Hi Arnd, > > +static void sdhci_arasan_writel(struct sdhci_host *host, u32 val, int reg) > > +{ > > + struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host); > > + struct sdhci_arasan_data *sdhci_arasan = pltfm_host->priv; > > + > > + if (reg == SDHCI_DMA_ADDRESS) { > > + if (sdhci_arasan->ahb_ops && sdhci_arasan->ahb_ops->xlat_addr) > > + sdhci_arasan->ahb_ops->xlat_addr(sdhci_arasan, > > + sg_dma_address(host->data->sg)); > > + } > > + writel(val, host->ioaddr + reg); > > +} > > + > > static struct sdhci_ops sdhci_arasan_ops = { > > + .write_l = sdhci_arasan_writel, > > .get_max_clock = sdhci_pltfm_clk_get_max_clock, > > .get_timeout_clock = sdhci_arasan_get_timeout_clock, > > }; > > This looks like you are doing it at the wrong place. From what I understand, > you are using the AHB inbound window as a minimal IOMMU. Why don't you make > this a proper IOMMU driver instead and leave the SDHCI driver unchanged? I looked at the existent iommu driver. It seems like one can install an IO MMU. The question that I have is what happen when we enable the actual IO mmu? Would this be an issue if both existed as there is only one pointer in the bus_type structure? -Loc -- 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