Re: [PATCH] tty: serial: add missing spin_lock_init for SiFive UART

[Greg KH <gregkh@linuxfoundation.org>]

On Wed, May 06, 2020 at 12:26:52PM -0700, Sagar Shrikant Kadam wrote:
> Uninitialised spin lock on sifive serial port (ssp) raised
> a race condition and resulted in spin lock bad magic as
> reported and discussed here [1]
> Initialising the spin lock resolves the issue.
> 
> The fix is tested on HiFive Unleashed A00 board with Linux 5.7-rc4
> and OpenSBI v0.7
> 
> [1] http://lists.infradead.org/pipermail/linux-riscv/2020-May/009713.html

In the future, can you link to lore.kernel.org instead?

> 
> Fixes: 45c054d0815b ("tty: serial: add driver for the SiFive UART")

So this should also go to stable kernels, right?

> Reported-by: Atish Patra <Atish.Patra@wdc.com>
> Signed-off-by: Sagar Shrikant Kadam <sagar.kadam@sifive.com>
> ---
>  drivers/tty/serial/sifive.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/tty/serial/sifive.c b/drivers/tty/serial/sifive.c
> index 13eadcb..b061bdb 100644
> --- a/drivers/tty/serial/sifive.c
> +++ b/drivers/tty/serial/sifive.c
> @@ -989,6 +989,7 @@ static int sifive_serial_probe(struct platform_device *pdev)
>  	ssp->clk = clk;
>  	ssp->clk_notifier.notifier_call = sifive_serial_clk_notifier;
>  
> +	spin_lock_init(&ssp->port.lock);

Shouldn't the port lock be initialized by the tty core instead?  I think
this is the second time I've seen this type of fix needed recently...

thanks,

greg k-h