo - only one cpu being used
Scott Long wrote: Erich Dollansky wrote: Hi, Eric Anderson wrote: Erich Dollansky wrote: Hi, Eric Anderson wrote: PID USERNAMETHR PRI NICE SIZERES STATE C TIME WCPU COMMAND 11 root 1 171 52 0K 8K CPU1 0 0:00 99.02% idle: cpu1 2653 root 1 1280 18564K 17560K RUN0 0:01 34.00% cc1plus could it be that it is just a problem with top itself? It cannot be that CPU1 uses 99% for the idle process and 34% for the compiler. Play with the other sort options. You might find the the idle process for CPU0. Is this what you want: $ ps -auxw | grep idle root11 99.0 0.0 0 8 ?? RL 7:45PM 0:00.00 [idle: cpu1] root12 0.0 0.0 0 8 ?? RL 7:45PM 51:04.57 [idle: cpu0] something is really wrong here. CPU1 gets 99% of the time but uses then only 0 seconds while CPU0 gets 0% of the time but uses 51 hours? CPU1 is being treated as a hyperthreading core instead of a real core, and is being disabled per our policy on Intel hyperthreading. By 'disabled' I mean that it is started, but it is being excluded from scheduling decisions, and thus is only running its idle proc. It's also handling any interrupts that come to it, such as timer and IPI interrupts, so it's at 99% instead of 100% for the idle proc. There is nothing broken about the number you are seeing, your system is just running under a scheduling policy that it should not be. This should have been fixed a week or so ago by a commit to HEAD, RELENG_6, and RELENG_6_1 by Colin Percival. How old is kernel? 6.1-RC FreeBSD 6.1-RC #13: Thu Apr 27 08:33:14 CDT 2006 So I probably just missed it. I'll rebuilt and try it tomorrow morning, and report back. Thanks for all the help and a good description. Eric -- Eric AndersonSr. Systems AdministratorCentaur Technology Anything that works is better than anything that doesn't. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]