2020-09-03 16:00:16

by Jiada Wang

[permalink] [raw]
Subject: [PATCH v2 1/1] Input: atmel_mxt_ts - implement I2C retries

From: Nick Dyer <[email protected]>

Some maXTouch chips (eg mXT1386) will not respond on the first I2C request
when they are in a sleep state. It must be retried after a delay for the
chip to wake up.

Signed-off-by: Nick Dyer <[email protected]>
Acked-by: Yufeng Shen <[email protected]>
(cherry picked from ndyer/linux/for-upstream commit 63fd7a2cd03c3a572a5db39c52f4856819e1835d)
[gdavis: Forward port and fix conflicts.]
Signed-off-by: George G. Davis <[email protected]>
[jiada: return exact errno when i2c_transfer & i2c_master_send fails]
Signed-off-by: Jiada Wang <[email protected]>
---
drivers/input/touchscreen/atmel_mxt_ts.c | 45 ++++++++++++++++--------
1 file changed, 30 insertions(+), 15 deletions(-)

diff --git a/drivers/input/touchscreen/atmel_mxt_ts.c b/drivers/input/touchscreen/atmel_mxt_ts.c
index a2189739e30f..5d4cb15d21dc 100644
--- a/drivers/input/touchscreen/atmel_mxt_ts.c
+++ b/drivers/input/touchscreen/atmel_mxt_ts.c
@@ -196,6 +196,7 @@ enum t100_type {
#define MXT_CRC_TIMEOUT 1000 /* msec */
#define MXT_FW_RESET_TIME 3000 /* msec */
#define MXT_FW_CHG_TIMEOUT 300 /* msec */
+#define MXT_WAKEUP_TIME 25 /* msec */

/* Command to unlock bootloader */
#define MXT_UNLOCK_CMD_MSB 0xaa
@@ -626,6 +627,7 @@ static int __mxt_read_reg(struct i2c_client *client,
struct i2c_msg xfer[2];
u8 buf[2];
int ret;
+ bool retry = false;

buf[0] = reg & 0xff;
buf[1] = (reg >> 8) & 0xff;
@@ -642,17 +644,22 @@ static int __mxt_read_reg(struct i2c_client *client,
xfer[1].len = len;
xfer[1].buf = val;

- ret = i2c_transfer(client->adapter, xfer, 2);
- if (ret == 2) {
- ret = 0;
- } else {
- if (ret >= 0)
- ret = -EIO;
- dev_err(&client->dev, "%s: i2c transfer failed (%d)\n",
- __func__, ret);
+retry_read:
+ ret = i2c_transfer(client->adapter, xfer, ARRAY_SIZE(xfer));
+ if (ret != ARRAY_SIZE(xfer)) {
+ if (!retry) {
+ dev_dbg(&client->dev, "%s: i2c retry\n", __func__);
+ msleep(MXT_WAKEUP_TIME);
+ retry = true;
+ goto retry_read;
+ } else {
+ dev_err(&client->dev, "%s: i2c transfer failed (%d)\n",
+ __func__, ret);
+ return ret < 0 ? ret : -EIO;
+ }
}

- return ret;
+ return 0;
}

static int __mxt_write_reg(struct i2c_client *client, u16 reg, u16 len,
@@ -661,6 +668,7 @@ static int __mxt_write_reg(struct i2c_client *client, u16 reg, u16 len,
u8 *buf;
size_t count;
int ret;
+ bool retry = false;

count = len + 2;
buf = kmalloc(count, GFP_KERNEL);
@@ -671,14 +679,21 @@ static int __mxt_write_reg(struct i2c_client *client, u16 reg, u16 len,
buf[1] = (reg >> 8) & 0xff;
memcpy(&buf[2], val, len);

+retry_write:
ret = i2c_master_send(client, buf, count);
- if (ret == count) {
- ret = 0;
+ if (ret != count) {
+ if (!retry) {
+ dev_dbg(&client->dev, "%s: i2c retry\n", __func__);
+ msleep(MXT_WAKEUP_TIME);
+ retry = true;
+ goto retry_write;
+ } else {
+ dev_err(&client->dev, "%s: i2c send failed (%d)\n",
+ __func__, ret);
+ ret = ret < 0 ? ret : -EIO;
+ }
} else {
- if (ret >= 0)
- ret = -EIO;
- dev_err(&client->dev, "%s: i2c send failed (%d)\n",
- __func__, ret);
+ ret = 0;
}

kfree(buf);
--
2.17.1


2020-09-05 14:09:09

by Dmitry Osipenko

[permalink] [raw]
Subject: Re: [PATCH v2 1/1] Input: atmel_mxt_ts - implement I2C retries

03.09.2020 18:59, Jiada Wang пишет:
> From: Nick Dyer <[email protected]>
>
> Some maXTouch chips (eg mXT1386) will not respond on the first I2C request
> when they are in a sleep state. It must be retried after a delay for the
> chip to wake up.
>
> Signed-off-by: Nick Dyer <[email protected]>
> Acked-by: Yufeng Shen <[email protected]>
> (cherry picked from ndyer/linux/for-upstream commit 63fd7a2cd03c3a572a5db39c52f4856819e1835d)
> [gdavis: Forward port and fix conflicts.]
> Signed-off-by: George G. Davis <[email protected]>
> [jiada: return exact errno when i2c_transfer & i2c_master_send fails]
> Signed-off-by: Jiada Wang <[email protected]>
> ---
> drivers/input/touchscreen/atmel_mxt_ts.c | 45 ++++++++++++++++--------
> 1 file changed, 30 insertions(+), 15 deletions(-)

Hello, Jiada!

Everything works well on Acer A500 tablet device that uses mXT1386 for
the touchscreen controller! Thank you very much!

Reviewed-by: Dmitry Osipenko <[email protected]>
Tested-by: Dmitry Osipenko <[email protected]>

2020-09-05 18:59:22

by Dmitry Osipenko

[permalink] [raw]
Subject: Re: [PATCH v2 1/1] Input: atmel_mxt_ts - implement I2C retries

05.09.2020 21:02, Andy Shevchenko пишет:
...
>  #define MXT_CRC_TIMEOUT                1000    /* msec */
>  #define MXT_FW_RESET_TIME      3000    /* msec */
>  #define MXT_FW_CHG_TIMEOUT     300     /* msec */
> +#define MXT_WAKEUP_TIME                25      /* msec */
>
>
> Can we simple add _MS unit suffix to the definition?

I'd expect this

>  /* Command to unlock bootloader */
>  #define MXT_UNLOCK_CMD_MSB     0xaa
> @@ -626,6 +627,7 @@ static int __mxt_read_reg(struct i2c_client *client,
>         struct i2c_msg xfer[2];
>         u8 buf[2];
>         int ret;
> +       bool retry = false;
>
>
> Keep this ordered by length.

and this to be separate patches that are cleaning whole driver code,
otherwise there are no much benefits.

2020-09-12 00:51:37

by Jiada Wang

[permalink] [raw]
Subject: Re: [PATCH v2 1/1] Input: atmel_mxt_ts - implement I2C retries

Hi Andy

Thanks for your comment

On 2020/09/06 3:02, Andy Shevchenko wrote:
>
>
> On Thursday, September 3, 2020, Jiada Wang <[email protected]
> <mailto:[email protected]>> wrote:
>
> From: Nick Dyer <[email protected] <mailto:[email protected]>>
>
> Some maXTouch chips (eg mXT1386) will not respond on the first I2C
> request
> when they are in a sleep state. It must be retried after a delay for the
> chip to wake up.
>
> Signed-off-by: Nick Dyer <[email protected]
> <mailto:[email protected]>>
> Acked-by: Yufeng Shen <[email protected]
> <mailto:[email protected]>>
>
>
> (cherry picked from ndyer/linux/for-upstream commit
> 63fd7a2cd03c3a572a5db39c52f4856819e1835d)
>
>
> It’s a noise for upstream.
>
sure, I will remove it

> [gdavis: Forward port and fix conflicts.]
> Signed-off-by: George G. Davis <[email protected]
> <mailto:[email protected]>>
> [jiada: return exact errno when i2c_transfer & i2c_master_send fails]
> Signed-off-by: Jiada Wang <[email protected]
> <mailto:[email protected]>>
> ---
>  drivers/input/touchscreen/atmel_mxt_ts.c | 45 ++++++++++++++++--------
>  1 file changed, 30 insertions(+), 15 deletions(-)
>
> diff --git a/drivers/input/touchscreen/atmel_mxt_ts.c
> b/drivers/input/touchscreen/atmel_mxt_ts.c
> index a2189739e30f..5d4cb15d21dc 100644
> --- a/drivers/input/touchscreen/atmel_mxt_ts.c
> +++ b/drivers/input/touchscreen/atmel_mxt_ts.c
> @@ -196,6 +196,7 @@ enum t100_type {
>  #define MXT_CRC_TIMEOUT                1000    /* msec */
>  #define MXT_FW_RESET_TIME      3000    /* msec */
>  #define MXT_FW_CHG_TIMEOUT     300     /* msec */
> +#define MXT_WAKEUP_TIME                25      /* msec */
>
>
> Can we simple add _MS unit suffix to the definition?
As Dmitry commented, I'd like to keep it as-is, probably a separate
patch to update all these together.
>
>
>  /* Command to unlock bootloader */
>  #define MXT_UNLOCK_CMD_MSB     0xaa
> @@ -626,6 +627,7 @@ static int __mxt_read_reg(struct i2c_client *client,
>         struct i2c_msg xfer[2];
>         u8 buf[2];
>         int ret;
> +       bool retry = false;
>
>
> Keep this ordered by length.
I will move "retry" upper.
>
>
>         buf[0] = reg & 0xff;
>         buf[1] = (reg >> 8) & 0xff;
> @@ -642,17 +644,22 @@ static int __mxt_read_reg(struct i2c_client
> *client,
>         xfer[1].len = len;
>         xfer[1].buf = val;
>
> -       ret = i2c_transfer(client->adapter, xfer, 2);
> -       if (ret == 2) {
> -               ret = 0;
> -       } else {
> -               if (ret >= 0)
> -                       ret = -EIO;
> -               dev_err(&client->dev, "%s: i2c transfer failed (%d)\n",
> -                       __func__, ret);
> +retry_read:
> +       ret = i2c_transfer(client->adapter, xfer, ARRAY_SIZE(xfer));
> +       if (ret != ARRAY_SIZE(xfer)) {
> +               if (!retry) {
>
>
> Why not positive conditional?
to me, it's not much different either positive or negative conditional,
can you elaborate more about this?


>
> +                       dev_dbg(&client->dev, "%s: i2c retry\n",
> __func__);
>
>
> __func__ is redundant for dev_dbg().
>
> +                       msleep(MXT_WAKEUP_TIME);
> +                       retry = true;
> +                       goto retry_read;
>
> +               } else {
>
>
> Redundant in either case of conditional. Allows to drop indentation level.
I will remove the redundant conditional

Thanks,
Jiada
>
> +                       dev_err(&client->dev, "%s: i2c transfer
> failed (%d)\n",
> +                               __func__, ret);
> +                       return ret < 0 ? ret : -EIO;
> +               }
>         }
>
> -       return ret;
> +       return 0;
>  }
>
>
> Same comments about below.
>
>  static int __mxt_write_reg(struct i2c_client *client, u16 reg, u16
> len,
> @@ -661,6 +668,7 @@ static int __mxt_write_reg(struct i2c_client
> *client, u16 reg, u16 len,
>         u8 *buf;
>         size_t count;
>         int ret;
> +       bool retry = false;
>
>         count = len + 2;
>         buf = kmalloc(count, GFP_KERNEL);
> @@ -671,14 +679,21 @@ static int __mxt_write_reg(struct i2c_client
> *client, u16 reg, u16 len,
>         buf[1] = (reg >> 8) & 0xff;
>         memcpy(&buf[2], val, len);
>
> +retry_write:
>         ret = i2c_master_send(client, buf, count);
> -       if (ret == count) {
> -               ret = 0;
> +       if (ret != count) {
> +               if (!retry) {
> +                       dev_dbg(&client->dev, "%s: i2c retry\n",
> __func__);
> +                       msleep(MXT_WAKEUP_TIME);
> +                       retry = true;
> +                       goto retry_write;
> +               } else {
> +                       dev_err(&client->dev, "%s: i2c send failed
> (%d)\n",
> +                               __func__, ret);
> +                       ret = ret < 0 ? ret : -EIO;
> +               }
>         } else {
> -               if (ret >= 0)
> -                       ret = -EIO;
> -               dev_err(&client->dev, "%s: i2c send failed (%d)\n",
> -                       __func__, ret);
> +               ret = 0;
>         }
>
>         kfree(buf);
> --
> 2.17.1
>
>
>
> --
> With Best Regards,
> Andy Shevchenko
>
>