ufs: core: Initialize a value of an attribute as returned by uic cmd

[PATCH v2] ufs: core: Initialize a value of an attribute as returned by uic cmd

Posted by Wonkon Kim 2 months ago

From: wkon-kim <wkon.kim@samsung.com>

If ufshcd_send_cmd() fails, *mib_val has a garbage value.
A value of an attribute can have an unintended result.
ufshcd_dme_get_attr() always initializes *mib_val.

Signed-off-by: Wonkon Kim <wkon.kim@samsung.com>
---
v2:
It is better to check ufshcd_dme_get() return rather than
to initialize argument.
And ufshcd_dme_get_attr() always initializes *mib_val.

 drivers/ufs/core/ufshcd.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
index 9a43102b2b21..6858f005cc8b 100644
--- a/drivers/ufs/core/ufshcd.c
+++ b/drivers/ufs/core/ufshcd.c
@@ -4273,8 +4273,8 @@ int ufshcd_dme_get_attr(struct ufs_hba *hba, u32 attr_sel,
 			get, UIC_GET_ATTR_ID(attr_sel),
 			UFS_UIC_COMMAND_RETRIES - retries);
 
-	if (mib_val && !ret)
-		*mib_val = uic_cmd.argument3;
+	if (mib_val)
+		*mib_val = ret == 0 ? uic_cmd.argument3 : 0;
 
 	if (peer && (hba->quirks & UFSHCD_QUIRK_DME_PEER_ACCESS_AUTO_MODE)
 	    && pwr_mode_change)
@@ -4990,7 +4990,7 @@ EXPORT_SYMBOL_GPL(ufshcd_hba_enable);
 
 static int ufshcd_disable_tx_lcc(struct ufs_hba *hba, bool peer)
 {
-	int tx_lanes = 0, i, err = 0;
+	int tx_lanes, i, err = 0;
 
 	if (!peer)
 		ufshcd_dme_get(hba, UIC_ARG_MIB(PA_CONNECTEDTXDATALANES),
-- 
2.34.1

Re: [PATCH v2] ufs: core: Initialize a value of an attribute as returned by uic cmd

Posted by Bart Van Assche 2 months ago

On 10/13/25 9:40 PM, Wonkon Kim wrote:
> diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
> index 9a43102b2b21..6858f005cc8b 100644
> --- a/drivers/ufs/core/ufshcd.c
> +++ b/drivers/ufs/core/ufshcd.c
> @@ -4273,8 +4273,8 @@ int ufshcd_dme_get_attr(struct ufs_hba *hba, u32 attr_sel,
>   			get, UIC_GET_ATTR_ID(attr_sel),
>   			UFS_UIC_COMMAND_RETRIES - retries);
>   
> -	if (mib_val && !ret)
> -		*mib_val = uic_cmd.argument3;
> +	if (mib_val)
> +		*mib_val = ret == 0 ? uic_cmd.argument3 : 0;
>   
>   	if (peer && (hba->quirks & UFSHCD_QUIRK_DME_PEER_ACCESS_AUTO_MODE)
>   	    && pwr_mode_change)
> @@ -4990,7 +4990,7 @@ EXPORT_SYMBOL_GPL(ufshcd_hba_enable);
>   
>   static int ufshcd_disable_tx_lcc(struct ufs_hba *hba, bool peer)
>   {
> -	int tx_lanes = 0, i, err = 0;
> +	int tx_lanes, i, err = 0;
>   
>   	if (!peer)
>   		ufshcd_dme_get(hba, UIC_ARG_MIB(PA_CONNECTEDTXDATALANES),

Please split this patch into two patches: a first patch with the
ufshcd_dme_get_attr() change and a second patch with the
ufshcd_disable_tx_lcc() change. Please also add a Fixes: tag to the
first patch. That will cause the first patch to be included
automatically in stable kernels and also in Android kernels.

Thanks,

Bart.

RE: [PATCH v2] ufs: core: Initialize a value of an attribute as returned by uic cmd

Posted by Wonkon Kim 2 months ago

> On 10/13/25 9:40 PM, Wonkon Kim wrote:
> > diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
> > index 9a43102b2b21..6858f005cc8b 100644
> > --- a/drivers/ufs/core/ufshcd.c
> > +++ b/drivers/ufs/core/ufshcd.c
> > @@ -4273,8 +4273,8 @@ int ufshcd_dme_get_attr(struct ufs_hba *hba, u32
> attr_sel,
> >   			get, UIC_GET_ATTR_ID(attr_sel),
> >   			UFS_UIC_COMMAND_RETRIES - retries);
> >
> > -	if (mib_val && !ret)
> > -		*mib_val = uic_cmd.argument3;
> > +	if (mib_val)
> > +		*mib_val = ret == 0 ? uic_cmd.argument3 : 0;
> >
> >   	if (peer && (hba->quirks & UFSHCD_QUIRK_DME_PEER_ACCESS_AUTO_MODE)
> >   	    && pwr_mode_change)
> > @@ -4990,7 +4990,7 @@ EXPORT_SYMBOL_GPL(ufshcd_hba_enable);
> >
> >   static int ufshcd_disable_tx_lcc(struct ufs_hba *hba, bool peer)
> >   {
> > -	int tx_lanes = 0, i, err = 0;
> > +	int tx_lanes, i, err = 0;
> >
> >   	if (!peer)
> >   		ufshcd_dme_get(hba, UIC_ARG_MIB(PA_CONNECTEDTXDATALANES),
> 
> Please split this patch into two patches: a first patch with the
> ufshcd_dme_get_attr() change and a second patch with the
> ufshcd_disable_tx_lcc() change. Please also add a Fixes: tag to the first
> patch. That will cause the first patch to be included automatically in
> stable kernels and also in Android kernels.
> 
> Thanks,
> 
> Bart.


Sure. I'll split and update it.

Thanks,
Wonkon Kim.

Re: [PATCH v2] ufs: core: Initialize a value of an attribute as returned by uic cmd

Posted by Peter Wang (王信友) 2 months ago

On Tue, 2025-10-14 at 13:40 +0900, Wonkon Kim wrote:
> 
> From: wkon-kim <wkon.kim@samsung.com>
> 
> If ufshcd_send_cmd() fails, *mib_val has a garbage value.
> A value of an attribute can have an unintended result.
> ufshcd_dme_get_attr() always initializes *mib_val.
> 
> Signed-off-by: Wonkon Kim <wkon.kim@samsung.com>

Reviewed-by: Peter Wang <peter.wang@mediatek.com>