Messages in this thread | | | Date | Fri, 10 Feb 2012 12:14:05 +0000 | From | Mark Brown <> | Subject | Re: [PATCH V1] regmap: Support for caching in reg_raw_write() |
| |
On Fri, Feb 10, 2012 at 01:01:26PM +0100, Lars-Peter Clausen wrote:
> In my opinion it makes more sense to introduce a regmap_bulk_write function > for this. regmap_raw_write was not really meant to write registers but
This isn't exclusive to that, I think Laxman planned to send a patch doing that on top of this one which would be what most users would end up using.
> rather binary blobs, like firmware. Also this keeps things consistent with > the read part of the regmap API.
See the previous discussion on this in the past day or so - bulk_write() is more complicated to implement by itself since it's going to end up boling down to a raw_write() internally anyway (as does reg_write()) and it seems nicer to just do the right thing if people ask for it.
We can always refactor later to improve performance if needed but this is a simple starting point. [unhandled content-type:application/pgp-signature] | |