Re: [PATCH v4] sched/fair: Fix CPU bandwidth limit bypass during CPU hotplug

[samir]

On 2024-12-12 10:01, Vishal Chourasia wrote:
> CPU controller limits are not properly enforced during CPU hotplug
> operations, particularly during CPU offline. When a CPU goes offline,
> throttled processes are unintentionally being unthrottled across all 
> CPUs
> in the system, allowing them to exceed their assigned quota limits.
>
> Consider below for an example,
>
> Assigning 6.25% bandwidth limit to a cgroup
> in a 8 CPU system, where, workload is running 8 threads for 20 seconds 
> at
> 100% CPU utilization, expected (user+sys) time = 10 seconds.
>
> $ cat /sys/fs/cgroup/test/cpu.max
> 50000 100000
>
> $ ./ebizzy -t 8 -S 20        // non-hotplug case
> real 20.00 s
> user 10.81 s                 // intended behaviour
> sys   0.00 s
>
> $ ./ebizzy -t 8 -S 20        // hotplug case
> real 20.00 s
> user 14.43 s                 // Workload is able to run for 14 secs
> sys   0.00 s                 // when it should have only run for 10 
> secs
>
> During CPU hotplug, scheduler domains are rebuilt and cpu_attach_domain
> is called for every active CPU to update the root domain. That ends up
> calling rq_offline_fair which un-throttles any throttled hierarchies.
>
> Unthrottling should only occur for the CPU being hotplugged to allow 
> its
> throttled processes to become runnable and get migrated to other CPUs.
>
> With current patch applied,
> $ ./ebizzy -t 8 -S 20        // hotplug case
> real 21.00 s
> user 10.16 s                 // intended behaviour
> sys   0.00 s
>
> This also has another symptom, when a CPU goes offline, and if the 
> cfs_rq
> is not in throttled state and the runtime_remaining still had plenty
> remaining, it gets reset to 1 here, causing the runtime_remaining of
> cfs_rq to be quickly depleted.
>
> Note: hotplug operation (online, offline) was performed in while(1) 
> loop
>
> Signed-off-by: Vishal Chourasia <vishalc@xxxxxxxxxxxxx>
> Suggested-by: Zhang Qiao <zhangqiao22@xxxxxxxxxx>
> Tested-by: Madadi Vineeth Reddy <vineethr@xxxxxxxxxxxxx>
>
> v3: 
> https://lore.kernel.org/all/20241210102346.228663-2-vishalc@xxxxxxxxxxxxx
> v2: 
> https://lore.kernel.org/all/20241207052730.1746380-2-vishalc@xxxxxxxxxxxxx
> v1: 
> https://lore.kernel.org/all/20241126064812.809903-2-vishalc@xxxxxxxxxxxxx
>
> ---
>  kernel/sched/fair.c | 20 +++++++++++++-------
>  1 file changed, 13 insertions(+), 7 deletions(-)
>
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index aa0238ee4857..72746e75700c 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -6679,6 +6679,10 @@ static void __maybe_unused
> unthrottle_offline_cfs_rqs(struct rq *rq)
>
>  	lockdep_assert_rq_held(rq);
>
> +	// Do not unthrottle for an active CPU
> +	if (cpumask_test_cpu(cpu_of(rq), cpu_active_mask))
> +		return;
> +
>  	/*
>  	 * The rq clock has already been updated in the
>  	 * set_rq_offline(), so we should skip updating
> @@ -6693,19 +6697,21 @@ static void __maybe_unused
> unthrottle_offline_cfs_rqs(struct rq *rq)
>  		if (!cfs_rq->runtime_enabled)
>  			continue;
>
> -		/*
> -		 * clock_task is not advancing so we just need to make sure
> -		 * there's some valid quota amount
> -		 */
> -		cfs_rq->runtime_remaining = 1;
>  		/*
>  		 * Offline rq is schedulable till CPU is completely disabled
>  		 * in take_cpu_down(), so we prevent new cfs throttling here.
>  		 */
>  		cfs_rq->runtime_enabled = 0;
>
> -		if (cfs_rq_throttled(cfs_rq))
> -			unthrottle_cfs_rq(cfs_rq);
> +		if (!cfs_rq_throttled(cfs_rq))
> +			continue;
> +
> +		/*
> +		 * clock_task is not advancing so we just need to make sure
> +		 * there's some valid quota amount
> +		 */
> +		cfs_rq->runtime_remaining = 1;
> +		unthrottle_cfs_rq(cfs_rq);
>  	}
>  	rcu_read_unlock();
>
>
> base-commit: 231825b2e1ff6ba799c5eaf396d3ab2354e37c6b

Hello,

I have verified this issue using the ebizzy workload and a Podman 
container. The tests confirm that the problem is resolved, and the 
provided fix is working as expected. Below are the results for 
reference, where the ebizzy workload was executed within the container 
with --cpu-quota=50000 allocated.
Additionally, I tested the patch under load conditions both with and 
without hot-plug operations. Observations are as follows:

Test Results
Without Hot-Plug Operation:
Command: ./ebizzy -t 64 -S 20
	Performance: 43,506 records/s
	Execution Time:
      	Real: 20.00 s
	User: 10.46 s
	Sys: 0.00 s
With Hot-Plug Operation:
Command: ./ebizzy -t 64 -S 20
	Performance: 35,642 records/s
	Execution Time:
	Real: 20.00 s
	User: 10.45 s
	Sys: 0.01 s

Tested-by: Samir Mulani <samir@xxxxxxxxxxxxx>

Thanks for the fix!