public inbox for devel@edk2.groups.io
 help / color / mirror / Atom feed
From: "Lendacky, Thomas" <thomas.lendacky@amd.com>
To: Brijesh Singh <brijesh.singh@amd.com>, devel@edk2.groups.io
Cc: James Bottomley <jejb@linux.ibm.com>, Min Xu <min.m.xu@intel.com>,
	Jiewen Yao <jiewen.yao@intel.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 3/3] OvmfPkg/ResetVector: move the GHCB page setup in AmdSev.asm
Date: Mon, 9 Aug 2021 12:03:25 -0500	[thread overview]
Message-ID: <298e16b9-f44c-a10d-a42e-92988e4725c1@amd.com> (raw)
In-Reply-To: <20210804202003.17543-4-brijesh.singh@amd.com>

On 8/4/21 3:20 PM, Brijesh Singh wrote:
> BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=3429
> 
> While build the initial page table, the SetCr3ForPageTables64 checks
> whether SEV-ES is enabled. If so, clear the page encryption mask from the
> GHCB page. Move the logic to clear the page encryption mask in the
> AmdSev.asm.
> 
> Cc: James Bottomley <jejb@linux.ibm.com>
> Cc: Min Xu <min.m.xu@intel.com>
> Cc: Jiewen Yao <jiewen.yao@intel.com>
> Cc: Tom Lendacky <thomas.lendacky@amd.com>
> Cc: Jordan Justen <jordan.l.justen@intel.com>
> Cc: Ard Biesheuvel <ardb+tianocore@kernel.org>
> Cc: Erdem Aktas <erdemaktas@google.com>
> Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
> ---
>  OvmfPkg/ResetVector/Ia32/AmdSev.asm       | 113 +++++++++++++++++-----
>  OvmfPkg/ResetVector/Ia32/PageTables64.asm |  53 ++--------
>  2 files changed, 94 insertions(+), 72 deletions(-)
> 
> diff --git a/OvmfPkg/ResetVector/Ia32/AmdSev.asm b/OvmfPkg/ResetVector/Ia32/AmdSev.asm
> index 87d81b01e263..fd2e6abcd4a0 100644
> --- a/OvmfPkg/ResetVector/Ia32/AmdSev.asm
> +++ b/OvmfPkg/ResetVector/Ia32/AmdSev.asm
> @@ -44,6 +44,27 @@ BITS    32
>  ; The unexpected response code
>  %define TERM_UNEXPECTED_RESP_CODE   2
>  
> +%define PAGE_PRESENT            0x01
> +%define PAGE_READ_WRITE         0x02
> +%define PAGE_USER_SUPERVISOR    0x04
> +%define PAGE_WRITE_THROUGH      0x08
> +%define PAGE_CACHE_DISABLE     0x010
> +%define PAGE_ACCESSED          0x020
> +%define PAGE_DIRTY             0x040
> +%define PAGE_PAT               0x080
> +%define PAGE_GLOBAL           0x0100
> +%define PAGE_2M_MBO            0x080
> +%define PAGE_2M_PAT          0x01000
> +
> +%define PAGE_4K_PDE_ATTR (PAGE_ACCESSED + \
> +                          PAGE_DIRTY + \
> +                          PAGE_READ_WRITE + \
> +                          PAGE_PRESENT)
> +
> +%define PAGE_PDP_ATTR (PAGE_ACCESSED + \
> +                       PAGE_READ_WRITE + \
> +                       PAGE_PRESENT)
> +
>  
>  ; Macro is used to issue the MSR protocol based VMGEXIT. The caller is
>  ; responsible to populate values in the EDX:EAX registers. After the vmmcall
> @@ -117,6 +138,72 @@ BITS    32
>  SevEsUnexpectedRespTerminate:
>      TerminateVmgExit    TERM_UNEXPECTED_RESP_CODE
>  
> +; If SEV-ES is enabled then initialize the make the GHCB page shared

s/the make/and make/ ?

> +SevClearPageEncMaskFromGHCBPage:

Just a nit, maybe SevClearPageEncMaskForGhcbPage?

> +    ; Check if SEV is enabled
> +    cmp       byte[WORK_AREA_GUEST_TYPE], 1
> +    jnz       SevClearPageEncMaskFromGHCBPageExit
> +
> +    ; Check if SEV-ES is enabled
> +    cmp       byte[SEV_ES_WORK_AREA], 1
> +    jnz       SevClearPageEncMaskFromGHCBPageExit
> +
> +    ;
> +    ; The initial GHCB will live at GHCB_BASE and needs to be un-encrypted.
> +    ; This requires the 2MB page for this range be broken down into 512 4KB
> +    ; pages.  All will be marked encrypted, except for the GHCB.
> +    ;
> +    mov     ecx, (GHCB_BASE >> 21)
> +    mov     eax, GHCB_PT_ADDR + PAGE_PDP_ATTR
> +    mov     [ecx * 8 + PT_ADDR (0x2000)], eax
> +
> +    ;
> +    ; Page Table Entries (512 * 4KB entries => 2MB)
> +    ;
> +    mov     ecx, 512
> +pageTableEntries4kLoop:
> +    mov     eax, ecx
> +    dec     eax
> +    shl     eax, 12
> +    add     eax, GHCB_BASE & 0xFFE0_0000
> +    add     eax, PAGE_4K_PDE_ATTR
> +    mov     [ecx * 8 + GHCB_PT_ADDR - 8], eax
> +    mov     [(ecx * 8 + GHCB_PT_ADDR - 8) + 4], edx
> +    loop    pageTableEntries4kLoop
> +
> +    ;
> +    ; Clear the encryption bit from the GHCB entry
> +    ;
> +    mov     ecx, (GHCB_BASE & 0x1F_FFFF) >> 12
> +    mov     [ecx * 8 + GHCB_PT_ADDR + 4], strict dword 0
> +
> +    mov     ecx, GHCB_SIZE / 4
> +    xor     eax, eax
> +clearGhcbMemoryLoop:
> +    mov     dword[ecx * 4 + GHCB_BASE - 4], eax
> +    loop    clearGhcbMemoryLoop
> +
> +SevClearPageEncMaskFromGHCBPageExit:
> +    OneTimeCallRet SevClearPageEncMaskFromGHCBPage
> +
> +; Check if SEV is enabled, and get the C-bit mask above 31.
> +; Modified: EDX
> +;
> +; The value is returned in the EDX
> +GetSevCBitMaskAbove31:
> +    ; Check if SEV is enabled
> +    cmp       byte[WORK_AREA_GUEST_TYPE], 1
> +    jnz       NoCbitValue
> +
> +    mov       edx, dword[SEV_ES_WORK_AREA_ENC_MASK + 4]
> +    jmp       GetSevCBitMaskAbove31Exit
> +
> +NoCbitValue:
> +    xor       edx, edx

How about moving the xor as the first line of this routine and jumping to
GetSevCBitMaskAbove31Exit if the first cmp is non-zero. Then you can just
do the move from SEV_ES_WORK_AREA_ENC_MASK + 4 and eliminate the extra jmp
statement and NoCbitValue label.

Thanks,
Tom

> +
> +GetSevCBitMaskAbove31Exit:
> +    OneTimeCallRet GetSevCBitMaskAbove31
> +
>  ; Check if Secure Encrypted Virtualization (SEV) features are enabled.
>  ;
>  ; Register usage is tight in this routine, so multiple calls for the
> @@ -249,32 +336,6 @@ SevExit:
>  
>      OneTimeCallRet CheckSevFeatures
>  
> -; Check if Secure Encrypted Virtualization - Encrypted State (SEV-ES) feature
> -; is enabled.
> -;
> -; Modified:  EAX
> -;
> -; If SEV-ES is enabled then EAX will be non-zero.
> -; If SEV-ES is disabled then EAX will be zero.
> -;
> -IsSevEsEnabled:
> -    xor       eax, eax
> -
> -    ; During CheckSevFeatures, the WORK_AREA_GUEST_TYPE is set
> -    ; to 1 if SEV is enabled.
> -    cmp       byte[WORK_AREA_GUEST_TYPE], 1
> -    jne       SevEsDisabled
> -
> -    ; During CheckSevFeatures, the SEV_ES_WORK_AREA was set to 1 if
> -    ; SEV-ES is enabled.
> -    cmp       byte[SEV_ES_WORK_AREA], 1
> -    jne       SevEsDisabled
> -
> -    mov       eax, 1
> -
> -SevEsDisabled:
> -    OneTimeCallRet IsSevEsEnabled
> -
>  ; Start of #VC exception handling routines
>  ;
>  
> diff --git a/OvmfPkg/ResetVector/Ia32/PageTables64.asm b/OvmfPkg/ResetVector/Ia32/PageTables64.asm
> index f688909f1c7d..0e8ba4dde534 100644
> --- a/OvmfPkg/ResetVector/Ia32/PageTables64.asm
> +++ b/OvmfPkg/ResetVector/Ia32/PageTables64.asm
> @@ -46,16 +46,13 @@ SetCr3ForPageTables64:
>      ; work area when detected.
>      mov     byte[WORK_AREA_GUEST_TYPE], 0
>  
> +    ; Check whether the SEV is active and populate the SevEsWorkArea
>      OneTimeCall   CheckSevFeatures
> -    xor     edx, edx
> -    test    eax, eax
> -    jz      SevNotActive
>  
> -    ; If SEV is enabled, C-bit is always above 31
> -    sub     eax, 32
> -    bts     edx, eax
> -
> -SevNotActive:
> +    ; If SEV is enabled, the C-bit position is always above 31.
> +    ; The mask will be saved in the EDX and applied during the
> +    ; the page table build below.
> +    OneTimeCall   GetSevCBitMaskAbove31
>  
>      ;
>      ; For OVMF, build some initial page tables at
> @@ -105,44 +102,8 @@ pageTableEntriesLoop:
>      mov     [(ecx * 8 + PT_ADDR (0x2000 - 8)) + 4], edx
>      loop    pageTableEntriesLoop
>  
> -    OneTimeCall   IsSevEsEnabled
> -    test    eax, eax
> -    jz      SetCr3
> -
> -    ;
> -    ; The initial GHCB will live at GHCB_BASE and needs to be un-encrypted.
> -    ; This requires the 2MB page for this range be broken down into 512 4KB
> -    ; pages.  All will be marked encrypted, except for the GHCB.
> -    ;
> -    mov     ecx, (GHCB_BASE >> 21)
> -    mov     eax, GHCB_PT_ADDR + PAGE_PDP_ATTR
> -    mov     [ecx * 8 + PT_ADDR (0x2000)], eax
> -
> -    ;
> -    ; Page Table Entries (512 * 4KB entries => 2MB)
> -    ;
> -    mov     ecx, 512
> -pageTableEntries4kLoop:
> -    mov     eax, ecx
> -    dec     eax
> -    shl     eax, 12
> -    add     eax, GHCB_BASE & 0xFFE0_0000
> -    add     eax, PAGE_4K_PDE_ATTR
> -    mov     [ecx * 8 + GHCB_PT_ADDR - 8], eax
> -    mov     [(ecx * 8 + GHCB_PT_ADDR - 8) + 4], edx
> -    loop    pageTableEntries4kLoop
> -
> -    ;
> -    ; Clear the encryption bit from the GHCB entry
> -    ;
> -    mov     ecx, (GHCB_BASE & 0x1F_FFFF) >> 12
> -    mov     [ecx * 8 + GHCB_PT_ADDR + 4], strict dword 0
> -
> -    mov     ecx, GHCB_SIZE / 4
> -    xor     eax, eax
> -clearGhcbMemoryLoop:
> -    mov     dword[ecx * 4 + GHCB_BASE - 4], eax
> -    loop    clearGhcbMemoryLoop
> +    ; Clear the C-bit from the GHCB page if the SEV-ES is enabled.
> +    OneTimeCall   SevClearPageEncMaskFromGHCBPage
>  
>  SetCr3:
>      ;
> 

  reply	other threads:[~2021-08-09 17:03 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-08-04 20:20 [PATCH 0/3] reuse the SevEsWork area Brijesh Singh
2021-08-04 20:20 ` [PATCH 1/3] OvmfPkg: introduce a common work area Brijesh Singh
2021-08-09 16:40   ` Lendacky, Thomas
2021-08-09 17:18     ` Brijesh Singh
2021-08-04 20:20 ` [PATCH 2/3] OvmfPkg/ResetVector: update SEV support to use new work area format Brijesh Singh
2021-08-09 16:54   ` Lendacky, Thomas
2021-08-09 17:21     ` Brijesh Singh
2021-08-04 20:20 ` [PATCH 3/3] OvmfPkg/ResetVector: move the GHCB page setup in AmdSev.asm Brijesh Singh
2021-08-09 17:03   ` Lendacky, Thomas [this message]
2021-08-05  2:18 ` [edk2-devel] [PATCH 0/3] reuse the SevEsWork area Yao, Jiewen
2021-08-05 11:31   ` Brijesh Singh
2021-08-05 14:43   ` Brijesh Singh
2021-08-05 15:07     ` Yao, Jiewen

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=298e16b9-f44c-a10d-a42e-92988e4725c1@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