[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-Id: <db744500-85f0-4df0-9d00-0e4b9fa18df1@app.fastmail.com>
Date: Thu, 18 May 2023 15:23:39 +0200
From: "Arnd Bergmann" <arnd@...db.de>
To: "Andrew Morton" <akpm@...ux-foundation.org>,
"Arnd Bergmann" <arnd@...nel.org>
Cc: "Nick Terrell" <terrelln@...com>, linux-mm@...ck.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH] decompressor: provide missing prototypes
On Wed, May 17, 2023, at 22:49, Andrew Morton wrote:
> On Wed, 17 May 2023 15:19:31 +0200 Arnd Bergmann <arnd@...nel.org> wrote:
>
>> From: Arnd Bergmann <arnd@...db.de>
>>
>> The entry points for the decompressor don't always have a prototype
>> included in the .c file:
>>
>> lib/decompress_inflate.c:42:17: error: no previous prototype for '__gunzip' [-Werror=missing-prototypes]
>> lib/decompress_unxz.c:251:17: error: no previous prototype for 'unxz' [-Werror=missing-prototypes]
>> lib/decompress_unzstd.c:331:17: error: no previous prototype for 'unzstd' [-Werror=missing-prototypes]
>>
>> include the correct headers for unxz and unzstd, and mark the
>> inflate function above as unconditionally 'static' to avoid
>> these warnings.
>
> These are errors, not warnings?
>
> Under what circumstances do they occur?
Running "make W=1" turns these errors on as warnings, enabling
CONFIG_WERROR turns all warnings into errors.
> Shouldn't we cc:stable? If so, do we have a suitable Fixes: target?
I have sent 140 patches for these, there is probably no point in backporting
them all. My hope is that we can enable -Wmissing-prototypes by default
after these are all merged, but that patch would not get backported
either.
I meant to include a link with the explanations to
https://people.kernel.org/arnd/missing-prototype-warnings-in-the-kernel
in my series but didn't have that in the separate patches.
Arnd
Powered by blists - more mailing lists