On 12 May 2015 at 11:53, Lu Y.B. <yangbo.lu@xxxxxxxxxxxxx> wrote: > Thanks Uffe. > I also think checking every time is not better idea but am not so clear about your method. > I think the plan you wanting should be one of the bellowing, which one is the plan you want? > Plan 1, > Define sdhci_be32bs_readl for big-endian > Define sdhci_le32bs_readl for little-endian > Define a function pointer sdhci_32bs_readl to be assigned when checking endian mode in probe. No, thanks. > Plan 2, > For esdhc, > Define struct sdhci_ops sdhci_esdhc_ops_1 for big-endian > Define struct sdhci_ops sdhci_esdhc_ops_2 for little-endian > Define struct sdhci_ops sdhci_esdhc_ops to be assigned when checking endian mode in probe. This is what I had in mind. Though I now realize that it then doesn't make sense to extend the common MMC DT parser, mmc_of_parse(), to fetch the endian mode. That's because the endian mode needs to be known to pick the correct sdhci_ops, which is the parameter you provide to sdhci_pltfm_init(). Forgive me for pushing you back an forth. So, yes I like you to implement plan2 and thus patch1 should only update the documentation of the DT bindings. Kind regards Uffe -- To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
