From: "Brijesh Singh" <brijesh.singh@amd.com>
To: Gerd Hoffmann <kraxel@redhat.com>
Cc: brijesh.singh@amd.com, devel@edk2.groups.io,
James Bottomley <jejb@linux.ibm.com>, Min Xu <min.m.xu@intel.com>,
Jiewen Yao <jiewen.yao@intel.com>,
Tom Lendacky <thomas.lendacky@amd.com>,
Jordan Justen <jordan.l.justen@intel.com>,
Ard Biesheuvel <ardb+tianocore@kernel.org>,
Erdem Aktas <erdemaktas@google.com>,
Michael Roth <Michael.Roth@amd.com>
Subject: Re: [PATCH v9 30/32] OvmfPkg/PlatformPei: mark cpuid and secrets memory reserved in EFI map
Date: Tue, 19 Oct 2021 09:36:20 -0500 [thread overview]
Message-ID: <01ffc274-edec-3cda-7801-a3451dc081e2@amd.com> (raw)
In-Reply-To: <20211018060105.dbmomij5xcft6jpn@sirius.home.kraxel.org>
On 10/18/21 1:01 AM, Gerd Hoffmann wrote:
> Hi,
>
>> The MEMFD range is outside of the firmware image map, MEMFD begins with
>> 0x800000 [1] and in my boots I don't see it reserved in e820.
> Ah, ok.
>
>> Here is the snippet.
>>
>> [ ... ]
>> [ 0.000000] BIOS-e820: [mem 0x0000000000800000-0x0000000000807fff] ACPI NVS
>> [ 0.000000] BIOS-e820: [mem 0x0000000000808000-0x000000000080afff] usable
>> [ 0.000000] BIOS-e820: [mem 0x000000000080b000-0x000000000080bfff] ACPI NVS
>> [ 0.000000] BIOS-e820: [mem 0x000000000080c000-0x000000000080ffff] usable
>> [ 0.000000] BIOS-e820: [mem 0x0000000000810000-0x00000000008fffff] ACPI NVS
>> [ ... ]
> Hmm. Confused. memfd size is 0xD00000, so should the block from 800000
> to 8cffff be reserved? Why does it end at 8fffff instead?
There is no strong reason for block all of the MEMFD. What I see in the
current code is some selective pages gets marked reserved or other
memory type. As system boots some pages may get released as a system RAM.
> The first hole is this:
>
> 0x008000|0x001000
> gUefiOvmfPkgTokenSpaceGuid.PcdOvmfSecGhcbPageTableBase|gUefiOvmfPkgTokenSpaceGuid.PcdOvmfSecGhcbPageTableSize
> 0x009000|0x002000
> gUefiOvmfPkgTokenSpaceGuid.PcdOvmfSecGhcbBase|gUefiOvmfPkgTokenSpaceGuid.PcdOvmfSecGhcbSize
>
> The second hole is this (git master) ...
>
> 0x00C000|0x001000
> gUefiOvmfPkgTokenSpaceGuid.PcdOvmfSecGhcbBackupBase|gUefiOvmfPkgTokenSpaceGuid.PcdOvmfSecGhcbBackupSize
>
> And IIRC the cpuid + secrets pages are added there.
>
> So, yes, they must be reserved indeed. What about the other pages?
> Shouldn't they be reserved too? Or will they not be used any more
> at runtime?
As I indicated above, the other part of the code (such MemDetect.c)
makes the pages reserved as system boot. Some page can be may not be
used at all during the runtime and thus gets released.
thanks
next prev parent reply other threads:[~2021-10-19 14:36 UTC|newest]
Thread overview: 76+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-13 16:56 [PATCH v9 00/32] Add AMD Secure Nested Paging (SEV-SNP) support Brijesh Singh
2021-10-13 16:56 ` [PATCH v9 01/32] OvmfPkg/SecMain: move SEV specific routines in AmdSev.c Brijesh Singh
2021-10-14 6:59 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 02/32] UefiCpuPkg/MpInitLib: " Brijesh Singh
2021-10-14 7:00 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 03/32] OvmfPkg/ResetVector: move clearing GHCB in SecMain Brijesh Singh
2021-10-14 7:02 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 04/32] OvmfPkg/ResetVector: introduce SEV metadata descriptor for VMM use Brijesh Singh
2021-10-14 7:08 ` Gerd Hoffmann
2021-10-14 19:51 ` Brijesh Singh
2021-10-15 5:08 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 05/32] OvmfPkg: reserve SNP secrets page Brijesh Singh
2021-10-14 7:09 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 06/32] OvmfPkg: reserve CPUID page Brijesh Singh
2021-10-14 7:10 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 07/32] OvmfPkg/ResetVector: pre-validate the data pages used in SEC phase Brijesh Singh
2021-10-14 7:13 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 08/32] OvmfPkg/ResetVector: use SEV-SNP-validated CPUID values Brijesh Singh
2021-10-14 7:15 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 09/32] OvmfPkg/MemEncryptSevLib: add MemEncryptSevSnpEnabled() Brijesh Singh
2021-10-14 7:42 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 10/32] OvmfPkg/SecMain: register GHCB gpa for the SEV-SNP guest Brijesh Singh
2021-10-14 7:44 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 11/32] OvmfPkg/VmgExitLib: use SEV-SNP-validated CPUID values Brijesh Singh
2021-10-14 7:47 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 12/32] OvmfPkg/PlatformPei: register GHCB gpa for the SEV-SNP guest Brijesh Singh
2021-10-14 7:48 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 13/32] OvmfPkg/AmdSevDxe: do not use extended PCI config space Brijesh Singh
2021-10-14 7:50 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 14/32] OvmfPkg/MemEncryptSevLib: add support to validate system RAM Brijesh Singh
2021-10-14 8:02 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 15/32] OvmfPkg/MemEncryptSevLib: add function to check the VMPL0 Brijesh Singh
2021-10-14 8:03 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 16/32] OvmfPkg/BaseMemEncryptSevLib: skip the pre-validated system RAM Brijesh Singh
2021-10-14 8:19 ` Gerd Hoffmann
2021-10-13 16:56 ` [PATCH v9 17/32] OvmfPkg/MemEncryptSevLib: add support to validate > 4GB memory in PEI phase Brijesh Singh
2021-10-14 8:31 ` Gerd Hoffmann
2021-10-14 19:59 ` Brijesh Singh
2021-10-13 16:56 ` [PATCH v9 18/32] OvmfPkg/SecMain: validate the memory used for decompressing Fv Brijesh Singh
2021-10-14 8:33 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 19/32] OvmfPkg/PlatformPei: validate the system RAM when SNP is active Brijesh Singh
2021-10-14 8:34 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 20/32] UefiCpuPkg: Define ConfidentialComputingGuestAttr Brijesh Singh
2021-10-14 8:38 ` Gerd Hoffmann
2021-10-14 20:11 ` Brijesh Singh
2021-10-13 16:57 ` [PATCH v9 21/32] OvmfPkg/PlatformPei: set PcdConfidentialComputingAttr when SEV is active Brijesh Singh
2021-10-14 8:39 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 22/32] UefiCpuPkg/MpInitLib: use PcdConfidentialComputingAttr to check SEV status Brijesh Singh
2021-10-14 8:41 ` Gerd Hoffmann
2021-10-14 21:58 ` [edk2-devel] " Brijesh Singh
2021-10-13 16:57 ` [PATCH v9 23/32] UefiCpuPkg: add PcdGhcbHypervisorFeatures Brijesh Singh
2021-10-14 8:42 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 24/32] OvmfPkg/PlatformPei: set the Hypervisor Features PCD Brijesh Singh
2021-10-14 8:43 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 25/32] MdePkg/GHCB: increase the GHCB protocol max version Brijesh Singh
2021-10-14 8:43 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 26/32] UefiCpuPkg/MpLib: add support to register GHCB GPA when SEV-SNP is enabled Brijesh Singh
2021-10-14 8:44 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 27/32] UefiCpuPkg/MpInitLib: use BSP to do extended topology check Brijesh Singh
2021-10-14 8:48 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 28/32] OvmfPkg/MemEncryptSevLib: change the page state in the RMP table Brijesh Singh
2021-10-14 8:52 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 29/32] OvmfPkg/MemEncryptSevLib: skip page state change for Mmio address Brijesh Singh
2021-10-14 8:53 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 30/32] OvmfPkg/PlatformPei: mark cpuid and secrets memory reserved in EFI map Brijesh Singh
2021-10-14 8:58 ` Gerd Hoffmann
2021-10-14 22:11 ` Brijesh Singh
2021-10-15 5:26 ` Gerd Hoffmann
2021-10-15 15:46 ` Brijesh Singh
2021-10-18 6:01 ` Gerd Hoffmann
2021-10-19 14:36 ` Brijesh Singh [this message]
2021-10-20 4:46 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 31/32] OvmfPkg/AmdSev: expose the SNP reserved pages through configuration table Brijesh Singh
2021-10-14 8:59 ` Gerd Hoffmann
2021-10-13 16:57 ` [PATCH v9 32/32] UefiCpuPkg/MpInitLib: Use SEV-SNP AP Creation NAE event to launch APs Brijesh Singh
2021-10-14 9:02 ` Gerd Hoffmann
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-list from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=01ffc274-edec-3cda-7801-a3451dc081e2@amd.com \
--to=devel@edk2.groups.io \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox