Le 08/03/2022 à 13:54, Dan Carpenter a écrit :
On Sun, Mar 06, 2022 at 06:44:56PM +0100, Christophe JAILLET wrote:
A dma_free_coherent() call is missing in the error handling path of the
probe, as already done in the remove function.
Fixes: 3a96dff0f828 ("mmc: SD/MMC Host Controller for Wondermedia WM8505/WM8650")
Signed-off-by: Christophe JAILLET <christophe.jaillet@xxxxxxxxxx>
---
I've not been able to find a Fixes tag because of the renaming of
function and files.
However, it looks old (before 2008)
You did add a fixes tag and it's from 2012. :P
Well, a bit too much of copy&paste from [1] :(
[1]:
https://lore.kernel.org/all/242ebc5e7dedc6b0d7f47cbf7768326c127f955b.1646584729.git.christophe.jaillet@xxxxxxxxxx/
---
drivers/mmc/host/wmt-sdmmc.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/mmc/host/wmt-sdmmc.c b/drivers/mmc/host/wmt-sdmmc.c
index 163ac9df8cca..8e18f01c0938 100644
--- a/drivers/mmc/host/wmt-sdmmc.c
+++ b/drivers/mmc/host/wmt-sdmmc.c
@@ -863,6 +863,8 @@ static int wmt_mci_probe(struct platform_device *pdev)
return 0;
fail6:
clk_put(priv->clk_sdmmc);
+ dma_free_coherent(&pdev->dev, mmc->max_blk_count * 16,
+ priv->dma_desc_buffer, priv->dma_desc_device_addr);
fail5:
free_irq(dma_irq, priv);
This isn't quite right. If of_clk_get() fails it should call
dma_free_coherent() but it does not. You need to add:
Agreed, thx for the review.
If no one fix it in the meanwhile, I'll send a v2 this WE or next week.
CJ
fail6:
clk_put(priv->clk_sdmmc);
+fail5_and_a_half:
+ dma_free_coherent(&pdev->dev, mmc->max_blk_count * 16,
+ priv->dma_desc_buffer, priv->dma_desc_device_addr);
fail5:
free_irq(dma_irq, priv);
regards,
dan carpenter
