On Wed, Oct 24, 2018 at 8:18 AM Dan Carpenter <dan.carpenter@xxxxxxxxxx> wrote:
>
> On Wed, Oct 24, 2018 at 07:59:45AM -0600, Spencer E. Olson wrote:
> > Changes implementation of INSN_CONFIG_GET_CMD_TIMING_CONSTRAINTS for
> > ni_mio devices to scale the result by the number of channels being used.
>
> I really can't understand this statement at all. What changes the
> implementation?
>
> > The user is already required to indicate which channels (and how many
> > obviously) are intended to be used. There is no point of not using this
> > information--the analog input cards already similarly scale the timing
> > results based on the number of channels.
>
> This sounds like it's just an optimization but I think this patch is a
> behavior change or a bug fix, right? It's not totally clear to me from
> the patch description what the user visible effect of this patch is.
> Could you spell that out a little bit more clearly and resend the
> patch? (It's also possible that I just don't understand Comedi well
> enough to understand the patch description).
>
> >
> > Signed-off-by: Spencer E. Olson <olsonse@xxxxxxxxx>
> > ---
> > This patch is made in reference to the last set of patches adding the timing
> > constraint facility in pci_mio_common
> > (51fd3673838396844f15de0e906be5333bfbbc8d).
>
>
> I feel like we should be using the Fixes tag for this. Like so:
>
> Fixes: 51fd36738383 ("staging: comedi: ni_mio_common: implement INSN_CONFIG_GET_CMD_TIMING_CONSTRAINTS")
> Signed-off-by: Your <email.com>
I did consider submitting it as a fix instead, but I couldn't actually
remember what my intentions were when I wrote the original code from
the earlier patch. On the other hand, after reading more on the
interface documentation and user code that I have written (not yet
submitted to the comedi community), I think you are right and that
this should be submitted as a fix. Thanks
>
> regards,
> dan carpenter
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
