On Fri, May 25, 2018 at 12:02:03PM -0700, Andrew Jeddeloh wrote:
> Oops, my bad, didn't mean to drop the ML.
No problem. Sometimes it happens.
> So you're saying the setup header could expand to include some of pad7
> in the future, but we error if it goes beyond that (into the e820
Yep!
> map)
Oops, my bad, didn't mean to drop the ML.
So you're saying the setup header could expand to include some of pad7
in the future, but we error if it goes beyond that (into the e820
map)? Looking at bootparam.h it looks like the _most_ correct thing
would be to stop at edd_mbr_sig_buffer, but grub do
Re-added grub-devel. Next time please do not drop GRUB ML from the addresses.
On Thu, May 17, 2018 at 03:40:32PM -0700, Andrew Jeddeloh wrote:
> Sorry about the long delay, I agree with all the sugguestions, except
> shouldn't
No problem.
> if (len > &linux_params.e820_map - &linux_params)
>
>
On Wed, May 09, 2018 at 10:46:47AM -0700, Andrew Jeddeloh wrote:
> This patch is prompted from a question I asked a while ago about why
> the disk read is necessary. See the thread here [1].
>
> This changes the disk read to use the length of the setup header as
> calculated by the x86 32 bit linux
This patch is prompted from a question I asked a while ago about why
the disk read is necessary. See the thread here [1].
This changes the disk read to use the length of the setup header as
calculated by the x86 32 bit linux boot protocol [1]. I'm not 100%
sure its patch that's wanted however. The
