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  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]
Date:   Mon, 3 Aug 2020 14:54:35 -0700
From:   Jakub Kicinski <kuba@...nel.org>
To:     Edward Cree <ecree@...arflare.com>
Cc:     <linux-net-drivers@...arflare.com>, <davem@...emloft.net>,
        <netdev@...r.kernel.org>
Subject: Re: [PATCH v2 net-next 03/11] sfc_ef100: read Design Parameters at
 probe time

On Mon, 3 Aug 2020 15:33:39 +0100 Edward Cree wrote:
> On 31/07/2020 21:18, Jakub Kicinski wrote:
> > On Fri, 31 Jul 2020 13:58:35 +0100 Edward Cree wrote:  
> >> +	default:
> >> +		/* Host interface says "Drivers should ignore design parameters
> >> +		 * that they do not recognise."
> >> +		 */
> >> +		netif_info(efx, probe, efx->net_dev,
> >> +			   "Ignoring unrecognised design parameter %u\n",
> >> +			   reader->type);  
> > 
> > Is this really important enough to spam the logs with?  
>
> Well, it implies your NIC (FPGA image) is newer than your driver,
>  and saying things the driver doesn't understand; I feel like that
>  should be recorded somewhere.

There are scenarios in which the driver may legitimately be older
 - bootloader kernel may not be updated as often as the production one
 - the driver doesn't actually need the feature, because it's for a
   different OS / workaround that doesn't apply. So all the kernel
   would be missing is a patch to ignore the TLV.

At scale FW and kernel are also maintained by different teams, not 
to mention applications. The FW may very well be newer while the
application team validates and moves to the latest kernel release.

But since it's just at info level I guess its more of a noise situation
than an annoyance.

> Maybe this should be a netif_dbg() instead?  (Or is this a subtle
>  way of telling me "you should implement devlink health"?)

Not devlink health.

Powered by blists - more mailing lists