While part of the original AMD64 spec, Long Mode Segment Limit Enable was a
feature not picked up by Intel, and therefore didn't see much adoption in
software.  AMD have finally dropped the feature from hardware, and allocated a
CPUID bit to indicate its absence.

Xen has never supported the feature for guests, even when running on capable
hardware, so advertise the feature's absence unilaterally.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
---
CC: Jan Beulich <JBeulich@suse.com>
CC: Roger Pau Monné <roger.pau@citrix.com>
CC: Wei Liu <wl@xen.org>

There is nothing specifically wrong with exposing this bit to PV guests, but
the PV ABI doesn't include a working concept of MSR_EFER in the first place,
so exposing it to PV guests seems somewhat out-of-place.
---
 tools/libs/light/libxl_cpuid.c              | 1 +
 tools/misc/xen-cpuid.c                      | 1 +
 xen/arch/x86/cpuid.c                        | 6 ++++++
 xen/include/public/arch-x86/cpufeatureset.h | 1 +
 4 files changed, 9 insertions(+)

diff --git a/tools/libs/light/libxl_cpuid.c b/tools/libs/light/libxl_cpuid.c
index 289c59c742..be0bd81a9b 100644
--- a/tools/libs/light/libxl_cpuid.c
+++ b/tools/libs/light/libxl_cpuid.c
@@ -270,6 +270,7 @@ int libxl_cpuid_parse_config(libxl_cpuid_policy_list *cpuid, const char* str)
         {"rstr-fp-err-ptrs", 0x80000008, NA, CPUID_REG_EBX, 2, 1},
         {"wbnoinvd",     0x80000008, NA, CPUID_REG_EBX,  9,  1},
         {"ibpb",         0x80000008, NA, CPUID_REG_EBX, 12,  1},
+        {"no-lmsle",     0x80000008, NA, CPUID_REG_EBX, 20,  1},
         {"ppin",         0x80000008, NA, CPUID_REG_EBX, 23,  1},
 
         {"nc",           0x80000008, NA, CPUID_REG_ECX,  0,  8},
diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c
index 2d04162d8d..1a82089463 100644
--- a/tools/misc/xen-cpuid.c
+++ b/tools/misc/xen-cpuid.c
@@ -152,6 +152,7 @@ static const char *const str_e8b[32] =
 
     [12] = "ibpb",
 
+    [20] = "no-lmsle",
     /* [22] */                 [23] = "ppin",
 };
 
diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c
index 18b08d9b87..3ed0feac4a 100644
--- a/xen/arch/x86/cpuid.c
+++ b/xen/arch/x86/cpuid.c
@@ -456,6 +456,12 @@ static void __init calculate_hvm_max_policy(void)
     __set_bit(X86_FEATURE_X2APIC, hvm_featureset);
 
     /*
+     * We don't support EFER.LMSLE at all.  AMD has dropped the feature from
+     * hardware and allocated a CPUID bit to indicate its absence.
+     */
+    __set_bit(X86_FEATURE_NO_LMSLE, hvm_featureset);
+
+    /*
      * On AMD, PV guests are entirely unable to use SYSENTER as Xen runs in
      * long mode (and init_amd() has cleared it out of host capabilities), but
      * HVM guests are able if running in protected mode.
diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h
index a501479820..0a4690decc 100644
--- a/xen/include/public/arch-x86/cpufeatureset.h
+++ b/xen/include/public/arch-x86/cpufeatureset.h
@@ -255,6 +255,7 @@ XEN_CPUFEATURE(CLZERO,        8*32+ 0) /*A  CLZERO instruction */
 XEN_CPUFEATURE(RSTR_FP_ERR_PTRS, 8*32+ 2) /*A  (F)X{SAVE,RSTOR} always saves/restores FPU Error pointers */
 XEN_CPUFEATURE(WBNOINVD,      8*32+ 9) /*   WBNOINVD instruction */
 XEN_CPUFEATURE(IBPB,          8*32+12) /*A  IBPB support only (no IBRS, used by AMD) */
+XEN_CPUFEATURE(NO_LMSLE,      8*32+20) /*S  EFER.LMSLE no longer supported. */
 XEN_CPUFEATURE(AMD_PPIN,      8*32+23) /*   Protected Processor Inventory Number */
 
 /* Intel-defined CPU features, CPUID level 0x00000007:0.edx, word 9 */
-- 
2.11.0

On 12.04.2021 12:22, Andrew Cooper wrote:
> --- a/xen/arch/x86/cpuid.c
> +++ b/xen/arch/x86/cpuid.c
> @@ -456,6 +456,12 @@ static void __init calculate_hvm_max_policy(void)
>      __set_bit(X86_FEATURE_X2APIC, hvm_featureset);
>  
>      /*
> +     * We don't support EFER.LMSLE at all.  AMD has dropped the feature from
> +     * hardware and allocated a CPUID bit to indicate its absence.
> +     */
> +    __set_bit(X86_FEATURE_NO_LMSLE, hvm_featureset);

Why only for HVM? And shouldn't the LM: entry in the dependencies
table be adjusted such that !LM implies this bit clear?

Jan

On 12/04/2021 11:48, Jan Beulich wrote:
> On 12.04.2021 12:22, Andrew Cooper wrote:
>> --- a/xen/arch/x86/cpuid.c
>> +++ b/xen/arch/x86/cpuid.c
>> @@ -456,6 +456,12 @@ static void __init calculate_hvm_max_policy(void)
>>      __set_bit(X86_FEATURE_X2APIC, hvm_featureset);
>>  
>>      /*
>> +     * We don't support EFER.LMSLE at all.  AMD has dropped the feature from
>> +     * hardware and allocated a CPUID bit to indicate its absence.
>> +     */
>> +    __set_bit(X86_FEATURE_NO_LMSLE, hvm_featureset);
> Why only for HVM?

That was discussed.

> And shouldn't the LM: entry in the dependencies
> table be adjusted such that !LM implies this bit clear?

Probably.

~Andrew

On 12.04.2021 13:39, Andrew Cooper wrote:
> On 12/04/2021 11:48, Jan Beulich wrote:
>> On 12.04.2021 12:22, Andrew Cooper wrote:
>>> --- a/xen/arch/x86/cpuid.c
>>> +++ b/xen/arch/x86/cpuid.c
>>> @@ -456,6 +456,12 @@ static void __init calculate_hvm_max_policy(void)
>>>      __set_bit(X86_FEATURE_X2APIC, hvm_featureset);
>>>  
>>>      /*
>>> +     * We don't support EFER.LMSLE at all.  AMD has dropped the feature from
>>> +     * hardware and allocated a CPUID bit to indicate its absence.
>>> +     */
>>> +    __set_bit(X86_FEATURE_NO_LMSLE, hvm_featureset);
>> Why only for HVM?
> 
> That was discussed.

Oh, in a post-commit-message remark which I did manage to skip over.
Would you mind making this part of the commit message?

Jan