[PATCH v5 4/5] net: wan: fsl_qmc_hdlc: Add runtime timeslots changes support

Herve Codina herve.codina at bootlin.com
Fri Mar 1 03:30:43 AEDT 2024


Hi Andy,

On Thu, 29 Feb 2024 17:20:59 +0200
Andy Shevchenko <andriy.shevchenko at linux.intel.com> wrote:

> On Thu, Feb 29, 2024 at 03:15:52PM +0100, Herve Codina wrote:
> > QMC channels support runtime timeslots changes but nothing is done at
> > the QMC HDLC driver to handle these changes.
> > 
> > Use existing IFACE ioctl in order to configure the timeslots to use.  
> 
> ...
> 
> > +	bitmap_scatter(ts_mask, map, ts_mask_avail, 64);  
> 
> Wondering if we may have returned value more useful and hence having something like
> 
> 	n = bitmap_scatter(...);

I thought about it.

In bitmap_{scatter,gather}(dst, src, mask, nbits), only returning the
weight of the third parameter (i.e. mask) can be efficient regarding to the
for_each_set_bit() loop done in the functions.
For dst parameter, we need to add a counter in the loop to count the number
of bit set depending on the test_bit() result. Will this be more efficient
than a call to bitmap_weight() ?

Also, in my case, the third parameter is ts_mask_avail and I don't need
its weight.

I thing users that need to have the dst or src weight should call
bitmap_weight() themselves as this is users context dependent.

bitmap_{scatter,gather}(dst, src, mask, nbits) can be improved later with
no impact to current users (except performance).

That's why I concluded to return nothing from bitmap_{scatter,gather} when
I took the old existing patches.

> 
> > +	if (bitmap_weight(ts_mask, 64) != bitmap_weight(map, 64)) {  
> 
> 	if (n != ...) {
> 
> ?
> 
> > +		dev_err(qmc_hdlc->dev, "Cannot translate timeslots %64pb -> (%64pb, %64pb)\n",
> > +			map, ts_mask_avail, ts_mask);
> > +		return -EINVAL;
> > +	}  
> 
> ...
> 
> > +	bitmap_gather(map, ts_mask, ts_mask_avail, 64);
> > +
> > +	if (bitmap_weight(ts_mask, 64) != bitmap_weight(map, 64)) {
> > +		dev_err(qmc_hdlc->dev, "Cannot translate timeslots (%64pb, %64pb) -> %64pb\n",
> > +			ts_mask_avail, ts_mask, map);
> > +		return -EINVAL;
> > +	}  
> 
> Ditto.
> 

Best regards,
Hervé


More information about the Linuxppc-dev mailing list