Re: [PATCH v3 3/7] padata: dispatch works on different nodes

[Tim Chen <tim.c.chen@linux.intel.com>]

On Mon, 2024-01-15 at 16:57 +0800, Gang Li wrote:
> 
> On 2024/1/13 02:27, Tim Chen wrote:
> > On Fri, 2024-01-12 at 15:09 +0800, Gang Li wrote:
> > > On 2024/1/12 01:50, Tim Chen wrote:
> > > > On Tue, 2024-01-02 at 21:12 +0800, Gang Li wrote:
> > > > > When a group of tasks that access different nodes are scheduled on the
> > > > > same node, they may encounter bandwidth bottlenecks and access latency.
> > > > > 
> > > > > Thus, numa_aware flag is introduced here, allowing tasks to be
> > > > > distributed across different nodes to fully utilize the advantage of
> > > > > multi-node systems.
> > > > > 
> > > > > Signed-off-by: Gang Li <gang.li@linux.dev>
> > > > > ---
> > > > >    include/linux/padata.h | 3 +++
> > > > >    kernel/padata.c        | 8 ++++++--
> > > > >    mm/mm_init.c           | 1 +
> > > > >    3 files changed, 10 insertions(+), 2 deletions(-)
> > > > > 
> > > > > diff --git a/include/linux/padata.h b/include/linux/padata.h
> > > > > index 495b16b6b4d72..f79ccd50e7f40 100644
> > > > > --- a/include/linux/padata.h
> > > > > +++ b/include/linux/padata.h
> > > > > @@ -137,6 +137,8 @@ struct padata_shell {
> > > > >     *             appropriate for one worker thread to do at once.
> > > > >     * @max_threads: Max threads to use for the job, actual number may be less
> > > > >     *               depending on task size and minimum chunk size.
> > > > > + * @numa_aware: Dispatch jobs to different nodes. If a node only has memory but
> > > > > + *              no CPU, dispatch its jobs to a random CPU.
> > > > >     */
> > > > >    struct padata_mt_job {
> > > > >    	void (*thread_fn)(unsigned long start, unsigned long end, void *arg);
> > > > > @@ -146,6 +148,7 @@ struct padata_mt_job {
> > > > >    	unsigned long		align;
> > > > >    	unsigned long		min_chunk;
> > > > >    	int			max_threads;
> > > > > +	bool			numa_aware;
> > > > >    };
> > > > >    
> > > > >    /**
> > > > > diff --git a/kernel/padata.c b/kernel/padata.c
> > > > > index 179fb1518070c..1c2b3a337479e 100644
> > > > > --- a/kernel/padata.c
> > > > > +++ b/kernel/padata.c
> > > > > @@ -485,7 +485,7 @@ void __init padata_do_multithreaded(struct padata_mt_job *job)
> > > > >    	struct padata_work my_work, *pw;
> > > > >    	struct padata_mt_job_state ps;
> > > > >    	LIST_HEAD(works);
> > > > > -	int nworks;
> > > > > +	int nworks, nid = 0;
> > > > 
> > > > If we always start from 0, we may be biased towards the low numbered node,
> > > > and not use high numbered nodes at all.  Suggest you do
> > > > static nid = 0;
> > > > 
> > > 
> > > When we use `static`, if there are multiple parallel calls to
> > > `padata_do_multithreaded`, it may result in an uneven distribution of
> > > tasks for each padata_do_multithreaded.
> > > 
> > > We can make the following modifications to address this issue.
> > > 
> > > ```
> > > diff --git a/kernel/padata.c b/kernel/padata.c
> > > index 1c2b3a337479e..925e48df6dd8d 100644
> > > --- a/kernel/padata.c
> > > +++ b/kernel/padata.c
> > > @@ -485,7 +485,8 @@ void __init padata_do_multithreaded(struct
> > > padata_mt_job *job)
> > >           struct padata_work my_work, *pw;
> > >           struct padata_mt_job_state ps;
> > >           LIST_HEAD(works);
> > > -       int nworks, nid = 0;
> > > +       int nworks, nid;
> > > +       static volatile int global_nid = 0;
> > > 
> > >           if (job->size == 0)
> > >                   return;
> > > @@ -516,12 +517,15 @@ void __init padata_do_multithreaded(struct
> > > padata_mt_job *job)
> > >           ps.chunk_size = max(ps.chunk_size, job->min_chunk);
> > >           ps.chunk_size = roundup(ps.chunk_size, job->align);
> > > 
> > > +       nid = global_nid;
> > >           list_for_each_entry(pw, &works, pw_list)
> > > -               if (job->numa_aware)
> > > -                       queue_work_node((++nid % num_node_state(N_MEMORY)),
> > > -                                       system_unbound_wq, &pw->pw_work);
> > > -               else
> > > +               if (job->numa_aware) {
> > > +                       queue_work_node(nid, system_unbound_wq,
> > > &pw->pw_work);
> > > +                       nid = next_node(nid, node_states[N_CPU]);
> > > +               } else
> > >                           queue_work(system_unbound_wq, &pw->pw_work);
> > > +       if (job->numa_aware)
> > > +               global_nid = nid;
> > 
> > Thinking more about it, there could still be multiple threads working
> > at the same time with stale global_nid.  We should probably do a compare
> > exchange of global_nid with new nid only if the global nid was unchanged.
> > Otherwise we should go to the next node with the changed global nid before
> > we queue the job.
> > 
> > Tim
> > 
> How about:
> ```
> nid = global_nid;
> list_for_each_entry(pw, &works, pw_list)
> 	if (job->numa_aware) {
> 		int old_node = nid;
> 		queue_work_node(nid, system_unbound_wq, &pw->pw_work);
> 		nid = next_node(nid, node_states[N_CPU]);
> 		cmpxchg(&global_nid, old_node, nid);
> 	} else
> 		queue_work(system_unbound_wq, &pw->pw_work);
> 
> ```
> 

I am thinking something like

static volatile atomic_t last_used_nid;

list_for_each_entry(pw, &works, pw_list)
 	if (job->numa_aware) {
		int old_node = atomic_read(&last_used_nid);
		
		do {
			nid = next_node_in(old_node, node_states[N_CPU]);
		} while (!atomic_try_cmpxchg(&last_used_nid, &old_node, nid));
 		queue_work_node(nid, system_unbound_wq, &pw->pw_work);		
 	} else {
 		queue_work(system_unbound_wq, &pw->pw_work);
	}

Note that we need to use next_node_in so we'll wrap around the node mask.

Tim