linux-kernel - Re: [PATCH v2 08/10] x86-64: Emulate vsyscalls

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <BANLkTi=UvQ4fam8QtJyX2eVu9V5hFRoNNg@mail.gmail.com>
Date:	Mon, 30 May 2011 22:29:41 -0400
From:	Andrew Lutomirski <luto@....edu>
To:	Jan Beulich <JBeulich@...ell.com>
Cc:	Ingo Molnar <mingo@...e.hu>, Borislav Petkov <bp@...en8.de>,
	Jesper Juhl <jj@...osbits.net>,
	richard -rw- weinberger <richard.weinberger@...il.com>,
	Arjan van de Ven <arjan@...radead.org>,
	Mikael Pettersson <mikpe@...uu.se>, x86@...nel.org,
	Thomas Gleixner <tglx@...utronix.de>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 08/10] x86-64: Emulate vsyscalls

On Mon, May 30, 2011 at 3:51 AM, Jan Beulich <JBeulich@...ell.com> wrote:
>>>> On 30.05.11 at 05:48, Andy Lutomirski <luto@....EDU> wrote:
>> This causes vsyscalls to be a little more expensive than real
>> syscalls.  Fortunately sensible programs don't use them.
>
> Hmm - weren't vsyscalls there for performance reasons?
>
> Besides that, just a mostly cosmetic remark:
>
>> --- a/arch/x86/Kconfig
>> +++ b/arch/x86/Kconfig
>> @@ -1650,6 +1650,23 @@ config COMPAT_VDSO
>>
>>         If unsure, say Y.
>>
>> +config UNSAFE_VSYSCALLS
>> +     def_bool y
>> +     prompt "Unsafe fast legacy vsyscalls"
>> +     depends on X86_64
>> +     ---help---
>> +       Legacy user code expects to be able to issue three syscalls
>> +       by calling fixed addresses in kernel space.  If you say N,
>> +       then the kernel traps and emulates these calls.  If you say
>> +       Y, then there is actual executable code at a fixed address
>> +       to implement these calls efficiently.
>> +
>> +       On a system with recent enough glibc (probably 2.14 or
>> +       newer) and no static binaries, you can say N without a
>> +       performance penalty to improve security
>> +
>> +       If unsure, say Y.
>> +
>>  config CMDLINE_BOOL
>>       bool "Built-in kernel command line"
>>       ---help---
>> --- a/arch/x86/kernel/Makefile
>> +++ b/arch/x86/kernel/Makefile
>> @@ -42,6 +42,9 @@ obj-$(CONFIG_X86_32)        += probe_roms_32.o
>>  obj-$(CONFIG_X86_32) += sys_i386_32.o i386_ksyms_32.o
>>  obj-$(CONFIG_X86_64) += sys_x86_64.o x8664_ksyms_64.o
>>  obj-$(CONFIG_X86_64) += syscall_64.o vsyscall_64.o vread_tsc_64.o
>> +ifndef CONFIG_UNSAFE_VSYSCALLS
>> +     obj-$(CONFIG_X86_64)    += vsyscall_emu_64.o
>> +endif
>
> With the Kconfig dependency on X86_64 above and the new
> variable being a boolean one, these three lines can be written
> as just
>
> obj-$(CONFIG_UNSAFE_VSYSCALLS) += vsyscall_emu_64.o

I think that's backwards.  But in v3 that file will be included unconditionally.

--Andy
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/