[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20201118085153.5tgnzdwuywzsaxti@ws.flokli.de>
Date: Wed, 18 Nov 2020 09:51:53 +0100
From: Florian Klink <flokli@...kli.de>
To: Randy Dunlap <rdunlap@...radead.org>
Cc: Jakub Kicinski <kuba@...nel.org>, David Ahern <dsahern@...il.com>,
netdev@...r.kernel.org, "David S. Miller" <davem@...emloft.net>,
Alexey Kuznetsov <kuznet@....inr.ac.ru>,
Hideaki YOSHIFUJI <yoshfuji@...ux-ipv6.org>,
Kim Phillips <kim.phillips@....com>
Subject: Re: [PATCH] ipv4: use IS_ENABLED instead of ifdef
>>> I looked at this yesterday and got distracted diving into the generated
>>> file to see the difference:
>>>
>>> #define CONFIG_IPV6 1
>>>
>>> vs
>>>
>>> #define CONFIG_IPV6_MODULE 1
>
>Digging up ancient history. ;)
>
>> Interesting.
>>
>> drivers/net/ethernet/netronome/nfp/flower/action.c:#ifdef CONFIG_IPV6
>>
>> Oops.
>
>Notify the maintainer!
Yeah, this is super scary stuff - allmodyes-like configs are quite
common in generic distros, and I assume similar mistakes could have
happened in many other places in the kernel as well.
I wonder if the "ifdef CONFIG_…" pattern should be discouraged, and
"IS_ENABLED(CONFIG_…)" used instead, at least for all tristate config
options.
Florian
Powered by blists - more mailing lists