| Message ID | 2674c8df1d05ab312826b69bfe9559f81d125a0b.1744975624.git.christophe.jaillet@wanadoo.fr (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | spi: stm32-ospi: Fix an error handling path in stm32_ospi_probe() | expand |
On 4/18/2025 4:57 PM, Christophe JAILLET wrote: [...] > diff --git a/drivers/spi/spi-stm32-ospi.c b/drivers/spi/spi-stm32-ospi.c > index 668022098b1e..9ec9823409cc 100644 > --- a/drivers/spi/spi-stm32-ospi.c > +++ b/drivers/spi/spi-stm32-ospi.c > @@ -960,6 +960,10 @@ static int stm32_ospi_probe(struct platform_device *pdev) > err_pm_enable: > pm_runtime_force_suspend(ospi->dev); > mutex_destroy(&ospi->lock); > + if (ospi->dma_chtx) > + dma_release_channel(ospi->dma_chtx); why can't you move to devm_dma_request_chan ? No need to cleanup. > + if (ospi->dma_chrx) > + dma_release_channel(ospi->dma_chrx); > > return ret; > }
Le 18/04/2025 à 14:09, Mukesh Kumar Savaliya a écrit : > > > On 4/18/2025 4:57 PM, Christophe JAILLET wrote: > [...] >> diff --git a/drivers/spi/spi-stm32-ospi.c b/drivers/spi/spi-stm32-ospi.c >> index 668022098b1e..9ec9823409cc 100644 >> --- a/drivers/spi/spi-stm32-ospi.c >> +++ b/drivers/spi/spi-stm32-ospi.c >> @@ -960,6 +960,10 @@ static int stm32_ospi_probe(struct >> platform_device *pdev) >> err_pm_enable: >> pm_runtime_force_suspend(ospi->dev); >> mutex_destroy(&ospi->lock); >> + if (ospi->dma_chtx) >> + dma_release_channel(ospi->dma_chtx); > why can't you move to devm_dma_request_chan ? No need to cleanup. Unless I miss something obvious, this function does not exist. CJ >> + if (ospi->dma_chrx) >> + dma_release_channel(ospi->dma_chrx); >> return ret; >> } > > >
On 4/18/2025 5:50 PM, Christophe JAILLET wrote: > Le 18/04/2025 à 14:09, Mukesh Kumar Savaliya a écrit : >> >> >> On 4/18/2025 4:57 PM, Christophe JAILLET wrote: >> [...] >>> diff --git a/drivers/spi/spi-stm32-ospi.c b/drivers/spi/spi-stm32-ospi.c >>> index 668022098b1e..9ec9823409cc 100644 >>> --- a/drivers/spi/spi-stm32-ospi.c >>> +++ b/drivers/spi/spi-stm32-ospi.c >>> @@ -960,6 +960,10 @@ static int stm32_ospi_probe(struct >>> platform_device *pdev) >>> err_pm_enable: >>> pm_runtime_force_suspend(ospi->dev); >>> mutex_destroy(&ospi->lock); >>> + if (ospi->dma_chtx) >>> + dma_release_channel(ospi->dma_chtx); >> why can't you move to devm_dma_request_chan ? No need to cleanup. > > Unless I miss something obvious, this function does not exist. > > CJ > Yes, You are right. Seems the patch is yet not merged. https://lore.kernel.org/all/20250115160244.1102881-1-csokas.bence@prolan.hu/T/ dma: Add devm_dma_request_chan() Can ignore it for now. >>> + if (ospi->dma_chrx) >>> + dma_release_channel(ospi->dma_chrx); >>> return ret; >>> } >> >> >> >
diff --git a/drivers/spi/spi-stm32-ospi.c b/drivers/spi/spi-stm32-ospi.c index 668022098b1e..9ec9823409cc 100644 --- a/drivers/spi/spi-stm32-ospi.c +++ b/drivers/spi/spi-stm32-ospi.c @@ -960,6 +960,10 @@ static int stm32_ospi_probe(struct platform_device *pdev) err_pm_enable: pm_runtime_force_suspend(ospi->dev); mutex_destroy(&ospi->lock); + if (ospi->dma_chtx) + dma_release_channel(ospi->dma_chtx); + if (ospi->dma_chrx) + dma_release_channel(ospi->dma_chrx); return ret; }
If an error occurs after a successful stm32_ospi_dma_setup() call, some dma_release_channel() calls are needed to release some resources, as already done in the remove function. Fixes: 79b8a705e26c ("spi: stm32: Add OSPI driver") Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr> --- Compile tested-only --- drivers/spi/spi-stm32-ospi.c | 4 ++++ 1 file changed, 4 insertions(+)