From: Sam L. <ph...@gm...> - 2009-03-31 03:06:03
|
Hi all, I'm try run oprofile on redhat 5.2, and my cpu cycle frequency is about 2000MHz, when I set the counter to 1M, it supposed to get 2000 samples per second, but It seems I only get 200 sampls around per second.(as following log, I got 3600 samples in 10s) Where is the problem? Thanks. Best Regards Sam ------------------------------------------------------------------------------- #opcontrol --event=CPU_CLK_UNHALTED:1000000:0:1:1 #opcontrol --status [root@localhost 0]# opcontrol --status Daemon not running Event 0: CPU_CLK_UNHALTED:1000000:0:1:1 Separate options: library vmlinux file: none Image filter: none Call-graph depth: 64 Buffer size: 65536 [root@localhost 0]# opcontrol --reset;opcontrol --start;sleep 10s;opcontrol --dump;opcontrol --shutdown; opreport -a -l| tail; Using 2.6+ OProfile kernel interface. Using log file /var/lib/oprofile/samples/oprofiled.log Daemon started. Profiler running. Stopping profiling. Killing daemon. warning: /no-vmlinux could not be found. 1 3690 0.0270 99.7567 libpthread-2.5.so scim-panel-gtk read 1 3691 0.0270 99.7837 libpthread-2.5.so xfwm4 __pthread_mutex_unlock_usercnt 1 3692 0.0270 99.8108 libqt-mt.so.3.3.6 kded (no symbols) 1 3693 0.0270 99.8378 libscim-1.0.so.8.1.0 scim-panel-gtk (no symbols) 1 3694 0.0270 99.8648 libscim-gtkutils-1.0.so.8.1.0 scim-panel-gtk (no symbols) 1 3695 0.0270 99.8919 libstdc++.so.6.0.7 acroread (no symbols) 1 3696 0.0270 99.9189 libxfce4mixer.so.0.0.0 xfce4-mixer-plugin (no symbols) 1 3697 0.0270 99.9459 libxfcegui4.so.4.2.6 xfce-mcs-manager (no symbols) 1 3698 0.0270 99.9730 ophelp ophelp (no symbols) 1 3699 0.0270 100.000 oprofiled oprofiled (no symbols) [root@localhost 0]# cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Core(TM)2 CPU 6400 @ 2.13GHz stepping : 2 cpu MHz : 1600.000 cache size : 2048 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall nx lm constant_tsc pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm bogomips : 4259.83 clflush size : 64 cache_alignment : 64 address sizes : 36 bits physical, 48 bits virtual power management: processor : 1 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Core(TM)2 CPU 6400 @ 2.13GHz stepping : 2 cpu MHz : 1600.000 cache size : 2048 KB physical id : 0 siblings : 2 core id : 1 cpu cores : 2 fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall nx lm constant_tsc pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm bogomips : 4256.01 clflush size : 64 cache_alignment : 64 address sizes : 36 bits physical, 48 bits virtual power management: |
From: Monthadar Al J. <mon...@gm...> - 2009-03-31 07:36:50
|
Hi, I would suspect that you used CPU_CLK_UNHALTED? If that is so, then you need to boot the kernel with polling option, idle=poll. The manual describes it better than me: 2.2. Idle time Your kernel is likely to support halting the processor when a CPU is idle. As the typical hardware events like CPU_CLK_UNHALTED do not count when the CPU is halted, the kernel profile will not reflect the actual amount of time spent idle. You can change this behaviour by booting with the idle=poll option, which uses a different idle routine. This will appear as poll_idle() in your kernel profile. BR, On Tue, Mar 31, 2009 at 5:05 AM, Sam Liao <ph...@gm...> wrote: > Hi all, > > I'm try run oprofile on redhat 5.2, and my cpu cycle frequency is > about 2000MHz, when I set the counter to 1M, it supposed to get 2000 > samples per second, but It seems I only get 200 sampls around per > second.(as following log, I got 3600 samples in 10s) > Where is the problem? > > Thanks. > Best Regards > Sam > ------------------------------------------------------------------------------- > #opcontrol --event=CPU_CLK_UNHALTED:1000000:0:1:1 > #opcontrol --status > [root@localhost 0]# opcontrol --status > Daemon not running > Event 0: CPU_CLK_UNHALTED:1000000:0:1:1 > Separate options: library > vmlinux file: none > Image filter: none > Call-graph depth: 64 > Buffer size: 65536 > [root@localhost 0]# opcontrol --reset;opcontrol --start;sleep > 10s;opcontrol --dump;opcontrol --shutdown; opreport -a -l| tail; > Using 2.6+ OProfile kernel interface. > Using log file /var/lib/oprofile/samples/oprofiled.log > Daemon started. > Profiler running. > Stopping profiling. > Killing daemon. > warning: /no-vmlinux could not be found. > 1 3690 0.0270 99.7567 libpthread-2.5.so > scim-panel-gtk read > 1 3691 0.0270 99.7837 libpthread-2.5.so > xfwm4 __pthread_mutex_unlock_usercnt > 1 3692 0.0270 99.8108 libqt-mt.so.3.3.6 > kded (no symbols) > 1 3693 0.0270 99.8378 libscim-1.0.so.8.1.0 > scim-panel-gtk (no symbols) > 1 3694 0.0270 99.8648 > libscim-gtkutils-1.0.so.8.1.0 scim-panel-gtk (no symbols) > 1 3695 0.0270 99.8919 libstdc++.so.6.0.7 > acroread (no symbols) > 1 3696 0.0270 99.9189 libxfce4mixer.so.0.0.0 > xfce4-mixer-plugin (no symbols) > 1 3697 0.0270 99.9459 libxfcegui4.so.4.2.6 > xfce-mcs-manager (no symbols) > 1 3698 0.0270 99.9730 ophelp > ophelp (no symbols) > 1 3699 0.0270 100.000 oprofiled > oprofiled (no symbols) > > [root@localhost 0]# cat /proc/cpuinfo > processor : 0 > vendor_id : GenuineIntel > cpu family : 6 > model : 15 > model name : Intel(R) Core(TM)2 CPU 6400 @ 2.13GHz > stepping : 2 > cpu MHz : 1600.000 > cache size : 2048 KB > physical id : 0 > siblings : 2 > core id : 0 > cpu cores : 2 > fpu : yes > fpu_exception : yes > cpuid level : 10 > wp : yes > flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge > mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall > nx lm constant_tsc pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm > bogomips : 4259.83 > clflush size : 64 > cache_alignment : 64 > address sizes : 36 bits physical, 48 bits virtual > power management: > > processor : 1 > vendor_id : GenuineIntel > cpu family : 6 > model : 15 > model name : Intel(R) Core(TM)2 CPU 6400 @ 2.13GHz > stepping : 2 > cpu MHz : 1600.000 > cache size : 2048 KB > physical id : 0 > siblings : 2 > core id : 1 > cpu cores : 2 > fpu : yes > fpu_exception : yes > cpuid level : 10 > wp : yes > flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge > mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall > nx lm constant_tsc pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm > bogomips : 4256.01 > clflush size : 64 > cache_alignment : 64 > address sizes : 36 bits physical, 48 bits virtual > power management: > > ------------------------------------------------------------------------------ > _______________________________________________ > oprofile-list mailing list > opr...@li... > https://lists.sourceforge.net/lists/listinfo/oprofile-list > -- //Monthadar Al Jaberi |
From: Sam L. <ph...@gm...> - 2009-03-31 07:51:22
|
Thanks. It works now. Best Regards Sam On Tue, Mar 31, 2009 at 3:36 PM, Monthadar Al Jaberi <mon...@gm...> wrote: > Hi, > > I would suspect that you used CPU_CLK_UNHALTED? > > If that is so, then you need to boot the kernel with polling option, > idle=poll. The manual describes it better than me: > > 2.2. Idle time > > Your kernel is likely to support halting the processor when a CPU is > idle. As the typical hardware events like CPU_CLK_UNHALTED do not > count when the CPU is halted, the kernel profile will not reflect the > actual amount of time spent idle. You can change this behaviour by > booting with the idle=poll option, which uses a different idle > routine. This will appear as poll_idle() in your kernel profile. > > BR, > > On Tue, Mar 31, 2009 at 5:05 AM, Sam Liao <ph...@gm...> wrote: >> Hi all, >> >> I'm try run oprofile on redhat 5.2, and my cpu cycle frequency is >> about 2000MHz, when I set the counter to 1M, it supposed to get 2000 >> samples per second, but It seems I only get 200 sampls around per >> second.(as following log, I got 3600 samples in 10s) >> Where is the problem? >> >> Thanks. >> Best Regards >> Sam >> ------------------------------------------------------------------------------- >> #opcontrol --event=CPU_CLK_UNHALTED:1000000:0:1:1 >> #opcontrol --status >> [root@localhost 0]# opcontrol --status >> Daemon not running >> Event 0: CPU_CLK_UNHALTED:1000000:0:1:1 >> Separate options: library >> vmlinux file: none >> Image filter: none >> Call-graph depth: 64 >> Buffer size: 65536 >> [root@localhost 0]# opcontrol --reset;opcontrol --start;sleep >> 10s;opcontrol --dump;opcontrol --shutdown; opreport -a -l| tail; >> Using 2.6+ OProfile kernel interface. >> Using log file /var/lib/oprofile/samples/oprofiled.log >> Daemon started. >> Profiler running. >> Stopping profiling. >> Killing daemon. >> warning: /no-vmlinux could not be found. >> 1 3690 0.0270 99.7567 libpthread-2.5.so >> scim-panel-gtk read >> 1 3691 0.0270 99.7837 libpthread-2.5.so >> xfwm4 __pthread_mutex_unlock_usercnt >> 1 3692 0.0270 99.8108 libqt-mt.so.3.3.6 >> kded (no symbols) >> 1 3693 0.0270 99.8378 libscim-1.0.so.8.1.0 >> scim-panel-gtk (no symbols) >> 1 3694 0.0270 99.8648 >> libscim-gtkutils-1.0.so.8.1.0 scim-panel-gtk (no symbols) >> 1 3695 0.0270 99.8919 libstdc++.so.6.0.7 >> acroread (no symbols) >> 1 3696 0.0270 99.9189 libxfce4mixer.so.0.0.0 >> xfce4-mixer-plugin (no symbols) >> 1 3697 0.0270 99.9459 libxfcegui4.so.4.2.6 >> xfce-mcs-manager (no symbols) >> 1 3698 0.0270 99.9730 ophelp >> ophelp (no symbols) >> 1 3699 0.0270 100.000 oprofiled >> oprofiled (no symbols) >> >> [root@localhost 0]# cat /proc/cpuinfo >> processor : 0 >> vendor_id : GenuineIntel >> cpu family : 6 >> model : 15 >> model name : Intel(R) Core(TM)2 CPU 6400 @ 2.13GHz >> stepping : 2 >> cpu MHz : 1600.000 >> cache size : 2048 KB >> physical id : 0 >> siblings : 2 >> core id : 0 >> cpu cores : 2 >> fpu : yes >> fpu_exception : yes >> cpuid level : 10 >> wp : yes >> flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge >> mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall >> nx lm constant_tsc pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm >> bogomips : 4259.83 >> clflush size : 64 >> cache_alignment : 64 >> address sizes : 36 bits physical, 48 bits virtual >> power management: >> >> processor : 1 >> vendor_id : GenuineIntel >> cpu family : 6 >> model : 15 >> model name : Intel(R) Core(TM)2 CPU 6400 @ 2.13GHz >> stepping : 2 >> cpu MHz : 1600.000 >> cache size : 2048 KB >> physical id : 0 >> siblings : 2 >> core id : 1 >> cpu cores : 2 >> fpu : yes >> fpu_exception : yes >> cpuid level : 10 >> wp : yes >> flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge >> mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall >> nx lm constant_tsc pni monitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm >> bogomips : 4256.01 >> clflush size : 64 >> cache_alignment : 64 >> address sizes : 36 bits physical, 48 bits virtual >> power management: >> >> ------------------------------------------------------------------------------ >> _______________________________________________ >> oprofile-list mailing list >> opr...@li... >> https://lists.sourceforge.net/lists/listinfo/oprofile-list >> > > > > -- > //Monthadar Al Jaberi > |
From: Sam L. <ph...@gm...> - 2009-07-07 08:59:16
|
Hi all, I am on a core2 intel processor (Intel(R) Core(TM)2 CPU 6400 @ 2.13GHz), use CPU_CLK_UNHALTED:100000:0:1:1 event for test with different count value. I think that samples(per second) * count value = constant = 2.13G*2 ( two core). And I run some test, but I found that when the dowcount is bigger than 250000, its okay, match this constant. While it's less then 200000, we get less samples as expected. Why this happened? Also I have append kernel start parameter "idle=poll" to disable halt. 1. get samples: 20431 downcount: 100000 constant: 2043100000 2. get samples: 19613 downcount: 150000 constant: 2941950000 3. get samples: 19124 downcount: 200000 constant: 3824800000 4. get samples: 16882 downcount: 250000 constant: 4220500000 5. get samples: 14116 downcount: 300000 const: 4234800000 6. get samples: 12107 downcount: 350000 const: 4237450000 7 get samples: 10600 downcount: 400000 const: 4240000000 8 get samples: 9426 downcount: 450000 const: 4241700000 9 get samples: 8486 downcount: 500000 const: 4243000000 10 get samples: 7717 downcount: 550000 const: 4244350000 ------------------------------------------------------- Best Regards Sam |