Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

In the NUMA or memory hot-add case where system memory has been
partitioned up, we immediately run in to a situation where the existing
PMB entry doesn't cover the new range (primarily as a result of the entry
size being shrunk to match the node size early in the initialization). In
order to fix this up it's necessary to preload a PMB mapping for the new
range prior to activation in order to circumvent reset by MMU.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

The minimum section size for the PMB is 16M, so just always error
out early if the specified size is too small. This permits us to
unconditionally call in to pmb_bolt_mapping() with variable sizes
without wasting a TLB and cache flush for the range.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This reworks much of the bootmem setup and initialization code allowing
us to get rid of duplicate work between the NUMA and non-NUMA cases. The
end result is that we end up with a much more flexible interface for
supporting more complex topologies (fake NUMA, highmem, etc, etc.) which
is entirely LMB backed. This is an incremental step for more NUMA work as
well as gradually enabling migration off of bootmem entirely.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This reworks the memory limit handling to tie in through the available
LMB infrastructure. This requires a bit of reordering as we need to have
all of the LMB reservations taken care of prior to establishing the
limits.

While we're at it, the crash kernel reservation semantics are reworked
so that we allocate from the bottom up and reduce the risk of having
to disable the memory limit due to a clash with the crash kernel
reservation.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This plugs in a memory init callback in the machvec to permit boards to
wire up various bits of memory directly in to LMB. A generic machvec
implementation is provided that simply wraps around the normal
Kconfig-derived memory start/size.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This bumps up the extra LMB reservations in ordering so that they're
accounted for prior to iterating over the region list. This ensures that
reservations are visible both within the LMB and bootmem context.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This switches over from bootmem -> LMB for the initrd area reservation.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This migrates the crash kernel handling off of bootmem and over to LMB.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Various boot loaders go to various extents to thwart the initrd detection
logic (mostly on account of not being able to be bothered with adhering
to the established boot ABI), so we make the detection logic a bit more
robust. This makes it possible to work around the SDK7786's firmware's
attempts to thwart compressed image booting. Victory is mine.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

arch/sh/kernel/smp.c:164: error: conflicting types for 'native_cpu_disable'
/home/matt/src/kernels/sh-2.6/arch/sh/include/asm/smp.h:48: error: previous declaration of 'native_cpu_disable' was here
Signed-off-by: NMatt Fleming <matt@console-pimps.org>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Add CONFIG_VIRTUALIZATION to the SH architecture
and include the virtio code there. Used to enable
the virtio drivers under QEMU.
Signed-off-by: NMagnus Damm <damm@opensource.se>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

The UP dependency was inherited from ARM, which seems to have run in to
it due to the stacktrace code not being available for SMP in certain
cases, as we don't have this particular limitation there is no specific
need to block on the SMP dependency.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This wires up CPU hotplug for SH-X3 SMP CPUs. Presently only secondary
cores can be hotplugged given that the boot CPU has to contend with the
broadcast timer. When real local timers are implemented this restriction
can be lifted.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This adds preliminary support for CPU hotplug for SH SMP systems.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This provides a cache of the secondary CPUs idle loop for the cases where
hotplug simply enters a low power state instead of resetting or powering
off the core.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

smp_store_cpu_info() is presently flagged as __init, but is called by
start_secondary() which is __cpuinit, fix it up.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This provides percpu CPU states in preparation for CPU hotplug and the
associated notifier chains.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This iterates over the maximum number of CPUs we plan to support and
makes sure they're all set in the present CPU map.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This converts from cpu_set() for the online map to set_cpu_online().
The two online map modifiers were the last remaining manual map
manipulation bits, with this in place everything now goes through
cpumask.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

With the platform ops migration, the definitions still need to be
included in the CONFIG_SMP=n case, so make the asm/smp.h include
explicit.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

When registering CPUs in the topology initialization ensure that all of
the present CPUs are flagged as hotpluggable.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Now that compressed image loading is possible for sdk7786, drop the
vmlinux.bin default image target and update the defconfig accordingly.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This patch updates the sh7751 pci code to handle io ports
correctly. The code is based on the sh7788x implementation.

Tested on a R2D-1 board with CONFIG_8139TOO_PIO=y.
Signed-off-by: NMagnus Damm <damm@opensource.se>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Update the R2D defconfigs to bump up the maximum
number of SCIF ports on the system.

Fixes a broken serial console regression added
by cd5f1076.
Reported-by: NShin-ichiro KAWASAKI <kawasaki@juno.dti.ne.jp>
Signed-off-by: NMagnus Damm <damm@opensource.se>
Tested-by: NAlexandre Courbot <alex@dcl.info.waseda.ac.jp>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

CHCR_TS_HIGH_SHIFT is defined as a shift of TS high bits in CHCR register,
relative to low bits. The TS_INDEX2VAL() macro has to take this into account.
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Lists of DMA channels and slaves are not changed, make them constant. Besides,
SH7724 channel and slave configuration of both DMA controllers is identical,
remove the extra copy of the configuration data.
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

As explained in commit 1c0fe6e3, we want to call the architecture independent
oom killer when getting an unexplained OOM from handle_mm_fault, rather than
simply killing current.

Cc: linux-sh@vger.kernel.org
Cc: linux-arch@vger.kernel.org
Signed-off-by: NNick Piggin <npiggin@suse.de>
Acked-by: NDavid Rientjes <rientjes@google.com>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Previously we were masking the PMB DATA array values with the value of
__MEMORY_START | PMB_V, which misses some PFN bits off the mask.
Signed-off-by: NMatt Fleming <matt@console-pimps.org>

There is a logic error in pmb_merge() that means we will incorrectly try
to merge two 128MB PMB mappings into one mapping. However, 256MB isn't a
valid PMB map size and pmb_merge() will actually drop the second 128MB
mapping.

This patch allows my SDK7786 board to boot when configured with
CONFIG_MEMORY_SIZE=0x10000000.
Signed-off-by: NMatt Fleming <matt@console-pimps.org>

We can't necessarily use the P1SEG region to access RAM when running in
32BIT mode, so use CONFIG_MEMORY_START as the base address.
Signed-off-by: NMatt Fleming <matt@console-pimps.org>

When running in 32BIT mode the P1SEG region doesn't necessarily provide
a window onto RAM (it depends how the bootloader setup the PMB). The
correct location to place the decompressed kernel is the physical
address of _text.
Signed-off-by: NMatt Fleming <matt@console-pimps.org>

This patch defines ___pa and ___va which return the physical and virtual
address of an address, respectively. These macros are suitable for
calling from assembly because they don't include the C casts required by
__pa and __va.
Signed-off-by: NMatt Fleming <matt@console-pimps.org>

This is a standalone version of VMware Balloon driver.  Ballooning is a
technique that allows hypervisor dynamically limit the amount of memory
available to the guest (with guest cooperation).  In the overcommit
scenario, when hypervisor set detects that it needs to shuffle some
memory, it instructs the driver to allocate certain number of pages, and
the underlying memory gets returned to the hypervisor.  Later hypervisor
may return memory to the guest by reattaching memory to the pageframes and
instructing the driver to "deflate" balloon.

We are submitting a standalone driver because KVM maintainer (Avi Kivity)
expressed opinion (rightly) that our transport does not fit well into
virtqueue paradigm and thus it does not make much sense to integrate with
virtio.

There were also some concerns whether current ballooning technique is the
right thing.  If there appears a better framework to achieve this we are
prepared to evaluate and switch to using it, but in the meantime we'd like
to get this driver upstream.

We want to get the driver accepted in distributions so that users do not
have to deal with an out-of-tree module and many distributions have
"upstream first" requirement.

The driver has been shipping for a number of years and users running on
VMware platform will have it installed as part of VMware Tools even if it
will not come from a distribution, thus there should not be additional
risk in pulling the driver into mainline.  The driver will only activate
if host is VMware so everyone else should not be affected at all.
Signed-off-by: NDmitry Torokhov <dtor@vmware.com>
Cc: Avi Kivity <avi@redhat.com>
Cc: Jeremy Fitzhardinge <jeremy@goop.org>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This adds support for Memory24, Memory32, and Memory32Fixed descriptors in
PCI host bridge _CRS.

I experimentally determined that Windows (2008 R2) accepts these descriptors
and treats them as windows that are forwarded to the PCI bus, e.g., if
it finds any PCI devices with BARs outside the windows, it moves them into
the windows.

I don't know whether any machines actually use these descriptors in PCI
host bridge _CRS methods, but if any exist and they're new enough that we
automatically turn on "pci=use_crs", they will work with Windows but not
with Linux.

Here are the details: https://bugzilla.kernel.org/show_bug.cgi?id=15817Signed-off-by: NBjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Commit "timekeeping: Fix clock_gettime vsyscall time warp" (0696b711)
introduced the new parameter "mult" to update_vsyscall(). This parameter
contains the internal NTP adjusted clock multiplier.

The s390x vdso did not use this adjusted multiplier.  Instead, it used
the constant clock multiplier for gettimeofday() and clock_gettime()
variants.  This may result in observable time warps as explained in
commit 0696b711.

Make the NTP adjusted clock multiplier available to the s390x vdso
implementation and use it for time calculations.

Cc: <stable@kernel.org>
Signed-off-by: NHendrik Brueckner <brueckner@linux.vnet.ibm.com>
Signed-off-by: NMartin Schwidefsky <schwidefsky@de.ibm.com>

Reenable multiple subchannel sets after hibernation,
prior to the device callbacks.

Cc: <stable@kernel.org>
Signed-off-by: NSebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: NMartin Schwidefsky <schwidefsky@de.ibm.com>

A 16-bit TSS is only 44 bytes long. So make sure to test for the correct
size on task switch.
Signed-off-by: NJan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: NAvi Kivity <avi@redhat.com>

Fix tcdrain on coldfire uarts.
Currently with coldfire uarts tcdrain returns without waiting for txempty,
because (tx)fifosize is 0.  Fix that and call uart_update_timeout when
setting the baud rate, otherwise tcdrain will wait for an half our :)
Also constify mcf_uart_ops.
Signed-off-by: NPhilippe De Muyter <phdm@macqel.be>
Signed-off-by: NGreg Ungerer <gerg@uclinux.org>