linux-kernel - Re: [PATCH RFC 2/2] wcslen() prototype in string.h

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: <20250325165847.GA2603000@ax162>
Date: Tue, 25 Mar 2025 09:58:47 -0700
From: Nathan Chancellor <nathan@...nel.org>
To: Andy Shevchenko <andy@...nel.org>
Cc: Kees Cook <kees@...nel.org>,
	Nick Desaulniers <nick.desaulniers+lkml@...il.com>,
	Bill Wendling <morbo@...gle.com>,
	Justin Stitt <justinstitt@...gle.com>,
	Ard Biesheuvel <ardb@...nel.org>, linux-kernel@...r.kernel.org,
	linux-hardening@...r.kernel.org, llvm@...ts.linux.dev,
	linux-efi@...r.kernel.org
Subject: Re: [PATCH RFC 2/2] wcslen() prototype in string.h

On Tue, Mar 25, 2025 at 06:17:34PM +0200, Andy Shevchenko wrote:
> On Tue, Mar 25, 2025 at 08:45:19AM -0700, Nathan Chancellor wrote:
> > If this is desired, it should be squashed into the previous change. I
> > wrote it separately because it is slightly more invasive.
> > 
> > In order to export wcslen() to the rest of the kernel (should it ever be
> > necessary elsewhere), it needs to be added to string.h, along with nls.h
> > for the typedef of wchar_t. However, dragging in nls.h into string.h
> > causes an error in the efistub due to a conflicting function name:
> > 
> >   drivers/firmware/efi/libstub/printk.c:27:5: error: static declaration of 'utf8_to_utf32' follows non-static declaration
> >      27 | u32 utf8_to_utf32(const u8 **s8)
> >         |     ^
> >   include/linux/nls.h:55:12: note: previous declaration is here
> >      55 | extern int utf8_to_utf32(const u8 *s, int len, unicode_t *pu);
> >         |            ^
> >   drivers/firmware/efi/libstub/printk.c:85:26: error: too few arguments to function call, expected 3, have 1
> >      85 |                 c32 = utf8_to_utf32(&s8);
> >         |                       ~~~~~~~~~~~~~    ^
> >   include/linux/nls.h:55:12: note: 'utf8_to_utf32' declared here
> >      55 | extern int utf8_to_utf32(const u8 *s, int len, unicode_t *pu);
> >         |            ^             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> >   2 errors generated.
> > 
> > Rename the efi function to avoid the conflict.
> 
> Hmm... Why not split this to two, rename patch as a standalone makes sense to
> me even outside of this series.

How so? If nls.h is not included in printk.c via string.h, which does
not happen without this series, what value does the rename have? I do
not mind splitting it up that way to keep things cleaner, I am just
wondering what would be the justification in the changelog (I guess just
that nls.h may get included in the future for some reason)?

Cheers,
Nathan