netdev - Re: [PATCH net] ethtool: Fix context creation with no parameters

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]

Message-ID: <20240807072137.34d300a8@kernel.org>
Date: Wed, 7 Aug 2024 07:21:37 -0700
From: Jakub Kicinski <kuba@...nel.org>
To: Gal Pressman <gal@...dia.com>
Cc: "David S. Miller" <davem@...emloft.net>, <netdev@...r.kernel.org>, Eric
 Dumazet <edumazet@...gle.com>, Paolo Abeni <pabeni@...hat.com>, Dragos
 Tatulea <dtatulea@...dia.com>, Jianbo Liu <jianbol@...dia.com>
Subject: Re: [PATCH net] ethtool: Fix context creation with no parameters

At a glance I think you popped it into the wrong place.

On Wed, 7 Aug 2024 16:25:41 +0300 Gal Pressman wrote:
> -	if ((rxfh.indir_size &&
> +	if (!create && ((rxfh.indir_size &&
>  	     rxfh.indir_size != ETH_RXFH_INDIR_NO_CHANGE &&
>  	     rxfh.indir_size != dev_indir_size) ||

This condition just checks if indir_size matches the device
expectations, is reset or is zero. Even if we're creating the
context, at present, the indir table size is fixed for the device.

>  	    (rxfh.key_size && (rxfh.key_size != dev_key_size)) ||

similarly this checks the key size

>  	    (rxfh.indir_size == ETH_RXFH_INDIR_NO_CHANGE &&
>  	     rxfh.key_size == 0 && rxfh.hfunc == ETH_RSS_HASH_NO_CHANGE &&
> -	     rxfh.input_xfrm == RXH_XFRM_NO_CHANGE))
> +	     rxfh.input_xfrm == RXH_XFRM_NO_CHANGE)))

only this validates the "is this a nop", so you gotta add the &&
!create here

That's why I (perhaps not very clearly) suggested that we should split
this if into two. Cause the first two conditions check "sizes" while
the last chunk checks for "nop". And readability suffers.

Feel free to send the new version tomorrow without a full 24h wait.