linux-kernel - Re: Smarter Kconfig help

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: <1551865552.9298.24.camel@pengutronix.de>
Date:   Wed, 06 Mar 2019 10:45:52 +0100
From:   Lucas Stach <l.stach@...gutronix.de>
To:     Russell King - ARM Linux admin <linux@...linux.org.uk>,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Cc:     Peng Hao <peng.hao2@....com.cn>,
        Mark Rutland <mark.rutland@....com>,
        Andy Shevchenko <andy.shevchenko@...il.com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: Smarter Kconfig help

Hi Russell,

Am Dienstag, den 05.03.2019, 17:31 +0000 schrieb Russell King - ARM Linux admin:
> Guys,
> 
> We need to be smarter when writing Kconfig help.  I'm just going
> through updating my build trees with the results of 5.0 development,
> and a number of the help texts are next to useless.  For example,
> 
> PVPANIC - is this something that should be enabled for a host or
> guest kernel?  Answer: you have to read the driver code to find out.
> 
> IMX_IRQSTEER - which i.MX SoCs does this apply to?  The help text
> just says:
> 
>   "Support for the i.MX IRQSTEER interrupt multiplexer/remapper."
> 
> which doesn't say which SoCs this should be enabled for - it turns
> out that grepping for the driver's DT compatible string, none of
> the 32-bit ARM cores have support for this, yet we still default
> it to enabled there.  It seems the help text should at the very
> least tell the user that this is not applicable to i.MX SoCs with
> 32-bit ARM cores.
> 
> I'm sure there's many other instances of this... I suspect that
> it's caused by review concentrating mostly on the technical aspects
> of the code and the Kconfig help text just gets forgotten about.
> 
> Can we at least improve these two options please?

While I totally agree that the irqsteer driver should only default Y on
64bit i.MX SoCs, I'm not really sure what to do about the Kconfig help
text.

Enumerating SoCs in the Kconfig of a IP block driver is always prone to
get outdated. Just as the first example I've been able to come up with,
the MMC_SDHCI_ESDHC_IMX help text says: "This selects the Freescale
eSDHC/uSDHC controller support found on i.MX25, i.MX35 i.MX5x and
i.MX6x.", while the same IP block is to be found on i.MX7 and various
i.MX8.

For the Kconfig user, who needs to decide if an option is relevant,
outdated SoC information is probably just as bad as no information at
all.

Regards,
Lucas