linux-kernel - Re: [patch V2 3/5] lib/vdso/32: Provide legacy syscall fallbacks

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: <alpine.DEB.2.21.1907302212410.1786@nanos.tec.linutronix.de>
Date:   Tue, 30 Jul 2019 22:14:11 +0200 (CEST)
From:   Thomas Gleixner <tglx@...utronix.de>
To:     Andy Lutomirski <luto@...nel.org>
cc:     Sean Christopherson <sean.j.christopherson@...el.com>,
        LKML <linux-kernel@...r.kernel.org>, X86 ML <x86@...nel.org>,
        Vincenzo Frascino <vincenzo.frascino@....com>,
        Kees Cook <keescook@...omium.org>,
        Paul Bolle <pebolle@...cali.nl>, Will Deacon <will@...nel.org>
Subject: Re: [patch V2 3/5] lib/vdso/32: Provide legacy syscall fallbacks

On Tue, 30 Jul 2019, Andy Lutomirski wrote:

> On Tue, Jul 30, 2019 at 2:39 AM Thomas Gleixner <tglx@...utronix.de> wrote:
> >
> > To address the regression which causes seccomp to deny applications the
> > access to clock_gettime64() and clock_getres64() syscalls because they
> > are not enabled in the existing filters.
> >
> > That trips over the fact that 32bit VDSOs use the new clock_gettime64() and
> > clock_getres64() syscalls in the fallback path.
> >
> > Add a conditional to invoke the 32bit legacy fallback syscalls instead of
> > the new 64bit variants. The conditional can go away once all architectures
> > are converted.
> >
> 
> I haven't surveyed all the architectures, but once everything is
> converted, shouldn't we use the 32-bit fallback for exactly the same
> set of architectures that want clock_gettime32 at all in the vdso?

Yes. That's why I want to remove the conditional once all all converted
over, that's x86/aaarg64 in mainline and a few in next.

Thanks,

	tglx