[PATCH] arm: omap: introduce OMAP MCOP board file

[PATCH] arm: omap: introduce OMAP MCOP board file

[Felipe Balbi]

From: Michael Fillinger <m-fillinger@ti.com>

MCOP is an FPGA-based Silicon Validation platform
which is used to test particular IPs inside OMAP
before we have a real ASIC.

Signed-off-by: Michael Fillinger <m-fillinger@ti.com>

[ balbi at ti.com : few cleanups here an there and also
	removal of some unnecessary code. ]

Signed-off-by: Felipe Balbi <balbi@ti.com>
---
 arch/arm/mach-omap2/Kconfig                  |    7 +++
 arch/arm/mach-omap2/Makefile                 |    1 +
 arch/arm/mach-omap2/board-mcop.c             |   60 ++++++++++++++++++++++++++
 arch/arm/plat-omap/include/plat/uncompress.h |    1 +
 4 files changed, 69 insertions(+), 0 deletions(-)
 create mode 100644 arch/arm/mach-omap2/board-mcop.c

diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig
index 1a2cf62..6f5699f 100644
--- a/arch/arm/mach-omap2/Kconfig
+++ b/arch/arm/mach-omap2/Kconfig
@@ -116,6 +116,13 @@ config MACH_OMAP_H4
 	select OMAP_PACKAGE_ZAF
 	select OMAP_DEBUG_DEVICES
 
+config MACH_OMAP_MCOP
+	bool "OMAP MCOP Board"
+	depends on ARCH_OMAP2420
+	default y
+#	select OMAP_PACKAGE_ZAF
+#	select OMAP_DEBUG_DEVICES
+
 config MACH_OMAP_APOLLON
 	bool "OMAP 2420 Apollon board"
 	depends on ARCH_OMAP2420
diff --git a/arch/arm/mach-omap2/Makefile b/arch/arm/mach-omap2/Makefile
index 1c0c2b0..6db99c4 100644
--- a/arch/arm/mach-omap2/Makefile
+++ b/arch/arm/mach-omap2/Makefile
@@ -153,6 +153,7 @@ endif
 # Specific board support
 obj-$(CONFIG_MACH_OMAP_GENERIC)		+= board-generic.o
 obj-$(CONFIG_MACH_OMAP_H4)		+= board-h4.o
+obj-$(CONFIG_MACH_OMAP_MCOP)		+= board-mcop.o
 obj-$(CONFIG_MACH_OMAP_2430SDP)		+= board-2430sdp.o \
 					   hsmmc.o
 obj-$(CONFIG_MACH_OMAP_APOLLON)		+= board-apollon.o
diff --git a/arch/arm/mach-omap2/board-mcop.c b/arch/arm/mach-omap2/board-mcop.c
new file mode 100644
index 0000000..b9fe1a4
--- /dev/null
+++ b/arch/arm/mach-omap2/board-mcop.c
@@ -0,0 +1,60 @@
+/*
+ * board-mcop.c - OMAP pre-Silicon Validation Platform
+ *
+ * Copyright (C) 2011 Texas Instruments Incorporated - http://www.ti.com
+ * Author: Michael Fillinger <m-fillinger@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/io.h>
+
+#include <asm/mach-types.h>
+#include <asm/mach/arch.h>
+#include <asm/mach/map.h>
+
+#include <plat/common.h>
+#include <plat/irqs.h>
+
+#include <mach/hardware.h>
+
+static void __init omap_mcop_init_irq(void)
+{
+	omap2_init_common_infrastructure();
+	omap2_init_common_devices(NULL, NULL);
+	omap_init_irq();
+}
+
+static void __init omap_mcop_init(void)
+{
+	omap_serial_init();
+}
+
+static void __init omap_mcop_map_io(void)
+{
+	omap2_set_globals_242x();
+	omap242x_map_common_io();
+}
+
+MACHINE_START(OMAP_MCOP, "OMAP MCOP Board")
+	/* Maintainer: Michael Fillinger <m-fillinger@ti.com> */
+
+	/*
+	 * REVISIT What values to put here so that it works
+	 * on all versions of MCOP
+	 *
+	 * .phys_io	= 0x48000000,
+	 * .io_pg_offst	= ((0xfa000000) >> 18) & 0xfffc,
+	 */
+
+	.boot_params	= 0x80000100,
+	.map_io		= omap_mcop_map_io,
+	.reserve	= omap_reserve,
+	.init_irq	= omap_mcop_init_irq,
+	.init_machine	= omap_mcop_init,
+	.timer		= &omap_timer,
+MACHINE_END
diff --git a/arch/arm/plat-omap/include/plat/uncompress.h b/arch/arm/plat-omap/include/plat/uncompress.h
index ad98b85..dbedd6e 100644
--- a/arch/arm/plat-omap/include/plat/uncompress.h
+++ b/arch/arm/plat-omap/include/plat/uncompress.h
@@ -129,6 +129,7 @@ static inline void __arch_decomp_setup(unsigned long arch_id)
 		DEBUG_LL_OMAP2(1, omap_2430sdp);
 		DEBUG_LL_OMAP2(1, omap_apollon);
 		DEBUG_LL_OMAP2(1, omap_h4);
+		DEBUG_LL_OMAP2(1, omapmcop);
 
 		/* omap2 based boards using UART3 */
 		DEBUG_LL_OMAP2(3, nokia_n800);
-- 
1.7.4.1.343.ga91df

[PATCH] arm: omap: introduce OMAP MCOP board file

[Felipe Balbi]

On Tue, Apr 05, 2011 at 12:32:50PM +0300, Felipe Balbi wrote:
> From: Michael Fillinger <m-fillinger@ti.com>
> 
> MCOP is an FPGA-based Silicon Validation platform
> which is used to test particular IPs inside OMAP
> before we have a real ASIC.
> 
> Signed-off-by: Michael Fillinger <m-fillinger@ti.com>
> 
> [ balbi at ti.com : few cleanups here an there and also
> 	removal of some unnecessary code. ]
> 
> Signed-off-by: Felipe Balbi <balbi@ti.com>

I should have RFCed this one, but bear with me for a minute.

This is just the bare minimum board-file for MCOP, there's still a bunch
of changes needed to get it actually booting. The attached diff shows
many of them.

Now, we don't want to send that patch upstream for obvious reasons and
we also don't want to add ifdefs to clock data files as that would break
multi-omap. What do you guys suggest ? How should we handle detection of
MCOP so that we choose correct HWMODs and clock data files for it ?

I don't think Linus will like if we add yet another hwmod + clk data
file just for MCOP, so we need to re-use what's in tree.

-- 
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mcop_remaining.diff
Type: text/x-diff
Size: 30349 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20110405/ea0208f9/attachment-0001.bin>

[PATCH] arm: omap: introduce OMAP MCOP board file

[Fillinger, Michael]

On Tue, Apr 05, 2011 at 11:37 AM +0300, Felipe Balbi wrote:
> On Tue, Apr 05, 2011 at 12:32:50PM +0300, Felipe Balbi wrote:
> > From: Michael Fillinger <m-fillinger@ti.com>
> >
> > MCOP is an FPGA-based Silicon Validation platform
> > which is used to test particular IPs inside OMAP
> > before we have a real ASIC.
> >
> > Signed-off-by: Michael Fillinger <m-fillinger@ti.com>
> >
> > [ balbi at ti.com : few cleanups here an there and also
> >     removal of some unnecessary code. ]
> >
> > Signed-off-by: Felipe Balbi <balbi@ti.com>
>
> I should have RFCed this one, but bear with me for a minute.
>
> This is just the bare minimum board-file for MCOP, there's still a bunch
> of changes needed to get it actually booting. The attached diff shows
> many of them.
>
> Now, we don't want to send that patch upstream for obvious reasons and
> we also don't want to add ifdefs to clock data files as that would break
> multi-omap. What do you guys suggest ? How should we handle detection of
> MCOP so that we choose correct HWMODs and clock data files for it ?
>
> I don't think Linus will like if we add yet another hwmod + clk data
> file just for MCOP, so we need to re-use what's in tree.
>
> --
> Balbi

What you need to know about the MCOP system, based on OMAP2420, is that there is no PRCM and no complex clock trees, or power management. There is just 1 main driver clock. As such the "clock tree" can be extremely simplified.

Michael


Texas Instruments France SA, 821 Avenue Jack Kilby, 06270 Villeneuve Loubet. 036 420 040 R.C.S Antibes. Capital de EUR 753.920

[PATCH] arm: omap: introduce OMAP MCOP board file

[Russell King - ARM Linux]

On Tue, Apr 05, 2011 at 12:36:57PM +0300, Felipe Balbi wrote:
> On Tue, Apr 05, 2011 at 12:32:50PM +0300, Felipe Balbi wrote:
> > From: Michael Fillinger <m-fillinger@ti.com>
> > 
> > MCOP is an FPGA-based Silicon Validation platform
> > which is used to test particular IPs inside OMAP
> > before we have a real ASIC.
> > 
> > Signed-off-by: Michael Fillinger <m-fillinger@ti.com>
> > 
> > [ balbi at ti.com : few cleanups here an there and also
> > 	removal of some unnecessary code. ]
> > 
> > Signed-off-by: Felipe Balbi <balbi@ti.com>
> 
> I should have RFCed this one, but bear with me for a minute.
> 
> This is just the bare minimum board-file for MCOP, there's still a bunch
> of changes needed to get it actually booting. The attached diff shows
> many of them.
> 
> Now, we don't want to send that patch upstream for obvious reasons and
> we also don't want to add ifdefs to clock data files as that would break
> multi-omap. What do you guys suggest ? How should we handle detection of
> MCOP so that we choose correct HWMODs and clock data files for it ?
> 
> I don't think Linus will like if we add yet another hwmod + clk data
> file just for MCOP, so we need to re-use what's in tree.

I'd suggest holding fire on new stuff.

We *absolutely* *must* show that we're taking Linus' complaint seriously
and make headway towards consolidating some of the code.  I don't see that
activity as optional.

I've now made the decision (as I mentioned I may do in the thread) that
for the next merge window I'm only taking consolidations and bug fixes
through my tree, and I encourage everyone else to do the same.  At the
moment, I'm planning for this up until the next merge window, but if
sufficient progress hasn't been done, I'll extend it thereafter.

In other words, no new platform code and no new platform class code.

The longer it takes to get the consolidation effort producing results, the
longer we will have to keep the restriction in place, and the more painful
it'll be for people who want to have their platforms merged.

So I hope *no* one is thinking "this isn't my problem, someone else will
solve it" - if you're thinking that then we're doomed.

[PATCH] arm: omap: introduce OMAP MCOP board file

[Felipe Balbi]

Hi,

On Tue, Apr 05, 2011 at 10:51:19AM +0100, Russell King - ARM Linux wrote:
> > I don't think Linus will like if we add yet another hwmod + clk data
> > file just for MCOP, so we need to re-use what's in tree.
> 
> I'd suggest holding fire on new stuff.
> 
> We *absolutely* *must* show that we're taking Linus' complaint seriously
> and make headway towards consolidating some of the code.  I don't see that
> activity as optional.
> 
> I've now made the decision (as I mentioned I may do in the thread) that
> for the next merge window I'm only taking consolidations and bug fixes
> through my tree, and I encourage everyone else to do the same.  At the
> moment, I'm planning for this up until the next merge window, but if
> sufficient progress hasn't been done, I'll extend it thereafter.
> 
> In other words, no new platform code and no new platform class code.
> 
> The longer it takes to get the consolidation effort producing results, the
> longer we will have to keep the restriction in place, and the more painful
> it'll be for people who want to have their platforms merged.
> 
> So I hope *no* one is thinking "this isn't my problem, someone else will
> solve it" - if you're thinking that then we're doomed.

Ok Russell, we will hold on to this patch and try to help as we can on
the code consolidation effort.

Let's hope to get this sorted out soon as it would be really great for
us to start pre-silicon validation with mainline kernel with little to
no effort needed :-D

-- 
balbi

[PATCH] arm: omap: introduce OMAP MCOP board file

[Tony Lindgren]

* Russell King - ARM Linux <linux@arm.linux.org.uk> [110405 02:48]:
> 
> I'd suggest holding fire on new stuff.
> 
> We *absolutely* *must* show that we're taking Linus' complaint seriously
> and make headway towards consolidating some of the code.  I don't see that
> activity as optional.

I agree.
 
> I've now made the decision (as I mentioned I may do in the thread) that
> for the next merge window I'm only taking consolidations and bug fixes
> through my tree, and I encourage everyone else to do the same.  At the
> moment, I'm planning for this up until the next merge window, but if
> sufficient progress hasn't been done, I'll extend it thereafter.

OK
 
> In other words, no new platform code and no new platform class code.
> 
> The longer it takes to get the consolidation effort producing results, the
> longer we will have to keep the restriction in place, and the more painful
> it'll be for people who want to have their platforms merged.
> 
> So I hope *no* one is thinking "this isn't my problem, someone else will
> solve it" - if you're thinking that then we're doomed.

Yes we need more people to pitch in to the consolidation process,
so everybody please help where possible.

Regards,

Tony