Re: [PATCH v7 0/4] support reserving crashkernel above 4G on arm64 kdump

[Bhupesh Sharma <bhsharma@redhat.com>]

Hi John,

On Wed, May 20, 2020 at 1:53 AM John Donnelly
<john.p.donnelly@oracle.com> wrote:
>
>
>
> > On May 19, 2020, at 5:21 AM, Arnd Bergmann <arnd@arndb.de> wrote:
> >
> > On Thu, Mar 26, 2020 at 4:10 AM Chen Zhou <chenzhou10@huawei.com> wrote:
> >>
> >> Hi all,
> >>
> >> Friendly ping...
> >
> > I was asked about this patch series, and see that you last posted it in
> > December. I think you should rebase it to linux-5.7-rc6 and post the
> > entire series again to make progress, as it's unlikely that any maintainer
> > would pick up the patches from last year.
> >
> > For the contents, everything seems reasonable to me, but I noticed that
> > you are adding a property to the /chosen node without adding the
> > corresponding documentation to
> > Documentation/devicetree/bindings/chosen.txt
> >
> > Please add that, and Cc the devicetree maintainers on the updated
> > patch.
> >
> >         Arnd
> >
> >> On 2019/12/23 23:23, Chen Zhou wrote:
> >>> This patch series enable reserving crashkernel above 4G in arm64.
> >>>
> >>> There are following issues in arm64 kdump:
> >>> 1. We use crashkernel=X to reserve crashkernel below 4G, which will fail
> >>> when there is no enough low memory.
> >>> 2. Currently, crashkernel=Y@X can be used to reserve crashkernel above 4G,
> >>> in this case, if swiotlb or DMA buffers are required, crash dump kernel
> >>> will boot failure because there is no low memory available for allocation.
> >>>
> >>> The previous changes and discussions can be retrieved from:
> >>>
> >>> Changes since [v6]
> >>> - Fix build errors reported by kbuild test robot.
> > ...
>
>
>  Hi
>
> We found
>
> https://lkml.org/lkml/2020/4/30/1583
>
> Has cured our Out-Of-Memory kdump failures.
>
> From    Henry Willard
> Subject [PATCH] mm: Limit boost_watermark on small zones.
>
> I am currently not on linux-kernel@vger.kernel.org. dlist for all to see  this message so you may want to rebase and see if this cures your OoM issue and share the results.

This is a very interesting finding. Thanks a lot for sharing the same.
I am working on further avoiding OOM issues with arm64 kdump kernels.
I will experiment more with this patch and get back with more details.

Regards,
Bhupesh