On 11.06.2018 12:42, David Hildenbrand wrote:
>
>>> +static void test_iep(void)
>>> +{
>>> + uint16_t *code;
>>> + uint8_t *iepbuf = NULL;
>>> + void (*fn)(void);
>>> +
>>> + /* Enable IEP */
>>> + ctl_set_bit(0, 20);
>>> +
>>> + /* Get and protect a page with the IEP bit */
>>> + iepbuf = alloc_page();
>>> + protect_page(iepbuf, PAGE_ENTRY_IEP);
>>> +
>>> + /* Code branches into r14 which contains the return address. */
>>> + code = (uint16_t *)iepbuf;
>>> + *code = 0x07fe;
>>> + fn = (void *)code;
>>
>> Not sure if I've got Christian's comment wrt to ipte right, but if I did
>> (Christian, please correct me if I'm wrong), I think it's better to move
>> the "protect_page(iepbuf, PAGE_ENTRY_IEP)" here, so that the ipte is
>> called after you've modified the contents of the page.
>
> Why? When we replace the entry (protect), we do an ipte. Next access
> will reload the right page table entry including protection. What am I
> missing?
Never mind, as explained by Christian, it's not necessary here. I'm just
too paranoid after running into cache-flush issues on ppc in the past ;-)
Thomas
