[PATCH] Make H2 board file compile without irda warning

[PATCH] Make H2 board file compile without irda warning

[Daniel Petrini]

[-- Attachment #1: Type: text/plain, Size: 345 bytes --]

Hi,

The board file for H2 is generating a warning if compiled without
InfraRed selection.

This patch changes the location where the function h2_transceiver_mode
is assigned to structure h2_irda_data, thereby preventing the warning.

Signed-off-by Daniel Petrini <d.pensator_at_gmail.com>

Daniel Petrini
--
INdT - Manaus - Brazil

[-- Attachment #2: h2-irda-compile.patch --]
[-- Type: text/x-patch, Size: 1076 bytes --]

Index: linux-omap-2.6.git-q/arch/arm/mach-omap1/board-h2.c
===================================================================
--- linux-omap-2.6.git-q.orig/arch/arm/mach-omap1/board-h2.c	2006-02-21 10:11:01.000000000 -0400
+++ linux-omap-2.6.git-q/arch/arm/mach-omap1/board-h2.c	2006-02-21 11:28:18.000000000 -0400
@@ -252,6 +252,7 @@ static int h2_transceiver_mode(struct de
 
 static struct omap_irda_config h2_irda_data = {
 	.transceiver_cap	= IR_SIRMODE | IR_MIRMODE | IR_FIRMODE,
+	.transceiver_mode	= h2_transceiver_mode,
 };
 
 static struct resource h2_irda_resources[] = {
@@ -377,10 +378,8 @@ static void __init h2_init(void)
 	/* Irda */
 #if defined(CONFIG_OMAP_IR) || defined(CONFIG_OMAP_IR_MODULE)
 	omap_writel(omap_readl(FUNC_MUX_CTRL_A) | 7, FUNC_MUX_CTRL_A);
-	if (!(omap_request_gpio(H2_IRDA_FIRSEL_GPIO_PIN))) {
+	if (!(omap_request_gpio(H2_IRDA_FIRSEL_GPIO_PIN)))
 		omap_set_gpio_direction(H2_IRDA_FIRSEL_GPIO_PIN, 0);
-		h2_irda_data.transceiver_mode = h2_transceiver_mode;
-	}
 #endif
 
 	platform_add_devices(h2_devices, ARRAY_SIZE(h2_devices));




[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

Re: [PATCH] Make H2 board file compile without irda warning

[Komal Shah]

--- Daniel Petrini <d.pensator@gmail.com> wrote:

> Hi,
> 
> The board file for H2 is generating a warning if compiled without
> InfraRed selection.
> 
> This patch changes the location where the function
> h2_transceiver_mode
> is assigned to structure h2_irda_data, thereby preventing the
> warning.

True, the warning will be generated, but I added that check considering
that transceiver_mode function will be active only if we are
successfully able to request that GPIO. So, I am _not_ sure, but it is
better to #ifdef around the h2_transceiver_mode(...) function. What do
you think?

---Komal Shah
http://komalshah.blogspot.com/

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

Re: [PATCH] Make H2 board file compile without irda warning

[Komal Shah]

[-- Attachment #1: Type: text/plain, Size: 760 bytes --]

--- Komal Shah <komal_shah802003@yahoo.com> wrote:

> 
> True, the warning will be generated, but I added that check
> considering
> that transceiver_mode function will be active only if we are
> successfully able to request that GPIO. So, I am _not_ sure, but it
> is
> better to #ifdef around the h2_transceiver_mode(...) function. What
> do
> you think?
> 

And here comes the patch.

OMAP: Irda: Remove warning
arch/arm/mach-omap1/board-h2.c:244: warning: 'h2_transceiver_mode'
defined but not used.

Signed-off-by: Komal Shah <komal_shah802003@yahoo.com>

---Komal Shah
http://komalshah.blogspot.com/

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

[-- Attachment #2: 390387012-h2Irda.patch --]
[-- Type: application/octet-stream, Size: 634 bytes --]

diff --git a/arch/arm/mach-omap1/board-h2.c b/arch/arm/mach-omap1/board-h2.c
index 680e67e..cbc7674 100644
--- a/arch/arm/mach-omap1/board-h2.c
+++ b/arch/arm/mach-omap1/board-h2.c
@@ -240,6 +240,7 @@ static struct platform_device h2_kp_devi
 
 #define H2_IRDA_FIRSEL_GPIO_PIN	17
 
+#if defined(CONFIG_OMAP_IR) || defined(CONFIG_OMAP_IR_MODULE) 
 static int h2_transceiver_mode(struct device *dev, int state)
 {
 	if (state & IR_SIRMODE)
@@ -249,6 +250,7 @@ static int h2_transceiver_mode(struct de
 
 	return 0;
 }
+#endif
 
 static struct omap_irda_config h2_irda_data = {
 	.transceiver_cap	= IR_SIRMODE | IR_MIRMODE | IR_FIRMODE,

[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

Re: [PATCH] Make H2 board file compile without irda warning

[Daniel Petrini]

[-- Attachment #1: Type: text/plain, Size: 1345 bytes --]

Hi,

On 2/21/06, Komal Shah <komal_shah802003@yahoo.com> wrote:
> --- Komal Shah <komal_shah802003@yahoo.com> wrote:
>
> >
> > True, the warning will be generated, but I added that check
> > considering
> > that transceiver_mode function will be active only if we are
> > successfully able to request that GPIO. So, I am _not_ sure, but it
> > is
> > better to #ifdef around the h2_transceiver_mode(...) function. What
> > do
> > you think?
> >
>
> And here comes the patch.
>
> OMAP: Irda: Remove warning
> arch/arm/mach-omap1/board-h2.c:244: warning: 'h2_transceiver_mode'
> defined but not used.
>
> Signed-off-by: Komal Shah <komal_shah802003@yahoo.com>
>
> ---Komal Shah
> http://komalshah.blogspot.com/

Humm, my first idea was exactly like this but I tried to avoid another #ifdef..
So what about this new approach, which still avoids #ifdefs although
with expense of one flag ?

--

The board file for H2 is generating a warning if compiled without
InfraRed selection.

This patch changes the location where the function h2_transceiver_mode
is assigned to structure h2_irda_data, thereby preventing the warning
and also creates a flag variable telling when is safe to use h2
platform function for IR.

Signed-off-by Daniel Petrini <d.pensator_at_gmail.com>

Daniel
--
INdT - Manaus - Brazil

[-- Attachment #2: h2-irda-compile.patch --]
[-- Type: text/x-patch, Size: 1484 bytes --]

Index: linux-omap-2.6.git-q/arch/arm/mach-omap1/board-h2.c
===================================================================
--- linux-omap-2.6.git-q.orig/arch/arm/mach-omap1/board-h2.c	2006-02-21 10:11:01.000000000 -0400
+++ linux-omap-2.6.git-q/arch/arm/mach-omap1/board-h2.c	2006-02-22 08:54:16.000000000 -0400
@@ -252,6 +252,8 @@ static int h2_transceiver_mode(struct de
 
 static struct omap_irda_config h2_irda_data = {
 	.transceiver_cap	= IR_SIRMODE | IR_MIRMODE | IR_FIRMODE,
+	.transceiver_mode	= h2_transceiver_mode,
+	.irda_capable		= 0,
 };
 
 static struct resource h2_irda_resources[] = {
@@ -379,7 +381,7 @@ static void __init h2_init(void)
 	omap_writel(omap_readl(FUNC_MUX_CTRL_A) | 7, FUNC_MUX_CTRL_A);
 	if (!(omap_request_gpio(H2_IRDA_FIRSEL_GPIO_PIN))) {
 		omap_set_gpio_direction(H2_IRDA_FIRSEL_GPIO_PIN, 0);
-		h2_irda_data.transceiver_mode = h2_transceiver_mode;
+		h2_irda_data.irda_capable = 1;
 	}
 #endif
 
Index: linux-omap-2.6.git-q/include/asm-arm/arch-omap/irda.h
===================================================================
--- linux-omap-2.6.git-q.orig/include/asm-arm/arch-omap/irda.h	2006-02-22 09:00:30.000000000 -0400
+++ linux-omap-2.6.git-q/include/asm-arm/arch-omap/irda.h	2006-02-22 09:01:02.000000000 -0400
@@ -25,6 +25,8 @@ struct omap_irda_config {
 	 * having calls which can sleep in irda_set_speed.
 	 */
 	struct work_struct gpio_expa;
+	/* Very specific to H2 - success in obtain gpio */
+	int irda_capable;
 };
 
 #endif


[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

Re: [PATCH] Make H2 board file compile without irda warning

[Daniel Petrini]

[-- Attachment #1: Type: text/plain, Size: 1497 bytes --]

Ops, here goes the proper patch:

On 2/22/06, Daniel Petrini <d.pensator@gmail.com> wrote:
> Hi,
>
> On 2/21/06, Komal Shah <komal_shah802003@yahoo.com> wrote:
> > --- Komal Shah <komal_shah802003@yahoo.com> wrote:
> >
> > >
> > > True, the warning will be generated, but I added that check
> > > considering
> > > that transceiver_mode function will be active only if we are
> > > successfully able to request that GPIO. So, I am _not_ sure, but it
> > > is
> > > better to #ifdef around the h2_transceiver_mode(...) function. What
> > > do
> > > you think?
> > >
> >
> > And here comes the patch.
> >
> > OMAP: Irda: Remove warning
> > arch/arm/mach-omap1/board-h2.c:244: warning: 'h2_transceiver_mode'
> > defined but not used.
> >
> > Signed-off-by: Komal Shah <komal_shah802003@yahoo.com>
> >
> > ---Komal Shah
> > http://komalshah.blogspot.com/
>

 Humm, my first idea was exactly like this but I tried to avoid another #ifdef..
 So what about this new approach, which still avoids #ifdefs although
 with expense of one flag ?

 --

The board file for H2 is generating a warning if compiled without
InfraRed selection.

This patch changes the location where the function h2_transceiver_mode
is assigned to structure h2_irda_data, thereby preventing the warning
and also creates a flag variable telling when is safe to use h2
platform function for IR.

Signed-off-by Daniel Petrini <d.pensator_at_gmail.com>

Daniel
--
INdT - Manaus - Brazil

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: h2-irda-compile.patch --]
[-- Type: text/x-patch; name="h2-irda-compile.patch", Size: 2100 bytes --]

Index: linux-omap-2.6.git-q/arch/arm/mach-omap1/board-h2.c
===================================================================
--- linux-omap-2.6.git-q.orig/arch/arm/mach-omap1/board-h2.c	2006-02-21 10:11:01.000000000 -0400
+++ linux-omap-2.6.git-q/arch/arm/mach-omap1/board-h2.c	2006-02-22 08:54:16.000000000 -0400
@@ -252,6 +252,8 @@ static int h2_transceiver_mode(struct de
 
 static struct omap_irda_config h2_irda_data = {
 	.transceiver_cap	= IR_SIRMODE | IR_MIRMODE | IR_FIRMODE,
+	.transceiver_mode	= h2_transceiver_mode,
+	.irda_capable		= 0,
 };
 
 static struct resource h2_irda_resources[] = {
@@ -379,7 +381,7 @@ static void __init h2_init(void)
 	omap_writel(omap_readl(FUNC_MUX_CTRL_A) | 7, FUNC_MUX_CTRL_A);
 	if (!(omap_request_gpio(H2_IRDA_FIRSEL_GPIO_PIN))) {
 		omap_set_gpio_direction(H2_IRDA_FIRSEL_GPIO_PIN, 0);
-		h2_irda_data.transceiver_mode = h2_transceiver_mode;
+		h2_irda_data.irda_capable = 1;
 	}
 #endif
 
Index: linux-omap-2.6.git-q/include/asm-arm/arch-omap/irda.h
===================================================================
--- linux-omap-2.6.git-q.orig/include/asm-arm/arch-omap/irda.h	2006-02-22 09:00:30.000000000 -0400
+++ linux-omap-2.6.git-q/include/asm-arm/arch-omap/irda.h	2006-02-22 09:01:02.000000000 -0400
@@ -25,6 +25,8 @@ struct omap_irda_config {
 	 * having calls which can sleep in irda_set_speed.
 	 */
 	struct work_struct gpio_expa;
+	/* Very specific to H2 - success in obtain gpio */
+	int irda_capable;
 };
 
 #endif
Index: linux-omap-2.6.git-q/drivers/net/irda/omap-ir.c
===================================================================
--- linux-omap-2.6.git-q.orig/drivers/net/irda/omap-ir.c	2006-02-14 08:24:54.000000000 -0400
+++ linux-omap-2.6.git-q/drivers/net/irda/omap-ir.c	2006-02-22 09:02:27.000000000 -0400
@@ -336,7 +336,7 @@ static int omap_irda_startup(struct net_
 
 	/* Only for H2?
 	 */
-	if (si->pdata->transceiver_mode && machine_is_omap_h2()) {
+	if (si->pdata->irda_capable && machine_is_omap_h2()) {
 	    	/* Is it select_irda on H2 ? */
 		omap_writel(omap_readl(FUNC_MUX_CTRL_A) | 7,
 	 FUNC_MUX_CTRL_A);

[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

Re: [PATCH] Make H2 board file compile without irda warning

[Komal Shah]

--- Daniel Petrini <d.pensator@gmail.com> wrote:

> 
>  Humm, my first idea was exactly like this but I tried to avoid
> another #ifdef..
>  So what about this new approach, which still avoids #ifdefs although
>  with expense of one flag ?
> 

I am now in confusion :). I like your patch too, but I thought that
adding #ifdef in board-* files is no harm. So, let's Tony speak :)

---Komal Shah
http://komalshah.blogspot.com/

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

Re: [PATCH] Make H2 board file compile without irda warning

[Tony Lindgren]

* Komal Shah <komal_shah802003@yahoo.com> [060222 09:26]:
> --- Daniel Petrini <d.pensator@gmail.com> wrote:
> 
> > 
> >  Humm, my first idea was exactly like this but I tried to avoid
> > another #ifdef..
> >  So what about this new approach, which still avoids #ifdefs although
> >  with expense of one flag ?
> > 
> 
> I am now in confusion :). I like your patch too, but I thought that
> adding #ifdef in board-* files is no harm. So, let's Tony speak :)

Both work, but let's use the smaller codesize one at the expense of
one ifdef :) I'll push Komal's patch today.

Tony