| Message ID | 20171206163852.8532-1-flavio.ceolin@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi Flavio, Thank you for the patch. On Wednesday, 6 December 2017 18:38:50 EET Flavio Ceolin wrote: > pxa_camera_probe() was not calling pxa_camera_deactivate(), > responsible to call clk_disable_unprepare(), on the failure path. This > was leading to unbalancing source clock. > > Found by Linux Driver Verification project (linuxtesting.org). Any chance I could sign you up for more work on this driver ? :-) > Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> I expect Hans Verkuil to pick up the patch. > --- > drivers/media/platform/pxa_camera.c | 4 +++- > 1 file changed, 3 insertions(+), 1 deletion(-) > > diff --git a/drivers/media/platform/pxa_camera.c > b/drivers/media/platform/pxa_camera.c index 9d3f0cb..7877037 100644 > --- a/drivers/media/platform/pxa_camera.c > +++ b/drivers/media/platform/pxa_camera.c > @@ -2489,7 +2489,7 @@ static int pxa_camera_probe(struct platform_device > *pdev) dev_set_drvdata(&pdev->dev, pcdev); > err = v4l2_device_register(&pdev->dev, &pcdev->v4l2_dev); > if (err) > - goto exit_free_dma; > + goto exit_deactivate; > > pcdev->asds[0] = &pcdev->asd; > pcdev->notifier.subdevs = pcdev->asds; > @@ -2525,6 +2525,8 @@ static int pxa_camera_probe(struct platform_device > *pdev) v4l2_clk_unregister(pcdev->mclk_clk); > exit_free_v4l2dev: > v4l2_device_unregister(&pcdev->v4l2_dev); > +exit_deactivate: > + pxa_camera_deactivate(pcdev); > exit_free_dma: > dma_release_channel(pcdev->dma_chans[2]); > exit_free_dma_u:
Hi Laurent, > Hi Flavio, > > Thank you for the patch. > > On Wednesday, 6 December 2017 18:38:50 EET Flavio Ceolin wrote: >> pxa_camera_probe() was not calling pxa_camera_deactivate(), >> responsible to call clk_disable_unprepare(), on the failure path. This >> was leading to unbalancing source clock. >> >> Found by Linux Driver Verification project (linuxtesting.org). > > Any chance I could sign you up for more work on this driver ? :-) Definetely, this would be great :) > >> Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com> > > Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> > > I expect Hans Verkuil to pick up the patch. > >> --- >> drivers/media/platform/pxa_camera.c | 4 +++- >> 1 file changed, 3 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/media/platform/pxa_camera.c >> b/drivers/media/platform/pxa_camera.c index 9d3f0cb..7877037 100644 >> --- a/drivers/media/platform/pxa_camera.c >> +++ b/drivers/media/platform/pxa_camera.c >> @@ -2489,7 +2489,7 @@ static int pxa_camera_probe(struct platform_device >> *pdev) dev_set_drvdata(&pdev->dev, pcdev); >> err = v4l2_device_register(&pdev->dev, &pcdev->v4l2_dev); >> if (err) >> - goto exit_free_dma; >> + goto exit_deactivate; >> >> pcdev->asds[0] = &pcdev->asd; >> pcdev->notifier.subdevs = pcdev->asds; >> @@ -2525,6 +2525,8 @@ static int pxa_camera_probe(struct platform_device >> *pdev) v4l2_clk_unregister(pcdev->mclk_clk); >> exit_free_v4l2dev: >> v4l2_device_unregister(&pcdev->v4l2_dev); >> +exit_deactivate: >> + pxa_camera_deactivate(pcdev); >> exit_free_dma: >> dma_release_channel(pcdev->dma_chans[2]); >> exit_free_dma_u: > > -- > Regards, > > Laurent Pinchart Regards, Flavio Ceolin
Hi Flavio, On Monday, 11 December 2017 23:05:46 EET Flavio Ceolin wrote: > > On Wednesday, 6 December 2017 18:38:50 EET Flavio Ceolin wrote: > >> pxa_camera_probe() was not calling pxa_camera_deactivate(), > >> responsible to call clk_disable_unprepare(), on the failure path. This > >> was leading to unbalancing source clock. > >> > >> Found by Linux Driver Verification project (linuxtesting.org). > > > > Any chance I could sign you up for more work on this driver ? :-) > > Definetely, this would be great :) Actually it looks like the work I thought was needed has already been performed. The pxa-camera driver used to make use of the soc-camera framework, which we are trying to remove, and occurrences of soc_camera in the code gave me the wrong idea that the driver was still based on that framework. It seems this isn't the case. Thank you for making me happy :-) > >> Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com> > > > > Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> > > > > I expect Hans Verkuil to pick up the patch. > > > >> --- > >> > >> drivers/media/platform/pxa_camera.c | 4 +++- > >> 1 file changed, 3 insertions(+), 1 deletion(-) > >> > >> diff --git a/drivers/media/platform/pxa_camera.c > >> b/drivers/media/platform/pxa_camera.c index 9d3f0cb..7877037 100644 > >> --- a/drivers/media/platform/pxa_camera.c > >> +++ b/drivers/media/platform/pxa_camera.c > >> @@ -2489,7 +2489,7 @@ static int pxa_camera_probe(struct platform_device > >> *pdev) > >> dev_set_drvdata(&pdev->dev, pcdev); > >> err = v4l2_device_register(&pdev->dev, &pcdev->v4l2_dev); > >> if (err) > >> - goto exit_free_dma; > >> + goto exit_deactivate; > >> > >> pcdev->asds[0] = &pcdev->asd; > >> pcdev->notifier.subdevs = pcdev->asds; > >> @@ -2525,6 +2525,8 @@ static int pxa_camera_probe(struct platform_device > >> *pdev) > >> v4l2_clk_unregister(pcdev->mclk_clk); > >> exit_free_v4l2dev: > >> v4l2_device_unregister(&pcdev->v4l2_dev); > >> +exit_deactivate: > >> + pxa_camera_deactivate(pcdev); > >> exit_free_dma: > >> dma_release_channel(pcdev->dma_chans[2]); > >> exit_free_dma_u:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> writes: > Hi Flavio, > > On Monday, 11 December 2017 23:05:46 EET Flavio Ceolin wrote: >> > On Wednesday, 6 December 2017 18:38:50 EET Flavio Ceolin wrote: >> >> pxa_camera_probe() was not calling pxa_camera_deactivate(), >> >> responsible to call clk_disable_unprepare(), on the failure path. This >> >> was leading to unbalancing source clock. >> >> >> >> Found by Linux Driver Verification project (linuxtesting.org). >> > >> > Any chance I could sign you up for more work on this driver ? :-) >> >> Definetely, this would be great :) > > Actually it looks like the work I thought was needed has already been > performed. The pxa-camera driver used to make use of the soc-camera framework, > which we are trying to remove, and occurrences of soc_camera in the code gave > me the wrong idea that the driver was still based on that framework. It seems > this isn't the case. Thank you for making me happy :-) I'm glad to hear it :) > >> >> Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com> >> > >> > Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> >> > >> > I expect Hans Verkuil to pick up the patch. >> > >> >> --- >> >> >> >> drivers/media/platform/pxa_camera.c | 4 +++- >> >> 1 file changed, 3 insertions(+), 1 deletion(-) >> >> >> >> diff --git a/drivers/media/platform/pxa_camera.c >> >> b/drivers/media/platform/pxa_camera.c index 9d3f0cb..7877037 100644 >> >> --- a/drivers/media/platform/pxa_camera.c >> >> +++ b/drivers/media/platform/pxa_camera.c >> >> @@ -2489,7 +2489,7 @@ static int pxa_camera_probe(struct platform_device >> >> *pdev) >> >> dev_set_drvdata(&pdev->dev, pcdev); >> >> err = v4l2_device_register(&pdev->dev, &pcdev->v4l2_dev); >> >> if (err) >> >> - goto exit_free_dma; >> >> + goto exit_deactivate; >> >> >> >> pcdev->asds[0] = &pcdev->asd; >> >> pcdev->notifier.subdevs = pcdev->asds; >> >> @@ -2525,6 +2525,8 @@ static int pxa_camera_probe(struct platform_device >> >> *pdev) >> >> v4l2_clk_unregister(pcdev->mclk_clk); >> >> exit_free_v4l2dev: >> >> v4l2_device_unregister(&pcdev->v4l2_dev); >> >> +exit_deactivate: >> >> + pxa_camera_deactivate(pcdev); >> >> exit_free_dma: >> >> dma_release_channel(pcdev->dma_chans[2]); >> >> exit_free_dma_u: > > -- > Regards, > > Laurent Pinchart Regards, Flavio Ceolin
diff --git a/drivers/media/platform/pxa_camera.c b/drivers/media/platform/pxa_camera.c index 9d3f0cb..7877037 100644 --- a/drivers/media/platform/pxa_camera.c +++ b/drivers/media/platform/pxa_camera.c @@ -2489,7 +2489,7 @@ static int pxa_camera_probe(struct platform_device *pdev) dev_set_drvdata(&pdev->dev, pcdev); err = v4l2_device_register(&pdev->dev, &pcdev->v4l2_dev); if (err) - goto exit_free_dma; + goto exit_deactivate; pcdev->asds[0] = &pcdev->asd; pcdev->notifier.subdevs = pcdev->asds; @@ -2525,6 +2525,8 @@ static int pxa_camera_probe(struct platform_device *pdev) v4l2_clk_unregister(pcdev->mclk_clk); exit_free_v4l2dev: v4l2_device_unregister(&pcdev->v4l2_dev); +exit_deactivate: + pxa_camera_deactivate(pcdev); exit_free_dma: dma_release_channel(pcdev->dma_chans[2]); exit_free_dma_u:
pxa_camera_probe() was not calling pxa_camera_deactivate(), responsible to call clk_disable_unprepare(), on the failure path. This was leading to unbalancing source clock. Found by Linux Driver Verification project (linuxtesting.org). Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com> --- drivers/media/platform/pxa_camera.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)