Re: [PATCH] [v3] spi: spi-zynqmp-gqspi: Fix runtime PM imbalance in zynqmp_qspi_probe

From: "quanyang.wang" <quanyang.wang@windriver.com>
To: Dinghao Liu <dinghao.liu@zju.edu.cn>, kjlu@umn.edu
Cc: Mark Brown <broonie@kernel.org>,
	Michal Simek <michal.simek@xilinx.com>,
	Naga Sureshkumar Relli <naga.sureshkumar.relli@xilinx.com>,
	Shubhrajyoti Datta <shubhrajyoti.datta@xilinx.com>,
	linux-spi@vger.kernel.org, linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH] [v3] spi: spi-zynqmp-gqspi: Fix runtime PM imbalance in zynqmp_qspi_probe
Date: Thu, 15 Apr 2021 15:27:51 +0800	[thread overview]
Message-ID: <a9fd2d17-ff87-f4cb-e255-8023ff0bba8b@windriver.com> (raw)
In-Reply-To: <20210415071714.19334-1-dinghao.liu@zju.edu.cn>

Hi Dinghao,

On 4/15/21 3:17 PM, Dinghao Liu wrote:
> There is a PM usage counter decrement after zynqmp_qspi_init_hw()
> without any refcount increment, which leads to refcount leak.Add
> a refcount increment to balance the refcount. Also set
> auto_runtime_pm to resume suspended spi controller.
>
> Fixes: 9e3a000362aec ("spi: zynqmp: Add pm runtime support")
> Signed-off-by: Dinghao Liu <dinghao.liu@zju.edu.cn>
> ---
>
> Changelog:
>
> v2: - Add a refcount increment to fix refcout leak instead of the
>        refcount decrement on error.
>        Set ctlr->auto_runtime_pm = true.
>
> v3: - Add fix tag.
>        Add a return value check against pm_runtime_get_sync().
>        Move pm_runtime_{mark_last_busy & put_autosuspend} to the
>        end of current function.
> ---
>   drivers/spi/spi-zynqmp-gqspi.c | 12 ++++++++++--
>   1 file changed, 10 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/spi/spi-zynqmp-gqspi.c b/drivers/spi/spi-zynqmp-gqspi.c
> index c8fa6ee18ae7..781ef3fc76e2 100644
> --- a/drivers/spi/spi-zynqmp-gqspi.c
> +++ b/drivers/spi/spi-zynqmp-gqspi.c
> @@ -1160,11 +1160,14 @@ static int zynqmp_qspi_probe(struct platform_device *pdev)
>   	pm_runtime_set_autosuspend_delay(&pdev->dev, SPI_AUTOSUSPEND_TIMEOUT);
>   	pm_runtime_set_active(&pdev->dev);
>   	pm_runtime_enable(&pdev->dev);
> +
> +	ret = pm_runtime_get_sync(&pdev->dev);
> +	if (ret < 0)
> +		goto clk_dis_all;

Would you please add some error info output here? Just like:

+       if (ret < 0) {
+               dev_err(&pdev->dev, "failed to pm_runtime_get_sync: 
%d\n", ret);
+               goto clk_dis_all;
+       }
Because sometimes it is very helpful to analyse probe failure.

Thanks,

Quanyang

> +
>   	/* QSPI controller initializations */
>   	zynqmp_qspi_init_hw(xqspi);
>   
> -	pm_runtime_mark_last_busy(&pdev->dev);
> -	pm_runtime_put_autosuspend(&pdev->dev);
>   	xqspi->irq = platform_get_irq(pdev, 0);
>   	if (xqspi->irq <= 0) {
>   		ret = -ENXIO;
> @@ -1187,6 +1190,7 @@ static int zynqmp_qspi_probe(struct platform_device *pdev)
>   	ctlr->mode_bits = SPI_CPOL | SPI_CPHA | SPI_RX_DUAL | SPI_RX_QUAD |
>   			    SPI_TX_DUAL | SPI_TX_QUAD;
>   	ctlr->dev.of_node = np;
> +	ctlr->auto_runtime_pm = true;
>   
>   	ret = devm_spi_register_controller(&pdev->dev, ctlr);
>   	if (ret) {
> @@ -1194,9 +1198,13 @@ static int zynqmp_qspi_probe(struct platform_device *pdev)
>   		goto clk_dis_all;
>   	}
>   
> +	pm_runtime_mark_last_busy(&pdev->dev);
> +	pm_runtime_put_autosuspend(&pdev->dev);
> +
>   	return 0;
>   
>   clk_dis_all:
> +	pm_runtime_put_sync(&pdev->dev);
>   	pm_runtime_set_suspended(&pdev->dev);
>   	pm_runtime_disable(&pdev->dev);
>   	clk_disable_unprepare(xqspi->refclk);