[PATCH] pps: add an error check in parport_attach

[PATCH] pps: add an error check in parport_attach

[Ma Ke]

In parport_attach, the return value of ida_alloc is unchecked, witch leads
to the use of an invalid index value.

To address this issue, index should be checked. When the index value is
abnormal, the device should be freed.

Found by code review, compile tested only.

Cc: stable@vger.kernel.org
Fixes: 55dbc5b5174d ("pps: remove usage of the deprecated ida_simple_xx() API")
Signed-off-by: Ma Ke <make24@iscas.ac.cn>
---
 drivers/pps/clients/pps_parport.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/pps/clients/pps_parport.c b/drivers/pps/clients/pps_parport.c
index 63d03a0df5cc..9ab7f6961e42 100644
--- a/drivers/pps/clients/pps_parport.c
+++ b/drivers/pps/clients/pps_parport.c
@@ -149,6 +149,11 @@ static void parport_attach(struct parport *port)
 	}
 
 	index = ida_alloc(&pps_client_index, GFP_KERNEL);
+	if (index < 0) {
+		pr_err("failed to get index\n");
+		goto err_free_device;
+	}
+
 	memset(&pps_client_cb, 0, sizeof(pps_client_cb));
 	pps_client_cb.private = device;
 	pps_client_cb.irq_func = parport_irq;
@@ -159,7 +164,7 @@ static void parport_attach(struct parport *port)
 						    index);
 	if (!device->pardev) {
 		pr_err("couldn't register with %s\n", port->name);
-		goto err_free;
+		goto err_free_ida;
 	}
 
 	if (parport_claim_or_block(device->pardev) < 0) {
@@ -187,8 +192,9 @@ static void parport_attach(struct parport *port)
 	parport_release(device->pardev);
 err_unregister_dev:
 	parport_unregister_device(device->pardev);
-err_free:
+err_free_ida:
 	ida_free(&pps_client_index, index);
+err_free_device:
 	kfree(device);
 }
 
-- 
2.25.1

Re: [PATCH] pps: add an error check in parport_attach

[Greg KH]

On Tue, Aug 13, 2024 at 11:08:00AM +0800, Ma Ke wrote:
> In parport_attach, the return value of ida_alloc is unchecked, witch leads
> to the use of an invalid index value.
> 
> To address this issue, index should be checked. When the index value is
> abnormal, the device should be freed.
> 
> Found by code review, compile tested only.
> 
> Cc: stable@vger.kernel.org
> Fixes: 55dbc5b5174d ("pps: remove usage of the deprecated ida_simple_xx() API")
> Signed-off-by: Ma Ke <make24@iscas.ac.cn>
> ---
>  drivers/pps/clients/pps_parport.c | 10 ++++++++--
>  1 file changed, 8 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/pps/clients/pps_parport.c b/drivers/pps/clients/pps_parport.c
> index 63d03a0df5cc..9ab7f6961e42 100644
> --- a/drivers/pps/clients/pps_parport.c
> +++ b/drivers/pps/clients/pps_parport.c
> @@ -149,6 +149,11 @@ static void parport_attach(struct parport *port)
>  	}
>  
>  	index = ida_alloc(&pps_client_index, GFP_KERNEL);
> +	if (index < 0) {
> +		pr_err("failed to get index\n");

No need to be noisy, right?

thanks,

greg k-h

Re: [PATCH] pps: add an error check in parport_attach

[Ma Ke]

Greg KH<gregkh@linuxfoundation.org> wrote:
> On Tue, Aug 13, 2024 at 11:08:00AM +0800, Ma Ke wrote:
> > In parport_attach, the return value of ida_alloc is unchecked, witch leads
> > to the use of an invalid index value.
> > 
> > To address this issue, index should be checked. When the index value is
> > abnormal, the device should be freed.
> > 
> > Found by code review, compile tested only.
> > 
> > Cc: stable@vger.kernel.org
> > Fixes: 55dbc5b5174d ("pps: remove usage of the deprecated ida_simple_xx() API")
> > Signed-off-by: Ma Ke <make24@iscas.ac.cn>
> > ---
> >  drivers/pps/clients/pps_parport.c | 10 ++++++++--
> >  1 file changed, 8 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/pps/clients/pps_parport.c b/drivers/pps/clients/pps_parport.c
> > index 63d03a0df5cc..9ab7f6961e42 100644
> > --- a/drivers/pps/clients/pps_parport.c
> > +++ b/drivers/pps/clients/pps_parport.c
> > @@ -149,6 +149,11 @@ static void parport_attach(struct parport *port)
> >  	}
> >  
> >  	index = ida_alloc(&pps_client_index, GFP_KERNEL);
> > +	if (index < 0) {
> > +		pr_err("failed to get index\n");
> 
> No need to be noisy, right?
> 
> thanks,
> 
> greg k-h

Firstly, I would like to express my gratitude for your valuable suggestions
on the patch I submitted. Based on your feedback, I understand that it is 
unnecessary to output error messages in this function. If this 
interpretation is correct, I will make the necessary modifications and 
resubmit the patch v2. Thank you for your response.

Best regards,

Ma Ke