linux-kernel - Re: Build regressions/improvements in v5.9-rc8

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]

Date:   Wed, 7 Oct 2020 14:48:42 +0100
From:   Christoph Hellwig <hch@...radead.org>
To:     Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
Cc:     Paul Bolle <pebolle@...cali.nl>,
        Richard Weinberger <richard.weinberger@...il.com>,
        Geert Uytterhoeven <geert@...ux-m68k.org>,
        Richard Weinberger <richard@....at>,
        Anton Ivanov <anton.ivanov@...bridgegreys.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Bartosz Golaszewski <bgolaszewski@...libre.com>,
        Linus Walleij <linus.walleij@...aro.org>,
        Nick Hu <nickhu@...estech.com>,
        Greentime Hu <green.hu@...il.com>,
        Vincent Chen <deanbo422@...il.com>
Subject: Re: Build regressions/improvements in v5.9-rc8

On Wed, Oct 07, 2020 at 04:02:41PM +0300, Andy Shevchenko wrote:
> On Wed, Oct 07, 2020 at 02:45:30PM +0200, Paul Bolle wrote:
> > Richard Weinberger schreef op wo 07-10-2020 om 14:08 [+0200]:
> > > UML has no ia32 emulation and therefore no in_ia32_syscall().
> > > Maybe you can check for CONFIG_IA32_EMULATION too?
> > 
> > The pending fix is:
> >     #if defined(CONFIG_X86_64) && !defined(CONFIG_UML)
> > 
> > Since this check guards in_ia32_syscall() just checking CONFIG_IA32_EMULATION
> > should do too.
> > 
> > (Way outside my limited expertise, but anyway: is does look odd to see a call
> > to in_ia32_syscall() in drivers/. All other calls are in arch/x86/. Isn't this
> > a bit too x86 specific for an arch independent driver?)
> 
> According to Arnd who suggested that fix the problem is basically appears on
> x86. He also mentioned the new approach that Christoph Hellwig is working on.
> But we decided to go the ifdeffery for better backporting.

linux-next has compat_need_64bit_alignment_fixup(), which is the
proper thing to use.  But for now you'll need the ifdef mess.