Re: [XenPPC] [Patch] Detect non hypervisor hardware

[Hollis Blanchard <hollisb@xxxxxxxxxx>]

On Mon, 2006-04-17 at 15:41 -0400, Jimi Xenidis wrote:
> On Apr 17, 2006, at 12:12 PM, Hollis Blanchard wrote:
> >
> > I think you're missing my point: I'm talking about initializing  
> > opt_nohv
> > and opt_earlygdb. These are global variables, and so they will land in
> > the .bss section.
> >
> Ahh this is my fault.
> *Spec geek voice --ON--*
> The explicit zeroing of a "tentative definition with external  
> linkage" didn't come out until C99
> *Spec geek voice --OFF--*
> 
> This is an old habit of mine that costs _nothing_ since assigning it  
> to zero generates the same assmebler definition.

It is a style issue, but as I noted elsewhere it can have an (admittedly
minor) impact .data size with some compilers.

> Is it common Linux practice not to do this?
> The Xen guys assign to zero even with static decls :)

As I explained in my other mail:
- yes, it is standard Linux practice not to do this.
- other Xen code only sometimes assigns to zero, and in the code I saw
it was far more common not to.

I don't mean to be a jerk, but I think we all have better things to do
than discuss coding style. Is this conversation really worth continuing?

-- 
Hollis Blanchard
IBM Linux Technology Center


_______________________________________________
Xen-ppc-devel mailing list
Xen-ppc-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ppc-devel