Re: Why would cpu cgroup get disabled by the kernel?

[Zefan Li <lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>]

On 2016/9/2 17:44, Rosen, Rami wrote:
> Hi, Sverd,
> 
> Could it be that you are using the cgroup_nov1=all kernel option (or include cpu in cgroup_nov1=... bitmask)?
> What does cat /proc/cmdline show ? 
> 

If the kernel is valinna 4.4.19, there's no cgroup_nov1 option, and I think the only
possibility is cgroup_disable=cpu kernel option.

> Regards,
> Rami Rosen
> 
> 
> -----Original Message-----
> From: cgroups-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org [mailto:cgroups-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org] On Behalf Of Sverd Johnsen
> Sent: Friday, September 02, 2016 12:13
> To: cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> Subject: Why would cpu cgroup get disabled by the kernel?
> 
> Hi! I'am on 4.4.19 (arm)  and seem to have all requirements for CPU cgroups build into the kernel:
> 
>  # zgrep -P "CGROUP|GROUP|FAIR" /proc/config.gz CONFIG_CGROUPS=y # CONFIG_CGROUP_DEBUG is not set # CONFIG_CGROUP_FREEZER is not set CONFIG_CGROUP_PIDS=y CONFIG_CGROUP_DEVICE=y CONFIG_CGROUP_CPUACCT=y # CONFIG_CGROUP_PERF is not set CONFIG_CGROUP_SCHED=y CONFIG_FAIR_GROUP_SCHED=y CONFIG_RT_GROUP_SCHED=y CONFIG_BLK_CGROUP=y # CONFIG_DEBUG_BLK_CGROUP is not set # CONFIG_SCHED_AUTOGROUP is not set CONFIG_CFQ_GROUP_IOSCHED=y # CONFIG_NET_CLS_CGROUP is not set # CONFIG_CGROUP_NET_PRIO is not set # CONFIG_CGROUP_NET_CLASSID is not set # CONFIG_THERMAL_DEFAULT_GOV_FAIR_SHARE is not set # CONFIG_THERMAL_GOV_FAIR_SHARE is not set
> 
> Other cgroups also work:
> 
> # findmnt -a
> TARGET                       SOURCE
> FSTYPE     OPTIONS
> |-/sys                       sysfs
> sysfs      rw,nosuid,nodev,noexec,relatime
> | |-/sys/kernel/security     securityfs
> securityfs rw,nosuid,nodev,noexec,relatime
> | |-/sys/fs/cgroup           tmpfs
> tmpfs      ro,nosuid,nodev,noexec,mode=755
> | | |-/sys/fs/cgroup/systemd cgroup
> cgroup     rw,nosuid,nodev,noexec,relatime,xattr,release_agent=/usr/lib/systemd/systemd-cgroups-agent,name=systemd
> | | |-/sys/fs/cgroup/devices cgroup
> cgroup     rw,nosuid,nodev,noexec,relatime,devices
> | | |-/sys/fs/cgroup/blkio   cgroup
> cgroup     rw,nosuid,nodev,noexec,relatime,blkio
> | | |-/sys/fs/cgroup/pids    cgroup
> cgroup     rw,nosuid,nodev,noexec,relatime,pids
> | | `-/sys/fs/cgroup/cpuacct cgroup
> cgroup     rw,nosuid,nodev,noexec,relatime,cpuacct
> 
> However I get this:
> 
> # dmesg | egrep "cgroup|control group"
> [    0.000000] Initializing cgroup subsys cpu
> [    0.000000] Initializing cgroup subsys cpuacct
> [    0.001242] Disabling cpu control group subsystem
> [    0.001263] Initializing cgroup subsys io
> [    0.001289] Initializing cgroup subsys devices
> [    0.001304] Initializing cgroup subsys pids
> 
> # column -t /proc/cgroups
> #subsys_name  hierarchy  num_cgroups  enabled
> cpu           0          1            0
> cpuacct       5          59           1
> blkio         3          9            1
> devices       2          56           1
> pids          4          64           1
> 
> Why? really weird. I control the boot process (systemd 231) and nothing there apparently disables it. When I try to mount cpu cgroup manually I just get ENOENT
> --
> To unsubscribe from this list: send the line "unsubscribe cgroups" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at  http://vger.kernel.org/majordomo-info.html
> \x13��칻\x1c�&�~�&�\x18��+-��ݶ\x17��w��˛���m� ����)���w*\x1fjg���\x1e�����ݢj/���z�ޖ��2�ޙ���&�)ߡ�a��\x7f��\x1e�G���h�\x0f�j:+v���w�٥
>