Re: [PATCH 0/3 v3] add reserved e820 ranges to the kdump kernel e820 table

From: Dave Young <dyoung@redhat.com>
To: Lianbo Jiang <lijiang@redhat.com>
Cc: linux-kernel@vger.kernel.org, kexec@lists.infradead.org,
	tglx@linutronix.de, mingo@redhat.com, hpa@zytor.com,
	x86@kernel.org, akpm@linux-foundation.org,
	dan.j.williams@intel.com, thomas.lendacky@amd.com,
	bhelgaas@google.com, baiyaowei@cmss.chinamobile.com,
	tiwai@suse.de, bp@suse.de, brijesh.singh@amd.com, bhe@redhat.com
Subject: Re: [PATCH 0/3 v3] add reserved e820 ranges to the kdump kernel e820 table
Date: Tue, 16 Oct 2018 10:56:57 +0800	[thread overview]
Message-ID: <20181016025657.GB13006@dhcp-128-65.nay.redhat.com> (raw)
In-Reply-To: <20180921073211.20097-1-lijiang@redhat.com>

On 09/21/18 at 03:32pm, Lianbo Jiang wrote:
> E820 reserved ranges is useful in kdump kernel, we have added this in
> kexec-tools code.
> 
> One reason is PCI mmconf (extended mode) requires reserved region otherwise
> it falls back to legacy mode.
> 
> Furthermore, when AMD SME kdump support, it needs to map dmi table area as
> unencrypted. For normal boot, these ranges sit in e820 reserved ranges,
> thus the early ioremap code naturally map them as unencrypted. If we also
> have same e820 reserve setup in kdump kernel then it will just work like
> normal kernel.
> 
> Kdump uses walk_iomem_res_desc to iterate resources, then adds matched desc
> to e820 table for the kdump kernel.
> 
> But IORES_DESC_NONE resource type includes several different e820 types, we
> need add exact e820 type to the kdump kernel e820 table, thus it also needs
> an extra checking in memmap_entry_callback() to match the e820 type and
> resource name.
> 
> By the way, we also fix an error which walks through iomem resources, the
> values of the function parameter may be modified in the while loop of
> __walk_iomem_res_desc(), which will cause us to not get the desired result
> in some cases.
> 
> Changes since v2:
> 1. Modified the value of flags to "0", when walking through the whole
> tree for e820 reserved ranges.
> 2. Modified the invalid SOB chain issue.
> 
> Lianbo Jiang (3):
>   resource: fix an error which walks through iomem resources
>   x86/kexec_file: add e820 entry in case e820 type string matches to io
>     resource name
>   x86/kexec_file: add reserved e820 ranges to kdump kernel e820 table

Lianbo, since Bjorn has fixed the resource issue in another patchset,
can you rebase your patch 2 and 3 on top of his patches and resend?

> 
>  arch/x86/include/asm/e820/api.h |  2 ++
>  arch/x86/kernel/crash.c         | 11 ++++++++++-
>  arch/x86/kernel/e820.c          |  2 +-
>  kernel/resource.c               |  3 +++
>  4 files changed, 16 insertions(+), 2 deletions(-)
> 
> -- 
> 2.17.1
> 

Thanks
Dave