netdev - Re: [net-next patch v2] bnx2x: Add run-time CNIC support

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:	Tue, 10 Jul 2012 15:33:54 +0300
From:	"Eilon Greenstein" <eilong@...adcom.com>
To:	"David Miller" <davem@...emloft.net>
cc:	meravs@...adcom.com, netdev@...r.kernel.org, dmitry@...adcom.com
Subject: Re: [net-next patch v2] bnx2x: Add run-time CNIC support

On Tue, 2012-07-10 at 05:21 -0700, David Miller wrote:
> From: "Merav Sicron" <meravs@...adcom.com>
> Date: Tue, 10 Jul 2012 15:17:00 +0300
> 
> > There are still two advantages in disabling CNIC in bnx2x: Saving
> > resources (MSI-X vector and memory) as well as reducing some latency.
> 
> But, nobody does this.  No end user can do this easily, this
> is therefore of zero value to him.

Most do not, but I'm aware of two customers that play with their own
kernel that do that - they can play with the driver and tweak it to
disable this mode manually, but that is similar to supporting something
outside the tree.

> > While it is true that distributions enable the CNIC Kconfig option, some
> > users that care about resources and latency compile a kernel without it.
> 
> This, therefore, results in a terrible user experience.

We are using the Kconfig since it is meant for advanced users that
customize their kernel to their needs.

> > Can you please re-consider this patch?
> 
> Absolutely not.
> 
> Make it really dynamic, and properly configurable at run time, so
> people don't have to go through hoops to get the "advantages" you
> speak so highly of.

This is possible for the resources, but not for the latency - we cannot
change the HW mode once traffic started to run. Why is that so bad to
support Kconfig as a working mode like we did thus far? We are using it
specifically for users that wants to optimize the kernel, so Kconfig
does not sound that bad in that context.

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html