Re: [PATCH 2/9] perf/core: Add PERF_SAMPLE_CGROUP feature

[Peter Zijlstra]

On Fri, Dec 20, 2019 at 01:32:46PM +0900, Namhyung Kim wrote:
> The PERF_SAMPLE_CGROUP bit is to save (perf_event) cgroup information
> in the sample.  It will add a 64-bit id to identify current cgroup and
> it's the file handle in the cgroup file system.  Userspace should use
> this information with PERF_RECORD_CGROUP event to match which cgroup
> it belongs.
> 
> I put it before PERF_SAMPLE_AUX for simplicity since it just needs a
> 64-bit word.  But if we want bigger samples, I can work on that
> direction too.
> 
> Cc: Tejun Heo <tj@xxxxxxxxxx>
> Cc: Li Zefan <lizefan@xxxxxxxxxx>
> Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
> Signed-off-by: Namhyung Kim <namhyung@xxxxxxxxxx>
> ---

> @@ -6895,6 +6901,18 @@ void perf_prepare_sample(struct perf_event_header *header,
>  	if (sample_type & PERF_SAMPLE_PHYS_ADDR)
>  		data->phys_addr = perf_virt_to_phys(data->addr);
>  
> +	if (sample_type & PERF_SAMPLE_CGROUP) {
> +		u64 cgrp_id = 0;
> +#ifdef CONFIG_CGROUP_PERF
> +		struct cgroup *cgrp;
> +
> +		/* protected by RCU */
> +		cgrp = task_css_check(current, perf_event_cgrp_id, 1)->cgroup;
> +		cgrp_id = cgroup_id(cgrp);
> +#endif
> +		data->cgroup = cgrp_id;
> +	}

Would it make more sense to refuse SAMPLE_CGROUP if !CGROUP_PERF?