linux-kernel - Re: [PATCH 01/13] seqnum_ops: Introduce Sequence Number Ops

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

Date:   Wed, 11 Nov 2020 08:56:49 -0700
From:   Shuah Khan <skhan@...uxfoundation.org>
To:     Peter Zijlstra <peterz@...radead.org>
Cc:     corbet@....net, keescook@...omium.org, gregkh@...uxfoundation.org,
        linux-doc@...r.kernel.org, linux-kernel@...r.kernel.org,
        skhan@...uxfoundation.org
Subject: Re: [PATCH 01/13] seqnum_ops: Introduce Sequence Number Ops

On 11/11/20 1:23 AM, Peter Zijlstra wrote:
> On Tue, Nov 10, 2020 at 12:53:27PM -0700, Shuah Khan wrote:
> 
>> + * The interface provides:
>> + * seqnum32 & seqnum64 functions:
>> + *	initialization
>> + *	set
>> + *	read
>> + *	increment and no return
>> + *	decrement and no return
> 
> NAK, this is batshit insane again.

Gosh that is a bit much. Definitely will never be part of my kernel
review/response vocabulary.

If you want a sequence number, the
> one and _ONLY_ primitive you want to expose is inc_return.
> 
> No set, no read, no inc, and most certainly, not dec.
> 

Agree with you on removing dec(). It isn't needed or up counting.
set() can go and use just init instead of set.

read and inc are needed for sure though. The reason being numbers
could be incremented in one place and read in other places. In some
cases inc_return is used.

Why would you say no to read and inc?

init, read, inc, and inc_return are necessary to be able to implement
up counters.

thanks,
-- Shuah