Re: [Qemu-devel] [v3][PATCH 2/2] libxl: introduce gfx_passthru_kind

["Chen, Tiejun" <tiejun.chen@intel.com>]

> Perhaps add "With qemu-xen-traditional IGD is always assumed and other
> options than autodetect or explicit IGD will result in an error"?

Will do.

>
>> diff --git a/tools/libxl/libxl_dm.c b/tools/libxl/libxl_dm.c
>> index a8b08f2..4fd6310 100644
>> --- a/tools/libxl/libxl_dm.c
>> +++ b/tools/libxl/libxl_dm.c
>> @@ -325,7 +325,15 @@ static char **
>> libxl__build_device_model_args_old(libxl__gc *gc,
>>                flexarray_vappend(dm_args, "-net", "none", NULL);
>>            }
>>            if (libxl_defbool_val(b_info->u.hvm.gfx_passthru)) {
>> -            flexarray_append(dm_args, "-gfx_passthru");
>> +            switch (b_info->u.hvm.gfx_passthru_kind) {
>> +            case LIBXL_GFX_PASSTHRU_KIND_DEFAULT:
>> +            case LIBXL_GFX_PASSTHRU_KIND_IGD:
>> +                flexarray_append(dm_args, "-gfx_passthru");
>> +                break;
>> +            default:
>> +                LOG(ERROR, "unsupported gfx_passthru_kind.\n");
>
> Sorry, LOG should not get a \n like my example had, my fault.

Actually myself really should double check this.

>
> With that if you resend the series with git send-email (so it doesn't
> get whitespace mangled) I think we are good to go!
>

Currently Qemu maintainers are busy finalizing qemu 2.3, they don't 
complete to review all associated qemu patch set. Although that don't 
bring any change to our two patches on Xen side, I think we'd better 
merge these patches until qemu patches are really applied into qemu 
tree. So I will send this series again until we can really consume this 
with qemu upstream, right?

BTW, I really appreciate your all comments in this thread.

Thanks
Tiejun