On 24/05/18 16:42, Roger Pau Monné wrote:
> On Tue, May 22, 2018 at 12:20:45PM +0100, Andrew Cooper wrote:
>> Up until this point, the MSR load/save lists have only ever accumulated
>> content. Introduce vmx_del_msr() as a companion to vmx_add_msr().
>>
>> Signed-off-by: Andrew Cooper
>> ---
>> CC: Jan Beulich
>> CC: Jun Nakajima
>> CC: Kevin Tian
>> CC: Wei Liu
>> CC: Roger Pau Monné
>> ---
>> xen/arch/x86/hvm/vmx/vmcs.c| 68
>> ++
>> xen/include/asm-x86/hvm/vmx/vmcs.h | 1 +
>> 2 files changed, 69 insertions(+)
>>
>> diff --git a/xen/arch/x86/hvm/vmx/vmcs.c b/xen/arch/x86/hvm/vmx/vmcs.c
>> index 7bf19a0..e1a8f95 100644
>> --- a/xen/arch/x86/hvm/vmx/vmcs.c
>> +++ b/xen/arch/x86/hvm/vmx/vmcs.c
>> @@ -1465,6 +1465,74 @@ int vmx_add_msr(struct vcpu *v, uint32_t msr,
>> uint64_t val,
>> return rc;
>> }
>>
>> +int vmx_del_msr(struct vcpu *v, uint32_t msr, enum vmx_msr_list_type type)
>> +{
>> +struct arch_vmx_struct *arch_vmx = &v->arch.hvm_vmx;
>> +struct vmx_msr_entry *start = NULL, *ent, *end;
>> +unsigned int substart, subend, total;
>> +
>> +ASSERT(v == current || !vcpu_runnable(v));
>> +
>> +switch ( type )
>> +{
>> +case VMX_MSR_HOST:
>> +start= arch_vmx->host_msr_area;
>> +substart = 0;
>> +subend = arch_vmx->host_msr_count;
>> +total= subend;
>> +break;
>> +
>> +case VMX_MSR_GUEST:
>> +start= arch_vmx->msr_area;
>> +substart = 0;
>> +subend = arch_vmx->msr_save_count;
>> +total= arch_vmx->msr_load_count;
>> +break;
>> +
>> +case VMX_MSR_GUEST_LOADONLY:
>> +start= arch_vmx->msr_area;
>> +substart = arch_vmx->msr_save_count;
>> +subend = arch_vmx->msr_load_count;
>> +total= subend;
>> +break;
>> +
>> +default:
>> +ASSERT_UNREACHABLE();
>> +}
> The above chunk is already in vmx_find_msr and vmx_add_msr, maybe a
> static helper that sets start/substart/subend/total would be helpful
> here?
Its actually more than that. They are identical identical until after
the locate_msr_entry() call. The problem is that I can't find a clean
way of collecting the logic which is also readable.
I'd also like to fold together the add side, but that is further
complicated by the memory allocation logic.
~Andrew
___
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
