Re: [Qemu-devel] [PATCH 3/4] 9pfs: stat_to_qid: use device as input to qid.path

[Emilio G. Cota]

On Fri, Feb 09, 2018 at 16:13:26 +0100, Greg Kurz wrote:
> On Thu, 8 Feb 2018 19:00:18 +0100
> <address@hidden> wrote:
(snip)
> > +/* creative abuse of tb_hash_func7, which is based on xxhash */
> > +static uint32_t qpp_hash(QppEntry e)
> > +{
> > +    return tb_hash_func7(e.ino_prefix, e.dev, 0, 0, 0);
> 
> Hmm... Looking at git log include/exec/tb-hash-xx.h, I see that this hash
> function signature evolved according to TCG needs. It started with 3
> arguments, then 4 and we have 5 today.
> 
> So I don't think we should add another unrelated user. Probably best to
> have our own version. Also it seems it could be simpler since you always
> pass 0 for the third and later arguments.

It's fine to use it; the compiler will do the right thing
with those '0' arguments, because the function is always inlined.

That said, I have some cleanup patches to export this as xxhash and make
tb_hash a user of it. The patches also remove most of the ugly 0's from
callers. Those patches will have to wait for a bit though, so for now
just use tb_hash_func7.

(snip)
> > +        val = g_malloc0(sizeof(QppEntry));
> > +        if (!val) {
> > +            return -ENOMEM;
> > +        }

g_malloc0 and all other glib memory allocation functions only
return NULL if the size argument is 0.
If out of memory, the application is terminated, which is why
this check is not needed.
Reference:
  
https://developer.gnome.org/glib/stable/glib-Memory-Allocation.html#glib-Memory-Allocation.description

Thanks,

                Emilio