netdev - Re: [Q] page_pool stats and 32bit architectures.

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <1a779ee7-e776-416d-88f2-3080d33adbb9@huawei.com>
Date: Thu, 13 Feb 2025 19:35:22 +0800
From: Yunsheng Lin <linyunsheng@...wei.com>
To: Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
	<netdev@...r.kernel.org>
CC: Joe Damato <jdamato@...tly.com>, Jesper Dangaard Brouer <hawk@...nel.org>,
	Ilias Apalodimas <ilias.apalodimas@...aro.org>, Thomas Gleixner
	<tglx@...utronix.de>
Subject: Re: [Q] page_pool stats and 32bit architectures.

On 2025/2/13 17:39, Sebastian Andrzej Siewior wrote:
> Hi,
> 
> The statistics counter (page_pool_{alloc|recycle}_stats) use u64 and
> this_cpu_inc() + regular read.
> This is okay on 64bit architectures but on 32bit architectures inc +
> read is split into two operations. So the 32bit arch might observe a
> wrong value if the lower 32bit part of the counter overflows before the
> upper part is updated.
> We can ignore this or switch to u64_stats_t.
> Suggestions?

u64_stats_t seems like the right direction.