| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-id: <1374146275-5758-1-git-send-email-cw00.choi@samsung.com> Date: Thu, 18 Jul 2013 20:17:52 +0900 From: Chanwoo Choi <cw00.choi@...sung.com> To: viresh.kumar@...aro.org, rjw@...k.pl, linux-kernel@...r.kernel.org Cc: linux-pm@...r.kernel.org, cpufreq@...r.kernel.org, kyungmin.park@...sung.com, myungjoo.ham@...sung.com, Chanwoo Choi <cw00.choi@...sung.com> Subject: [PATCH 0/3 v6] cpufreq: Add 'load_table' debugfs file to show colleced CPUs load This patchset add 'load_table' debugfs file to provide collected CPUs data. The load_table debugfs file gives below CPU datas. - measured time - old CPU frequency - new CPU frequency - Per-CPU load These data will mean the change of CPU frequency according to CPUs load at specific measured time. Also, the user can determine the storage size of colleced CPUs data. The range is from 10 to 1000. The patch3 explain the detailed description of load_table debugfs file. Changes since v5: [patch1/3] - Refactoring patch v4 - Create again symbolic link of debugfs directory when first CPU dev is removed (In this case, many CPUs share only one cpufreq policy) [patch2/3] - Determine index value of policy->cpu_debugfs[] according to cpumask_weight(policy->cpus) value - Bug fix, store 'policy->cpu' to 'freq->cpu' before notify CPUFREQ_LOADCHECK notification [patch3/3] - Add description of test case Changes since v4: [patch2/3] - Reset the data of CPUs load when cpufreq governor is changed - Move code about creating debugfs directory to below first patch : [PATCH 1/2] cpufreq: Add debugfs directory for cpufreq -- I sent only one patch before v4 and then I divided one patch in 3 patches after v4. Changes since v3: - Extend a range of accumulated data (10 ~ 1000) - Add unit information of time/freq and align 'Time' field as left for readability - Use CONFIG_CPU_FREQ_STAT depdendency instead of CONFIG_CPU_FREQ_STAT_DETATILS - Initialize load of offline CPUx as zero(0) - Create/remove debugfs root directory on cpufreq_stats_init/exit() because debugfs root is used on all CPUs. Changes since v2: - Code clean according to Viresh Kumar's comment - Show both old frequency and new frequency on 'load_table' debugfs file - Change debufs file patch as below old: /sys/kernel/debugfs/cpufreq/load_table new: /sys/kernel/debugfs/cpufreq/cpuX/load_table Changes since v1: - Set maximum storage size to save CPUs load on Kconfig - Use spinlock to synchronize read/write operation for CPUs load - Use local variable instead of global variable(struct cpufreq_freqs *freqs) - Use pointer of data structure to get correct size of data structure in sizeof() macro instead of structure name : sizeof(struct cpufreq_freqs) -> sizeof(*stat->load_table) - Change time unit from nanosecond to microsecond - Remove unnecessary memory copy Thanks, Chanwoo Choi [Test Result] - condition : Samsung SoC Exynos4412 (all CPUs share one voltage/frequency) 1. [ONDEMAND governor, the number of online CPU is 4] -sh-4.1# ls -al /sys/kernel/debug/cpufreq/ total 0 drwxr-xr-x 3 root root 0 Jan 1 09:00 . drwx------ 28 root root 0 Jan 1 09:00 .. drwxr-xr-x 2 root root 0 Jan 1 09:00 cpu0 lrwxrwxrwx 1 root root 0 Jan 1 09:00 cpu1 -> ./cpu0 lrwxrwxrwx 1 root root 0 Jan 1 09:00 cpu2 -> ./cpu0 lrwxrwxrwx 1 root root 0 Jan 1 09:00 cpu3 -> ./cpu0 -sh-4.1# cat /sys/kernel/debug/cpufreq/cpu0/load_table Time(ms) Old Freq(Hz) New Freq(Hz) CPU0 CPU1 CPU2 CPU3 135720 400000 400000 56 52 53 91 135820 400000 400000 67 65 83 64 135920 400000 400000 67 66 77 67 136020 400000 1100000 40 41 98 37 136120 1100000 600000 21 24 43 32 136220 600000 600000 29 36 25 87 136320 600000 600000 83 45 44 44 136420 600000 300000 38 27 37 45 136520 300000 1100000 48 45 49 97 136620 1100000 900000 28 38 26 74 2. [ONDEMAND governor, the number of online CPU is 3] -sh-4.1# ls -al /sys/kernel/debug/cpufreq/ total 0 drwxr-xr-x 3 root root 0 Jan 1 09:02 . drwx------ 28 root root 0 Jan 1 09:00 .. drwxr-xr-x 2 root root 0 Jan 1 09:00 cpu0 lrwxrwxrwx 1 root root 0 Jan 1 09:00 cpu1 -> ./cpu0 lrwxrwxrwx 1 root root 0 Jan 1 09:00 cpu3 -> ./cpu0 -sh-4.1# cat /sys/kernel/debug/cpufreq/cpu0/load_table Time(ms) Old Freq(Hz) New Freq(Hz) CPU0 CPU1 CPU3 156720 200000 200000 60 60 60 156820 200000 200000 61 58 60 156930 200000 200000 61 61 59 157020 200000 200000 61 67 61 157120 200000 200000 63 66 63 157230 200000 200000 61 61 60 157320 200000 200000 60 63 61 157420 200000 200000 61 61 60 157520 200000 200000 69 69 69 157620 200000 200000 93 95 94 <s/devices/system/cpu/cpu0/cpufreq/scaling_governor 3. [PERFORMANCE governor, the number of online CPU is 3] -sh-4.1# echo performance > /sys/kernel/debug/cpufreq/cpu0/cpufreq/scaling_governor -sh-4.1# cat /sys/kernel/debug/cpufreq/cpu0/load_table Time(ms) Old Freq(Hz) New Freq(Hz) CPU0 CPU1 CPU3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Chanwoo Choi (3): cpufreq: Add debugfs directory for cpufreq cpufreq: stats: Add 'load_table' debugfs file to show accumulated data of CPUs Documentation: cpufreq: load_table: Update load_table debugfs file documentation Documentation/cpu-freq/cpufreq-stats.txt | 40 ++++- drivers/cpufreq/Kconfig | 6 + drivers/cpufreq/cpufreq.c | 181 ++++++++++++++++++++++ drivers/cpufreq/cpufreq_governor.c | 14 ++ drivers/cpufreq/cpufreq_stats.c | 258 ++++++++++++++++++++++++++++--- include/linux/cpufreq.h | 7 + 6 files changed, 478 insertions(+), 28 deletions(-) -- 1.8.0 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@...r.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists