On 15/11/2024 12:15 pm, Luca Fancellu wrote:
>
>> On 15 Nov 2024, at 12:00, Andrew Cooper <andrew.coop...@citrix.com> wrote:
>>
>> On 15/11/2024 10:50 am, Luca Fancellu wrote:
>>> diff --git a/xen/include/xen/xvmalloc.h b/xen/include/xen/xvmalloc.h
>>> index 440d85a284bb..802be6687085 100644
>>> --- a/xen/include/xen/xvmalloc.h
>>> +++ b/xen/include/xen/xvmalloc.h
>>> @@ -40,20 +40,46 @@
>>> ((typeof(ptr))_xvrealloc(ptr, offsetof(typeof(*(ptr)), field[nr]), \
>>> __alignof__(typeof(*(ptr)))))
>>>
>>> +#if defined(CONFIG_HAS_VMAP)
>>> +
>>> /* Free any of the above. */
>>> void xvfree(void *va);
>>>
>>> +/* Underlying functions */
>>> +void *_xvmalloc(size_t size, unsigned int align);
>>> +void *_xvzalloc(size_t size, unsigned int align);
>>> +void *_xvrealloc(void *va, size_t size, unsigned int align);
>>> +
>>> +#else
>>> +
>>> +static inline void xvfree(void *va)
>>> +{
>>> + xfree(va);
>>> +}
>>> +
>>> +void *_xvmalloc(size_t size, unsigned int align)
>>> +{
>>> + return _xmalloc(size, align);
>>> +}
>>> +
>>> +void *_xvzalloc(size_t size, unsigned int align)
>>> +{
>>> + return _xzalloc(size, align);
>>> +}
>>> +
>>> +void *_xvrealloc(void *va, size_t size, unsigned int align)
>>> +{
>>> + return _xrealloc(va, size, align);
>>> +}
>>> +
>>> +#endif
>> Does this really compile with the wrappers not being static inline ?
>>
>> That aside, could we not do this using conditional aliases, rather than
>> wrapping the functions? It would certainly be shorter, code wise.
> Do you mean something like below?
>
> #define xvfree xfree
> #define _xvmalloc _xmalloc
> […]
I mean __attribute__((__alias__("")))
There are two examples in tree already. See efi_compat_get_info() being
aliased to efi_get_info()
In this case, in the !HAS_VMAP case, we'd just declare _xmalloc() to
have an alias called _xvmalloc() too.
This avoids needing to wrap every function in the headers.
~Andrew
