Re: [x86/vsyscall] 3dc33bd30f Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b

[Kees Cook]

On Tue, Feb 21, 2017 at 9:23 PM, Andy Lutomirski <luto@xxxxxxxxxxxxxx> wrote:
> On Tue, Feb 21, 2017 at 4:46 PM, Kees Cook <keescook@xxxxxxxxxxxx> wrote:
>> On Tue, Feb 21, 2017 at 4:39 PM, Chunyu Hu <chuhu.ncepu@xxxxxxxxx> wrote:
>>> I hit the similar panic on Fedora 25. Is it the same issue?
>>>
>>>
>>> [    2.527391] Freeing unused kernel memory: 1688K (ffffffffbdf66000 -
>>> ffffffffbe10c000)
>>> [    2.535222] Write protecting the kernel read-only data: 14336k
>>> [    2.542679] Freeing unused kernel memory: 1892K (ffff94cb30827000 -
>>> ffff94cb30a00000)
>>> [    2.553482] Freeing unused kernel memory: 700K (ffff94cb30d51000 -
>>> ffff94cb30e00000)
>>> [    2.572048] x86/mm: Checked W+X mappings: passed, no W+X pages found.
>>> Fatal: [    2.598239] traps: init[1] general protection ip:7fc841bfc642
>>> sp:7ffc3fd85870 error:0no entropy gathering module dete[    2.606473]  in
>>> libc-2.24.so[7fc841bc5000+1bd000]
>>> cted
>>
>> I would not expect a vsyscall table problem if it's truly using glibc
>> 2.24. Maybe something regressed in glibc? That really looks like an
>> attempt to call vsyscalls?
>
> There should be a vsyscall error if so, and shouldn't it be a page
> fault, not a gpf?

Yeah, the earlier email from Fengguang included a vsyscall error, so
I'm not sure what this is, exactly. But if it bisects to the same
commit, we likely need to investigate a bit more.

-Kees

-- 
Kees Cook
Pixel Security