Re: [PATCH v2 1/6] ASoC: amd: acp: Add generic support for PDM controller on ACP

From: Ajit Kumar Pandey
Date: Mon Jan 17 2022 - 06:49:19 EST




On 1/14/2022 12:04 AM, Pierre-Louis Bossart wrote:
[CAUTION: External Email]

couple of nit-picks:


diff --git a/sound/soc/amd/acp/acp-pdm.c b/sound/soc/amd/acp/acp-pdm.c
new file mode 100644
index 000000000000..cb9bbd795eee
--- /dev/null
+++ b/sound/soc/amd/acp/acp-pdm.c
@@ -0,0 +1,181 @@
+// SPDX-License-Identifier: (GPL-2.0-only OR BSD-3-Clause)
+//
+// This file is provided under a dual BSD/GPLv2 license. When using or
+// redistributing this file, you may do so under either license.
+//
+// Copyright(c) 2021 Advanced Micro Devices, Inc.

2022?

+//
+// Authors: Ajit Kumar Pandey <AjitKumar.Pandey@xxxxxxx>
+// Vijendar Mukunda <Vijendar.Mukunda@xxxxxxx>
+//
+
+/*
+ * Generic Hardware interface for ACP Audio PDM controller
+ */
+
+#include <linux/platform_device.h>
+#include <linux/module.h>
+#include <linux/err.h>
+#include <linux/io.h>
+#include <sound/pcm_params.h>
+#include <sound/soc.h>
+#include <sound/soc-dai.h>
+#include <linux/dma-mapping.h>

alphabetical order?

+
+#include "amd.h"
+
+#define DRV_NAME "acp-pdm"
+
+#define PDM_DMA_STAT 0x10
+#define PDM_DMA_INTR_MASK 0x10000
+#define PDM_DEC_64 0x2
+#define PDM_CLK_FREQ_MASK 0x07
+#define PDM_MISC_CTRL_MASK 0x10
+#define PDM_ENABLE 0x01
+#define PDM_DISABLE 0x00
+#define DMA_EN_MASK 0x02
+#define DELAY_US 5
+#define PDM_TIMEOUT 1000
+
+static int acp_dmic_dai_trigger(struct snd_pcm_substream *substream,
+ int cmd, struct snd_soc_dai *dai)
+{
+ struct acp_stream *stream = substream->runtime->private_data;
+ struct device *dev = dai->component->dev;
+ struct acp_dev_data *adata = dev_get_drvdata(dev);
+ u32 physical_addr, size_dmic, period_bytes;
+ unsigned int dma_enable;
+ int ret = 0;
+
+ period_bytes = frames_to_bytes(substream->runtime,
+ substream->runtime->period_size);
+ size_dmic = frames_to_bytes(substream->runtime,
+ substream->runtime->buffer_size);
+
+ physical_addr = stream->reg_offset + MEM_WINDOW_START;
+
+ /* Init DMIC Ring buffer */
+ writel(physical_addr, adata->acp_base + ACP_WOV_RX_RINGBUFADDR);
+ writel(size_dmic, adata->acp_base + ACP_WOV_RX_RINGBUFSIZE);
+ writel(period_bytes, adata->acp_base + ACP_WOV_RX_INTR_WATERMARK_SIZE);
+ writel(0x01, adata->acp_base + ACPAXI2AXI_ATU_CTRL);

could this be done in a .prepare step?

+
+ switch (cmd) {
+ case SNDRV_PCM_TRIGGER_START:
+ case SNDRV_PCM_TRIGGER_RESUME:
+ case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:
+ dma_enable = readl(adata->acp_base + ACP_WOV_PDM_DMA_ENABLE);
+ if (!(dma_enable & DMA_EN_MASK)) {
+ writel(PDM_ENABLE, adata->acp_base + ACP_WOV_PDM_ENABLE);
+ writel(PDM_ENABLE, adata->acp_base + ACP_WOV_PDM_DMA_ENABLE);
+ }
+
+ ret = readl_poll_timeout_atomic(adata->acp_base + ACP_WOV_PDM_DMA_ENABLE,
+ dma_enable, (dma_enable & DMA_EN_MASK),
+ DELAY_US, PDM_TIMEOUT);
+ break;
+ case SNDRV_PCM_TRIGGER_STOP:
+ case SNDRV_PCM_TRIGGER_SUSPEND:
+ case SNDRV_PCM_TRIGGER_PAUSE_PUSH:
+ dma_enable = readl(adata->acp_base + ACP_WOV_PDM_DMA_ENABLE);
+ if ((dma_enable & DMA_EN_MASK)) {
+ writel(PDM_DISABLE, adata->acp_base + ACP_WOV_PDM_ENABLE);
+ writel(PDM_DISABLE, adata->acp_base + ACP_WOV_PDM_DMA_ENABLE);
+
+ }
+
+ ret = readl_poll_timeout_atomic(adata->acp_base + ACP_WOV_PDM_DMA_ENABLE,
+ dma_enable, !(dma_enable & DMA_EN_MASK),
+ DELAY_US, PDM_TIMEOUT);

is the _atomic needed?

+ break;
+ default:
+ ret = -EINVAL;
+ break;
+ }
+
+ return ret;
+}
+
+static int acp_dmic_hwparams(struct snd_pcm_substream *substream,
+ struct snd_pcm_hw_params *hwparams, struct snd_soc_dai *dai)
+{
+ struct device *dev = dai->component->dev;
+ struct acp_dev_data *adata = dev_get_drvdata(dev);
+ unsigned int dmic_ctrl, channels, ch_mask;
+
+ /* Enable default DMIC clk */
+ writel(PDM_CLK_FREQ_MASK, adata->acp_base + ACP_WOV_CLK_CTRL);
+ dmic_ctrl = readl(adata->acp_base + ACP_WOV_MISC_CTRL);
+ dmic_ctrl |= PDM_MISC_CTRL_MASK;
+ writel(dmic_ctrl, adata->acp_base + ACP_WOV_MISC_CTRL);

.hw_params can be called multiple times, should this clock handling be
done in a .prepare step?

Or alternatively in .startup - this doesn't seem to depend on hardware
parameters?

+
+ channels = params_channels(hwparams);
+ switch (channels) {
+ case 2:
+ ch_mask = 0;
+ break;
+ case 4:
+ ch_mask = 1;
+ break;
+ case 6:
+ ch_mask = 2;
+ break;
+ default:
+ dev_err(dev, "Invalid channels %d\n", channels);
+ return -EINVAL;
+ }
+
+ if (params_format(hwparams) != SNDRV_PCM_FORMAT_S32_LE) {
+ dev_err(dai->dev, "Invalid format:%d\n", params_format(hwparams));
+ return -EINVAL;
+ }
+
+ writel(ch_mask, adata->acp_base + ACP_WOV_PDM_NO_OF_CHANNELS);
+ writel(PDM_DEC_64, adata->acp_base + ACP_WOV_PDM_DECIMATION_FACTOR);
+
+ return 0;
+}

+MODULE_LICENSE("GPL v2");

"GPL" is enough


Thanks, will try to adopt commnets in v3 patch chain