Re: [PATCH 4.4 178/193] x86/syscall: Sanitize syscall table de-references under speculation

From: Jinpu Wang
Date: Tue Mar 06 2018 - 11:11:47 EST


On Tue, Mar 6, 2018 at 3:21 PM, Jiri Slaby <jslaby@xxxxxxx> wrote:
> On 02/23/2018, 07:26 PM, Greg Kroah-Hartman wrote:
>> 4.4-stable review patch. If anyone has any objections, please let me know.
>>
>> ------------------
>>
>> From: Dan Williams <dan.j.williams@xxxxxxxxx>
>>
>> (cherry picked from commit 2fbd7af5af8665d18bcefae3e9700be07e22b681)
>>
>> The syscall table base is a user controlled function pointer in kernel
>> space. Use array_index_nospec() to prevent any out of bounds speculation.
>>
>> While retpoline prevents speculating into a userspace directed target it
>> does not stop the pointer de-reference, the concern is leaking memory
>> relative to the syscall table base, by observing instruction cache
>> behavior.
>>
>> Reported-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
>> Signed-off-by: Dan Williams <dan.j.williams@xxxxxxxxx>
>> Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
>> Cc: linux-arch@xxxxxxxxxxxxxxx
>> Cc: kernel-hardening@xxxxxxxxxxxxxxxxxx
>> Cc: gregkh@xxxxxxxxxxxxxxxxxxx
>> Cc: Andy Lutomirski <luto@xxxxxxxxxx>
>> Cc: alan@xxxxxxxxxxxxxxx
>> Link: https://lkml.kernel.org/r/151727417984.33451.1216731042505722161.stgit@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>> Signed-off-by: David Woodhouse <dwmw@xxxxxxxxxxxx>
>> [jwang: port to 4.4, no syscall_64]
>
> This is not complete IMO, the syscall is indeed there, only written in
> assembly in 4.4 yet.
>
> So this patch looks like it is missing these two hunks (from my
> SLE12-SP2 backport):
>
>> --- a/arch/x86/entry/entry_64.S
>> +++ b/arch/x86/entry/entry_64.S
>> @@ -184,6 +184,8 @@ entry_SYSCALL_64_fastpath:
>> cmpl $__NR_syscall_max, %eax
>> #endif
>> ja 1f /* return -ENOSYS (already in pt_regs->ax) */
>> + sbb %rcx, %rcx /* array_index_mask_nospec() */
>> + and %rcx, %rax
>> movq %r10, %rcx
>> #ifdef CONFIG_RETPOLINE
>> movq sys_call_table(, %rax, 8), %rax
>> @@ -282,6 +284,8 @@ tracesys_phase2:
>> cmpl $__NR_syscall_max, %eax
>> #endif
>> ja 1f /* return -ENOSYS (already in pt_regs->ax) */
>> + sbb %rcx, %rcx /* array_index_mask_nospec() */
>> + and %rcx, %rax
>> movq %r10, %rcx /* fixup for C */
>> #ifdef CONFIG_RETPOLINE
>> movq sys_call_table(, %rax, 8), %rax
>
> Discovered by Jan Beulich.
>
> thanks,
> --
> js
> suse labs

Thanks Jiri, yes, indeed, could you send a formal patch of the fix?

Thanks!
Jack Wang