Re: [PATCH] x86: Kill E820_RESERVED_KERN

Re: [PATCH] x86: Kill E820_RESERVED_KERN

[Yu Chen]

Hi,
On 07/29/2015 03:21 AM, Yinghai Lu wrote:
> E820_RESERVED_KERN was introduced to do early allocation for
> setup_data when we were using original early_res with e820 map.
>
> Now we are using memblock to do early resource reserve/allocation, and
> setup_data is reserved in memblock early already.
>
> For kexec path, kexec generate setup_data (Now kexec-tools create SETUP_EFI
> and SETUP_E820_EXT), and pass pointer to second kernel, and
> second kernel reserve setup_data by their own without using e820 map.
>
> So we do not need to touch e820 map at all, and we can kill
> E820_RESERVED_KERN.
>
> That make the code simpler, and at same time that will fix bug with
> hibernation:
>    mark_nonsave_region that can not handle that case:
>      E820_RAM and E820_RESERVED_KERN ranges are continuous and
>      boundary is not page aligned.
>
> Link: https://bugzilla.opensuse.org/show_bug.cgi?id=913885
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=96111

I've tested Hibernation on latest 4.2.-rc7 and encountered panic when
resuming, so I guess this patch has not been merged upstream:

BUG: unable to handle kernel paging request at ffff880085894000
IP: [<ffffffff810c5dc2>] load_image_lzo+0x8c2/0xe70

With current patch and Lee, Chun-Yi's patch applied, the panic
disappeared, would someone please have a look at this patch,
thanks a lot.

Best Regards,
Yu