diff mbox series

[v1,1/1] pwrap: mediatek: fix wait time out issue

Message ID 20220329115824.13005-2-zhiyong.tao@mediatek.com (mailing list archive)
State New, archived
Headers show
Series Mediatek PMIC patch | expand

Commit Message

zhiyong.tao March 29, 2022, 11:58 a.m. UTC
From: "Zhiyong.Tao" <zhiyong.tao@mediatek.com>

add sleep delay before read "jiffies"

Signed-off-by: Zhiyong.Tao <zhiyong.tao@mediatek.com>
---
 drivers/soc/mediatek/mtk-pmic-wrap.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
 mode change 100644 => 100755 drivers/soc/mediatek/mtk-pmic-wrap.c

Comments

Matthias Brugger March 29, 2022, 2:44 p.m. UTC | #1
On 29/03/2022 13:58, Zhiyong Tao wrote:
> From: "Zhiyong.Tao" <zhiyong.tao@mediatek.com>
> 
> add sleep delay before read "jiffies"
> 
> Signed-off-by: Zhiyong.Tao <zhiyong.tao@mediatek.com>
> ---
>   drivers/soc/mediatek/mtk-pmic-wrap.c | 8 ++++++--
>   1 file changed, 6 insertions(+), 2 deletions(-)
>   mode change 100644 => 100755 drivers/soc/mediatek/mtk-pmic-wrap.c
> 
> diff --git a/drivers/soc/mediatek/mtk-pmic-wrap.c b/drivers/soc/mediatek/mtk-pmic-wrap.c
> old mode 100644
> new mode 100755
> index 952bc554f443..ac7139a67e87
> --- a/drivers/soc/mediatek/mtk-pmic-wrap.c
> +++ b/drivers/soc/mediatek/mtk-pmic-wrap.c
> @@ -12,6 +12,7 @@
>   #include <linux/platform_device.h>
>   #include <linux/regmap.h>
>   #include <linux/reset.h>
> +#include <linux/delay.h>
>   
>   #define PWRAP_MT8135_BRIDGE_IORD_ARB_EN		0x4
>   #define PWRAP_MT8135_BRIDGE_WACS3_EN		0x10
> @@ -1197,10 +1198,13 @@ static int pwrap_wait_for_state(struct pmic_wrapper *wrp,
>   	timeout = jiffies + usecs_to_jiffies(10000);
>   
>   	do {
> -		if (time_after(jiffies, timeout))
> -			return fp(wrp) ? 0 : -ETIMEDOUT;
>   		if (fp(wrp))
>   			return 0;
> +
> +		usleep_range(10, 11);
> +

You need to explain this change. Why is it needed, are you sure it does not 
break other platforms etc.

I can guess why you need that patch here. In any case in the code you provide, 
it would make sense to move the first if out of the do loop and only after this 
wait 10 us before reading again. Right now the code does usleep only after every 
other read.

Regards,
Matthias

> +		if (time_after(jiffies, timeout))
> +			return fp(wrp) ? 0 : -ETIMEDOUT;
>   	} while (1);
>   }
>
zhiyong.tao April 18, 2022, 1:29 a.m. UTC | #2
On Tue, 2022-03-29 at 16:44 +0200, Matthias Brugger wrote:
> 
> On 29/03/2022 13:58, Zhiyong Tao wrote:
> > From: "Zhiyong.Tao" <zhiyong.tao@mediatek.com>
> > 
> > add sleep delay before read "jiffies"
> > 
> > Signed-off-by: Zhiyong.Tao <zhiyong.tao@mediatek.com>
> > ---
> >   drivers/soc/mediatek/mtk-pmic-wrap.c | 8 ++++++--
> >   1 file changed, 6 insertions(+), 2 deletions(-)
> >   mode change 100644 => 100755 drivers/soc/mediatek/mtk-pmic-wrap.c
> > 
> > diff --git a/drivers/soc/mediatek/mtk-pmic-wrap.c
> > b/drivers/soc/mediatek/mtk-pmic-wrap.c
> > old mode 100644
> > new mode 100755
> > index 952bc554f443..ac7139a67e87
> > --- a/drivers/soc/mediatek/mtk-pmic-wrap.c
> > +++ b/drivers/soc/mediatek/mtk-pmic-wrap.c
> > @@ -12,6 +12,7 @@
> >   #include <linux/platform_device.h>
> >   #include <linux/regmap.h>
> >   #include <linux/reset.h>
> > +#include <linux/delay.h>
> >   
> >   #define PWRAP_MT8135_BRIDGE_IORD_ARB_EN		0x4
> >   #define PWRAP_MT8135_BRIDGE_WACS3_EN		0x10
> > @@ -1197,10 +1198,13 @@ static int pwrap_wait_for_state(struct
> > pmic_wrapper *wrp,
> >   	timeout = jiffies + usecs_to_jiffies(10000);
> >   
> >   	do {
> > -		if (time_after(jiffies, timeout))
> > -			return fp(wrp) ? 0 : -ETIMEDOUT;
> >   		if (fp(wrp))
> >   			return 0;
> > +
> > +		usleep_range(10, 11);
> > +
> 
> You need to explain this change. Why is it needed, are you sure it
> does not 
> break other platforms etc.
> 
> I can guess why you need that patch here. In any case in the code you
> provide, 
> it would make sense to move the first if out of the do loop and only
> after this 
> wait 10 us before reading again. Right now the code does usleep only
> after every 
> other read.
> 
> Regards,
> Matthias
> 
Hi Mattias,

we confirm the delay time with our pwrap designer.
we should add the 10us delay before reading again.
The patch is accord with all platforms pwrap design.
we will explain this in next version.

Thanks.
> > +		if (time_after(jiffies, timeout))
> > +			return fp(wrp) ? 0 : -ETIMEDOUT;
> >   	} while (1);
> >   }
> >
diff mbox series

Patch

diff --git a/drivers/soc/mediatek/mtk-pmic-wrap.c b/drivers/soc/mediatek/mtk-pmic-wrap.c
old mode 100644
new mode 100755
index 952bc554f443..ac7139a67e87
--- a/drivers/soc/mediatek/mtk-pmic-wrap.c
+++ b/drivers/soc/mediatek/mtk-pmic-wrap.c
@@ -12,6 +12,7 @@ 
 #include <linux/platform_device.h>
 #include <linux/regmap.h>
 #include <linux/reset.h>
+#include <linux/delay.h>
 
 #define PWRAP_MT8135_BRIDGE_IORD_ARB_EN		0x4
 #define PWRAP_MT8135_BRIDGE_WACS3_EN		0x10
@@ -1197,10 +1198,13 @@  static int pwrap_wait_for_state(struct pmic_wrapper *wrp,
 	timeout = jiffies + usecs_to_jiffies(10000);
 
 	do {
-		if (time_after(jiffies, timeout))
-			return fp(wrp) ? 0 : -ETIMEDOUT;
 		if (fp(wrp))
 			return 0;
+
+		usleep_range(10, 11);
+
+		if (time_after(jiffies, timeout))
+			return fp(wrp) ? 0 : -ETIMEDOUT;
 	} while (1);
 }