Re: [PATCH v2] x86: make Viridian support optional

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 24/03/2025 1:37 pm, Jan Beulich wrote:
> On 21.03.2025 10:26, Sergiy Kibrik wrote:
>> --- a/xen/arch/x86/hvm/hvm.c
>> +++ b/xen/arch/x86/hvm/hvm.c
>> @@ -695,9 +695,12 @@ int hvm_domain_initialise(struct domain *d,
>>      if ( hvm_tsc_scaling_supported )
>>          d->arch.hvm.tsc_scaling_ratio = hvm_default_tsc_scaling_ratio;
>>  
>> -    rc = viridian_domain_init(d);
>> -    if ( rc )
>> -        goto fail2;
>> +    if ( is_viridian_domain(d) )
>> +    {
>> +        rc = viridian_domain_init(d);
>> +        if ( rc )
>> +            goto fail2;
>> +    }
> It is, aiui, this conditional initialization which causes
>
> (XEN) [    9.136463] UBSAN: Undefined behaviour in 
> arch/x86/hvm/viridian/viridian.c:1118:43
> (XEN) [    9.136465] member access within null pointer of type 'const struct 
> viridian_domain'
> (XEN) [    9.136471] ----[ Xen-4.21-unstable  x86_64  debug=y ubsan=y  Not 
> tainted ]----
> (XEN) [    9.136473] CPU:    14
> (XEN) [    9.136475] RIP:    e008:[<ffff82d04032e8a8>] 
> common/ubsan/ubsan.c#ubsan_epilogue+0xa/0xd9
> (XEN) [    9.136480] RFLAGS: 0000000000010096   CONTEXT: hypervisor (d0v6)
> (XEN) [    9.136484] rax: 0000000000000000   rbx: ffff8304761afb40   rcx: 
> 0000000000000868
> (XEN) [    9.136486] rdx: ffff8304761affd0   rsi: 0000000000000004   rdi: 
> ffff8304761afb40
> (XEN) [    9.136488] rbp: ffff8304761afb30   rsp: ffff8304761afb20   r8:  
> 0000000000000006
> (XEN) [    9.136490] r9:  00000000ffffffff   r10: 0000000000000000   r11: 
> 0000000000000000
> (XEN) [    9.136493] r12: ffff82d040913d22   r13: 0000000000000003   r14: 
> ffff82d040b237a0
> (XEN) [    9.136495] r15: 000000000000000f   cr0: 0000000080050033   cr4: 
> 0000000000b526e0
> (XEN) [    9.136497] cr3: 0000000308528000   cr2: 00007f24a0c9de28
> (XEN) [    9.136499] fsb: 00007f24a3fd6b48   gsb: ffff8881b9b80000   gss: 
> 0000000000000000
> (XEN) [    9.136501] ds: 0000   es: 0000   fs: 0000   gs: 0000   ss: e010   
> cs: e008
> (XEN) [    9.136505] Xen code around <ffff82d04032e8a8> 
> (common/ubsan/ubsan.c#ubsan_epilogue+0xa/0xd9):
> (XEN) [    9.136507]  89 e5 41 54 53 48 89 fb <0f> 0b 48 8d 3d ff 92 37 00 e8 
> 9d 99 00 00 48 85
> (XEN) [    9.136572] Xen stack trace from rsp=ffff8304761afb20:
> (XEN) [    9.136574]    0000000000000003 ffff82d040913d22 ffff8304761afb70 
> ffff82d04032ef58
> (XEN) [    9.136578]    0000000000000202 ffff82d040911100 0000000000000000 
> 0000000000000000
> (XEN) [    9.136581]    ffff82d040b237a0 000000000000000f ffff8304761afbd0 
> ffff82d04032f3af
> (XEN) [    9.136585]    ffff82d0406b8900 0000002b8000045e ffff82d040913d1e 
> 0000000000000001
> (XEN) [    9.136588]    ffff82d040b23703 000000000000000f ffff8304761afbc0 
> ffff830405d65000
> (XEN) [    9.136592]    ffff8304761afd00 0000000000000000 ffff8304761afc18 
> ffff82d040418b9c
> (XEN) [    9.136595]    0000000000000000 0000000000000000 0000000000000000 
> 0000000000000000
> (XEN) [    9.136598]    ffff82d040b23a88 00000000000002e8 ffff830401d6cfe0 
> ffff8304761afc98
> (XEN) [    9.136602]    ffff82d04045fc14 00000000000002c0 ffff82d040b23a58 
> ffff8304761afd00
> (XEN) [    9.136605]    ffff82d040418a0a ffff830405d65000 ffff82d040b23a88 
> 0000000154381286
> (XEN) [    9.136609]    0000000000000000 00383ffe00090672 0000000000001c78 
> fffffffffffffff4
> (XEN) [    9.136612]    ffff830476161000 ffff8304761afdb0 ffff830405d65000 
> ffff8304761afd50
> (XEN) [    9.136616]    ffff82d040630690 800000030847e127 0000000000000206 
> 000000000000000e
> (XEN) [    9.136619]    ffff82d040a00b44 ffff82d040a00b40 ffff82d040a00b44 
> ffff83047618f000
> (XEN) [    9.136622]    00007f24a3c7a010 ffff8304761afef8 ffff8304761afef8 
> ffff8304761b5088
> (XEN) [    9.136626]    00001c7800000580 ffff830483a24000 0000000000000001 
> 00007f24a3c7a010
> (XEN) [    9.136629]    0000000000000001 ffff830405d65000 00007f24a3c7a010 
> 0000000000000001
> (XEN) [    9.136633]    0000000000000021 0000000000000000 ffff8304761afe68 
> ffff82d0402c662f
> (XEN) [    9.136636]    ffff8304761afd80 ffff82d0405a0fbb ffff830476161000 
> ffff8881b9b80000
> (XEN) [    9.136640]    ffff8304761afdb8 ffff82d04052fdef ffff830476161000 
> ffff8881b9b80000
> (XEN) [    9.136643] Xen call trace:
> (XEN) [    9.136646]    [<ffff82d04032e8a8>] R 
> common/ubsan/ubsan.c#ubsan_epilogue+0xa/0xd9
> (XEN) [    9.136648]    [<ffff82d04032ef58>] F 
> __ubsan_handle_type_mismatch+0x135/0x48e
> (XEN) [    9.136651]    [<ffff82d04032f3af>] F 
> __ubsan_handle_type_mismatch_v1+0xfe/0x100
> (XEN) [    9.136655]    [<ffff82d040418b9c>] F 
> arch/x86/hvm/viridian/viridian.c#viridian_save_domain_ctxt+0x192/0x204
> (XEN) [    9.136658]    [<ffff82d04045fc14>] F hvm_save+0x4d1/0x5c6
> (XEN) [    9.136661]    [<ffff82d040630690>] F arch_do_domctl+0x2f8/0x51aa
> (XEN) [    9.136664]    [<ffff82d0402c662f>] F do_domctl+0xc91/0x460d
> (XEN) [    9.136667]    [<ffff82d04053ab70>] F pv_hypercall+0x6d6/0x834
> (XEN) [    9.136669]    [<ffff82d0402012d3>] F lstar_enter+0x143/0x150
> (XEN) [    9.136671] 
> (XEN) [    9.136673] 
> ================================================================================
> (XEN) [    9.622282] 
> (XEN) [    9.624469] ****************************************
> (XEN) [    9.630471] Panic on CPU 14:
> (XEN) [    9.634182] UBSAN failure detected
> (XEN) [    9.638469] ****************************************
>
> throughout tests on actual hardware in CI. And really I don't think it's just 
> UB, but
> rather a real NULL deref:

This is a consequence of us turning on UBSAN generally in GitlabCI, and
will be the norm moving forward.

Nevertheless, "member access within null pointer of type $X" is very
clear cut.

~Andrew