Re: [PATCH 1/2] x86 boot: add E820_RESVD_KERN

["Yinghai Lu" <yhlu.kernel@gmail.com>]

On Mon, Jun 30, 2008 at 6:09 PM, Huang, Ying <ying.huang@intel.com> wrote:
> Hi, Yinghai,
>
> On Mon, 2008-06-30 at 16:20 -0700, Yinghai Lu wrote:
>> On Mon, Jun 30, 2008 at 3:53 PM, Yinghai Lu <yhlu.kernel@gmail.com> wrote:
>> > On Mon, Jun 30, 2008 at 12:16 PM, Ingo Molnar <mingo@elte.hu> wrote:
>> >>
>> >> * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
>> >>
>> >>> > For find_e820_area, this is safe enough. But what about conflict
>> >>> > between setup_data and ebda or ramdisk?
>> >>>
>> >>> can you have setup_data and ebda at the same time?
>> >>>
>> >>> setup_data and ramdisk should be ok, because bootloader is supposed to
>> >>> make them not to be conflicts.
>> >>
>> >> the more sanity checks we do before relying on some crutial data, the
>> >> better. It's easier to panic or sanitize data in some structured way and
>> >> complain about it in the syslog than to let things get corrupted. Boot
>> >> loaders are ... not unknown to be have bugs too, at times.
>> >
>> > to address Ying's concern, we could let reserve_setup_data
>> > call reserve_early in addition to e820_update_range...
>> >
>> > reserve_early will panic if RAMDISK overlap efi setup_data...
>> >
>>
>> Ying,
>> please check the attached patch
>
> Which git-x86 head should I use to test the patch? It can not be applied
> to x86/master.
>

tip/master

YH