[PATCH] spi: bitbang: reinitialize transfer parameters for every message

[PATCH] spi: bitbang: reinitialize transfer parameters for every message

[Brian Niebuhr]

This patch fixes the setup_transfer logic to account for the case where
multiple messages to different SPI devices are in the queue simultaneously.
With the current logic, the second message in the queue will end up
using the transfer parameters for the previous message in the queue.

The fix is to reinitialize the transfer parameters for each message
rather than only once on the first message.

Signed-off-by: Brian Niebuhr <bniebuhr-JaPwekKOx1yaMJb+Lgu22Q@public.gmane.org>
---
 drivers/spi/spi_bitbang.c |    9 +++------
 1 files changed, 3 insertions(+), 6 deletions(-)

diff --git a/drivers/spi/spi_bitbang.c b/drivers/spi/spi_bitbang.c
index 5265330..8b55724 100644
--- a/drivers/spi/spi_bitbang.c
+++ b/drivers/spi/spi_bitbang.c
@@ -259,7 +259,6 @@ static void bitbang_work(struct work_struct *work)
 	struct spi_bitbang	*bitbang =
 		container_of(work, struct spi_bitbang, work);
 	unsigned long		flags;
-	int			do_setup = -1;
 	int			(*setup_transfer)(struct spi_device *,
 					struct spi_transfer *);
 
@@ -275,6 +274,7 @@ static void bitbang_work(struct work_struct *work)
 		unsigned		tmp;
 		unsigned		cs_change;
 		int			status;
+		int			do_setup = -1;
 
 		m = container_of(bitbang->queue.next, struct spi_message,
 				queue);
@@ -307,6 +307,8 @@ static void bitbang_work(struct work_struct *work)
 				status = setup_transfer(spi, t);
 				if (status < 0)
 					break;
+				if (do_setup == -1)
+					do_setup = 0;
 			}
 
 			/* set up default clock polarity, and activate chip;
@@ -367,11 +369,6 @@ static void bitbang_work(struct work_struct *work)
 		m->status = status;
 		m->complete(m->context);
 
-		/* restore speed and wordsize if it was overridden */
-		if (do_setup == 1)
-			setup_transfer(spi, NULL);
-		do_setup = 0;
-
 		/* normally deactivate chipselect ... unless no error and
 		 * cs_change has hinted that the next message will probably
 		 * be for this chip too.
-- 
1.6.3.3


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first

Re: [PATCH] spi: bitbang: reinitialize transfer parameters for every message

[Brian Niebuhr]

> This patch fixes the setup_transfer logic to account for the 
> case where
> multiple messages to different SPI devices are in the queue 
> simultaneously.
> With the current logic, the second message in the queue will end up
> using the transfer parameters for the previous message in the queue.
> 
> The fix is to reinitialize the transfer parameters for each message
> rather than only once on the first message.
> 
> Signed-off-by: Brian Niebuhr <bniebuhr-JaPwekKOx1yaMJb+Lgu22Q@public.gmane.org>
> ---
>  drivers/spi/spi_bitbang.c |    9 +++------
>  1 files changed, 3 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/spi/spi_bitbang.c b/drivers/spi/spi_bitbang.c
> index 5265330..8b55724 100644
> --- a/drivers/spi/spi_bitbang.c
> +++ b/drivers/spi/spi_bitbang.c
> @@ -259,7 +259,6 @@ static void bitbang_work(struct work_struct *work)
>  	struct spi_bitbang	*bitbang =
>  		container_of(work, struct spi_bitbang, work);
>  	unsigned long		flags;
> -	int			do_setup = -1;
>  	int			(*setup_transfer)(struct spi_device *,
>  					struct spi_transfer *);
>  
> @@ -275,6 +274,7 @@ static void bitbang_work(struct work_struct *work)
>  		unsigned		tmp;
>  		unsigned		cs_change;
>  		int			status;
> +		int			do_setup = -1;
>  
>  		m = container_of(bitbang->queue.next, struct 
> spi_message,
>  				queue);
> @@ -307,6 +307,8 @@ static void bitbang_work(struct work_struct *work)
>  				status = setup_transfer(spi, t);
>  				if (status < 0)
>  					break;
> +				if (do_setup == -1)
> +					do_setup = 0;
>  			}
>  
>  			/* set up default clock polarity, and 
> activate chip;
> @@ -367,11 +369,6 @@ static void bitbang_work(struct 
> work_struct *work)
>  		m->status = status;
>  		m->complete(m->context);
>  
> -		/* restore speed and wordsize if it was overridden */
> -		if (do_setup == 1)
> -			setup_transfer(spi, NULL);
> -		do_setup = 0;
> -
>  		/* normally deactivate chipselect ... unless no 
> error and
>  		 * cs_change has hinted that the next message 
> will probably
>  		 * be for this chip too.
> -- 
> 1.6.3.3

Hopefully this version should address all of the previous comments.  I
tested it on my board with multiple spi devices, each using different
transfer parameters, and it still works.

Brian


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first

Re: [PATCH] spi: bitbang: reinitialize transfer parameters for every message

[Grant Likely]

On Fri, Jul 2, 2010 at 9:30 AM, Brian Niebuhr <bniebuhr3-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> This patch fixes the setup_transfer logic to account for the case where
> multiple messages to different SPI devices are in the queue simultaneously.
> With the current logic, the second message in the queue will end up
> using the transfer parameters for the previous message in the queue.
>
> The fix is to reinitialize the transfer parameters for each message
> rather than only once on the first message.
>
> Signed-off-by: Brian Niebuhr <bniebuhr-JaPwekKOx1yaMJb+Lgu22Q@public.gmane.org>

Picked up, thanks.

g.

------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first