[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <50aa7df3f34f55ec435c4f489d8f3df3@ispras.ru>
Date: Sun, 20 Nov 2022 18:37:40 +0300
From: Evgeniy Baskov <baskov@...ras.ru>
To: joeyli <jlee@...e.com>
Cc: Ard Biesheuvel <ardb@...nel.org>, Borislav Petkov <bp@...en8.de>,
Andy Lutomirski <luto@...nel.org>,
Dave Hansen <dave.hansen@...ux.intel.com>,
Ingo Molnar <mingo@...hat.com>,
Peter Zijlstra <peterz@...radead.org>,
Thomas Gleixner <tglx@...utronix.de>,
Alexey Khoroshilov <khoroshilov@...ras.ru>,
Peter Jones <pjones@...hat.com>, lvc-project@...uxtesting.org,
x86@...nel.org, linux-efi@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-hardening@...r.kernel.org
Subject: Re: [PATCH v2 00/23] x86_64: Improvements at compressed kernel stage
On 2022-11-20 04:49, joeyli wrote:
> Hi Evgeniy,
>
> Thanks for your effort!
>
> On Tue, Oct 25, 2022 at 05:12:38PM +0300, Evgeniy Baskov wrote:
...
>
> Because Peter Jones point out this patchset to me, so I tried it
> on OVMF, and I set the EfiLoaderData in DXE memory protection policy:
>
> Index: edk2/MdeModulePkg/MdeModulePkg.dec
> ===================================================================
> --- edk2.orig/MdeModulePkg/MdeModulePkg.dec
> +++ edk2/MdeModulePkg/MdeModulePkg.dec
> @@ -1392,7 +1392,7 @@
> # e.g. 0x7BD4 can be used for all memory except Code and
> ACPINVS/Reserved. <BR>
> #
> # @Prompt Set DXE memory protection policy.
> -
> gEfiMdeModulePkgTokenSpaceGuid.PcdDxeNxMemoryProtectionPolicy|0x0000000|UINT64|0x00001048^M
> +
> gEfiMdeModulePkgTokenSpaceGuid.PcdDxeNxMemoryProtectionPolicy|0x0000004|UINT64|0x00001048^M
>
> ## PCI Serial Device Info. It is an array of Device, Function, and
> Power Management
> # information that describes the path that contains zero or more
> PCI to PCI bridges
>
>
> I applied this v2 patch set on top of v6.1-rc5 kernel, and boot with a
> shim which
> set the PE NX-compatibility DLL Characteristic flag. I got a page
> fault exception:
>
> Loading Linux 6.1.0-rc5-default+ ...
> Loading initial ramdisk ...
> !!!! X64 Exception Type - 0E(#PF - Page-Fault) CPU Apic ID - 00000000
> !!!!
> ExceptionData - 0000000000000011 I:1 R:0 U:0 W:0 P:1 PK:0 SS:0 SGX:0
> RIP - 0000000076A3C390, CS - 0000000000000038, RFLAGS -
> 0000000000210202
> RAX - 000000007D8CCDF8, RCX - 0000000076A3C390, RDX - 000000007DE86000
> RBX - 0000000076A3C000, RSP - 000000007FF0D2C8, RBP - 000000007DE86000
> RSI - 000000007F9EE018, RDI - 000000007DFD1C18
> R8 - 0000000076A3C000, R9 - 0000000000000190, R10 - 000000007FF1D658
> R11 - 0000000000000004, R12 - 0000000000000190, R13 - 000000007D8CCE00
> R14 - 000000007D8C76B4, R15 - 000000007BF0CBD5
> DS - 0000000000000030, ES - 0000000000000030, FS - 0000000000000030
> GS - 0000000000000030, SS - 0000000000000030
> CR0 - 0000000080010033, CR2 - 0000000076A3C390, CR3 - 000000007FC01000
> CR4 - 0000000000000668, CR8 - 0000000000000000
> DR0 - 0000000000000000, DR1 - 0000000000000000, DR2 - 0000000000000000
> DR3 - 0000000000000000, DR6 - 00000000FFFF0FF0, DR7 - 0000000000000400
> GDTR - 000000007F9DE000 0000000000000047, LDTR - 0000000000000000
> IDTR - 000000007F2E9018 0000000000000FFF, TR - 0000000000000000
> FXSAVE_STATE - 000000007FF0CF20
> !!!! Find image based on IP(0x7BF0BAB5)
> /mnt/working/source_code-git/edk2/Build/OvmfX64/DEBUG_GCC5/X64/MdeModulePkg/Universal/Variable/RuntimeDxe/VariableRuntimeDxe/DEBUG/VariableRuntimeDxe.dll
> (ImageBase=0000000000F40E7C, EntryPoint=0000000000F767B8) !!!!
>
>
> My question is: Can I just set EfiLoaderData in DXE memory protection
> policy
> in EDK2/OVMF to test this patchset? Or which platform (virtual or
> physical)
> can we use for testing?
>
> Thanks a lot!
> Joey Lee
Hi,
Thank you for testing!
The EDK2 OVMF with adjusted PcdDxeNxMemoryProtectionPolicy
should work with the kernel itself.
As far as I can see from my testing with EDK2 OVMF, this
#PF occurred inside GRUB code before the first instruction
of the kernel. You can try using GRUB master branch, where
allocations are changed to use EfiLoaderCode type and with
PcdDxeNxMemoryProtectionPolicy=0x0000004 it will still be
executable.
Thanks,
Evgeniy Baskov
Powered by blists - more mailing lists