Re: [Xen-devel] [PATCH v6] libxl: provide libxl_bitmap_{or,and}

[Wei Liu <wei.liu2@xxxxxxxxxx>]

On Wed, Apr 15, 2015 at 02:29:10PM +0100, Ian Campbell wrote:
> On Wed, 2015-04-15 at 14:15 +0100, Wei Liu wrote:
> > On Wed, Apr 15, 2015 at 01:45:14PM +0100, Ian Campbell wrote:
> > > On Wed, 2015-04-15 at 05:45 -0600, Linda Jacobson wrote:
> > > > There are new functions to provide logical and and or of two bitmaps.
> > > 
> > > Please could you add a sentence or two on the intended use of these
> > > functions, since there are no callers being added here.
> > > 
> > 
> > Linda is our Outreachy applicant. This is a small task that Julien and I
> > assigned to her.
> 
> Sure, but that doesn't remove the need for the commit log to be a
> standalone justification for the patch in its own right.
> 
> > One user I can think of is in some of the vNUMA validation functions
> > that operate on bitmaps. But to keep this task small and simple I didn't
> > ask her to actually use the functions she introduce.
> > 
> > > In particular without that I can't tell if these need to be part of the
> > > public API or if they are going to be used by something internal.
> > > 
> > 
> > I think these functions should be public functions.
> 
> Sure, but the reasoning for why you^WLinda thinks that needs to be in
> the commit log. In particular because there are no users being added.
> 
> I could probably guess why you think these should be public, but I
> shouldn't have to guess and in any case that doesn't help in 6 months
> when someone asks "why do we have these functions".
> 

Linda, can you use something like this in your commit log:

Provide libxl_bitmap_{and,or} functions. These functions can be used in
vNUMA configuration check function. They are also generally useful so
they are made as public API.

> > > If the former then I think a LIBXL_HAVE_... #define is needed in libxl.h
> > > (as described by the comments in there, and there are many examples).
> > > 
> > 
> > We're a bit lax on these functions (there is no LIBXL_HAVE_BITMAP...) 
> 
> Weren't they always there and hence don't need it?
> 

OK, maybe I'm mistaken.

Linda, you need to add one more hunk to libxl.h.

#define LIBXL_HAVE_BITMAP_AND_OR 1

Search for other LIBXL_HAVE_ macro to get an idea how it is handled.

If you have any questions, just ask.

Wei.

> Ian.
> 
> > so
> > I didn't ask her to add one. We can add LIBXL_HAVE_BITMAP_AND_OR in
> > libxl.h if you think it is necessary.
> > 
> > Wei.
> > 
> > > If the latter then the prototype should differ. (I'll explain that
> > > if/when this turns out to be the case).
> > > 
> > > > diff --git a/tools/libxl/libxl_utils.h b/tools/libxl/libxl_utils.h
> > > > index acacdd9..a128a7c 100644
> > > > --- a/tools/libxl/libxl_utils.h
> > > > +++ b/tools/libxl/libxl_utils.h
> > > > @@ -90,6 +90,13 @@ int libxl_bitmap_test(const libxl_bitmap *bitmap, 
> > > > int bit);
> > > >  void libxl_bitmap_set(libxl_bitmap *bitmap, int bit);
> > > >  void libxl_bitmap_reset(libxl_bitmap *bitmap, int bit);
> > > >  int libxl_bitmap_count_set(const libxl_bitmap *cpumap);
> > > > +/* Or and and functions for two bitmaps */
> > > 
> > > This comment doesn't say anything which isn't already apparent from the
> > > names and prototypes of the functions in question. Just leave it out.
> > > 
> > > 
> > > > +int libxl_bitmap_or(libxl_ctx *ctx, libxl_bitmap *or_map,
> > > > +                    const libxl_bitmap *map1, 
> > > > +                    const libxl_bitmap *map2); 
> > > > +int libxl_bitmap_and(libxl_ctx *ctx, libxl_bitmap *and_map,
> > > > +                     const libxl_bitmap *map1, 
> > > > +                     const libxl_bitmap *map2);
> > > >  char *libxl_bitmap_to_hex_string(libxl_ctx *ctx, const libxl_bitmap 
> > > > *cpumap);
> > > >  static inline void libxl_bitmap_set_any(libxl_bitmap *bitmap)
> > > >  {
> > > 
> 

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel