[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <YCmrZ7i2TXZZXjLy@zeniv-ca.linux.org.uk>
Date: Sun, 14 Feb 2021 22:59:51 +0000
From: Al Viro <viro@...iv.linux.org.uk>
To: Randy Dunlap <rdunlap@...radead.org>
Cc: kernel test robot <lkp@...el.com>, kbuild-all@...ts.01.org,
linux-kernel@...r.kernel.org
Subject: Re: arch/alpha/lib/csum_partial_copy.c:328:1: error: no previous
prototype for 'csum_and_copy_from_user'
On Thu, Feb 11, 2021 at 06:52:36PM -0800, Randy Dunlap wrote:
> >>> arch/alpha/lib/csum_partial_copy.c:328:1: error: no previous prototype for 'csum_and_copy_from_user' [-Werror=missing-prototypes]
> > 328 | csum_and_copy_from_user(const void __user *src, void *dst, int len,
> > | ^~~~~~~~~~~~~~~~~~~~~~~
> > arch/alpha/lib/csum_partial_copy.c:375:1: error: no previous prototype for 'csum_partial_copy_nocheck' [-Werror=missing-prototypes]
> > 375 | csum_partial_copy_nocheck(const void *src, void *dst, int len, __wsum sum)
> > | ^~~~~~~~~~~~~~~~~~~~~~~~~
> > cc1: all warnings being treated as errors
>
> I can't reproduce this (wrong version of gcc) but it looks like adding
> #include <asm/checksum.h>
> to arch/alpha/lib/csum_partial_copy.c should fix the warnings.
I'd rather go with net/checksum.h; converting all includes outside of
net/checksum.h itself would have to wait, but let's not breed more of
those without a good reason...
Powered by blists - more mailing lists