From c8ebae37034c0ead62eb4df8ef88e999ddb8d5cf Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 11 Jan 2011 19:35:53 +0000 Subject: ARM: mmci: add dmaengine-based DMA support Based on a patch from Linus Walleij. Add dmaengine based support for DMA to the MMCI driver, using the Primecell DMA engine interface. The changes over Linus' driver are: - rename txsize_threshold to dmasize_threshold, as this reflects the purpose more. - use 'mmci_dma_' as the function prefix rather than 'dma_mmci_'. - clean up requesting of dma channels. - don't release a single channel twice when it's shared between tx and rx. - get rid of 'dma_enable' bool - instead check whether the channel is NULL. - detect incomplete DMA at the end of a transfer. Some DMA controllers (eg, PL08x) are unable to be configured for scatter DMA and also listen to all four DMA request signals [BREQ,SREQ,LBREQ,LSREQ] from the MMCI. They can do one or other but not both. As MMCI uses LBREQ/LSREQ for the final burst/words, PL08x does not transfer the last few words. - map and unmap DMA buffers using the DMA engine struct device, not the MMCI struct device - the DMA engine is doing the DMA transfer, not us. - avoid double-unmapping of the DMA buffers on MMCI data errors. - don't check for negative values from the dmaengine tx submission function - Dan says this must never fail. - use new dmaengine helper functions rather than using the ugly function pointers directly. - allow DMA code to be fully optimized away using dma_inprogress() which is defined to constant 0 if DMA engine support is disabled. - request maximum segment size from the DMA engine struct device and set this appropriately. - removed checking of buffer alignment - the DMA engine should deal with its own restrictions on buffer alignment, not the individual DMA engine users. - removed setting DMAREQCTL - this confuses some DMA controllers as it causes LBREQ to be asserted for the last seven transfers, rather than six SREQ and one LSREQ. - removed burst setting - the DMA controller should not burst past the transfer size required to complete the DMA operation. Tested-by: Linus Walleij Signed-off-by: Russell King --- include/linux/amba/mmci.h | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) (limited to 'include/linux/amba/mmci.h') diff --git a/include/linux/amba/mmci.h b/include/linux/amba/mmci.h index f4ee9acc9721..f60227088b7b 100644 --- a/include/linux/amba/mmci.h +++ b/include/linux/amba/mmci.h @@ -6,6 +6,9 @@ #include +/* Just some dummy forwarding */ +struct dma_chan; + /** * struct mmci_platform_data - platform configuration for the MMCI * (also known as PL180) block. @@ -27,6 +30,17 @@ * @cd_invert: true if the gpio_cd pin value is active low * @capabilities: the capabilities of the block as implemented in * this platform, signify anything MMC_CAP_* from mmc/host.h + * @dma_filter: function used to select an apropriate RX and TX + * DMA channel to be used for DMA, if and only if you're deploying the + * generic DMA engine + * @dma_rx_param: parameter passed to the DMA allocation + * filter in order to select an apropriate RX channel. If + * there is a bidirectional RX+TX channel, then just specify + * this and leave dma_tx_param set to NULL + * @dma_tx_param: parameter passed to the DMA allocation + * filter in order to select an apropriate TX channel. If this + * is NULL the driver will attempt to use the RX channel as a + * bidirectional channel */ struct mmci_platform_data { unsigned int f_max; @@ -38,6 +52,9 @@ struct mmci_platform_data { int gpio_cd; bool cd_invert; unsigned long capabilities; + bool (*dma_filter)(struct dma_chan *chan, void *filter_param); + void *dma_rx_param; + void *dma_tx_param; }; #endif -- cgit v1.2.3