Message ID | 20210408125648.136519-1-lihuafei1@huawei.com |
---|---|
State | New |
Headers | show |
Series | [-next] i2c: omap: fix PM reference leak in omap_i2c_probe() | expand |
Hi. On 4/8/21 6:26 PM, Li Huafei wrote: > pm_runtime_get_sync will increment pm usage counter even it failed. > Forgetting to putting operation will result in reference leak here. Fix > it by replacing it with pm_runtime_resume_and_get to keep usage counter > balanced. > > Reported-by: Hulk Robot <hulkci@huawei.com> > Signed-off-by: Li Huafei <lihuafei1@huawei.com> > --- > drivers/i2c/busses/i2c-omap.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c > index 12ac4212aded..edbe498d49b8 100644 > --- a/drivers/i2c/busses/i2c-omap.c > +++ b/drivers/i2c/busses/i2c-omap.c > @@ -1404,7 +1404,7 @@ omap_i2c_probe(struct platform_device *pdev) > pm_runtime_set_autosuspend_delay(omap->dev, OMAP_I2C_PM_TIMEOUT); > pm_runtime_use_autosuspend(omap->dev); > > - r = pm_runtime_get_sync(omap->dev); > + r = pm_runtime_resume_and_get(omap->dev); > if (r < 0) > goto err_free_mem; > > @@ -1525,7 +1525,7 @@ static int omap_i2c_remove(struct platform_device *pdev) > int ret; > > i2c_del_adapter(&omap->adapter); > - ret = pm_runtime_get_sync(&pdev->dev); > + ret = pm_runtime_resume_and_get(&pdev->dev); > if (ret < 0) > return ret; > > There is a similar patch at: lore.kernel.org/r/20210407033030.13419-1-dinghao.liu@zju.edu.cn But seems like this patch is better as it fixes module remove path as well. Would appreciate if you could work with author of above patch and come up with a single patch. Could you add stable tag and repost given that pm_runtime_resume_and_get() API is backported to stable kernels such as v5.4? Regards Vignesh
On Thu, Apr 08, 2021 at 08:56:48PM +0800, Li Huafei wrote: > pm_runtime_get_sync will increment pm usage counter even it failed. > Forgetting to putting operation will result in reference leak here. Fix > it by replacing it with pm_runtime_resume_and_get to keep usage counter > balanced. > > Reported-by: Hulk Robot <hulkci@huawei.com> > Signed-off-by: Li Huafei <lihuafei1@huawei.com> Thanks, yet I applied this series now: http://patchwork.ozlabs.org/project/linux-i2c/list/?series=217733&state=*
diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c index 12ac4212aded..edbe498d49b8 100644 --- a/drivers/i2c/busses/i2c-omap.c +++ b/drivers/i2c/busses/i2c-omap.c @@ -1404,7 +1404,7 @@ omap_i2c_probe(struct platform_device *pdev) pm_runtime_set_autosuspend_delay(omap->dev, OMAP_I2C_PM_TIMEOUT); pm_runtime_use_autosuspend(omap->dev); - r = pm_runtime_get_sync(omap->dev); + r = pm_runtime_resume_and_get(omap->dev); if (r < 0) goto err_free_mem; @@ -1525,7 +1525,7 @@ static int omap_i2c_remove(struct platform_device *pdev) int ret; i2c_del_adapter(&omap->adapter); - ret = pm_runtime_get_sync(&pdev->dev); + ret = pm_runtime_resume_and_get(&pdev->dev); if (ret < 0) return ret;
pm_runtime_get_sync will increment pm usage counter even it failed. Forgetting to putting operation will result in reference leak here. Fix it by replacing it with pm_runtime_resume_and_get to keep usage counter balanced. Reported-by: Hulk Robot <hulkci@huawei.com> Signed-off-by: Li Huafei <lihuafei1@huawei.com> --- drivers/i2c/busses/i2c-omap.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)