So select GENERIC_HARDIRQS_NO__DO_IRQ to disable it.
Signed-off-by: NMichael Ellerman <michael@ellerman.id.au>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Don't call __do_IRQ() directly in gatwick_action(), use
generic_handle_irq().
Signed-off-by: NMichael Ellerman <michael@ellerman.id.au>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

We should no longer have any irq code that needs __do_IRQ(), so
remove the fallback to __do_IRQ().
Signed-off-by: NMichael Ellerman <michael@ellerman.id.au>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

The guts of do_IRQ() isn't really the right place to be documenting
the ppc_md.get_irq() interface. So move the comment into machdep.h
Signed-off-by: NMichael Ellerman <michael@ellerman.id.au>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Makes do_IRQ() shorter and clearer.
Signed-off-by: NMichael Ellerman <michael@ellerman.id.au>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Rather than a giant ifdef in the body of do_IRQ(), including a
dangling else, move the irq stack logic into a separate routine and
do the ifdef there.
Signed-off-by: NMichael Ellerman <michael@ellerman.id.au>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

ps3 has 4 ipis per cpu and can use the new smp_request_message_ipi to
reduce path length when receiving an ipi.

This has the side effect of setting IRQF_PERCPU.
Signed-off-by: NMilton Miller <miltonm@bga.com>
Acked-by: NGeoff Levand <geoffrey.levand@am.sony.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Adds support for the "unused" page hint which can be used in shared
memory partitions to flag pages not in use, which will then be stolen
before active pages by the hypervisor when memory needs to be moved to
LPARs in need of additional memory.  Failure to mark pages as 'unused'
makes the LPAR slower to give up unused memory to other partitions.

This adds the kernel parameter 'cmo_free_hint' to disable this
functionality.
Signed-off-by: NBrian King <brking@linux.vnet.ibm.com>
Signed-off-by: NRobert Jennings <rcj@linux.vnet.ibm.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

mpic_find() was overloaded to do two things, finding the mpic instance
for a given interrupt and returning if it's an IPI. Instead we introduce
mpic_is_ipi() and simplify mpic_find() to just return the mpic instance

Also silences the warning:
arch/powerpc/sysdev/mpic.c: In function 'mpic_irq_set_priority':
arch/powerpc/sysdev/mpic.c:1382: warning: 'is_ipi' may be used uninitialized in this function
Signed-off-by: NTony Breeds <tony@bakeyournoodle.com>
Acked-by: NMichael Ellerman <michael@ellerman.id.au>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Now that leds-gpio is a proper OF platform driver, the Warp can use
the leds-gpio driver rather than the old out-of-kernel driver.

One side-effect is the leds-gpio driver always turns the leds off
while the old driver left them alone. So we have to set them back to
the correct settings.
Signed-off-by: NSean MacLennan <smaclennan@pikatech.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Previouslly we just always set the inbound window to 2G.  This was
broken for systems with >2G.  If a system has >=4G we will need
SWIOTLB support to handle that case.

We now allocate PCICSRBAR/PEXCSRBAR right below the lowest PCI outbound
address for MMIO or the 4G boundary (if the lowest PCI address is above
4G).
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

The P2020 is a dual e500v2 core based SOC with:
* 3 PCIe controllers
* 2 General purpose DMA controllers
* 2 sRIO controllers
* 3 eTSECS
* USB 2.0
* SDHC
* SPI, I2C, DUART
* enhanced localbus
* and optional Security (P2020E) security w/XOR acceleration

The p2020 DS reference board is pretty similar to the existing MPC85xx
DS boards and has a ULI 1575 connected on one of the PCIe controllers.
Signed-off-by: NTed Peters <Ted.Peters@freescale.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Instead of fixed address in old code.
Signed-off-by: NLi Yang <leoli@freescale.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

We we build with resource_size_t as a 64-bit quantity we get:

arch/powerpc/sysdev/fsl_rio.c: In function 'fsl_rio_setup':
arch/powerpc/sysdev/fsl_rio.c:1029: warning: format '%08x' expects type 'unsigned int', but argument 4 has type 'resource_size_t'
arch/powerpc/sysdev/fsl_rio.c:1029: warning: format '%08x' expects type 'unsigned int', but argument 5 has type 'resource_size_t'
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

In these configuration we expect to have large amounts of memory (> 4G)
and thus will bounce via swiotlb some region of PCI address space.

The outbound windows were wasting 512M of address space by leaving a
gap between the top of the outbound window and the 4G boundary.  By
moving the top of the outbound window up to the 4G boundary we can
reclaim the vast majority of the 512M (minus space needed for PEXCSRBAR)
and thus reduces the amount of memory we have to bounce.
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Its feasible based on how the PCI address map is setup that the region
of PCI address space used for MSIs differs for each PHB on the same SoC.

Instead of assuming that the address mappes to CCSRBAR 1:1 we read
PEXCSRBAR (BAR0) for the PHB that the given pci_dev is on.
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

The MPC8568/9 chips support MSIs on PCIe so no reason not to enable them.
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

For serial flash support we need to:

- Add QE Par IO Bank E device tree node, a GPIO from this bank is
  used for SPI chip-select line;
- Add serial-flash node;
- Add proper module alias into of/base.c.
Signed-off-by: NAnton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Select HAS_RAPIDIO symbol and add rio nodes for MPC8568E-MDS
and MPC8569E-MDS boards.
Signed-off-by: NAnton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Note that eSDHC and DUART0 are mutually exclusive on MPC8569E-MDS
boards. Default option is DUART0, so eSDHC is disabled by default.
U-Boot will fixup device tree if eSDHC should be used instead of
DUART0.
Signed-off-by: NAnton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

This patch fixes bogus reg = <> property in the localbus node,
and fixes interrupt property (should be "interrupts").

Also add node for NAND support.
Signed-off-by: NAnton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

fsl,exec-units-mask should be 0xbfe to include SNOW unit in
MPC8569E's security engine.
Signed-off-by: NAnton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

This patch adds PCI IDs for MPC8569 and MPC8569E processors,
plus adds appropriate quirks for these IDs, and thus makes
PCI-E actually work on MPC8569E-MDS boards.
Signed-off-by: NAnton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Between the addition of the ecm/mcm law nodes and the fact that the
get_immrbase() has been using the range property of the SoC to determine
the base address of CCSR space we no longer need the reg property at
the soc node level.  It has been ill specified and varied between device
trees to cover either the {e,m}cm-law node, some odd subset of CCSR
space or all of CCSR space.
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Add fsl,qe-num-riscs and fsl,qe-num-snums to all the devices trees which
have qe node.
Signed-off-by: NHaiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

The MPC8569 is similiar to the MPC8568.  It doubles the number of
QUICC Engine RISC cores from 2 to 4.  Removes eTSECs, TLU and adds
the eSDHC controller.
Signed-off-by: NHaiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

The latest QE chip may have more Serial Number(SNUM)s of thread to use. We
will get the number of SNUMs from device tree by reading the new property
"fsl,qe-num-snums", and set 28 as the default number of SNUMs so that it is
compatible with the old QE chips' device trees which don't have this new
property. The macro QE_NUM_OF_SNUM is defined as the maximum number in QE
snum table which is 256.

Also we update the snum_init[] array with 18 more new SNUMs which are
confirmed to be useful on new chip.
Signed-off-by: NHaiying Wang <Haiying.Wang@freescale.com>
Acked-by: NTimur Tabi <timur@freescale.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Change the RISC allocation to macros instead of enum, add function to read
the number of risc engines from the new property "fsl,qe-num-riscs" under
the qe node in dts. Add new property "fsl,qe-num-riscs" description in
qe.txt
Signed-off-by: NHaiying Wang <Haiying.Wang@freescale.com>
Acked-by: NTimur Tabi <timur@freescale.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Keep an unique machine def for the MPC8568 MDS board to handle some
subtle differences between the future MDS board. Also set the bcsrs in
setup_arch() only for mpc8568_mds because other mds has different bcsr
settings.
Signed-off-by: NHaiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Removed the need for asm/mpc86xx.h as it was only used in mpc86xx_smp.c
and just moved the defines it cared about into there.  Also fixed up
the ioremap to only map the one 4k page we need access to and to iounmap
when we are done.
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Kill of some old defines and macros that we no longer use like
CPM_MAP_ADDR and CPM_IRQ_OFFSET.
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Also, convert them to resource_size_t (which is unsigned long
on 64-bit, so it's not a change there).

We will be using these on fsl 32b to indicate the start and size
address of memory that the pci controller can actually reach - this
is needed to determine if an address requires bounce buffering.  For
now, initialize them to a standard value; in the near future, the
value will be calculated based on how the inbound windows are
programmed.
Signed-off-by: NBecky Bruce <beckyb@kernel.crashing.org>
Acked-by: NBen Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

The rstcr register mapping code was written sometime ago before
of_iomap() existed.  We can use it and clean up the code a bit
and get rid of one user of get_immrbase() in the process.
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

The new dts places most of the devices in physical address space
above 32-bits, which allows us to have more than 4GB of RAM present.
Signed-off-by: NBecky Bruce <beckyb@kernel.crashing.org>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

Refactor the check to determine if the quirk is applicable to the boards
into one inline function so we only have to change one place to add more
boards that the quirks might be applicable to.

Also removed a warning related to unused temp variable.
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

The cell-index property isn't used on PCI nodes and is ill defined.
Remove it for now and if someone comes up with a good reason and
consistent definition for it we can add it back
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

It's still in the git history if anyone wants it.
Signed-off-by: NMichael Ellerman <michael@ellerman.id.au>
Signed-off-by: NKumar Gala <galak@kernel.crashing.org>

A couple of issues crept in since about 2.6.27 related to accessing PCI
device ROMs on various powerpc machines.

First, historically, we don't allocate the ROM resource in the resource
tree. I'm not entirely certain of why, I susepct they often contained
garbage on x86 but it's hard to tell. This causes the current generic
code to always call pci_assign_resource() when trying to access the said
ROM from sysfs, which will try to re-assign some new address regardless
of what the ROM BAR was already set to at boot time. This can be a
problem on hypervisor platforms like pSeries where we aren't supposed
to move PCI devices around (and in fact probably can't).

Second, our code that generates the PCI tree from the OF device-tree
(instead of doing config space probing) which we mostly use on pseries
at the moment, didn't set the (new) flag IORESOURCE_SIZEALIGN on any
resource. That means that any attempt at re-assigning such a resource
with pci_assign_resource() would fail due to resource_alignment()
returning 0.

This fixes this by doing these two things:

 - The code that calculates resource flags based on the OF device-node
is improved to set IORESOURCE_SIZEALIGN on any valid BAR, and while at
it also set IORESOURCE_READONLY for ROMs since we were lacking that too

 - We now allocate ROM resources as part of the resource tree. However
to limit the chances of nasty conflicts due to busted firmwares, we
only do it on the second pass of our two-passes allocation scheme,
so that all valid and enabled BARs get precedence.

This brings pSeries back the ability to access PCI ROMs via sysfs (and
thus initialize various video cards from X etc...).
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>