Hello, Le 29/11/2016 à 22:17, Wolfram Sang a écrit :
+ if (of_device_is_compatible(np, "marvell,armada-3700-i2c")) { + i2c->fm_mask = ICR_BUSMODE_FM; + i2c->hs_mask = ICR_BUSMODE_HS; + } else { + i2c->fm_mask = ICR_FM; + i2c->hs_mask = ICR_HS; + } *i2c_types = (enum pxa_i2c_types)(of_id->data); @@ -1181,6 +1194,13 @@ static int i2c_pxa_probe_pdata(struct platform_device *pdev, i2c->master_code = 0xe; i2c->rate = plat->rate; } + if (!strcmp(id->name, "armada-3700-i2c")) { + i2c->fm_mask = ICR_BUSMODE_FM; + i2c->hs_mask = ICR_BUSMODE_HS; + } else { + i2c->fm_mask = ICR_FM; + i2c->hs_mask = ICR_HS; + }Okay, having the same code twice is not nice as well. Sorry for missing this in the first review and going a step back, but I think now the best solution is to have again a REGS_A3700 struct, but we should extend it with new entries for the shifted bits. Then in the init code, you can do something like: i2c->fm_mask = pxa_reg_layout[i2c_type].fm_mask ?: ICR_FM; Makes sense?
Mhhhh... makes sense yes, it is simpler and would remove the duplicated code, yes (no no need to modify probe_dt and probe_pdata in this case). What do you prefer everything in one commit or two seperated commit ? (one including the new fields for fm_mask and another one to add support for a3700-i2c).
Thanks, Romain -- Romain Perier, Free Electrons Embedded Linux and Kernel engineering http://free-electrons.com -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
