qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH 02/14] scsi: add a qdev property for the disk's


From: Blue Swirl
Subject: Re: [Qemu-devel] [PATCH 02/14] scsi: add a qdev property for the disk's WWN
Date: Sat, 7 Jul 2012 07:48:07 +0000

On Fri, Jul 6, 2012 at 7:05 AM, Paolo Bonzini <address@hidden> wrote:
> Il 05/07/2012 20:03, Blue Swirl ha scritto:
>> > > > +            if (s->wwn) {
>> > > > +                outbuf[buflen++] = 0x1; // Binary
>> > > > +                outbuf[buflen++] = 0x3; // NAA
>> > > > +                outbuf[buflen++] = 0;   // reserved
>> > >
>> > > C99 comments.
>> >
>> > Just following the style of this code.  Feel free to send a patch to
>> > replace with #defines.
>>
>> That's not how we should work. New code should be compliant with our
>> goals. Pushing the responsibility for fixing issues to other people
>> does not scale.
>
> I believe the coding style are attacking the wrong problem.  It's
> end-of-line comments that should be avoided in favor of #defines, packed
> structs, designated initializers, etc.

This applies to C89 comments too. Perhaps HACKING file should mention this.

>  But for end-of-line comments,
> C++ comments are superior to /* */ comments.

Well, submit a change to CODING_STYLE then. I think the only places
where C99 is better are lines like
//#define DEBUG_XYZ

>
> I do plan to fix the clarity issue with SCSI data structures and
> constants.  But for now, the best compromise is to keep C++ comments
> IMHO.  I'm not pushing the responsibility to other people in general,
> but if they think C++ comments are a major issue they can send patches.

When you add them in new code, that is a problem. New code should
comply with CODING_STYLE,  code with other styles should be converted
as it is touched.

It shouldn't be much effort from your part to change the comments in
this patch to C89 style.

>
> Paolo



reply via email to

[Prev in Thread] Current Thread [Next in Thread]