netdev - Re: [syzbot] [afs?] [net?] KCSAN: data-race in rxrpc_peer_keepalive_worker / rxrpc_send_data

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: <2744876.1767882492@warthog.procyon.org.uk>
Date: Thu, 08 Jan 2026 14:28:12 +0000
From: David Howells <dhowells@...hat.com>
To: Eric Dumazet <edumazet@...gle.com>
Cc: dhowells@...hat.com,
    syzbot <syzbot+6182afad5045e6703b3d@...kaller.appspotmail.com>,
    davem@...emloft.net, horms@...nel.org, kuba@...nel.org,
    linux-afs@...ts.infradead.org, linux-kernel@...r.kernel.org,
    marc.dionne@...istor.com, netdev@...r.kernel.org, pabeni@...hat.com,
    syzkaller-bugs@...glegroups.com
Subject: Re: [syzbot] [afs?] [net?] KCSAN: data-race in rxrpc_peer_keepalive_worker / rxrpc_send_data_packet

Eric Dumazet <edumazet@...gle.com> wrote:

> LGTM, but potential load and store tearing should be avoided, using
> READ_ONCE() and WRITE_ONCE().

Fair point.

> last_tx_at being time64_t, this would still be racy on 32bit arches.
> 
> last_tx_at could probably be an "unsigned long" (in jiffies units)...

I've tried avoiding jiffies where possible.  We have way too many different
clocks with different granularities and uses in the kernel, but you might be
right.

David