On 21 May 2015 at 11:39, Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote: > On Thu, May 21, 2015 at 11:31 AM, Rafał Miłecki <zajec5@xxxxxxxxx> wrote: >> On 21 May 2015 at 10:58, Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote: >>> On Thu, May 21, 2015 at 10:50 AM, Rafał Miłecki <zajec5@xxxxxxxxx> wrote: >>>>>>> I think your comments suggest that I shouldn't be removing "spi-nor" >>>>>>> from m25p_ids[] nor from this block: >>>>>>> >>>>>>> if (data && data->type) >>>>>>> flash_name = data->type; >>>>>>> else if (!strcmp(spi->modalias, "spi-nor")) >>>>>>> flash_name = NULL; /* auto-detect */ >>>>>>> else >>>>>>> flash_name = spi->modalias; >>>>>>> >>>>>>> So it stays in both m25p_ids[] and .of_match_table. >>>>>>> >>>>>>> I suppose that can work. It then allows people to do weird stuff like: >>>>>>> >>>>>>> compatible = "idontknowwhatimdoing,spi-nor"; >>>>>>> >>>>>>> in their device tree. But other than that, there's not much downside I don't >>>>>>> think. >>>>>> >>>>>> It sounds like a reasonable solution. I guess there isn't a perfect >>>>>> one. Even if we decide to go for sth like "jedec-spi-nor", there >>>>>> always will be a chance of someone using >>>>>> compatible = "idontknowwhatimdoing,jedec-spi-nor"; >>>>>> So if you rework your patch to leave "spi-nor" support in m25p_ids and >>>>>> conditions block, it should be OK. >>>>> >>>>> Typically platform devices just use the driver's name. Hence IMHO there's >>>>> no need to add a shiny new spi-nor device name. >>>>> >>>>> So what's wrong with using "m25p80", and treating that as auto-detect iff >>>>> !spi->dev.of_node? >>>> >>>> Treating "m25p80" as auto-detect triggering string won't allow >>>> platform to *force* "m25p80" flash type if there ever appears to be >>>> needed. Maybe it's unlikely, but it still sounds like a bit bad design >>>> for me. >>> >>> To force m25p80 flash, you set flash_platform_data.type to "m25p80"? >> >> Oh, I think I got lost in the way m25p80 is probed. Is it handled by >> spi_board_info and its "modalias"? > > Indeed. > >> Could I leave flash_platform_data.type set to NULL and still have m25p80 probed? > > Yes, that case is the final "else" branch above. Or wait a second, I see the problem. So the last "else" branch does: flash_name = spi->modalias; which will result in flash_name getting "m25p80" which will result in passing "m25p80" model to spi-nor. So some sort of solution is to replace if (data && data->type) flash_name = data->type; with if (data) flash_name = data->type; which will allow platform stuff to *not* specify type and trigger auto (JEDEC RD ID) detection. The only minor problem is we will still need specifying spi_board_info.platform_data, but 99.9% of code does as it's requires for setting partitions anyway. Brian: so I think your initial patch was OK, sorry for all this noise :| -- Rafał -- To unsubscribe from this list: send the line "unsubscribe linux-spi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html