Re: [PATCH 02/23] lib: move 64-bit div/mod compiler helpers

[Julien Grall <julien@xxxxxxx>]

On 07/04/2021 09:33, Jan Beulich wrote:
> On 06.04.2021 21:34, Julien Grall wrote:
> > Hi Jan,
> >
> > On 01/04/2021 16:23, Jan Beulich wrote:
> > > On 01.04.2021 16:56, Julien Grall wrote:
> > > > On 01/04/2021 11:19, Jan Beulich wrote:
> > > > > --- a/xen/common/lib.c
> > > > > +++ b/xen/lib/divmod.c
> > > > > @@ -40,7 +40,6 @@
> > > > >      * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
> > > > >      * SUCH DAMAGE.
> > > > >      */
> > > > > -#if BITS_PER_LONG == 32
> > > > In theory BITS_PER_LONG == 32 is not quite the same as 32-bit arch. Can
> > > > you clarify whether this code is necessary on 64-bit arch where long is
> > > > only 32-bit.
> > > Likely the compiler can get away without invoking these helpers, so
> > > the code would remain unused. I'm uncertain whether CONFIG_64BIT
> > > ought to be set for such an architecture, as we assume sizeof(long)
> > > == sizeof(void*), and hence pointers would then need to be 32-bit
> > > as well there.
> > This is a fair point. Would you mind to add a sentence explaining that
> > in the commit message?
> I've added
>
> "Note that we imply "32-bit arch" to be the same as BITS_PER_LONG == 32,
>   i.e. we aren't (not just here) prepared to have a 64-bit arch with
>   BITS_PER_LONG == 32. Yet even if we supported such, likely the compiler
>   would get away there without invoking these helpers, so the code would
>   remain unused in practice."
Sounds fine to me.

> > With that:
> >
> > Acked-by: Julien Grall <jgrall@xxxxxxxxxx>
> Thanks. Any chance to also get an ack on patch 1, so at least these two
> (or three, seeing that you also did ack patch 3) could go in before my
> re-posting of the series to add the one line commit message additions
> that you did ask for on all the str* and mem* patches? (Alternatively I
> could take the time and re-order the two.)
I didn't ack #1 because I am not very familiar with the x86 constraints.
If anyone with x86 background (maybe Roger?) is willing to review it, 
then I would be happy to give my ack.
Cheers,

--
Julien Grall