mirror of
https://github.com/opencontainers/runc.git
synced 2026-07-11 06:03:57 +08:00
Cédric Clerget
afc23e3397
This handles a corner case when joining a container having all the processes running exclusively on isolated CPU cores to force the kernel to schedule runc process on the first CPU core within the cgroups cpuset. The introduction of the kernel commit 46a87b3851f0d6eb05e6d83d5c5a30df0eca8f76 has affected this deterministic scheduling behavior by distributing tasks across CPU cores within the cgroups cpuset. Some intensive real-time application are relying on this deterministic behavior and use the first CPU core to run a slow thread while other CPU cores are fully used by real-time threads with SCHED_FIFO policy. Such applications prevents runc process from joining a container when the runc process is randomly scheduled on a CPU core owned by a real-time thread. Introduces isolated CPU affinity transition OCI runtime annotation org.opencontainers.runc.exec.isolated-cpu-affinity-transition to restore the behavior during runc exec. Fix issue with kernel >= 6.2 not resetting CPU affinity for container processes. Signed-off-by: Cédric Clerget <cedric.clerget@gmail.com>
79 lines
2.5 KiB
Go
79 lines
2.5 KiB
Go
package cgroups
|
|
|
|
import (
|
|
"errors"
|
|
|
|
"github.com/opencontainers/runc/libcontainer/configs"
|
|
)
|
|
|
|
var (
|
|
// ErrDevicesUnsupported is an error returned when a cgroup manager
|
|
// is not configured to set device rules.
|
|
ErrDevicesUnsupported = errors.New("cgroup manager is not configured to set device rules")
|
|
|
|
// DevicesSetV1 and DevicesSetV2 are functions to set devices for
|
|
// cgroup v1 and v2, respectively. Unless libcontainer/cgroups/devices
|
|
// package is imported, it is set to nil, so cgroup managers can't
|
|
// manage devices.
|
|
DevicesSetV1 func(path string, r *configs.Resources) error
|
|
DevicesSetV2 func(path string, r *configs.Resources) error
|
|
)
|
|
|
|
type Manager interface {
|
|
// Apply creates a cgroup, if not yet created, and adds a process
|
|
// with the specified pid into that cgroup. A special value of -1
|
|
// can be used to merely create a cgroup.
|
|
Apply(pid int) error
|
|
|
|
// GetPids returns the PIDs of all processes inside the cgroup.
|
|
GetPids() ([]int, error)
|
|
|
|
// GetAllPids returns the PIDs of all processes inside the cgroup
|
|
// any all its sub-cgroups.
|
|
GetAllPids() ([]int, error)
|
|
|
|
// GetStats returns cgroups statistics.
|
|
GetStats() (*Stats, error)
|
|
|
|
// Freeze sets the freezer cgroup to the specified state.
|
|
Freeze(state configs.FreezerState) error
|
|
|
|
// Destroy removes cgroup.
|
|
Destroy() error
|
|
|
|
// Path returns a cgroup path to the specified controller/subsystem.
|
|
// For cgroupv2, the argument is unused and can be empty.
|
|
Path(string) string
|
|
|
|
// Set sets cgroup resources parameters/limits. If the argument is nil,
|
|
// the resources specified during Manager creation (or the previous call
|
|
// to Set) are used.
|
|
Set(r *configs.Resources) error
|
|
|
|
// GetPaths returns cgroup path(s) to save in a state file in order to
|
|
// restore later.
|
|
//
|
|
// For cgroup v1, a key is cgroup subsystem name, and the value is the
|
|
// path to the cgroup for this subsystem.
|
|
//
|
|
// For cgroup v2 unified hierarchy, a key is "", and the value is the
|
|
// unified path.
|
|
GetPaths() map[string]string
|
|
|
|
// GetCgroups returns the cgroup data as configured.
|
|
GetCgroups() (*configs.Cgroup, error)
|
|
|
|
// GetFreezerState retrieves the current FreezerState of the cgroup.
|
|
GetFreezerState() (configs.FreezerState, error)
|
|
|
|
// Exists returns whether the cgroup path exists or not.
|
|
Exists() bool
|
|
|
|
// OOMKillCount reports OOM kill count for the cgroup.
|
|
OOMKillCount() (uint64, error)
|
|
|
|
// GetEffectiveCPUs returns the effective CPUs of the cgroup, an empty
|
|
// value means that the cgroups cpuset subsystem/controller is not enabled.
|
|
GetEffectiveCPUs() string
|
|
}
|