Re: [PATCH v4 4/9] i2c: use client addresses directly in ATR interface

[Romain Gantois <romain.gantois@bootlin.com>]

[-- Attachment #1: Type: text/plain, Size: 1594 bytes --]

Hi,

On mercredi 8 janvier 2025 14:38:11 heure normale d’Europe centrale Tomi 
Valkeinen wrote:
> Hi,
> 
> On 08/01/2025 15:31, Romain Gantois wrote:
> > Hi,
> > 
> > On lundi 6 janvier 2025 10:51:20 heure normale d’Europe centrale Tomi
> > 
> > Valkeinen wrote:
> >> Hi,
> >> 
> >> On 30/12/2024 15:22, Romain Gantois wrote:
> >>> The I2C Address Translator (ATR) module defines mappings from i2c_client
> >>> structs to aliases. However, only the physical address of each
> >>> i2c_client
> > 
> > ...
> > 
> >>> -	dev_dbg(atr->dev, "chan%u: client 0x%02x mapped at alias 0x%02x (%s)
> > 
> > \n",
> > 
> >>> -		chan->chan_id, client->addr, alias, client->name);
> >>> +	dev_dbg(atr->dev, "chan%u: addr 0x%02x mapped at alias 0x%02x\n",
> >>> +		chan->chan_id, addr, alias);
> >> 
> >> This, and the dev_dbg() below, sound a bit odd to my ear. But I'm not
> >> sure what would be a good print... "added alias 0x12 for address 0x34"?
> > 
> > Maybe "assigned address 0x34 to alias 0x12"? Since the alias doesn't
> > really go anywhere, we just assign different downstream addresses to it.
> 
> I guess it's how you look at this =). I like to think it (in this
> order): alias -> address, as it's basically a mapping. So a debug print
> that prints the alias first and address second feels more natural.
> "using alias 0x12 for address 0x34"?

This seems like a valid view as well. Since you're the ATR maintainer I'll go 
with your suggestion.

Thanks,

-- 
Romain Gantois, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 833 bytes --]