Re: [PATCH v4] Avoid crash calling PrintErrMesg from efi_multiboot2

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 26/09/2024 2:26 pm, Marek Marczykowski-Górecki wrote:
> On Mon, Aug 19, 2024 at 03:29:52PM +0100, Frediano Ziglio wrote:
>> Although code is compiled with -fpic option data is not position
>> independent. This causes data pointer to become invalid if
>> code is not relocated properly which is what happens for
>> efi_multiboot2 which is called by multiboot entry code.
>>
>> Code tested adding
>>    PrintErrMesg(L"Test message", EFI_BUFFER_TOO_SMALL);
>> in efi_multiboot2 before calling efi_arch_edd (this function
>> can potentially call PrintErrMesg).
>>
>> Before the patch (XenServer installation on Qemu, xen replaced
>> with vanilla xen.gz):
>>   Booting `XenServer (Serial)'Booting `XenServer (Serial)'
>>   Test message: !!!! X64 Exception Type - 0E(#PF - Page-Fault)  CPU Apic ID 
>> - 00000000 !!!!
>>   ExceptionData - 0000000000000000  I:0 R:0 U:0 W:0 P:0 PK:0 SS:0 SGX:0
>>   RIP  - 000000007DC29E46, CS  - 0000000000000038, RFLAGS - 0000000000210246
>>   RAX  - 0000000000000000, RCX - 0000000000000050, RDX - 0000000000000000
>>   RBX  - 000000007DAB4558, RSP - 000000007EFA1200, RBP - 0000000000000000
>>   RSI  - FFFF82D040467A88, RDI - 0000000000000000
>>   R8   - 000000007EFA1238, R9  - 000000007EFA1230, R10 - 0000000000000000
>>   R11  - 000000007CF42665, R12 - FFFF82D040467A88, R13 - 000000007EFA1228
>>   R14  - 000000007EFA1225, R15 - 000000007DAB45A8
>>   DS   - 0000000000000030, ES  - 0000000000000030, FS  - 0000000000000030
>>   GS   - 0000000000000030, SS  - 0000000000000030
>>   CR0  - 0000000080010033, CR2 - FFFF82D040467A88, CR3 - 000000007EC01000
>>   CR4  - 0000000000000668, CR8 - 0000000000000000
>>   DR0  - 0000000000000000, DR1 - 0000000000000000, DR2 - 0000000000000000
>>   DR3  - 0000000000000000, DR6 - 00000000FFFF0FF0, DR7 - 0000000000000400
>>   GDTR - 000000007E9E2000 0000000000000047, LDTR - 0000000000000000
>>   IDTR - 000000007E4E5018 0000000000000FFF,   TR - 0000000000000000
>>   FXSAVE_STATE - 000000007EFA0E60
>>   !!!! Find image based on IP(0x7DC29E46) (No PDB)  
>> (ImageBase=000000007DC28000, EntryPoint=000000007DC2B917) !!!!
>>
>> After the patch:
>>   Booting `XenServer (Serial)'Booting `XenServer (Serial)'
>>   Test message: Buffer too small
>>   BdsDxe: loading Boot0000 "UiApp" from 
>> Fv(7CB8BDC9-F8EB-4F34-AAEA-3EE4AF6516A1)/FvFile(462CAA21-7614-4503-836E-8AB6F4662331)
>>   BdsDxe: starting Boot0000 "UiApp" from 
>> Fv(7CB8BDC9-F8EB-4F34-AAEA-3EE4AF6516A1)/FvFile(462CAA21-7614-4503-836E-8AB6F4662331)
>>
>> Fixes: 9180f5365524 ("x86: add multiboot2 protocol support for EFI 
>> platforms")
>> Signed-off-by: Frediano Ziglio <frediano.ziglio@xxxxxxxxx>
> I was hoping it would fix also an issue with xen.efi as the crash is
> pretty similar
> (https://github.com/QubesOS/qubes-issues/issues/8206#issuecomment-2366835136),
> but it seems to be something different.
>
> Anyway, 
>
> Acked-by: Marek Marczykowski-Górecki <marmarek@xxxxxxxxxxxxxxxxxxxxxx>

While I hate to drag this on further, there's a very relevant observation in

https://lore.kernel.org/xen-devel/20240925150059.3955569-31-ardb+git@xxxxxxxxxx/T/#u

which was posted yesterday.  Exactly the same is true of the early MB2
code calling into regular EFI code, and I wonder if it's causing the
other issue too.

I suspect following that pattern will be rather more robust.  Thoughts?

~Andrew