We can only use page_address on memory that has been mapped using kmap, when the buffer passed to the SPI has been allocated by vmalloc the page has not necessarily been mapped through kmap. This means sometimes page_address will return NULL causing the pointer we pass to sg_init_one to be invalid. Currently, this issue doesn't show up on the MXS architecture as the defconfig defines CONFIG_HIGHMEM=n which means all pages are mapped. For the sake of robustness though it is best to correct the issue. As we only call page_address so that we can pass a virtual address to sg_init_one which will eventually call virt_to_page on it, fix this by calling sg_set_page directly rather then relying on the sg_init_one helper. Note this patch is only build tested as I don't have an MXS system to test on. Signed-off-by: Charles Keepax <ckeepax@xxxxxxxxxxxxxxxxxxxxxxxxxxx> --- drivers/spi/spi-mxs.c | 10 +++++----- 1 files changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/spi/spi-mxs.c b/drivers/spi/spi-mxs.c index a9e72f9..06a1154 100644 --- a/drivers/spi/spi-mxs.c +++ b/drivers/spi/spi-mxs.c @@ -182,7 +182,6 @@ static int mxs_spi_txrx_dma(struct mxs_spi *spi, int min, ret; u32 ctrl0; struct page *vm_page; - void *sg_buf; struct { u32 pio[4]; struct scatterlist sg; @@ -232,13 +231,14 @@ static int mxs_spi_txrx_dma(struct mxs_spi *spi, ret = -ENOMEM; goto err_vmalloc; } - sg_buf = page_address(vm_page) + - ((size_t)buf & ~PAGE_MASK); + + sg_init_table(&dma_xfer[sg_count].sg, 1); + sg_set_page(&dma_xfer[sg_count].sg, vm_page, + min, offset_in_page(buf)); } else { - sg_buf = buf; + sg_init_one(&dma_xfer[sg_count].sg, buf, min); } - sg_init_one(&dma_xfer[sg_count].sg, sg_buf, min); ret = dma_map_sg(ssp->dev, &dma_xfer[sg_count].sg, 1, (flags & TXRX_WRITE) ? DMA_TO_DEVICE : DMA_FROM_DEVICE); -- 1.7.2.5 -- 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