On 28.01.2016 23:08, Eric Anholt wrote:
kernel@xxxxxxxxxxxxxxxx writes:From: Martin Sperl <kernel@xxxxxxxxxxxxxxxx> Since the move to the new clock framework with commit 94cb7f76caa0 ("ARM: bcm2835: Switch to using the new clock driver support.") this driver was no longer functional as it was manipulating the clock registers locally without going true the framework. This patch moves to use the new clock framework and also moves away from the hardcoded address offsets for DMA getting the dma-address directly from the device tree. Note that the optimal bclk_ratio selection to avoid jitter due to the use of fractional dividers, which is in the current version has been removed, because not all devices support these non power of 2 sized transfers, which resulted in lots of (downstream) modules that use: snd_soc_dai_set_bclk_ratio(cpu_dai, sample_bits * 2); Signed-off-by: Martin Sperl <kernel@xxxxxxxxxxxxxxxx> --- sound/soc/bcm/bcm2835-i2s.c | 284 ++++++++++--------------------------------- 1 file changed, 64 insertions(+), 220 deletions(-) diff --git a/sound/soc/bcm/bcm2835-i2s.c b/sound/soc/bcm/bcm2835-i2s.c index 3303d5f..1c1f221 100644 --- a/sound/soc/bcm/bcm2835-i2s.c +++ b/sound/soc/bcm/bcm2835-i2s.c- dev->i2s_regmap = regmap[0]; - dev->clk_regmap = regmap[1]; + /* get the clock */ + dev->clk_prepared = false; + dev->clk = devm_clk_get(&pdev->dev, NULL); + if (IS_ERR(dev->clk)) { + dev_err(&pdev->dev, "could not get clk: %ld\n", + PTR_ERR(dev->clk)); + return PTR_ERR(dev->clk); + } + + /* Request ioarea */ + mem = platform_get_resource(pdev, IORESOURCE_MEM, 0); + base = devm_ioremap_resource(&pdev->dev, mem); + if (IS_ERR(base)) + return PTR_ERR(base); + + dev->i2s_regmap = devm_regmap_init_mmio(&pdev->dev, base, + &bcm2835_regmap_config); + if (IS_ERR(dev->i2s_regmap)) + return PTR_ERR(dev->i2s_regmap); + + /* Set the DMA address - we have to parse DT ourselves */ + addr = of_get_address(pdev->dev.of_node, 0, NULL, NULL); + if (!addr) { + dev_err(&pdev->dev, "could not get DMA-register address\n"); + return -EINVAL; + } + dma_base = be32_to_cpup(addr);Why aren't we just using mem->start like before? That seems like an independent change that should be justified on its own. I'd be ready to ack the patch if that change is removed.
Problem is that we need the VC4 bus-address (0x7e203000), which is the actual <reg> value from the device tree without any mapping. Not the ARM MMU visible address mappings that mem->start provides (typically 0x20203000 or 0x3f203000 for bcm2836) Nor the mapped address (base) available in the kernel (typically 0xdc......). This is the only way to get it correctly and has been done the same way with spi-bcm2835. -- 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
