linux-kernel - RE: [RFC net-next] net: ena: Use u64_stats_t with u64_stats

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>] [day] [month] [year] [list]

Message-ID: <FFDD6CD9-5100-49E6-B3B8-F38099043E87@amazon.com>
Date: Tue, 27 Jan 2026 06:40:34 +0000
From: "Arinzon, David" <darinzon@...zon.com>
To: "mmyangfl@...il.com" <mmyangfl@...il.com>, "netdev@...r.kernel.org"
	<netdev@...r.kernel.org>
CC: "mmyangfl@...il.com" <mmyangfl@...il.com>, "Allen, Neil"
	<shayagr@...zon.com>, "Kiyanovski, Arthur" <akiyano@...zon.com>, "Bshara,
 Saeed" <saeedb@...zon.com>, "andrew+netdev@...n.ch" <andrew+netdev@...n.ch>,
	"davem@...emloft.net" <davem@...emloft.net>, "edumazet@...gle.com"
	<edumazet@...gle.com>, "kuba@...nel.org" <kuba@...nel.org>,
	"pabeni@...hat.com" <pabeni@...hat.com>, "ast@...nel.org" <ast@...nel.org>,
	"daniel@...earbox.net" <daniel@...earbox.net>, "hawk@...nel.org"
	<hawk@...nel.org>, "john.fastabend@...il.com" <john.fastabend@...il.com>,
	"sdf@...ichev.me" <sdf@...ichev.me>, "Bernstein, Amit" <amitbern@...zon.com>,
	"horms@...nel.org" <horms@...nel.org>, "bhelgaas@...gle.com"
	<bhelgaas@...gle.com>, "leitao@...ian.org" <leitao@...ian.org>, "Enju, Kohei"
	<enjuk@...zon.co.jp>, "ahmed.zaki@...el.com" <ahmed.zaki@...el.com>,
	"mingo@...nel.org" <mingo@...nel.org>, "tglx@...nel.org" <tglx@...nel.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	"bpf@...r.kernel.org" <bpf@...r.kernel.org>
Subject: RE: [RFC net-next] net: ena: Use u64_stats_t with u64_stats_sync
 properly

Hi David,

Thank you for submitting the patch.

> On 64bit arches, struct u64_stats_sync is empty and provides no help
> against load/store tearing. Convert to u64_stats_t to ensure atomic
> operations.
> 

We would like to run some quick performance tests internally before signing off.

> Signed-off-by: David Yang <mmyangfl@...il.com>
> ---
> RFC Comment:
> 
> The write side of u64_stats should ensure mutual exclusion, however I couldn't
> find the synchronization mechanism in use (for example, ena_up / ena_io_poll /
> ena_start_xmit). Should this be considered an issue?

We reviewed other drivers and it looks like none add extra synchronization on the write side.
Adding extra synchronization would degrade datapath performance to fix extremely rare missed counts.

You've submitted similar patches for other drivers but only commented on write-side
synchronization for ENA. Is there something specific to ENA that differs from other drivers?