netdev - Re: [PATCH v20 net-next 1/6] sched: Struct definition and parsing of dualpi2 qdisc

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: <20250701163053.22113645@kernel.org>
Date: Tue, 1 Jul 2025 16:30:53 -0700
From: Jakub Kicinski <kuba@...nel.org>
To: "Chia-Yu Chang (Nokia)" <chia-yu.chang@...ia-bell-labs.com>
Cc: "alok.a.tiwari@...cle.com" <alok.a.tiwari@...cle.com>,
 "pctammela@...atatu.com" <pctammela@...atatu.com>, "horms@...nel.org"
 <horms@...nel.org>, "donald.hunter@...il.com" <donald.hunter@...il.com>,
 "xandfury@...il.com" <xandfury@...il.com>, "netdev@...r.kernel.org"
 <netdev@...r.kernel.org>, "dave.taht@...il.com" <dave.taht@...il.com>,
 "pabeni@...hat.com" <pabeni@...hat.com>, "jhs@...atatu.com"
 <jhs@...atatu.com>, "stephen@...workplumber.org"
 <stephen@...workplumber.org>, "xiyou.wangcong@...il.com"
 <xiyou.wangcong@...il.com>, "jiri@...nulli.us" <jiri@...nulli.us>,
 "davem@...emloft.net" <davem@...emloft.net>, "edumazet@...gle.com"
 <edumazet@...gle.com>, "andrew+netdev@...n.ch" <andrew+netdev@...n.ch>,
 "ast@...erby.net" <ast@...erby.net>, "liuhangbin@...il.com"
 <liuhangbin@...il.com>, "shuah@...nel.org" <shuah@...nel.org>,
 "linux-kselftest@...r.kernel.org" <linux-kselftest@...r.kernel.org>,
 "ij@...nel.org" <ij@...nel.org>, "ncardwell@...gle.com"
 <ncardwell@...gle.com>, "Koen De Schepper (Nokia)"
 <koen.de_schepper@...ia-bell-labs.com>, "g.white@...lelabs.com"
 <g.white@...lelabs.com>, "ingemar.s.johansson@...csson.com"
 <ingemar.s.johansson@...csson.com>, "mirja.kuehlewind@...csson.com"
 <mirja.kuehlewind@...csson.com>, "cheshire@...le.com" <cheshire@...le.com>,
 "rs.ietf@....at" <rs.ietf@....at>, "Jason_Livingood@...cast.com"
 <Jason_Livingood@...cast.com>, "vidhi_goel@...le.com"
 <vidhi_goel@...le.com>
Subject: Re: [PATCH v20 net-next 1/6] sched: Struct definition and parsing
 of dualpi2 qdisc

On Tue, 1 Jul 2025 05:41:53 +0000 Chia-Yu Chang (Nokia) wrote:
> > > I don't see we access the same atomic variable multiple times in a 
> > > single expression, the 2 WRITE_ONCE() are in different expressions.
> > >
> > > And, in the last WRITE_ONCE(), what we access are local variables:
> > > "step_pkt" "step_th", will it create problem?  
> > 
> > Not really a problem, but what I'm saying is that I don't understand why all the writes are sprinkled with WRITE_ONCE(). You take
> > 
> >         sch_tree_lock(sch);
> > 
> > to block data path and the control path is under rtnl_lock. So why the WRITE_ONCE()? WRITE_ONCE() is used to annotate writes which can be read concurrently without holding relevant locks.  
> 
> This follows series https://lore.kernel.org/netdev/20240415132054.3822230-1-edumazet@google.com/ to use READ_ONCE in dual2_dump() and WRITE_ONCE in dualpi2_change.
> 
> And because of one previous chnage request (https://lore.kernel.org/all/26f2f366-aa14-4879-978a-58d46f9d83a4@redhat.com/) to split single dualpi2.c into 3 patches, so you can find them in the 2nd patch of this series.
> 
> What I can do is either move these WRITE_ONCE also to the next patch or merge these patches.
> 
> I would prefer the first approach to make it still 3 patches, any other suggestion?

I see. With the change to use two separate fields for packets and time 
I think this should be fine. We will hopefully report one or the other,
but not one misinterpreted as the other.