LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: David Hildenbrand <david@redhat.com>
To: Shile Zhang <shile.zhang@linux.alibaba.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Kirill Tkhai <ktkhai@virtuozzo.com>,
	Pavel Tatashin <pasha.tatashin@soleen.com>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v3] mm: fix tick timer stall during deferred page init
Date: Wed, 1 Apr 2020 16:26:06 +0200	[thread overview]
Message-ID: <1ca342ec-23e7-7884-5758-2b14675266e4@redhat.com> (raw)
In-Reply-To: <20200311123848.118638-1-shile.zhang@linux.alibaba.com>

On 11.03.20 13:38, Shile Zhang wrote:
> When 'CONFIG_DEFERRED_STRUCT_PAGE_INIT' is set, 'pgdatinit' kthread will
> initialise the deferred pages with local interrupts disabled. It is
> introduced by commit 3a2d7fa8a3d5 ("mm: disable interrupts while
> initializing deferred pages").
> 
> On machine with NCPUS <= 2, the 'pgdatinit' kthread could be bound to
> the boot CPU, which could caused the tick timer long time stall, system
> jiffies not be updated in time.
> 
> The dmesg shown that:
> 
>     [    0.197975] node 0 initialised, 32170688 pages in 1ms
> 
> Obviously, 1ms is unreasonable.
> 
> Now, fix it by restore in the pending interrupts for every 32*1204 pages
> (128MB) initialized, give the chance to update the systemd jiffies.
> The reasonable demsg shown likes:
> 
>     [    1.069306] node 0 initialised, 32203456 pages in 894ms
> 
> Fixes: 3a2d7fa8a3d5 ("mm: disable interrupts while initializing deferred pages").
> 
> Co-developed-by: Kirill Tkhai <ktkhai@virtuozzo.com>
> Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
> Signed-off-by: Shile Zhang <shile.zhang@linux.alibaba.com>
> ---
>  mm/page_alloc.c | 25 ++++++++++++++++++++++---
>  1 file changed, 22 insertions(+), 3 deletions(-)
> 
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 3c4eb750a199..a3a47845e150 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -1763,12 +1763,17 @@ deferred_init_maxorder(u64 *i, struct zone *zone, unsigned long *start_pfn,
>  	return nr_pages;
>  }
>  
> +/*
> + * Release the pending interrupts for every TICK_PAGE_COUNT pages.
> + */
> +#define TICK_PAGE_COUNT	(32 * 1024)
> +
>  /* Initialise remaining memory on a node */
>  static int __init deferred_init_memmap(void *data)
>  {
>  	pg_data_t *pgdat = data;
>  	const struct cpumask *cpumask = cpumask_of_node(pgdat->node_id);
> -	unsigned long spfn = 0, epfn = 0, nr_pages = 0;
> +	unsigned long spfn = 0, epfn = 0, nr_pages = 0, prev_nr_pages = 0;
>  	unsigned long first_init_pfn, flags;
>  	unsigned long start = jiffies;
>  	struct zone *zone;
> @@ -1779,6 +1784,7 @@ static int __init deferred_init_memmap(void *data)
>  	if (!cpumask_empty(cpumask))
>  		set_cpus_allowed_ptr(current, cpumask);
>  
> +again:
>  	pgdat_resize_lock(pgdat, &flags);
>  	first_init_pfn = pgdat->first_deferred_pfn;
>  	if (first_init_pfn == ULONG_MAX) {
> @@ -1790,7 +1796,6 @@ static int __init deferred_init_memmap(void *data)
>  	/* Sanity check boundaries */
>  	BUG_ON(pgdat->first_deferred_pfn < pgdat->node_start_pfn);
>  	BUG_ON(pgdat->first_deferred_pfn > pgdat_end_pfn(pgdat));
> -	pgdat->first_deferred_pfn = ULONG_MAX;
>  
>  	/* Only the highest zone is deferred so find it */
>  	for (zid = 0; zid < MAX_NR_ZONES; zid++) {
> @@ -1809,9 +1814,23 @@ static int __init deferred_init_memmap(void *data)
>  	 * that we can avoid introducing any issues with the buddy
>  	 * allocator.
>  	 */
> -	while (spfn < epfn)
> +	while (spfn < epfn) {
>  		nr_pages += deferred_init_maxorder(&i, zone, &spfn, &epfn);
> +		/*
> +		 * Release the interrupts for every TICK_PAGE_COUNT pages
> +		 * (128MB) to give tick timer the chance to update the

Nit: 128MB is only true for 4k pages.

> +		 * system jiffies.
> +		 */
> +		if ((nr_pages - prev_nr_pages) > TICK_PAGE_COUNT) {
> +			prev_nr_pages = nr_pages;
> +			pgdat->first_deferred_pfn = spfn;
> +			pgdat_resize_unlock(pgdat, &flags);
> +			goto again;
> +		}
> +	}
> +

I find that deferred page init code horribly complicated to understand,
but that's a different story.

Your change looks sane to me and survives my basic testing. Thanks!

Acked-by: David Hildenbrand <david@redhat.com>

-- 
Thanks,

David / dhildenb


  parent reply	other threads:[~2020-04-01 14:26 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-03-11 12:38 Shile Zhang
2020-03-11 17:45 ` Pavel Tatashin
2020-03-11 20:16   ` Kirill Tkhai
2020-03-11 20:25     ` Pavel Tatashin
2020-03-19 19:05 ` Daniel Jordan
2020-03-26 18:58   ` Daniel Jordan
2020-03-26 19:36     ` Pavel Tatashin
2020-03-27  4:39       ` Shile Zhang
2020-03-28  0:17         ` Daniel Jordan
2020-04-01 14:26 ` David Hildenbrand [this message]
2020-04-01 15:42 ` Michal Hocko
2020-04-01 15:50   ` David Hildenbrand
2020-04-01 16:00     ` Michal Hocko
2020-04-01 16:09       ` Daniel Jordan
2020-04-01 16:12         ` Michal Hocko
2020-04-01 16:15           ` David Hildenbrand
2020-04-01 16:18           ` Daniel Jordan
2020-04-01 16:26             ` Michal Hocko
2020-04-01 16:41               ` Pavel Tatashin
2020-04-01 16:46                 ` Michal Hocko
2020-04-01 16:51                   ` Pavel Tatashin
2020-04-01 17:13                     ` Daniel Jordan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1ca342ec-23e7-7884-5758-2b14675266e4@redhat.com \
    --to=david@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=ktkhai@virtuozzo.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=pasha.tatashin@soleen.com \
    --cc=shile.zhang@linux.alibaba.com \
    --subject='Re: [PATCH v3] mm: fix tick timer stall during deferred page init' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).