Re: [PATCH v7u1 22/31] x86, boot: add fields to support load bzImageand ramdisk above 4G

[Yinghai Lu]

On Mon, Jan 14, 2013 at 9:49 AM, H. Peter Anvin <hpa@xxxxxxxxx> wrote:
> On 01/13/2013 09:37 PM, Yinghai Lu wrote:
>>>
>>>
>>> This is kinda missing from the mechanism of the sentinel and it should
>>> be documented too.
>>
>>
>> No, we should have too much duplicated info.
>>
>
> That is not duplicating info... that is basic documentation.  As you show in
> the post further on, it took a very simple description, and it *is* a very
> subtle thing that is inherently different from how the other fields operate.

please check if following is enough?

+       /*
+        * kernel have sentinel to set as 0xff in setup link scripts,
+        * so if bootloader just copy whole page from kernel image to
+        * get setup_header instead of clearing boot_param buffer and
+        * copying setup_header only, will leave sentinel as 0xff.
+        * With that, we can tell some fields in boot_param have
+        * invalid values, and we need to zero them in kernel.
+        */
+       __u8  sentinel;                                 /* 0x1ef */


>
> It doesn't help that you didn't, despite repeated requests, implement what I
> *asked for*, which is:
>
> If the sentinel is flagged, zero *all fields not explicitly set by the
> broken versions of kexec*, not just your new "ext" fields.

other fields are pad* fields, so do we zero out them
with memset with exact address?
so next times, when someone change pad fields to other ext_*,
they don't need to change code again here.


>
> Yinghai, I understand you're frustrated, but please understand that Borislav
> is not in any shape, way, or form "some guys that do not know the code well
> keep sending comments out to waste others time".  Rather, he has spent a
> huge amount of time giving you an awful lot of good feedback  A lot of them
> have centered on documentation and code maintainability, both of which are
> vitally important part of a long-lived codebase.
>
> Having someone doing line-by-line review of your code is enormously
> time-consuming and not something most people enjoy doing.  Borislav is doing
> you -- and me -- a huge favor here.

yes.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/