All that remains is the EBUS DMA programming library for
sparc64.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

These have no dependencies on the EBUS probing layer, the clients
setup the registers and all of those details.  The EBUS DMA layer
just programs and manages the DMA controller found in EBUS.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In order to convert EBUS drivers doing DMA into pure OF drivers,
we need the of_device->dev.archdata setup properly.

EBUS instances that can provide DMA for device nodes sit on PCI,
so detect and propagate the information there.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Also, delete the mini RTC driver, no longer used.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Based largely upon a patch by Krzysztof Helt <krzysztof.h1@poczta.fm>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Based largely upon a patch by Krzysztof Helt <krzysztof.h1@poczta.fm>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

As we convert the various by-hand sparc64 RTC drivers to use the
generic RTC framework and drivers, we need to keep the NTP
set_rtc_mmss() support via update_persistent_clock() working.

In the end, after all the RTC device cases are converted, this
local set_rtc_mmss() function will be deleted.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In order to make this week I also had to add an include
of linux/dma-mapping.h to asm/pci_32.h because drivers/pci/pci.c
really depends upon getting this header somehow.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

No longer used.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The individual SBUS IOMMU arch code now sets the IOMMU information
directly into the OF device objects.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add a helper function that, given a bus of_device node, propagates
all iommu, stc, and host_controller values down to the child nodes.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

32-bit sparc just needed it to register the ioport procfs bits, do this
via an arch_initcall() instead.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

No drivers or code uses this stuff any more, every driver has been
converted over to OF device probing.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

All the drivers use OF device objects now for this
information.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

IRQs are obtained by drivers from the of_device struct.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is in preparation for the subsequent asm/sbus.h removal.

Also, make these routines take a "struct device" or no
arguments, as appropriate.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

And all the SBUS dma interfaces are deleted.

A private implementation remains inside of the 32-bit sparc port which
exists only for the sake of the implementation of dma_*().
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is the first step in converting all the SBUS drivers
over to generic dma_*().
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This thing was completely pointless.

Just find the OF device in the parent of drivers that want to program
this device, and map the DMA regs inside such drivers too.

This also moves the dummy claim_dma_lock() and release_dma_lock()
implementation to floppy_32.h, which makes it handle this issue
just like floppy_64.h does.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Stephen Rothwell noticed that I committed an earlier version
of the patch that didn't have two things fixed:

1) irq_of_parse_and_map() should return "unsigned int" not "int"
   and it should return zero for "no irq"

2) irq_dispose_mapping() should be an inline function, not a macro,
   for type checking

With feedback and suggestions from Anton Vorontsov.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This also allows arch/sparc64/kernel/pci.c to be properly CONFIG_PCI
conditional compiled in the Makefile.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The way to do this varies by platform type and the exact memory
controller the cpu uses.

For Spitfire cpus we currently just use prom_getunumber() and hope
that works.

For Cheetah cpus we have a memory controller driver that can
compute this information.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

sys32_pause() is identical to the generically provided
sys_pause() in kernel/signal.c

Noticed by Christoph Hellwig.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Christoph Hellwig noticed that having both entry and exit
logic in one function no longer makes sense, and having
seperate ones simplifies things a lot.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This allows more OF layer code to be shared between powerpc and
sparc.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

On some platforms, the I2C controller is shared between the OS and
OBP.  OBP uses this I2C controller to access the EEPROM, and thus is
programmed when the kernel calls prom_setprop().

Wrap such calls with the new of_set_property_mutex.

Relevant I2C bus drivers can grab this mutex around top-level I2C
operations to provide the proper protection.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

First, lmb_enforce_memory_limit() interprets it's argument
(mostly, heh) as a size limit not an address limit.  So pass
the raw cmdline_memory_size value into it.  And we don't
need to check it against zero, lmb_enforce_memory_limit() does
that for us.

Next, free_initmem() needs special handling when the kernel
command line trims the available memory.  The problem case is
if the trimmed out memory is where the kernel image itself
resides.

When that memory is trimmed out, we don't add those physical
ram areas to the sparsemem active ranges, amongst other things.
Which means that this free_initmem() code will free up invalid
page structs, resulting in either crashes or hangs.

Just quick fix this by not freeing initmem at all if "mem="
was given on the boot command line.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If 'start' does not begin on a page boundary, we can overshoot
past 'end'.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Things like lockdep can try to do stack backtraces before
the irqstack blocks have been setup.  So don't try to match
their ranges so early on.

Also, remove unused variable in save_stack_trace().
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>