Re: [PATCH v2 04/15] x86, kaslr: get kaslr_enabled back correctly

[Yinghai Lu <yinghai@kernel.org>]

On Wed, Mar 4, 2015 at 2:16 AM, Borislav Petkov <bp@alien8.de> wrote:
> On Wed, Mar 04, 2015 at 12:00:37AM -0800, Yinghai Lu wrote:
>> commit f47233c2d34f ("x86/mm/ASLR: Propagate base load address calculation")
>> is using address as value for kaslr_enabled.
>>
>> That will random kaslr_enabled get that set or cleared.
>> Will have problem for system really have kaslr enabled.
>>
>> -v2: update changelog.
>
> This is still not good enough. Please do this:
>
> In commit f47233c2d34f we did A. The problem with that is B. Change the
> code to do C.
>
> Now you only have to fill out the A,B and C variables with the
> respective text which is understandable even for people who don't know
> this code.
>

I don't know, that is trivial and obvious.

the old code use address as value instead of using reference...


>>
>>
>>  static void __init parse_kaslr_setup(u64 pa_data, u32 data_len)
>>  {
>> -     kaslr_enabled = (bool)(pa_data + sizeof(struct setup_data));
>> +     /* kaslr_setup_data is defined in aslr.c */
>> +     unsigned char *data;
>> +     unsigned long offset = sizeof(struct setup_data);
>> +
>> +     data = early_memremap(pa_data, offset + 1);
>
> early_memremap() needs its retval checked before accessing it.
>

will fix that.