[PATCH 1/2] OMAP: Kill compile warnings

[PATCH 1/2] OMAP: Kill compile warnings

[nishanth menon]

Got the subject wrong last time.. Sorry..
kill warning1: remove link to .init.text
omap_timer refers to __init omap2_gp_timer_init function causing a warning
while the linking is done

Signed-off-by: Nishanth Menon <menon.nishanth@gmail.com>
---
 arch/arm/mach-omap2/timer-gp.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/arm/mach-omap2/timer-gp.c b/arch/arm/mach-omap2/timer-gp.c
index 8f380a1..f4cf6fe 100644
--- a/arch/arm/mach-omap2/timer-gp.c
+++ b/arch/arm/mach-omap2/timer-gp.c
@@ -175,6 +175,6 @@ static void __init omap2_gp_timer_init(void)
 	omap2_gp_clocksource_init();
 }

-struct sys_timer omap_timer = {
-	.init	= omap2_gp_timer_init,
+struct sys_timer omap_timer __initdata = {
+	.init	 = omap2_gp_timer_init,
 };
-- 
1.5.0.6

Re: [PATCH 1/2] OMAP: Kill compile warnings

[Tony Lindgren]

* nishanth menon <menon.nishanth@gmail.com> [070525 16:05]:
> Got the subject wrong last time.. Sorry..
> kill warning1: remove link to .init.text
> omap_timer refers to __init omap2_gp_timer_init function causing a warning
> while the linking is done
>
> Signed-off-by: Nishanth Menon <menon.nishanth@gmail.com>
> ---
> arch/arm/mach-omap2/timer-gp.c |    4 ++--
> 1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/arch/arm/mach-omap2/timer-gp.c 
> b/arch/arm/mach-omap2/timer-gp.c
> index 8f380a1..f4cf6fe 100644
> --- a/arch/arm/mach-omap2/timer-gp.c
> +++ b/arch/arm/mach-omap2/timer-gp.c
> @@ -175,6 +175,6 @@ static void __init omap2_gp_timer_init(void)
> 	omap2_gp_clocksource_init();
> }
>
> -struct sys_timer omap_timer = {
> -	.init	= omap2_gp_timer_init,
> +struct sys_timer omap_timer __initdata = {
> +	.init	 = omap2_gp_timer_init,
> };
> -- 

This warning is wrong, and we cannot mark sys_timer as initdata. See
other mails related to this on this list. I'd assume this will get fixed
once sys_timer is added to the list of things to ignore.

Tony

RE: [PATCH 1/2] OMAP: Kill compile warnings

[Syed Mohammed, Khasim]

>
>This warning is wrong, and we cannot mark sys_timer as initdata. See
>other mails related to this on this list. I'd assume this will get fixed
>once sys_timer is added to the list of things to ignore.
>
>Tony

Yes, I had a similar discussion with Dirk, he passed me few pointers yesterday, please see through if it helps....


-----Original Message-----
From: Dirk Behme [mailto:dirk.behme@googlemail.com] 
Sent: Friday, May 25, 2007 7:56 AM
To: Syed Mohammed, Khasim
Subject: Re: Compiler warning in omap-twl4030keypad.c

Syed Mohammed, Khasim wrote:
> Hi Dirk,
>  
> Thanks for the info. I was kind of busy with other stuff. Will look into 
> this later this week and submit a patch.

I got the following stuff from Felipe Balbi regarding this in private 
mail:

1) -- cut --
I don't think the problem is with TI's code for the omap-twl4030. If
you check the definition of struct input_dev, at
include/linux/input.h:

     union {            /* temporarily so while we switching to struct 
device */
         struct device *parent;
     } dev;

Probably, we just need to wait until they switch to struct device 
completely.
Or maybe this code will help if we assume that omap_twl4030->cdev->dev
points to the same place as omap_twl4030->dev. I would need any of you
guys to test it:

diff --git a/drivers/input/keyboard/omap-twl4030keypad.c
b/drivers/input/keyboard/omap-twl4030keypad.c
index b96e84f..3776f97 100644
--- a/drivers/input/keyboard/omap-twl4030keypad.c
+++ b/drivers/input/keyboard/omap-twl4030keypad.c
@@ -254,7 +254,7 @@ static int __init omap_kp_probe(struct
platform_device *pdev)
        omap_twl4030kp->keycodemax      = pdata->keymapsize;

        /* Get the debug Device to omap keypad device */
-       dbg_dev = &(omap_twl4030kp->dev);
+       dbg_dev = omap_twl4030kp->dev.parent;

        ret = input_register_device(omap_twl4030kp);
        if (ret < 0) {
-- cut --

2) -- cut --
I forgot to change the text... sorry...
I'm not using the struct class_device here... I'm just accessing the
struct device member inside the union dev.

This is an ugly hack :-P ... but it seems to work.
-- cut --

3) -- cut --
Still a hack, but probably looks better:

diff --git a/drivers/input/keyboard/omap-twl4030keypad.c
b/drivers/input/keyboard/omap-twl4030keypad.c
index b96e84f..5eb3ece 100644
--- a/drivers/input/keyboard/omap-twl4030keypad.c
+++ b/drivers/input/keyboard/omap-twl4030keypad.c
@@ -50,6 +50,8 @@
#define KEYNUM_MASK            0xFF000000
#define ROWCOL_MASK            0x00FFFFFF

+#define get_dbg_dev(d)         (d).parent
+
static char *switch_name[NUM_ROWS][NUM_COLS] = {
        {"S2_L", "S2_D", "S2_S", "S3", "S4", "S23"},
        {"S2_R", "S2_U", "S5", "S6", "S7", "S24"},
@@ -254,7 +256,7 @@ static int __init omap_kp_probe(struct
platform_device *pdev)
        omap_twl4030kp->keycodemax      = pdata->keymapsize;

        /* Get the debug Device to omap keypad device */
-       dbg_dev = &(omap_twl4030kp->dev);
+       dbg_dev = get_dbg_dev(omap_twl4030kp->dev);

        ret = input_register_device(omap_twl4030kp);
        if (ret < 0) {

-- cut --

Regards

Dirk

Re: [PATCH 1/2] OMAP: Kill compile warnings

[nishanth menon]

Khasim,
On 5/25/07, Syed Mohammed, Khasim <x0khasim@ti.com> wrote:
> Yes, I had a similar discussion with Dirk, he passed me few pointers yesterday, please see through if it helps....
>        /* Get the debug Device to omap keypad device */
> -       dbg_dev = &(omap_twl4030kp->dev);
> +       dbg_dev = omap_twl4030kp->dev.parent;
That would be redundant:
220         /* Get the debug Device */
221         dbg_dev = &(pdev->dev);
already did that for you.. only, now the messages will show
"input:xyz.." instead of "omap_twl4030:xyz"..
Regards,
Nishanth Menon

RE: [PATCH 1/2] OMAP: Kill compile warnings

[Syed Mohammed, Khasim]

>That would be redundant:
>220         /* Get the debug Device */
>221         dbg_dev = &(pdev->dev);
>already did that for you.. only, now the messages will show
>"input:xyz.." instead of "omap_twl4030:xyz"..

OK, thanks. 

Acked-by: x0khasim@ti.com

Regards,
Khasim

Re: [PATCH 1/2] OMAP: Kill compile warnings

[Dirk Behme]

Syed Mohammed, Khasim wrote:
>> That would be redundant:
>> 220         /* Get the debug Device */
>> 221         dbg_dev = &(pdev->dev);
>> already did that for you.. only, now the messages will show
>> "input:xyz.." instead of "omap_twl4030:xyz"..
> 
> OK, thanks. 
> 
> Acked-by: x0khasim@ti.com

Okay, after 2 minutes thinking about what's so confusing here, I have
it now. We have two different patches with nearly identical subject.
The first one should fix a warning we already discussed under an other
subject. A answer to the second patch was sent under the subject of
the first one ;) Therefore some clarification as I understand it now:

We have two warnings:

1) Section mismatch with sys_timer

This is what "[PATCH 1/2] OMAP: Kill compile warnings" is about.

As mentioned by Tony this patch isn't correct and we will wait (as
discussed already earlier) until this is fixed by ignore list.

http://linux.omap.com/pipermail/linux-omap-open-source/2007-May/010087.html

http://linux.omap.com/pipermail/linux-omap-open-source/2007-May/010061.html

2) Warning in twl4030 keypad about dbg_dev

This is what "[PATCH 2/2] OMAP: Kill compile warnings" is about.

For this we already had an other subject in the list archives as well
"Compiler warning in omap-twl4030keypad.c".

Above Khasim acked it (with misleading subject ;) ) and I also think
it's okay

http://linux.omap.com/pipermail/linux-omap-open-source/2007-May/010086.html

Acked-by: Dirk Behme <dirk.behme@gmail.com>

Conclusion: For (1) we have to wait, for (2) please apply.

Hope that helps a little,

Dirk

Re: [PATCH 1/2] OMAP: Kill compile warnings

[Tony Lindgren]

* Dirk Behme <dirk.behme@googlemail.com> [070525 23:02]:
> Syed Mohammed, Khasim wrote:
>>> That would be redundant:
>>> 220         /* Get the debug Device */
>>> 221         dbg_dev = &(pdev->dev);
>>> already did that for you.. only, now the messages will show
>>> "input:xyz.." instead of "omap_twl4030:xyz"..
>> OK, thanks. Acked-by: x0khasim@ti.com
>
> Okay, after 2 minutes thinking about what's so confusing here, I have
> it now. We have two different patches with nearly identical subject.
> The first one should fix a warning we already discussed under an other
> subject. A answer to the second patch was sent under the subject of
> the first one ;) Therefore some clarification as I understand it now:
>
> We have two warnings:
>
> 1) Section mismatch with sys_timer
>
> This is what "[PATCH 1/2] OMAP: Kill compile warnings" is about.
>
> As mentioned by Tony this patch isn't correct and we will wait (as
> discussed already earlier) until this is fixed by ignore list.
>
> http://linux.omap.com/pipermail/linux-omap-open-source/2007-May/010087.html
>
> http://linux.omap.com/pipermail/linux-omap-open-source/2007-May/010061.html
>
> 2) Warning in twl4030 keypad about dbg_dev
>
> This is what "[PATCH 2/2] OMAP: Kill compile warnings" is about.
>
> For this we already had an other subject in the list archives as well
> "Compiler warning in omap-twl4030keypad.c".
>
> Above Khasim acked it (with misleading subject ;) ) and I also think
> it's okay
>
> http://linux.omap.com/pipermail/linux-omap-open-source/2007-May/010086.html
>
> Acked-by: Dirk Behme <dirk.behme@gmail.com>
>
> Conclusion: For (1) we have to wait, for (2) please apply.
>
> Hope that helps a little,

Yeah, it's getting complex :) Pushing Felipe's patch posted by Nishant
today.

Tony

Re: [PATCH 1/2] OMAP: Kill compile warnings

[Tony Lindgren]

* Tony Lindgren <tony@atomide.com> [070605 03:16]:
> * Dirk Behme <dirk.behme@googlemail.com> [070525 23:02]:
> > Syed Mohammed, Khasim wrote:
> >>> That would be redundant:
> >>> 220         /* Get the debug Device */
> >>> 221         dbg_dev = &(pdev->dev);
> >>> already did that for you.. only, now the messages will show
> >>> "input:xyz.." instead of "omap_twl4030:xyz"..
> >> OK, thanks. Acked-by: x0khasim@ti.com
> >
> > Okay, after 2 minutes thinking about what's so confusing here, I have
> > it now. We have two different patches with nearly identical subject.
> > The first one should fix a warning we already discussed under an other
> > subject. A answer to the second patch was sent under the subject of
> > the first one ;) Therefore some clarification as I understand it now:
> >
> > We have two warnings:
> >
> > 1) Section mismatch with sys_timer
> >
> > This is what "[PATCH 1/2] OMAP: Kill compile warnings" is about.
> >
> > As mentioned by Tony this patch isn't correct and we will wait (as
> > discussed already earlier) until this is fixed by ignore list.
> >
> > http://linux.omap.com/pipermail/linux-omap-open-source/2007-May/010087.html
> >
> > http://linux.omap.com/pipermail/linux-omap-open-source/2007-May/010061.html
> >
> > 2) Warning in twl4030 keypad about dbg_dev
> >
> > This is what "[PATCH 2/2] OMAP: Kill compile warnings" is about.
> >
> > For this we already had an other subject in the list archives as well
> > "Compiler warning in omap-twl4030keypad.c".
> >
> > Above Khasim acked it (with misleading subject ;) ) and I also think
> > it's okay
> >
> > http://linux.omap.com/pipermail/linux-omap-open-source/2007-May/010086.html
> >
> > Acked-by: Dirk Behme <dirk.behme@gmail.com>
> >
> > Conclusion: For (1) we have to wait, for (2) please apply.
> >
> > Hope that helps a little,
> 
> Yeah, it's getting complex :) Pushing Felipe's patch posted by Nishant
> today.

Correction Felipe's patch posted by Khasim :)

Tony