[PATCH 1/2] mmc: atmel-mci: DMA can be used with other controller
ludovic.desroches
ludovic.desroches at atmel.com
Tue Aug 28 09:25:58 EDT 2012
Hi Hein,
Thanks for your investigation about the MCI issue with AVR chips.
Le 08/26/2012 10:20 PM, Hein Tibosch a écrit :
> After the latest changes to atmel-mci, it could not be used with
> DMA on the AVR32 platform. This patch will allow to use DMA again
> and it will avoid access to MCI register ATMCI_DMA.
>
I think you should add more details about the issue:
Even if the IP version is lower than v3xx and doesn't have the DMA
configuration register, DMA transfers can be used with a different
controller than the Atmel AHB DMA one. For instance, some AVR chips use
the Synopsys DesignWare AHB DMA controller.
> Signed-off-by: Hein Tibosch <hein_tibosch at yahoo.es>
>
Some comments below...
> ---
> drivers/mmc/host/atmel-mci.c | 21 +++++++++------------
> 1 files changed, 9 insertions(+), 12 deletions(-)
>
> diff --git a/drivers/mmc/host/atmel-mci.c b/drivers/mmc/host/atmel-mci.c
> index f2c115e..66e9fdc 100644
> --- a/drivers/mmc/host/atmel-mci.c
> +++ b/drivers/mmc/host/atmel-mci.c
> @@ -71,7 +71,7 @@ enum atmci_pdc_buf {
> };
>
> struct atmel_mci_caps {
> - bool has_dma;
> + bool has_dma_conf_reg;
> bool has_pdc;
> bool has_cfg_reg;
> bool has_cstor_reg;
> @@ -411,7 +411,7 @@ static int atmci_regs_show(struct seq_file *s, void *v)
> atmci_show_status_reg(s, "SR", buf[ATMCI_SR / 4]);
> atmci_show_status_reg(s, "IMR", buf[ATMCI_IMR / 4]);
>
> - if (host->caps.has_dma) {
> + if (host->caps.has_dma_conf_reg) {
> u32 val;
>
> val = buf[ATMCI_DMA / 4];
> @@ -767,7 +767,7 @@ static void atmci_dma_complete(void *arg)
>
> dev_vdbg(&host->pdev->dev, "DMA complete\n");
>
> - if (host->caps.has_dma)
> + if (host->caps.has_dma_conf_reg)
> /* Disable DMA hardware handshaking on MCI */
> atmci_writel(host, ATMCI_DMA, atmci_readl(host, ATMCI_DMA) & ~ATMCI_DMAEN);
>
> @@ -954,7 +954,9 @@ atmci_prepare_data_dma(struct atmel_mci *host, struct mmc_data *data)
> maxburst = atmci_convert_chksize(host->dma_conf.dst_maxburst);
> }
>
> - atmci_writel(host, ATMCI_DMA, ATMCI_DMA_CHKSIZE(maxburst) | ATMCI_DMAEN);
> + if (host->caps.has_dma_conf_reg)
> + atmci_writel(host, ATMCI_DMA, ATMCI_DMA_CHKSIZE(maxburst) |
> + ATMCI_DMAEN);
>
> sglen = dma_map_sg(chan->device->dev, data->sg,
> data->sg_len, direction);
> @@ -2202,7 +2204,7 @@ static void __init atmci_get_cap(struct atmel_mci *host)
> dev_info(&host->pdev->dev,
> "version: 0x%x\n", version);
>
> - host->caps.has_dma = 0;
> + host->caps.has_dma_conf_reg = 0;
> host->caps.has_pdc = 1;
> host->caps.has_cfg_reg = 0;
> host->caps.has_cstor_reg = 0;
> @@ -2219,12 +2221,7 @@ static void __init atmci_get_cap(struct atmel_mci *host)
> host->caps.has_odd_clk_div = 1;
> case 0x400:
> case 0x300:
> -#ifdef CONFIG_AT_HDMAC
> - host->caps.has_dma = 1;
> -#else
> - dev_info(&host->pdev->dev,
> - "has dma capability but dma engine is not selected, then use pio\n");
> -#endif
> + host->caps.has_dma_conf_reg = 1;
> host->caps.has_pdc = 0;
> host->caps.has_cfg_reg = 1;
> host->caps.has_cstor_reg = 1;
> @@ -2298,7 +2295,7 @@ static int __init atmci_probe(struct platform_device *pdev)
>
> /* Get MCI capabilities and set operations according to it */
> atmci_get_cap(host);
> - if (host->caps.has_dma && atmci_configure_dma(host)) {
> + if (atmci_configure_dma(host)) {
> host->prepare_data = &atmci_prepare_data_dma;
> host->submit_data = &atmci_submit_data_dma;
> host->stop_transfer = &atmci_stop_transfer_dma;
>
I have tested this patch on at91sam9m10g45, at91sam9g10 and at91sam9261.
For chips without MCI support for Atmel AHB DMA, removing the
host->caps.has_dma test involves to use pdata->dma_slave with more
caution to avoid some NULL pointer dereferences
@@ -2168,16 +2168,21 @@ static bool atmci_configure_dma(struct atmel_mci
*host)
return false;
pdata = host->pdev->dev.platform_data;
+ if (!pdata)
+ return false;
- if (pdata && find_slave_dev(pdata->dma_slave)) {
+ if (pdata->dma_slave && find_slave_dev(pdata->dma_slave)) {
dma_cap_mask_t mask;
Regards
Ludovic
More information about the linux-arm-kernel
mailing list