Re: [PATCH] spi: spi-zynqmp-gqspi: unlock on error path in exec_op

["quanyang.wang" <quanyang.wang@xxxxxxxxxxxxx>]

Hi Dan,

On 4/13/21 6:50 PM, Dan Carpenter wrote:
> If the allocation fails then we need to drop the &xqspi->op_lock mutex
> before returning.
>
> Fixes: a0f65be6e880 ("spi: spi-zynqmp-gqspi: add mutex locking for exec_op")
> Signed-off-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
> ---
>   drivers/spi/spi-zynqmp-gqspi.c | 6 ++++--
>   1 file changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/spi/spi-zynqmp-gqspi.c b/drivers/spi/spi-zynqmp-gqspi.c
> index 9494da9aea40..d08ca6d5670c 100644
> --- a/drivers/spi/spi-zynqmp-gqspi.c
> +++ b/drivers/spi/spi-zynqmp-gqspi.c
> @@ -965,8 +965,10 @@ static int zynqmp_qspi_exec_op(struct spi_mem *mem,
>   
>   	if (op->cmd.opcode) {
>   		tmpbuf = kzalloc(op->cmd.nbytes, GFP_KERNEL | GFP_DMA);
> -		if (!tmpbuf)
> -			return -ENOMEM;
> +		if (!tmpbuf) {
> +			err = -ENOMEM;
> +			goto return_err;
> +		}
>   		tmpbuf[0] = op->cmd.opcode;
>   		reinit_completion(&xqspi->data_completion);
>   		xqspi->txbuf = tmpbuf;

Thank you for your patch. But similar patch has been applied:

https://www.spinics.net/lists/arm-kernel/msg888807.html

Thanks,

Quanyang