[libvirt] [PATCH 2/6] conf: new functions to check if PCI address is wanted/present

[Laine Stump]

On 05/19/2016 01:21 PM, Ján Tomko wrote:
> On Wed, May 18, 2016 at 03:23:50PM -0400, Laine Stump wrote:
>> In order to allow <address type='pci'/> with no other attributes to
>> mean "I want a PCI address, but any PCI address will do" (just as
>> having no <address> at all usually indicates), we will need to change
>> several places in the code from a simple "info->type == (or !=)
>> VIR_DOMAIN_DEVICE_ADDRESS_TYPE_(PCI|NONE)" into something slightly
>> more complex, this patch adds to new functions that take a
>> virDomainDeviceInfoPtr and return true/false depending on 1) whether
>> the current state of the info indicates that we "want" a PCI address
>> for this device (virDeviceInfoPCIAddressWanted()) and 2) whether this
>> device already has a valid PCI address
>> (virDeviceInfoPCIAddressPresent()).
>>
>> Both of these functions required the simpler check for whether a pci
>> address is "empty" (i.e. all of its attributes are 0, which can never
>> happen in a real PCI address, since slot 0 of bus 0 of domain 0 is
>> always reserved), so that function is also added.
>> ---
>>   src/conf/device_conf.h | 21 +++++++++++++++++++++
>>   1 file changed, 21 insertions(+)
>>
>> diff --git a/src/conf/device_conf.h b/src/conf/device_conf.h
>> index 46c720d..847564b 100644
>> --- a/src/conf/device_conf.h
>> +++ b/src/conf/device_conf.h
>> @@ -184,6 +184,27 @@ typedef struct _virDomainDeviceInfo {
>>   int virPCIDeviceAddressIsValid(virPCIDeviceAddressPtr addr,
>>                                  bool report);
>>   
>> +static inline bool
> I think inline keywords are pointless nowadays. Some compilers will
> inline the functions even if you don't want them to.

If you remove "inline" from a function defined in a .h file, then it 
becomes a simple "static bool f()...", which gives you this error when 
compiling any .c file that includes the .h but doesn't use the function:

In file included from util/virnetdev.h:34:0,
from util/virnetdevveth.c:36:
./conf/device_conf.h:194:1: error: 'virDeviceInfoPCIAddressWanted' 
defined but not used [-Werror=unused-function] 
virDeviceInfoPCIAddressWanted(const virDomainDeviceInfo *info) ^


And if you remove both static *and* inline, you get this:

In file included from ./conf/domain_conf.h:48:0,
                  from ./conf/virdomainobjlist.h:28,
                  from util/virclosecallbacks.h:28,
                  from util/virclosecallbacks.c:28:
./conf/device_conf.h:188:1: error: no previous prototype for 
'virPCIDeviceAddressIsEmpty' [-Werror=missing-prototypes]
  virPCIDeviceAddressIsEmpty(const virPCIDeviceAddress *addr)
  ^

(and if you disabled that warning and got far enough to link, you could 
end up with a linker error due to multiple definitions of the same 
function in multiple .o's).