Re: [PATCH 6/6] x86: kaslr: report kernel offset on panic

["H. Peter Anvin" <hpa@zytor.com>]

On 04/26/2013 03:13 PM, Borislav Petkov wrote:
> On Fri, Apr 26, 2013 at 12:03:25PM -0700, Kees Cook wrote:
>> When the system panics, include the kernel offset in the report to assist
>> in debugging.
>>
>> Signed-off-by: Kees Cook <keescook@chromium.org>
>> ---
>>  arch/x86/kernel/setup.c |   24 ++++++++++++++++++++++++
>>  1 file changed, 24 insertions(+)
>>
>> diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c
>> index fae9134..95a33b1 100644
>> --- a/arch/x86/kernel/setup.c
>> +++ b/arch/x86/kernel/setup.c
>> @@ -830,6 +830,18 @@ static void __init trim_low_memory_range(void)
>>  }
>>  	
>>  /*
>> + * Dump out kernel offset information on panic.
>> + */
>> +static int
>> +dump_kernel_offset(struct notifier_block *self, unsigned long v, void *p)
>> +{
>> +	pr_emerg("Kernel Offset: 0x%lx\n",
>> +		 (unsigned long)&_text - __START_KERNEL);
> 
> So what's wrong with subtracting the offset from the function addresses
> on the stack so that traces can show the addresses as they are in
> vmlinux, completely agnostic of any randomization?
> 

That really makes it harder to figure out what the heck the register
content may mean...

	-hpa