Re: [PATCH 1/2] mm: page_alloc: speed up fallbacks in rmqueue_bulk()

[Johannes Weiner]

On Wed, Apr 09, 2025 at 04:02:39PM +0800, Yunsheng Lin wrote:
> On 2025/4/8 2:01, Johannes Weiner wrote:
> > @@ -2934,6 +2981,7 @@ struct page *rmqueue_buddy(struct zone *preferred_zone, struct zone *zone,
> >  {
> >  	struct page *page;
> >  	unsigned long flags;
> > +	enum rmqueue_mode rmqm = RMQUEUE_NORMAL;
> >  
> >  	do {
> >  		page = NULL;
> > @@ -2945,7 +2993,7 @@ struct page *rmqueue_buddy(struct zone *preferred_zone, struct zone *zone,
> >  		if (alloc_flags & ALLOC_HIGHATOMIC)
> >  			page = __rmqueue_smallest(zone, order, MIGRATE_HIGHATOMIC);
> >  		if (!page) {
> > -			page = __rmqueue(zone, order, migratetype, alloc_flags);
> > +			page = __rmqueue(zone, order, migratetype, alloc_flags, &rmqm);
> >  
> >  			/*
> >  			 * If the allocation fails, allow OOM handling and
> 
> It was not in the diff, but it seems the zone->lock is held inside the do..while loop,
> doesn't it mean that the freelists are subject to outside changes and rmqm is stale?

Yes. Note that it only loops when there is a bug/corrupted page, so it
won't make much difference in practice. But it's still kind of weird.

Thanks for your review, Yunsheng!

Andrew, could you please fold the below fixlet?

---