[PATCH] i2c-omap: Set latency requirements only once for several messages

[PATCH] i2c-omap: Set latency requirements only once for several messages

[Samu Onkalo]

Ordinary I2C read consist of two messages. First a write operation
to tell register address and then read operation to get data.
CPU wake up latency is set and removed twice in read case.
Set latency requirement before the message processing loop
and remove the requirement after the loop to remove latency
adjustment operations between the messages.

Signed-off-by: Samu Onkalo <samu.p.onkalo@nokia.com>
---
 drivers/i2c/busses/i2c-omap.c |   10 ++++++----
 1 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index b33c785..3e9323e 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -616,12 +616,8 @@ static int omap_i2c_xfer_msg(struct i2c_adapter *adap,
 	 * REVISIT: We should abort the transfer on signals, but the bus goes
 	 * into arbitration and we're currently unable to recover from it.
 	 */
-	if (dev->set_mpu_wkup_lat != NULL)
-		dev->set_mpu_wkup_lat(dev->dev, dev->latency);
 	r = wait_for_completion_timeout(&dev->cmd_complete,
 					OMAP_I2C_TIMEOUT);
-	if (dev->set_mpu_wkup_lat != NULL)
-		dev->set_mpu_wkup_lat(dev->dev, -1);
 	dev->buf_len = 0;
 	if (r < 0)
 		return r;
@@ -672,12 +668,18 @@ omap_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num)
 	if (r < 0)
 		goto out;
 
+	if (dev->set_mpu_wkup_lat != NULL)
+		dev->set_mpu_wkup_lat(dev->dev, dev->latency);
+
 	for (i = 0; i < num; i++) {
 		r = omap_i2c_xfer_msg(adap, &msgs[i], (i == (num - 1)));
 		if (r != 0)
 			break;
 	}
 
+	if (dev->set_mpu_wkup_lat != NULL)
+		dev->set_mpu_wkup_lat(dev->dev, -1);
+
 	if (r == 0)
 		r = num;
 
-- 
1.6.0.4

Re: [PATCH] i2c-omap: Set latency requirements only once for several messages

[Paul Walmsley]

On Tue, 16 Nov 2010, Samu Onkalo wrote:

> Ordinary I2C read consist of two messages. First a write operation
> to tell register address and then read operation to get data.
> CPU wake up latency is set and removed twice in read case.
> Set latency requirement before the message processing loop
> and remove the requirement after the loop to remove latency
> adjustment operations between the messages.
> 
> Signed-off-by: Samu Onkalo <samu.p.onkalo@nokia.com>

Acked-by: Paul Walmsley <paul@pwsan.com>


- Paul

Re: [PATCH] i2c-omap: Set latency requirements only once for several messages

[Kevin Hilman]

Samu Onkalo <samu.p.onkalo@nokia.com> writes:

> Ordinary I2C read consist of two messages. First a write operation
> to tell register address and then read operation to get data.
> CPU wake up latency is set and removed twice in read case.
> Set latency requirement before the message processing loop
> and remove the requirement after the loop to remove latency
> adjustment operations between the messages.
>
> Signed-off-by: Samu Onkalo <samu.p.onkalo@nokia.com>

Acked-by: Kevin Hilman <khilman@deeprootsystems.com>

Samu,

Can you collect the acks and repost to linux-i2c@vger.kernel.org 
and Cc Ben Dooks <ben-linux@fluff.org>.

This one should go upstream via Ben.

Thanks,

Kevin

> ---
>  drivers/i2c/busses/i2c-omap.c |   10 ++++++----
>  1 files changed, 6 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
> index b33c785..3e9323e 100644
> --- a/drivers/i2c/busses/i2c-omap.c
> +++ b/drivers/i2c/busses/i2c-omap.c
> @@ -616,12 +616,8 @@ static int omap_i2c_xfer_msg(struct i2c_adapter *adap,
>  	 * REVISIT: We should abort the transfer on signals, but the bus goes
>  	 * into arbitration and we're currently unable to recover from it.
>  	 */
> -	if (dev->set_mpu_wkup_lat != NULL)
> -		dev->set_mpu_wkup_lat(dev->dev, dev->latency);
>  	r = wait_for_completion_timeout(&dev->cmd_complete,
>  					OMAP_I2C_TIMEOUT);
> -	if (dev->set_mpu_wkup_lat != NULL)
> -		dev->set_mpu_wkup_lat(dev->dev, -1);
>  	dev->buf_len = 0;
>  	if (r < 0)
>  		return r;
> @@ -672,12 +668,18 @@ omap_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num)
>  	if (r < 0)
>  		goto out;
>  
> +	if (dev->set_mpu_wkup_lat != NULL)
> +		dev->set_mpu_wkup_lat(dev->dev, dev->latency);
> +
>  	for (i = 0; i < num; i++) {
>  		r = omap_i2c_xfer_msg(adap, &msgs[i], (i == (num - 1)));
>  		if (r != 0)
>  			break;
>  	}
>  
> +	if (dev->set_mpu_wkup_lat != NULL)
> +		dev->set_mpu_wkup_lat(dev->dev, -1);
> +
>  	if (r == 0)
>  		r = num;

[PATCH] i2c-omap: Set latency requirements only once for several messages

[Samu Onkalo]

Ordinary I2C read consist of two messages. First a write operation
to tell register address and then read operation to get data.
CPU wake up latency is set and removed twice in read case.
Set latency requirement before the message processing loop
and remove the requirement after the loop to remove latency
adjustment operations between the messages.

Signed-off-by: Samu Onkalo <samu.p.onkalo-xNZwKgViW5gAvxtiuMwx3w@public.gmane.org>
Acked-by: Kevin Hilman <khilman-1D3HCaltpLuhEniVeURVKkEOCMrvLtNR@public.gmane.org>
Acked-by: Paul Walmsley <paul-DWxLp4Yu+b8AvxtiuMwx3w@public.gmane.org>

---
 drivers/i2c/busses/i2c-omap.c |   10 ++++++----
 1 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index b33c785..3e9323e 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -616,12 +616,8 @@ static int omap_i2c_xfer_msg(struct i2c_adapter *adap,
 	 * REVISIT: We should abort the transfer on signals, but the bus goes
 	 * into arbitration and we're currently unable to recover from it.
 	 */
-	if (dev->set_mpu_wkup_lat != NULL)
-		dev->set_mpu_wkup_lat(dev->dev, dev->latency);
 	r = wait_for_completion_timeout(&dev->cmd_complete,
 					OMAP_I2C_TIMEOUT);
-	if (dev->set_mpu_wkup_lat != NULL)
-		dev->set_mpu_wkup_lat(dev->dev, -1);
 	dev->buf_len = 0;
 	if (r < 0)
 		return r;
@@ -672,12 +668,18 @@ omap_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num)
 	if (r < 0)
 		goto out;
 
+	if (dev->set_mpu_wkup_lat != NULL)
+		dev->set_mpu_wkup_lat(dev->dev, dev->latency);
+
 	for (i = 0; i < num; i++) {
 		r = omap_i2c_xfer_msg(adap, &msgs[i], (i == (num - 1)));
 		if (r != 0)
 			break;
 	}
 
+	if (dev->set_mpu_wkup_lat != NULL)
+		dev->set_mpu_wkup_lat(dev->dev, -1);
+
 	if (r == 0)
 		r = num;
 
-- 
1.6.0.4

Re: [PATCH] i2c-omap: Set latency requirements only once for several messages

[Ben Dooks]

On Thu, Nov 18, 2010 at 12:04:20PM +0200, Samu Onkalo wrote:
> Ordinary I2C read consist of two messages. First a write operation
> to tell register address and then read operation to get data.
> CPU wake up latency is set and removed twice in read case.
> Set latency requirement before the message processing loop
> and remove the requirement after the loop to remove latency
> adjustment operations between the messages.
> 
> Signed-off-by: Samu Onkalo <samu.p.onkalo-xNZwKgViW5gAvxtiuMwx3w@public.gmane.org>
> Acked-by: Kevin Hilman <khilman-1D3HCaltpLuhEniVeURVKkEOCMrvLtNR@public.gmane.org>
> Acked-by: Paul Walmsley <paul-DWxLp4Yu+b8AvxtiuMwx3w@public.gmane.org>

I'll queue this for -rc unless it is going in via the omap list
or it should wait for merge window?

-- 
Ben

Q:      What's a light-year?
A:      One-third less calories than a regular year.

RE: [PATCH] i2c-omap: Set latency requirements only once for several messages

[samu.p.onkalo-xNZwKgViW5gAvxtiuMwx3w]

>-----Original Message-----
>From: ext Ben Dooks [mailto:ben-i2c-elnMNo+KYs3YtjvyW6yDsg@public.gmane.org]
>Sent: 03 December, 2010 04:38
>To: Onkalo Samu.P (Nokia-MS/Tampere)
>Cc: ben-linux-elnMNo+KYs3YtjvyW6yDsg@public.gmane.org; linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org;
>khilman-1D3HCaltpLuhEniVeURVKkEOCMrvLtNR@public.gmane.org; paul-DWxLp4Yu+b8AvxtiuMwx3w@public.gmane.org; linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>Subject: Re: [PATCH] i2c-omap: Set latency requirements only once for
>several messages
>
>On Thu, Nov 18, 2010 at 12:04:20PM +0200, Samu Onkalo wrote:
>> Ordinary I2C read consist of two messages. First a write operation
>> to tell register address and then read operation to get data.
>> CPU wake up latency is set and removed twice in read case.
>> Set latency requirement before the message processing loop
>> and remove the requirement after the loop to remove latency
>> adjustment operations between the messages.
>>
>> Signed-off-by: Samu Onkalo <samu.p.onkalo-xNZwKgViW5gAvxtiuMwx3w@public.gmane.org>
>> Acked-by: Kevin Hilman <khilman-1D3HCaltpLuhEniVeURVKkEOCMrvLtNR@public.gmane.org>
>> Acked-by: Paul Walmsley <paul-DWxLp4Yu+b8AvxtiuMwx3w@public.gmane.org>
>
>I'll queue this for -rc unless it is going in via the omap list
>or it should wait for merge window?
>
Hi Ben,

I first sent this to omap-list and I got instructions to send this to you.
So I assume that you should queue this. Kevin, Paul, is -rc ok for you?

-Samu

RE: [PATCH] i2c-omap: Set latency requirements only once for several messages

[Paul Walmsley]

Hello Ben, Samu,

On Fri, 3 Dec 2010, samu.p.onkalo@nokia.com wrote:

> >-----Original Message-----
> >From: ext Ben Dooks [mailto:ben-i2c@fluff.org]
> >Sent: 03 December, 2010 04:38
> >To: Onkalo Samu.P (Nokia-MS/Tampere)
> >Cc: ben-linux@fluff.org; linux-i2c@vger.kernel.org;
> >khilman@deeprootsystems.com; paul@pwsan.com; linux-omap@vger.kernel.org
> >Subject: Re: [PATCH] i2c-omap: Set latency requirements only once for
> >several messages
> >
> >On Thu, Nov 18, 2010 at 12:04:20PM +0200, Samu Onkalo wrote:
> >> Ordinary I2C read consist of two messages. First a write operation
> >> to tell register address and then read operation to get data.
> >> CPU wake up latency is set and removed twice in read case.
> >> Set latency requirement before the message processing loop
> >> and remove the requirement after the loop to remove latency
> >> adjustment operations between the messages.
> >>
> >> Signed-off-by: Samu Onkalo <samu.p.onkalo@nokia.com>
> >> Acked-by: Kevin Hilman <khilman@deeprootsystems.com>
> >> Acked-by: Paul Walmsley <paul@pwsan.com>
> >
> >I'll queue this for -rc unless it is going in via the omap list
> >or it should wait for merge window?
> >
> 
> I first sent this to omap-list and I got instructions to send this to you.
> So I assume that you should queue this. Kevin, Paul, is -rc ok for you?

I'd suggest that this go in through Ben, and go in as part of the 2.6.38 
merge window, rather than the .37-rc.  The patch doesn't fix any 
regressions introduced in the 2.6.37 merge window, and there's no impact 
on system stability, as far as I know.


- Paul