Re: [PATCH v3 3/3] firmware: mediatek: add adsp ipc protocol interface - Tzung-Bi Shih

From: Tzung-Bi Shih <tzungbi@google.com>
To: "allen-kh.cheng" <allen-kh.cheng@mediatek.com>
Cc: devicetree@vger.kernel.org,
	Linux-ALSA <alsa-devel@alsa-project.org>,
	Kai Vehmanen <kai.vehmanen@linux.intel.com>,
	cujomalainey@google.com, linux-kernel@vger.kernel.org,
	Mark Brown <broonie@kernel.org>,
	Jassi Brar <jassisinghbrar@gmail.com>,
	Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>,
	Liam Girdwood <lgirdwood@gmail.com>,
	Project_Global_Chrome_Upstream_Group@mediatek.com,
	Rob Herring <robh+dt@kernel.org>,
	linux-mediatek@lists.infradead.org,
	Ranjani Sridharan <ranjani.sridharan@linux.intel.com>,
	Matthias Brugger <matthias.bgg@gmail.com>,
	Takashi Iwai <tiwai@suse.com>,
	Daniel Baluta <daniel.baluta@nxp.com>,
	linux-arm-kernel@lists.infradead.org,
	sound-open-firmware@alsa-project.org
Subject: Re: [PATCH v3 3/3] firmware: mediatek: add adsp ipc protocol interface
Date: Wed, 24 Nov 2021 18:25:30 +0800	[thread overview]
Message-ID: <YZ4TGvtwO5ySCEkT@google.com> (raw)
In-Reply-To: <20211124084514.28002-4-allen-kh.cheng@mediatek.com>

On Wed, Nov 24, 2021 at 04:45:14PM +0800, allen-kh.cheng wrote:
>  drivers/firmware/Kconfig                      |   1 +
>  drivers/firmware/Makefile                     |   1 +
>  drivers/firmware/mediatek/Kconfig             |  10 ++
>  drivers/firmware/mediatek/Makefile            |   2 +
>  drivers/firmware/mediatek/mtk-adsp-ipc.c      | 130 ++++++++++++++++++
>  .../linux/firmware/mediatek/mtk-adsp-ipc.h    |  72 ++++++++++
>  6 files changed, 216 insertions(+)
>  create mode 100644 drivers/firmware/mediatek/Kconfig
>  create mode 100644 drivers/firmware/mediatek/Makefile
>  create mode 100644 drivers/firmware/mediatek/mtk-adsp-ipc.c
>  create mode 100644 include/linux/firmware/mediatek/mtk-adsp-ipc.h

The patch should move before the 2nd patch in the series as the 2nd patch uses mtk-adsp-ipc.h.

> diff --git a/drivers/firmware/mediatek/mtk-adsp-ipc.c b/drivers/firmware/mediatek/mtk-adsp-ipc.c
[...]
> +int adsp_ipc_send(struct mtk_adsp_ipc *ipc, unsigned int idx, uint32_t op)
> +{
> +	struct mtk_adsp_chan *dsp_chan = &ipc->chans[idx];
> +	struct adsp_mbox_ch_info *ch_info = dsp_chan->ch->con_priv;
> +	int ret;
> +
> +	if (idx >= MTK_ADSP_MBOX_NUM)
> +		return -EINVAL;

If idx >= MTK_ADSP_MBOX_NUM, the invalid memory access has occurred at beginning of the function.

> +static int mtk_adsp_ipc_probe(struct platform_device *pdev)
> +{
[...]
> +	device_set_of_node_from_dev(&pdev->dev, pdev->dev.parent);

Why does it need to call device_set_of_node_from_dev()?

> +	for (i = 0; i < MTK_ADSP_MBOX_NUM; i++) {
> +		chan_name = kasprintf(GFP_KERNEL, "mbox%d", i);
> +		if (!chan_name)
> +			return -ENOMEM;
> +
> +		dsp_chan = &dsp_ipc->chans[i];
> +		cl = &dsp_chan->cl;
> +		cl->dev = dev->parent;
> +		cl->tx_block = false;
> +		cl->knows_txdone = false;
> +		cl->tx_prepare = NULL;
> +		cl->rx_callback = adsp_ipc_recv;
> +
> +		dsp_chan->ipc = dsp_ipc;
> +		dsp_chan->idx = i;
> +		dsp_chan->ch = mbox_request_channel_byname(cl, chan_name);
> +		if (IS_ERR(dsp_chan->ch)) {
> +			ret = PTR_ERR(dsp_chan->ch);
> +			if (ret != -EPROBE_DEFER)
> +				dev_err(dev, "Failed to request mbox chan %d ret %d\n",
> +					i, ret);

If ret == -EPROBE_DEFER, wouldn't it need to return -EPROBE_DEFER?  It doesn't retry later if -EPROBE_DEFER.