Re: [PATCH v6 10/11] tools, libxl: handle the iomem parameter with the memory_mapping hcall

[Julien Grall <julien.grall@linaro.org>]

On 04/28/2014 02:25 PM, Ian Campbell wrote:
> On Tue, 2014-04-22 at 21:03 +0100, Julien Grall wrote:
>> Hi Arianna,
>>
>> On 21/04/14 14:45, Arianna Avanzini wrote:
>>> diff --git a/tools/libxl/libxl_create.c b/tools/libxl/libxl_create.c
>>> index 6aa630e..cdf03cd 100644
>>> --- a/tools/libxl/libxl_create.c
>>> +++ b/tools/libxl/libxl_create.c
>>> @@ -1113,6 +1113,19 @@ static void domcreate_launch_dm(libxl__egc *egc, libxl__multidev *multidev,
>>>                    "failed give dom%d access to iomem range %"PRIx64"-%"PRIx64,
>>>                    domid, io->start, io->start + io->number - 1);
>>>               ret = ERROR_FAIL;
>>> +            continue;
>>> +        }
>>> +        if (!libxl__arch_auto_translated_physmap(d_config))
>>> +            continue;
>>> +        ret = xc_domain_memory_mapping(CTX->xch, domid,
>>> +                                       io->gfn, io->start,
>>> +                                       io->number, 1);
>>> +        if (ret < 0) {
>>> +            LOGE(ERROR,
>>> +                 "failed to map to dom%d iomem range %"PRIx64"-%"PRIx64
>>> +                 " to guest address %"PRIx64,
>>> +                 domid, io->start, io->start + io->number - 1, io->gfn);
>>> +            ret = ERROR_FAIL;
>>
>> I know I've asked you to create the small helper 
>> "libxl__arch_auto_translated_physmap" in libxl. I'm wondering if we can 
>> replace xc_domain_iomem_permission by xc_domain_memory_mapping.
> 
> The path we've decided to go down is to separate the granting/revoking
> of permissions from the establishment of mappings, isn't it?
> 
> Doing as you suggest would be contrary to that plan I think.

Hmmm right. Sorry, I start to get lost with all the mail.

In another case, as the {,un}map code do nothing for translated domain
(return 0). Shall we let the toolstack call this function unconditionally?

Regards,


-- 
Julien Grall