Re: [Xen-devel] [PATCH 08/10] pvh/acpi: Handle ACPI accesses for PVH guests

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 07.11.16 at 16:55, <konrad.wilk@xxxxxxxxxx> wrote:
> On Sun, Nov 06, 2016 at 04:42:41PM -0500, Boris Ostrovsky wrote:
>> Signed-off-by: Boris Ostrovsky <boris.ostrovsky@xxxxxxxxxx>
>> ---
>> CC: Paul Durrant <paul.durrant@xxxxxxxxxx>
>> ---
>>  xen/arch/x86/hvm/ioreq.c | 66 
> ++++++++++++++++++++++++++++++++++++++++++++++++
>>  1 file changed, 66 insertions(+)
>> 
>> diff --git a/xen/arch/x86/hvm/ioreq.c b/xen/arch/x86/hvm/ioreq.c
>> index 171ea82..ced7c92 100644
>> --- a/xen/arch/x86/hvm/ioreq.c
>> +++ b/xen/arch/x86/hvm/ioreq.c
>> @@ -1392,6 +1392,72 @@ void hvm_ioreq_init(struct domain *d)
>>  static int acpi_ioaccess(
>>      int dir, unsigned int port, unsigned int bytes, uint32_t *val)
>>  {
>> +    unsigned int i;
>> +    unsigned int bits = bytes * 8;
>> +    uint8_t *reg = NULL;
>> +    unsigned idx = port & 3;
>> +    bool is_cpu_map = 0;
>> +    struct domain *currd = current->domain;
>> +
>> +    BUILD_BUG_ON((ACPI_PM1A_EVT_BLK_LEN != 4) ||
>> +                 (ACPI_GPE0_BLK_LEN_V1 != 4));
>> +
>> +    switch (port)
>> +    {
>> +    case ACPI_PM1A_EVT_BLK_ADDRESS_V1 ...
>> +        (ACPI_PM1A_EVT_BLK_ADDRESS_V1 + ACPI_PM1A_EVT_BLK_LEN - 1):
>> +        reg = currd->arch.hvm_domain.acpi_io.pm1a;
>> +        break;
>> +    case ACPI_GPE0_BLK_ADDRESS_V1 ...
>> +        (ACPI_GPE0_BLK_ADDRESS_V1 + ACPI_GPE0_BLK_LEN_V1 - 1):
>> +        reg = currd->arch.hvm_domain.acpi_io.gpe;
>> +        break;
>> +    case 0xaf00 ... (0xaf00 + HVM_MAX_VCPUS/8 - 1):
> 
> That may need some documentation or a #define perhaps?
> 
> Also just in case somebody decided it was funny and compile Xen with
> HVM_MAX_VCPUS set to say 4, won't this go in 0xfffffff region?
> 
> You may want to add a BUILD_BUG_ON for the HVM_MAX_VCPUS, like:
> 
> BUILD_BUGON(HVM_MAX_VCPUS > 8)?

No. The expression wants to be 0xaf00 + (HVM_MAX_VCPUS - 1) / 8.
And a BUILD_BUG_ON() should check the range doesn't spill into
anything later (presumably starting at 0xb000).

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel