2020-10-11 08:31:53

by Ran Wang

[permalink] [raw]
Subject: [PATCH] usb: host: fsl-mph-dr-of: check return of dma_set_mask()

fsl_usb2_device_register() should stop init if dma_set_mask() return
error.

Fixes: cae058610465 ("drivers/usb/host: fsl: Set DMA_MASK of usb platform device")
Signed-off-by: Ran Wang <[email protected]>
---
drivers/usb/host/fsl-mph-dr-of.c | 9 ++++++---
1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/usb/host/fsl-mph-dr-of.c b/drivers/usb/host/fsl-mph-dr-of.c
index ae8f60f..44a7e58 100644
--- a/drivers/usb/host/fsl-mph-dr-of.c
+++ b/drivers/usb/host/fsl-mph-dr-of.c
@@ -94,10 +94,13 @@ static struct platform_device *fsl_usb2_device_register(

pdev->dev.coherent_dma_mask = ofdev->dev.coherent_dma_mask;

- if (!pdev->dev.dma_mask)
+ if (!pdev->dev.dma_mask) {
pdev->dev.dma_mask = &ofdev->dev.coherent_dma_mask;
- else
- dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
+ } else {
+ retval = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
+ if (retval)
+ goto error;
+ }

retval = platform_device_add_data(pdev, pdata, sizeof(*pdata));
if (retval)
--
2.7.4


2020-10-11 18:50:36

by Peter Chen

[permalink] [raw]
Subject: Re: [PATCH] usb: host: fsl-mph-dr-of: check return of dma_set_mask()

On 20-10-10 14:03:08, Ran Wang wrote:
> fsl_usb2_device_register() should stop init if dma_set_mask() return
> error.
>
> Fixes: cae058610465 ("drivers/usb/host: fsl: Set DMA_MASK of usb platform device")
> Signed-off-by: Ran Wang <[email protected]>
> ---
> drivers/usb/host/fsl-mph-dr-of.c | 9 ++++++---
> 1 file changed, 6 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/usb/host/fsl-mph-dr-of.c b/drivers/usb/host/fsl-mph-dr-of.c
> index ae8f60f..44a7e58 100644
> --- a/drivers/usb/host/fsl-mph-dr-of.c
> +++ b/drivers/usb/host/fsl-mph-dr-of.c
> @@ -94,10 +94,13 @@ static struct platform_device *fsl_usb2_device_register(
>
> pdev->dev.coherent_dma_mask = ofdev->dev.coherent_dma_mask;
>
> - if (!pdev->dev.dma_mask)
> + if (!pdev->dev.dma_mask) {
> pdev->dev.dma_mask = &ofdev->dev.coherent_dma_mask;
> - else
> - dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
> + } else {
> + retval = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
> + if (retval)
> + goto error;
> + }
>
> retval = platform_device_add_data(pdev, pdata, sizeof(*pdata));
> if (retval)
> --
> 2.7.4
>

Reviewed-by: Peter Chen <[email protected]>

One more place need to fix, if platform_device_alloc returns NULL,
it should not call platform_device_put to release platform
device memory.

pdev = platform_device_alloc(name, id);
if (!pdev) {
retval = -ENOMEM;
goto error;
}
...
error:
platform_device_put(pdev);
return ERR_PTR(retval);
--

Thanks,
Peter Chen

2020-10-14 13:29:14

by Ran Wang

[permalink] [raw]
Subject: RE: [PATCH] usb: host: fsl-mph-dr-of: check return of dma_set_mask()

Hi Peter,

On Sunday, October 11, 2020 7:41 AM Peter Chen wrote:
>
> On 20-10-10 14:03:08, Ran Wang wrote:
> > fsl_usb2_device_register() should stop init if dma_set_mask() return
> > error.
> >
> > Fixes: cae058610465 ("drivers/usb/host: fsl: Set DMA_MASK of usb
> > platform device")
> > Signed-off-by: Ran Wang <[email protected]>
> > ---
> > drivers/usb/host/fsl-mph-dr-of.c | 9 ++++++---
> > 1 file changed, 6 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/usb/host/fsl-mph-dr-of.c
> > b/drivers/usb/host/fsl-mph-dr-of.c
> > index ae8f60f..44a7e58 100644
> > --- a/drivers/usb/host/fsl-mph-dr-of.c
> > +++ b/drivers/usb/host/fsl-mph-dr-of.c
> > @@ -94,10 +94,13 @@ static struct platform_device
> > *fsl_usb2_device_register(
> >
> > pdev->dev.coherent_dma_mask = ofdev->dev.coherent_dma_mask;
> >
> > - if (!pdev->dev.dma_mask)
> > + if (!pdev->dev.dma_mask) {
> > pdev->dev.dma_mask = &ofdev->dev.coherent_dma_mask;
> > - else
> > - dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
> > + } else {
> > + retval = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
> > + if (retval)
> > + goto error;
> > + }
> >
> > retval = platform_device_add_data(pdev, pdata, sizeof(*pdata));
> > if (retval)
> > --
> > 2.7.4
> >
>
> Reviewed-by: Peter Chen <[email protected]>
>
> One more place need to fix, if platform_device_alloc returns NULL,
> it should not call platform_device_put to release platform
> device memory.
>
> pdev = platform_device_alloc(name, id);
> if (!pdev) {
> retval = -ENOMEM;
> goto error;
> }
> ...
> error:
> platform_device_put(pdev);
> return ERR_PTR(retval);

Got it, let me check this later.

Thanks & Regards,
Ran