Instead of a busy waiting loop while loop using udelay
use readl_poll_timeout function to check the condition
is met or timeout occurs in crport_handshake function.
Fixes: d8c80bb3b55b ("phy: exynos5-usbdrd: Calibrate LOS levels for exynos5420/5800")
Signed-off-by: Anand Moon <[email protected]>
---
Changes v3:
--Fix the commit message.
--Drop the variable, used the value directly.
Changes v2:
--used the default timeout values.
--Added missing Fixed tags.
---
drivers/phy/samsung/phy-exynos5-usbdrd.c | 35 +++++++-----------------
1 file changed, 10 insertions(+), 25 deletions(-)
diff --git a/drivers/phy/samsung/phy-exynos5-usbdrd.c b/drivers/phy/samsung/phy-exynos5-usbdrd.c
index e510732afb8b..fa75fa88da33 100644
--- a/drivers/phy/samsung/phy-exynos5-usbdrd.c
+++ b/drivers/phy/samsung/phy-exynos5-usbdrd.c
@@ -16,6 +16,7 @@
#include <linux/of.h>
#include <linux/of_address.h>
#include <linux/of_device.h>
+#include <linux/iopoll.h>
#include <linux/phy/phy.h>
#include <linux/platform_device.h>
#include <linux/mutex.h>
@@ -556,40 +557,24 @@ static int exynos5_usbdrd_phy_power_off(struct phy *phy)
static int crport_handshake(struct exynos5_usbdrd_phy *phy_drd,
u32 val, u32 cmd)
{
- u32 usec = 100;
unsigned int result;
+ int err;
writel(val | cmd, phy_drd->reg_phy + EXYNOS5_DRD_PHYREG0);
- do {
- result = readl(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1);
- if (result & PHYREG1_CR_ACK)
- break;
-
- udelay(1);
- } while (usec-- > 0);
-
- if (!usec) {
- dev_err(phy_drd->dev,
- "CRPORT handshake timeout1 (0x%08x)\n", val);
+ err = readl_poll_timeout(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1,
+ result, (result & PHYREG1_CR_ACK), 1, 100);
+ if (err) {
+ dev_err(phy_drd->dev, "CRPORT handshake timeout1 (0x%08x)\n", val);
return -ETIME;
}
- usec = 100;
-
writel(val, phy_drd->reg_phy + EXYNOS5_DRD_PHYREG0);
- do {
- result = readl(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1);
- if (!(result & PHYREG1_CR_ACK))
- break;
-
- udelay(1);
- } while (usec-- > 0);
-
- if (!usec) {
- dev_err(phy_drd->dev,
- "CRPORT handshake timeout2 (0x%08x)\n", val);
+ err = readl_poll_timeout(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1,
+ result, !(result & PHYREG1_CR_ACK), 1, 100);
+ if (err) {
+ dev_err(phy_drd->dev, "CRPORT handshake timeout2 (0x%08x)\n", val);
return -ETIME;
}
--
2.27.0
On Tue, Jul 07, 2020 at 09:59:08AM +0000, Anand Moon wrote:
> Instead of a busy waiting loop while loop using udelay
You doubled "loop".
> use readl_poll_timeout function to check the condition
> is met or timeout occurs in crport_handshake function.
Still you did not mention that you convert the function to use sleeping
primitive. You also did not mention whether it is actually allowed in
this context and I am not sure if you investigated it.
Best regards,
Krzysztof
>
> Fixes: d8c80bb3b55b ("phy: exynos5-usbdrd: Calibrate LOS levels for exynos5420/5800")
> Signed-off-by: Anand Moon <[email protected]>
> ---
> Changes v3:
> --Fix the commit message.
> --Drop the variable, used the value directly.
> Changes v2:
> --used the default timeout values.
> --Added missing Fixed tags.
> ---
> drivers/phy/samsung/phy-exynos5-usbdrd.c | 35 +++++++-----------------
> 1 file changed, 10 insertions(+), 25 deletions(-)
>
> diff --git a/drivers/phy/samsung/phy-exynos5-usbdrd.c b/drivers/phy/samsung/phy-exynos5-usbdrd.c
> index e510732afb8b..fa75fa88da33 100644
> --- a/drivers/phy/samsung/phy-exynos5-usbdrd.c
> +++ b/drivers/phy/samsung/phy-exynos5-usbdrd.c
> @@ -16,6 +16,7 @@
> #include <linux/of.h>
> #include <linux/of_address.h>
> #include <linux/of_device.h>
> +#include <linux/iopoll.h>
> #include <linux/phy/phy.h>
> #include <linux/platform_device.h>
> #include <linux/mutex.h>
> @@ -556,40 +557,24 @@ static int exynos5_usbdrd_phy_power_off(struct phy *phy)
> static int crport_handshake(struct exynos5_usbdrd_phy *phy_drd,
> u32 val, u32 cmd)
> {
> - u32 usec = 100;
> unsigned int result;
> + int err;
>
> writel(val | cmd, phy_drd->reg_phy + EXYNOS5_DRD_PHYREG0);
>
> - do {
> - result = readl(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1);
> - if (result & PHYREG1_CR_ACK)
> - break;
> -
> - udelay(1);
> - } while (usec-- > 0);
> -
> - if (!usec) {
> - dev_err(phy_drd->dev,
> - "CRPORT handshake timeout1 (0x%08x)\n", val);
> + err = readl_poll_timeout(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1,
> + result, (result & PHYREG1_CR_ACK), 1, 100);
> + if (err) {
> + dev_err(phy_drd->dev, "CRPORT handshake timeout1 (0x%08x)\n", val);
> return -ETIME;
> }
>
> - usec = 100;
> -
> writel(val, phy_drd->reg_phy + EXYNOS5_DRD_PHYREG0);
>
> - do {
> - result = readl(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1);
> - if (!(result & PHYREG1_CR_ACK))
> - break;
> -
> - udelay(1);
> - } while (usec-- > 0);
> -
> - if (!usec) {
> - dev_err(phy_drd->dev,
> - "CRPORT handshake timeout2 (0x%08x)\n", val);
> + err = readl_poll_timeout(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1,
> + result, !(result & PHYREG1_CR_ACK), 1, 100);
> + if (err) {
> + dev_err(phy_drd->dev, "CRPORT handshake timeout2 (0x%08x)\n", val);
> return -ETIME;
> }
>
> --
> 2.27.0
>
Hi Krzysztof,
On Tue, 7 Jul 2020 at 17:06, Krzysztof Kozlowski <[email protected]> wrote:
>
> On Tue, Jul 07, 2020 at 09:59:08AM +0000, Anand Moon wrote:
> > Instead of a busy waiting loop while loop using udelay
>
> You doubled "loop".
>
I will fix this in the next version.
> > use readl_poll_timeout function to check the condition
> > is met or timeout occurs in crport_handshake function.
>
> Still you did not mention that you convert the function to use sleeping
> primitive. You also did not mention whether it is actually allowed in
> this context and I am not sure if you investigated it.
>
OK, I am not sure how to resolve your query.
I learned some new things.
So here are some points.
-- Yes read_poll_timeout internally used might_sleep if sleep_us != 0
under some condition.
-- None of the code in phy-exynos5-usbdrd.c is called using kernel
synchronization
methods like spinlock / mutex.
-- I have checked this function is called non atomic context.
see below.
[ 6.006395] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.013042] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.019646] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.026174] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.032699] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.039246] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.045707] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.052276] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.058760] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.065189] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.071631] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.078033] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.084433] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.090812] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.097102] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.103413] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.109670] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
[ 6.115871] exynos5_usb3drd_phy 12500000.phy: Not In atomic context
> Best regards,
> Krzysztof
>
> >
> > Fixes: d8c80bb3b55b ("phy: exynos5-usbdrd: Calibrate LOS levels for exynos5420/5800")
> > Signed-off-by: Anand Moon <[email protected]>
> > ---
> > Changes v3:
> > --Fix the commit message.
> > --Drop the variable, used the value directly.
> > Changes v2:
> > --used the default timeout values.
> > --Added missing Fixed tags.
> > ---
> > drivers/phy/samsung/phy-exynos5-usbdrd.c | 35 +++++++-----------------
> > 1 file changed, 10 insertions(+), 25 deletions(-)
> >
> > diff --git a/drivers/phy/samsung/phy-exynos5-usbdrd.c b/drivers/phy/samsung/phy-exynos5-usbdrd.c
> > index e510732afb8b..fa75fa88da33 100644
> > --- a/drivers/phy/samsung/phy-exynos5-usbdrd.c
> > +++ b/drivers/phy/samsung/phy-exynos5-usbdrd.c
> > @@ -16,6 +16,7 @@
> > #include <linux/of.h>
> > #include <linux/of_address.h>
> > #include <linux/of_device.h>
> > +#include <linux/iopoll.h>
> > #include <linux/phy/phy.h>
> > #include <linux/platform_device.h>
> > #include <linux/mutex.h>
> > @@ -556,40 +557,24 @@ static int exynos5_usbdrd_phy_power_off(struct phy *phy)
> > static int crport_handshake(struct exynos5_usbdrd_phy *phy_drd,
> > u32 val, u32 cmd)
> > {
> > - u32 usec = 100;
> > unsigned int result;
> > + int err;
> >
> > writel(val | cmd, phy_drd->reg_phy + EXYNOS5_DRD_PHYREG0);
> >
> > - do {
> > - result = readl(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1);
> > - if (result & PHYREG1_CR_ACK)
> > - break;
> > -
> > - udelay(1);
> > - } while (usec-- > 0);
> > -
> > - if (!usec) {
> > - dev_err(phy_drd->dev,
> > - "CRPORT handshake timeout1 (0x%08x)\n", val);
> > + err = readl_poll_timeout(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1,
> > + result, (result & PHYREG1_CR_ACK), 1, 100);
> > + if (err) {
> > + dev_err(phy_drd->dev, "CRPORT handshake timeout1 (0x%08x)\n", val);
> > return -ETIME;
Here return should be -ETIMEDOUT;
> > }
> >
> > - usec = 100;
> > -
> > writel(val, phy_drd->reg_phy + EXYNOS5_DRD_PHYREG0);
> >
> > - do {
> > - result = readl(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1);
> > - if (!(result & PHYREG1_CR_ACK))
> > - break;
> > -
> > - udelay(1);
> > - } while (usec-- > 0);
> > -
> > - if (!usec) {
> > - dev_err(phy_drd->dev,
> > - "CRPORT handshake timeout2 (0x%08x)\n", val);
> > + err = readl_poll_timeout(phy_drd->reg_phy + EXYNOS5_DRD_PHYREG1,
> > + result, !(result & PHYREG1_CR_ACK), 1, 100);
> > + if (err) {
> > + dev_err(phy_drd->dev, "CRPORT handshake timeout2 (0x%08x)\n", val);
> > return -ETIME;
> > }
> >
Best Regards
-Anand
On Wed, Jul 08, 2020 at 01:59:46PM +0530, Anand Moon wrote:
> > Still you did not mention that you convert the function to use sleeping
> > primitive. You also did not mention whether it is actually allowed in
> > this context and I am not sure if you investigated it.
> >
> OK, I am not sure how to resolve your query.
> I learned some new things.
>
> So here are some points.
> -- Yes read_poll_timeout internally used might_sleep if sleep_us != 0
> under some condition.
> -- None of the code in phy-exynos5-usbdrd.c is called using kernel
> synchronization
> methods like spinlock / mutex.
More important is rather the call to calibrare() as this is the place
where affected code is used.
It is not only about synchronisation primitives used in the driver but
also in the phy core. I guess there should not be a problem. I just
stated the fact that you did not mention anything about it.
> -- I have checked this function is called non atomic context.
Great!
Best regards,
Krzysztof