Re: [Qemu-devel] [PATCH v2 12/29] Include hw/irq.h a lot less

[Eric Blake <eblake@redhat.com>]

[-- Attachment #1.1: Type: text/plain, Size: 1654 bytes --]

On 8/7/19 11:27 PM, Markus Armbruster wrote:

>>>>  /*
>>>>   * Function types
>>>>   */
>>>>  typedef void SaveStateHandler(QEMUFile *f, void *opaque);
>>>>  typedef int LoadStateHandler(QEMUFile *f, void *opaque, int version_id);
>>>> +typedef void (*qemu_irq_handler)(void *opaque, int n, int level);
>>
>> Should we prefer a consistent form for function pointer typedefs?  Here,
>> we've mixed 'rettype Name(params)' with 'rettype (*name)(params)'.
> 
> Which of the two difference I can see do you mean?

Both :)

> 
> CamelCase vs. lower_case_with_underscore?

Indeed, choosing a consistent naming convention to make it obvious we
are referring to a typedef naming a function signature may be useful.

> 
> Parenthesis around the type name?  I wouldn't call that inconsistent, we
> simply use parenthesis only when needed.

And the question of whether the * is implied by the typedef itself
(which we tend to avoid for typedefs to structures) or must be supplied
at the point of use.  It gets odd since C permits both 'FuncType f =
myfunc' and 'FuncType f = &myfunc', and since invoking through a
function pointer can be done as both 'f(args)' and '(*f)(args)'.
However, since we only have 3 uses total that declare a use of
SaveStateHandler and LoadStateHandler, compared to a LOT of use of
qemu_irq_handler, moving the '*' out of the declarations and into the
typedef of the former is a lot easier than moving the '*' out of the
typedef and into the declarations of the latter.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]