Patchwork [6/8] ARM: OMAP2+: Disable code that currently does not work with multiplaform

login
register
mail settings
Submitter Tony Lindgren
Date Jan. 3, 2013, 8:15 p.m.
Message ID <20130103201521.GI25633@atomide.com>
Download mbox | patch
Permalink /patch/209307/
State New
Headers show

Comments

Tony Lindgren - Jan. 3, 2013, 8:15 p.m.
Hi all,

Looks like we need to also disable mailbox to avoid dependencies to
the mailbox series as it's being moved to live under drivers.

And we need to disable omap3isp as it tries to use the now private
include <plat/*.h> headers.

Updated patch below.

Tony


From: Tony Lindgren <tony@atomide.com>
Date: Wed, 2 Jan 2013 10:24:09 -0800
Subject: [PATCH] ARM: OMAP2+: Disable code that currently does not work with multiplaform
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

We still need to fix up few places for multiplatform support,
but that can proceed separately. Fix the issue by making the
problem drivers depends !ARCH_MULTIPLATFORM for now.

The remaining pieces that are not multiplatform compatible
for omap2+ SoCs are:

1. Some drivers are using custom omap_dm_timer calls

There are two drivers that are directly usign omap hardware
timers for PWM and DSP clocking: drivers/media/rc/ir-rx51.c and
drivers/staging/tidspbridge/core/dsp-clock.c. These can be
fixed for multiplatform by allowing a minimal set of hardware
timers to be accessed, and for some functionality by using the
hrtimer framework.

2. Hardware OMAP4_ERRATA_I688 needs to be fixed up

This can't be enabled for multiplatform configurations in
it's current form. It may be possible to fix it up to do
instruction replacement early on during init. Luckily it
looks like this errata does not seem to get hit with
mainline kernel code alone at least currently.

3. Legacy header needed for omap-sham.c

Looks like it still needs mach/irqs.h for omap1 that
does not exist for multiplatform systems. Just ifdef
it for now.

4. Mailbox is waiting to get moved to drivers

Disable it for now to avoid adding a dependency to the
mailbox patches.

5. Camera omap3isp needs to be disabled

The include <plat/*.h> files won't be available to drivers
after multiplatform support is enabled.

Cc: Timo Kokkonen <timo.t.kokkonen@iki.fi>
Cc: Sean Young <sean@mess.org>
Cc: "Víctor Manuel Jáquez Leal" <vjaquez@igalia.com>
Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc: Mauro Carvalho Chehab <mchehab@redhat.com>
Cc: Omar Ramirez Luna <omar.ramirez@ti.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Santosh Shilimkar <santosh.shilimkar@ti.com>
Tested-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
[tony@atomide.com: updated to disable mailbox and omap3isp]
Signed-off-by: Tony Lindgren <tony@atomide.com>
Laurent Pinchart - Jan. 3, 2013, 9:23 p.m.
Hi Tony,

On Thursday 03 January 2013 12:15:21 Tony Lindgren wrote:
> Hi all,
> 
> Looks like we need to also disable mailbox to avoid dependencies to
> the mailbox series as it's being moved to live under drivers.
> 
> And we need to disable omap3isp as it tries to use the now private
> include <plat/*.h> headers.

I've posted a patch to the linux-media mailing list to remove plat/*.h headers 
that are not present anymore in v3.8. It will be pushed to v3.8, as this 
breaks compilation of the driver.

The driver still includes the plat/cpu.h header that can be removed as well. 
I've just posted a patch that remove the header and will push it to v3.9. The 
OMAP3 ISP part of this patch can thus be dropped.

> Updated patch below.
> 
> Tony
> 
> 
> From: Tony Lindgren <tony@atomide.com>
> Date: Wed, 2 Jan 2013 10:24:09 -0800
> Subject: [PATCH] ARM: OMAP2+: Disable code that currently does not work with
> multiplaform MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
> 
> We still need to fix up few places for multiplatform support,
> but that can proceed separately. Fix the issue by making the
> problem drivers depends !ARCH_MULTIPLATFORM for now.
> 
> The remaining pieces that are not multiplatform compatible
> for omap2+ SoCs are:
> 
> 1. Some drivers are using custom omap_dm_timer calls
> 
> There are two drivers that are directly usign omap hardware
> timers for PWM and DSP clocking: drivers/media/rc/ir-rx51.c and
> drivers/staging/tidspbridge/core/dsp-clock.c. These can be
> fixed for multiplatform by allowing a minimal set of hardware
> timers to be accessed, and for some functionality by using the
> hrtimer framework.
> 
> 2. Hardware OMAP4_ERRATA_I688 needs to be fixed up
> 
> This can't be enabled for multiplatform configurations in
> it's current form. It may be possible to fix it up to do
> instruction replacement early on during init. Luckily it
> looks like this errata does not seem to get hit with
> mainline kernel code alone at least currently.
> 
> 3. Legacy header needed for omap-sham.c
> 
> Looks like it still needs mach/irqs.h for omap1 that
> does not exist for multiplatform systems. Just ifdef
> it for now.
> 
> 4. Mailbox is waiting to get moved to drivers
> 
> Disable it for now to avoid adding a dependency to the
> mailbox patches.
> 
> 5. Camera omap3isp needs to be disabled
> 
> The include <plat/*.h> files won't be available to drivers
> after multiplatform support is enabled.
> 
> Cc: Timo Kokkonen <timo.t.kokkonen@iki.fi>
> Cc: Sean Young <sean@mess.org>
> Cc: "Víctor Manuel Jáquez Leal" <vjaquez@igalia.com>
> Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
> Cc: Mauro Carvalho Chehab <mchehab@redhat.com>
> Cc: Omar Ramirez Luna <omar.ramirez@ti.com>
> Cc: Herbert Xu <herbert@gondor.apana.org.au>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Santosh Shilimkar <santosh.shilimkar@ti.com>
> Tested-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
> [tony@atomide.com: updated to disable mailbox and omap3isp]
> Signed-off-by: Tony Lindgren <tony@atomide.com>
> 
> diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig
> index 41b581f..492d764 100644
> --- a/arch/arm/mach-omap2/Kconfig
> +++ b/arch/arm/mach-omap2/Kconfig
> @@ -397,7 +397,7 @@ config OMAP3_SDRC_AC_TIMING
> 
>  config OMAP4_ERRATA_I688
>  	bool "OMAP4 errata: Async Bridge Corruption"
> -	depends on ARCH_OMAP4
> +	depends on ARCH_OMAP4 && !ARCH_MULTIPLATFORM
>  	select ARCH_HAS_BARRIERS
>  	help
>  	  If a data is stalled inside asynchronous bridge because of back
> diff --git a/arch/arm/plat-omap/Kconfig b/arch/arm/plat-omap/Kconfig
> index 665870d..03f1252 100644
> --- a/arch/arm/plat-omap/Kconfig
> +++ b/arch/arm/plat-omap/Kconfig
> @@ -118,7 +118,7 @@ config OMAP_MUX_WARNINGS
> 
>  config OMAP_MBOX_FWK
>  	tristate "Mailbox framework support"
> -	depends on ARCH_OMAP
> +	depends on ARCH_OMAP && !ARCH_MULTIPLATFORM
>  	help
>  	  Say Y here if you want to use OMAP Mailbox framework support for
>  	  DSP, IVA1.0 and IVA2 in OMAP1/2/3.
> diff --git a/drivers/crypto/omap-sham.c b/drivers/crypto/omap-sham.c
> index 1d75e6f..d65f22c 100644
> --- a/drivers/crypto/omap-sham.c
> +++ b/drivers/crypto/omap-sham.c
> @@ -38,7 +38,10 @@
>  #include <crypto/internal/hash.h>
> 
>  #include <linux/omap-dma.h>
> +
> +#ifdef CONFIG_ARCH_OMAP1
>  #include <mach/irqs.h>
> +#endif
> 
>  #define SHA_REG_DIGEST(x)		(0x00 + ((x) * 0x04))
>  #define SHA_REG_DIN(x)			(0x1C + ((x) * 0x04))
> diff --git a/drivers/media/platform/Kconfig b/drivers/media/platform/Kconfig
> index 3dcfea6..eff53c9 100644
> --- a/drivers/media/platform/Kconfig
> +++ b/drivers/media/platform/Kconfig
> @@ -99,7 +99,7 @@ config VIDEO_OMAP2
> 
>  config VIDEO_OMAP3
>  	tristate "OMAP 3 Camera support (EXPERIMENTAL)"
> -	depends on OMAP_IOVMM && VIDEO_V4L2 && I2C && VIDEO_V4L2_SUBDEV_API &&
> ARCH_OMAP3 && EXPERIMENTAL +	depends on OMAP_IOVMM && VIDEO_V4L2 && I2C &&
> VIDEO_V4L2_SUBDEV_API && ARCH_OMAP3 && EXPERIMENTAL && !ARCH_MULTIPLATFORM
> ---help---
>  	  Driver for an OMAP 3 camera controller.
> 
> diff --git a/drivers/media/rc/Kconfig b/drivers/media/rc/Kconfig
> index 79ba242..19f3563 100644
> --- a/drivers/media/rc/Kconfig
> +++ b/drivers/media/rc/Kconfig
> @@ -291,7 +291,7 @@ config IR_TTUSBIR
> 
>  config IR_RX51
>  	tristate "Nokia N900 IR transmitter diode"
> -	depends on OMAP_DM_TIMER && LIRC
> +	depends on OMAP_DM_TIMER && LIRC && !ARCH_MULTIPLATFORM
>  	---help---
>  	   Say Y or M here if you want to enable support for the IR
>  	   transmitter diode built in the Nokia N900 (RX51) device.
> diff --git a/drivers/staging/tidspbridge/Kconfig
> b/drivers/staging/tidspbridge/Kconfig index 0dd479f..60848f1 100644
> --- a/drivers/staging/tidspbridge/Kconfig
> +++ b/drivers/staging/tidspbridge/Kconfig
> @@ -4,7 +4,7 @@
> 
>  menuconfig TIDSPBRIDGE
>  	tristate "DSP Bridge driver"
> -	depends on ARCH_OMAP3
> +	depends on ARCH_OMAP3 && !ARCH_MULTIPLATFORM
>  	select OMAP_MBOX_FWK
>  	help
>  	  DSP/BIOS Bridge is designed for platforms that contain a GPP and
Tony Lindgren - Jan. 3, 2013, 10:49 p.m.
* Laurent Pinchart <laurent.pinchart@ideasonboard.com> [130103 13:25]:
> Hi Tony,
> 
> On Thursday 03 January 2013 12:15:21 Tony Lindgren wrote:
> > Hi all,
> > 
> > Looks like we need to also disable mailbox to avoid dependencies to
> > the mailbox series as it's being moved to live under drivers.
> > 
> > And we need to disable omap3isp as it tries to use the now private
> > include <plat/*.h> headers.
> 
> I've posted a patch to the linux-media mailing list to remove plat/*.h headers 
> that are not present anymore in v3.8. It will be pushed to v3.8, as this 
> breaks compilation of the driver.

Great, thanks for doing that.
 
> The driver still includes the plat/cpu.h header that can be removed as well. 
> I've just posted a patch that remove the header and will push it to v3.9. The 
> OMAP3 ISP part of this patch can thus be dropped.

OK. I suggest you merge the plat/cpu.h during the -rc cycle as well.
With v3.8-rc2 plat/cpu.h no longer contains anything for omap2+, and
we could make it now to do #error for omap2+ if the current users
are fixed.

Will drop the omap3isp part from this patch as you suggested though.

Regards,

Tony
Laurent Pinchart - Jan. 3, 2013, 10:53 p.m.
Hi Tony,

On Thursday 03 January 2013 14:49:54 Tony Lindgren wrote:
> * Laurent Pinchart <laurent.pinchart@ideasonboard.com> [130103 13:25]:
> > Hi Tony,
> > 
> > On Thursday 03 January 2013 12:15:21 Tony Lindgren wrote:
> > > Hi all,
> > > 
> > > Looks like we need to also disable mailbox to avoid dependencies to
> > > the mailbox series as it's being moved to live under drivers.
> > > 
> > > And we need to disable omap3isp as it tries to use the now private
> > > include <plat/*.h> headers.
> > 
> > I've posted a patch to the linux-media mailing list to remove plat/*.h
> > headers that are not present anymore in v3.8. It will be pushed to v3.8,
> > as this breaks compilation of the driver.
> 
> Great, thanks for doing that.
> 
> > The driver still includes the plat/cpu.h header that can be removed as
> > well. I've just posted a patch that remove the header and will push it to
> > v3.9. The OMAP3 ISP part of this patch can thus be dropped.
> 
> OK. I suggest you merge the plat/cpu.h during the -rc cycle as well.
> With v3.8-rc2 plat/cpu.h no longer contains anything for omap2+, and
> we could make it now to do #error for omap2+ if the current users
> are fixed.

If Mauro is fine with the patch going to v3.8-rc I have no issue with that. Is 
this patch targetted at v3.8 or v3.9 ?

> Will drop the omap3isp part from this patch as you suggested though.

Thank you.
Tony Lindgren - Jan. 3, 2013, 11 p.m.
* Laurent Pinchart <laurent.pinchart@ideasonboard.com> [130103 14:54]:
> On Thursday 03 January 2013 14:49:54 Tony Lindgren wrote:
> > * Laurent Pinchart <laurent.pinchart@ideasonboard.com> [130103 13:25]:
> > >
> > > The driver still includes the plat/cpu.h header that can be removed as
> > > well. I've just posted a patch that remove the header and will push it to
> > > v3.9. The OMAP3 ISP part of this patch can thus be dropped.
> > 
> > OK. I suggest you merge the plat/cpu.h during the -rc cycle as well.
> > With v3.8-rc2 plat/cpu.h no longer contains anything for omap2+, and
> > we could make it now to do #error for omap2+ if the current users
> > are fixed.
> 
> If Mauro is fine with the patch going to v3.8-rc I have no issue with that. Is 
> this patch targetted at v3.8 or v3.9 ?

Well ideally ASAP to prevent new drivers including plat/cpu.h as otherwise
build breakage will occur when we flip on the multiplatform support :)

Regards,

Tony

Patch

diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig
index 41b581f..492d764 100644
--- a/arch/arm/mach-omap2/Kconfig
+++ b/arch/arm/mach-omap2/Kconfig
@@ -397,7 +397,7 @@  config OMAP3_SDRC_AC_TIMING
 
 config OMAP4_ERRATA_I688
 	bool "OMAP4 errata: Async Bridge Corruption"
-	depends on ARCH_OMAP4
+	depends on ARCH_OMAP4 && !ARCH_MULTIPLATFORM
 	select ARCH_HAS_BARRIERS
 	help
 	  If a data is stalled inside asynchronous bridge because of back
diff --git a/arch/arm/plat-omap/Kconfig b/arch/arm/plat-omap/Kconfig
index 665870d..03f1252 100644
--- a/arch/arm/plat-omap/Kconfig
+++ b/arch/arm/plat-omap/Kconfig
@@ -118,7 +118,7 @@  config OMAP_MUX_WARNINGS
 
 config OMAP_MBOX_FWK
 	tristate "Mailbox framework support"
-	depends on ARCH_OMAP
+	depends on ARCH_OMAP && !ARCH_MULTIPLATFORM
 	help
 	  Say Y here if you want to use OMAP Mailbox framework support for
 	  DSP, IVA1.0 and IVA2 in OMAP1/2/3.
diff --git a/drivers/crypto/omap-sham.c b/drivers/crypto/omap-sham.c
index 1d75e6f..d65f22c 100644
--- a/drivers/crypto/omap-sham.c
+++ b/drivers/crypto/omap-sham.c
@@ -38,7 +38,10 @@ 
 #include <crypto/internal/hash.h>
 
 #include <linux/omap-dma.h>
+
+#ifdef CONFIG_ARCH_OMAP1
 #include <mach/irqs.h>
+#endif
 
 #define SHA_REG_DIGEST(x)		(0x00 + ((x) * 0x04))
 #define SHA_REG_DIN(x)			(0x1C + ((x) * 0x04))
diff --git a/drivers/media/platform/Kconfig b/drivers/media/platform/Kconfig
index 3dcfea6..eff53c9 100644
--- a/drivers/media/platform/Kconfig
+++ b/drivers/media/platform/Kconfig
@@ -99,7 +99,7 @@  config VIDEO_OMAP2
 
 config VIDEO_OMAP3
 	tristate "OMAP 3 Camera support (EXPERIMENTAL)"
-	depends on OMAP_IOVMM && VIDEO_V4L2 && I2C && VIDEO_V4L2_SUBDEV_API && ARCH_OMAP3 && EXPERIMENTAL
+	depends on OMAP_IOVMM && VIDEO_V4L2 && I2C && VIDEO_V4L2_SUBDEV_API && ARCH_OMAP3 && EXPERIMENTAL && !ARCH_MULTIPLATFORM
 	---help---
 	  Driver for an OMAP 3 camera controller.
 
diff --git a/drivers/media/rc/Kconfig b/drivers/media/rc/Kconfig
index 79ba242..19f3563 100644
--- a/drivers/media/rc/Kconfig
+++ b/drivers/media/rc/Kconfig
@@ -291,7 +291,7 @@  config IR_TTUSBIR
 
 config IR_RX51
 	tristate "Nokia N900 IR transmitter diode"
-	depends on OMAP_DM_TIMER && LIRC
+	depends on OMAP_DM_TIMER && LIRC && !ARCH_MULTIPLATFORM
 	---help---
 	   Say Y or M here if you want to enable support for the IR
 	   transmitter diode built in the Nokia N900 (RX51) device.
diff --git a/drivers/staging/tidspbridge/Kconfig b/drivers/staging/tidspbridge/Kconfig
index 0dd479f..60848f1 100644
--- a/drivers/staging/tidspbridge/Kconfig
+++ b/drivers/staging/tidspbridge/Kconfig
@@ -4,7 +4,7 @@ 
 
 menuconfig TIDSPBRIDGE
 	tristate "DSP Bridge driver"
-	depends on ARCH_OMAP3
+	depends on ARCH_OMAP3 && !ARCH_MULTIPLATFORM
 	select OMAP_MBOX_FWK
 	help
 	  DSP/BIOS Bridge is designed for platforms that contain a GPP and