All of lore.kernel.org
 help / color / mirror / Atom feed
From: Mike Rapoport <rppt@kernel.org>
To: Meijing Zhao <zhaomeijing100@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org,
	Wandun Chen <chenwandun@lixiang.com>,
	Meijing Zhao <zhaomeijing@lixiang.com>
Subject: Re: [PATCH v2 RESEND 3/3] mm: memblock: synchronize debugfs reads with memory hotplug
Date: Tue, 25 Aug 2026 12:35:53 +0300	[thread overview]
Message-ID: <ao1h-Vxca7yPCin2@kernel.org> (raw)
In-Reply-To: <20260821020910.3428585-4-zhaomeijing100@gmail.com>

Hi,

On Fri, Aug 21, 2026 at 10:09:10AM +0800, Meijing Zhao wrote:
> From: Meijing Zhao <zhaomeijing@lixiang.com>
> 
> memblock_debug_show() walks a memblock region array without
> synchronization. With CONFIG_ARCH_KEEP_MEMBLOCK, memory hotplug can
> concurrently add a region. If the array has to grow,
> memblock_double_array() replaces type->regions and frees the old
> allocation while the debugfs reader may still be using it.
> 
> Hold mem_hotplug_lock in read mode while producing the debugfs output.
> Memory hotplug updates already hold the write side of this lock, so the
> region array remains stable throughout the walk. The helpers are no-ops
> when CONFIG_MEMORY_HOTPLUG is disabled.

Hmm, I'm not sure it's the right lock to take here. With MEMORY_HOTPLUG and
ARCH_KEEP_MEMBLOCK there potentially other races in memblock and we'd
better to design proper locking for them.
 
> Fixes: f9126ab9241f ("memory-hotplug: fix wrong edge when hot add a new node")
> Signed-off-by: Meijing Zhao <zhaomeijing@lixiang.com>
> ---
>  mm/memblock.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/mm/memblock.c b/mm/memblock.c
> index 36a8d2a9378d..f1ddbcce47ac 100644
> --- a/mm/memblock.c
> +++ b/mm/memblock.c
> @@ -14,6 +14,7 @@
>  #include <linux/pfn.h>
>  #include <linux/debugfs.h>
>  #include <linux/kmemleak.h>
> +#include <linux/memory_hotplug.h>
>  #include <linux/seq_file.h>
>  #include <linux/memblock.h>
>  #include <linux/mutex.h>
> @@ -2899,6 +2900,7 @@ static int memblock_debug_show(struct seq_file *m, void *private)
>  	phys_addr_t end;
>  	bool first;
>  
> +	get_online_mems();
>  	for (i = 0; i < type->cnt; i++) {
>  		reg = &type->regions[i];
>  		end = reg->base + reg->size - 1;
> @@ -2927,6 +2929,7 @@ static int memblock_debug_show(struct seq_file *m, void *private)
>  			seq_puts(m, "NONE\n");
>  		}
>  	}
> +	put_online_mems();
>  	return 0;
>  }
>  DEFINE_SHOW_ATTRIBUTE(memblock_debug);
> -- 
> 2.25.1
> 

-- 
Sincerely yours,
Mike.


  reply	other threads:[~2026-08-25  9:36 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-21  2:09 [PATCH v2 RESEND 0/3] mm: memblock: fix debugfs flag reporting and synchronization Meijing Zhao
2026-08-21  2:09 ` [PATCH v2 RESEND 1/3] mm: memblock: add missing HugeTLB flag name Meijing Zhao
2026-08-21  2:09 ` [PATCH v2 RESEND 2/3] mm: memblock: show all region flags in debugfs Meijing Zhao
2026-08-24 15:15   ` Mike Rapoport
2026-08-25  1:56     ` Meijing Zhao
2026-08-21  2:09 ` [PATCH v2 RESEND 3/3] mm: memblock: synchronize debugfs reads with memory hotplug Meijing Zhao
2026-08-25  9:35   ` Mike Rapoport [this message]
2026-08-23 14:43 ` [PATCH v2 RESEND 0/3] mm: memblock: fix debugfs flag reporting and synchronization Mike Rapoport

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=ao1h-Vxca7yPCin2@kernel.org \
    --to=rppt@kernel.org \
    --cc=akpm@linux-foundation.org \
    --cc=chenwandun@lixiang.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=zhaomeijing100@gmail.com \
    --cc=zhaomeijing@lixiang.com \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.