| Message ID | 20240822213645.1125016-7-bvanassche@acm.org (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | Simplify the UFS driver initialization code | expand |
On Thu, Aug 22, 2024 at 02:36:07PM -0700, Bart Van Assche wrote: > Move the ufshcd_device_init(hba, true) call from ufshcd_async_scan() > into ufshcd_init(). This patch prepares for moving both scsi_add_host() > calls into ufshcd_add_scsi_host(). Calling ufshcd_device_init() from > ufshcd_init() without holding hba->host_sem is safe because > hba->host_sem serializes core code with sysfs callback code and because > the ufshcd_device_init() is moved before the scsi_add_host() call. I think this last sentence is not complete. > > Signed-off-by: Bart Van Assche <bvanassche@acm.org> One nitpick below. Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> > --- > drivers/ufs/core/ufshcd.c | 10 ++++++---- > 1 file changed, 6 insertions(+), 4 deletions(-) > > diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c > index 5c8751672bc7..0fdf19889191 100644 > --- a/drivers/ufs/core/ufshcd.c > +++ b/drivers/ufs/core/ufshcd.c > @@ -8933,10 +8933,7 @@ static void ufshcd_async_scan(void *data, async_cookie_t cookie) > int ret; > > down(&hba->host_sem); > - /* Initialize hba, detect and initialize UFS device */ > - ret = ufshcd_device_init(hba, /*init_dev_params=*/true); > - if (ret == 0) > - ret = ufshcd_probe_hba(hba); > + ret = ufshcd_probe_hba(hba); > up(&hba->host_sem); > if (ret) > goto out; > @@ -10632,6 +10629,11 @@ int ufshcd_init(struct ufs_hba *hba, void __iomem *mmio_base, unsigned int irq) > */ > ufshcd_set_ufs_dev_active(hba); > > + /* Initialize hba, detect and initialize UFS device */ > + err = ufshcd_device_init(hba, /*init_dev_params=*/true); I think this inline comment is not really needed. It is rather decreasing code readability. - Mani > + if (err) > + goto out_disable; > + > err = ufshcd_add_scsi_host(hba); > if (err) > goto out_disable;
On 8/25/24 1:33 AM, Manivannan Sadhasivam wrote: > On Thu, Aug 22, 2024 at 02:36:07PM -0700, Bart Van Assche wrote: >> Move the ufshcd_device_init(hba, true) call from ufshcd_async_scan() >> into ufshcd_init(). This patch prepares for moving both scsi_add_host() >> calls into ufshcd_add_scsi_host(). Calling ufshcd_device_init() from >> ufshcd_init() without holding hba->host_sem is safe because >> hba->host_sem serializes core code with sysfs callback code and because >> the ufshcd_device_init() is moved before the scsi_add_host() call. > > I think this last sentence is not complete. It is complete but your feedback means to me that it is hard to understand. I will reword it. >> @@ -10632,6 +10629,11 @@ int ufshcd_init(struct ufs_hba *hba, void __iomem *mmio_base, unsigned int irq) >> */ >> ufshcd_set_ufs_dev_active(hba); >> >> + /* Initialize hba, detect and initialize UFS device */ >> + err = ufshcd_device_init(hba, /*init_dev_params=*/true); > > I think this inline comment is not really needed. It is rather decreasing code > readability. I will remove it. Thanks, Bart.
diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c index 5c8751672bc7..0fdf19889191 100644 --- a/drivers/ufs/core/ufshcd.c +++ b/drivers/ufs/core/ufshcd.c @@ -8933,10 +8933,7 @@ static void ufshcd_async_scan(void *data, async_cookie_t cookie) int ret; down(&hba->host_sem); - /* Initialize hba, detect and initialize UFS device */ - ret = ufshcd_device_init(hba, /*init_dev_params=*/true); - if (ret == 0) - ret = ufshcd_probe_hba(hba); + ret = ufshcd_probe_hba(hba); up(&hba->host_sem); if (ret) goto out; @@ -10632,6 +10629,11 @@ int ufshcd_init(struct ufs_hba *hba, void __iomem *mmio_base, unsigned int irq) */ ufshcd_set_ufs_dev_active(hba); + /* Initialize hba, detect and initialize UFS device */ + err = ufshcd_device_init(hba, /*init_dev_params=*/true); + if (err) + goto out_disable; + err = ufshcd_add_scsi_host(hba); if (err) goto out_disable;
Move the ufshcd_device_init(hba, true) call from ufshcd_async_scan() into ufshcd_init(). This patch prepares for moving both scsi_add_host() calls into ufshcd_add_scsi_host(). Calling ufshcd_device_init() from ufshcd_init() without holding hba->host_sem is safe because hba->host_sem serializes core code with sysfs callback code and because the ufshcd_device_init() is moved before the scsi_add_host() call. Signed-off-by: Bart Van Assche <bvanassche@acm.org> --- drivers/ufs/core/ufshcd.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)