Re: [PATCH v2 03/13] x86: Add macro for 64bit entry startup_64

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

We probably should add a flag for those, but that is more of a boot_flag...

Yinghai Lu <yinghai@kernel.org> wrote:

>On Mon, Nov 19, 2012 at 3:13 PM, H. Peter Anvin <hpa@zytor.com> wrote:
>> On 11/19/2012 02:53 PM, Yinghai Lu wrote:
>>>
>>> any other field, in header struct field that we can use to tell
>>> bzImage could be used that
>>> 0x200 directly?
>>>
>>> hardware_subarch?
>>>
>>
>> There isn't one... this dates back all the way to the original x86-64
>> kernels.
>>
>> Are you asking if we can tell this is a 64-bit kernel (as opposed to
>a
>> 32-bit kernel, which obviously doesn't have a 64-bit entry point)?
>> Unfortunately there isn't an intentional one that I know of.  There
>> might be an accidental such indicator, but we'd have to go back to
>look
>> at 8+ years of kernels.  We can't even rely on a jmp instruction at
>the
>> address...
>
>So we could add one field to tell that bzImage could be used with
>64bit?
>
>current in this patchset, I added
>
>0268/4  2.12+   ext_ramdisk_image ramdisk_image 32 bits
>026C/4  2.12+   ext_ramdisk_size ramdisk_size high 32 bits
>0270/4  2.12+   code64_start_offset 64bit start offset for bzImage
>0274/4  2.12+   ext_cmd_line_ptr cmd_line_ptr high 32 bits
>
>so you don't like code64_start_offset.
>
>how about other three?
>
>can we use bits 31 of hardware_subarch to tell it is bzImage for
>x86_64?
>
>Thanks
>
>Yinghai

-- 
Sent from my mobile phone. Please excuse brevity and lack of formatting.