This adds the low level irq tracing hooks to the powerpc architecture
needed to enable full lockdep functionality.
This is partly based on Johannes Berg's initial version. I removed
the asm trampoline that isn't needed (thus improving performance) and
modified all sorts of bits and pieces, reworking most of the assembly,
etc...
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
This adds stacktrace support for powerpc, which will be needed for
lockdep.
Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Lockdep found out that we can occasionally take the device-tree
lock for reading from softirq time (from rtas_token called
by the rtas real time clock code called by the NTP code),
while we take it occasionally for writing without masking
interrupts. The combination of those two can thus deadlock.
While some of those cases of interrupt read lock could be fixed
(such as caching the RTAS tokens) I figured that taking the
lock for writing is so rare (device-tree modification) that we
may as well penalize that case and allow reading from interrupts.
Thus, this turns all the writers to take the lock with irqs
masked to avoid the situation.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] Make pci_bus_to_host()'s struct pci_bus * argument const
A) It's not modified and so it can be made const. const is good.
B) If one has a function that was given a const pci_bus pointer and you
want to get a pointer to its pci_controller, you'll get a warning from gcc
when you use pci_bus_to_host(). This is the right way to stop that
warning.
Signed-off-by: Trent Piepho <tpiepho@freescale.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
The xics code currently has a direct and lpar variant of
xics_host_map, the only difference being which irq_chip they use. If
we remember which irq_chip we're using we can combine these two
routines. That also allows us to have a single irq_host_ops instead
of two.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] Use pseries_setup_i8259_cascade() in pseries_mpic_init_IRQ()
pseries_mpic_init_IRQ() implements the same logic as the xics code did to
find the i8259 cascade irq. Now that we've pulled that logic out into
pseries_setup_i8259_cascade() we can use it in the mpic code.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] Turn xics_setup_8259_cascade() into a generic pseries_setup_i8259_cascade()
Remove the xics references from xics_setup_8259_cascade(), and merge the
good bits from the almost identical logic in pseries_mpic_init_IRQ().
Signed-off-by: Michael Ellerman <michael@ellerman.id.au> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] Move xics_setup_8259_cascade() into platforms/pseries/setup.c
The code in xics.c to setup the i8259 cascaded irq handler is not really
xics specific, so move it into setup.c - we will clean this up further in
a subsequent patch.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] Use asm-generic/bitops/find.h in bitops.h
Powerpc and ppc have some code in their bitops.h that is exactly the
same as asm-generic/bitops/find.h. Include this header instead of the
private implementation.
Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm> Signed-off-by: Paul Mackerras <paulus@samba.org>
Anton Vorontsov [Wed, 9 Apr 2008 13:59:25 +0000 (17:59 +0400)]
[POWERPC] 83xx: mpc8315 - fix USB UTMI Host setup
Currently USB Host isn't functional on the MPC8315E boards, for two
reasons as described below.
MPC8315 Reference Manual says:
"The USB DR unit must have the same clock ratio as the encryption core
unit, unless one of them has its clock disabled."
The encryption core also drives I2C clock, so it is enabled and is equal
to 01. That means USBDRCM should be 01 here.
Plus, according to MPC8315E-RDB schematics, USB unit consumes CLK_IN
clock from the 24.00MHz oscillator, which means we must adjust REFSEL
bits as well.
p.s.
Idially we should rework whole 83xx/usb.c code, in two steps:
1. Move SCCR code to the U-Boot;
2. Implement fsl,usb-clock property in the device tree, so usb.c could
decide what clock exactly to use on per-board basis.
Though, today we're not in a hurry since there is just one 8315e board
out there.
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Paul Gortmaker [Wed, 16 Apr 2008 17:53:07 +0000 (13:53 -0400)]
[POWERPC] 86xx: mpc86xx_hpcn - Temporarily accept old dts node identifier.
As suggested by Timur Tabi, we match on the old compat node ID for one
version and warn accordingly. If we don't do this, we plunge people who
try to use an old DTB into silent boot death with no clear indication of
what the problem is.
This patch should be removed at the beginning of the 2.6.27 dev cycle.
It is only meant to ease the transition in the short term.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Paul Gortmaker [Wed, 16 Apr 2008 17:53:06 +0000 (13:53 -0400)]
[POWERPC] 86xx: mark functions static, other minor cleanups
Cleanups as suggested by Stephen Rothwell and Dale Farnsworth, which
incudes marking a bunch of functions static and add a vendor prefix to
the compat node check for uniqueness.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Timur Tabi [Mon, 14 Apr 2008 15:43:38 +0000 (10:43 -0500)]
[POWERPC] Make rheap safe for spinlocks
The rheap allocation function, rh_alloc, could call kmalloc with GFP_KERNEL.
This can sleep, which means you couldn't hold a spinlock while called rh_alloc.
Change all kmalloc calls to use GFP_ATOMIC so that it won't sleep. This is
safe because only small blocks are allocated.
Signed-off-by: Timur Tabi <timur@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Kumar Gala [Wed, 16 Apr 2008 04:35:47 +0000 (23:35 -0500)]
[POWERPC] 85xx: Fix compile warning
arch/powerpc/platforms/85xx/mpc85xx_ads.c: In function ‘init_ioports’:
arch/powerpc/platforms/85xx/mpc85xx_ads.c:168: warning: initialization discards qualifiers from pointer target type
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Paul Gortmaker [Tue, 15 Apr 2008 22:41:31 +0000 (18:41 -0400)]
[POWERPC] 86xx: Add device tree source for Wind River SBC8641D
This adds in the device tree source for the SBC8641D, based
largely on the mpc8641_hpcn.dts. The biggest differences are
the lack of a complex IRQ mapping (since no Uli/i8259 cascade)
and the different layout of devices on the localbus node.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Paul Gortmaker [Tue, 15 Apr 2008 22:41:30 +0000 (18:41 -0400)]
[POWERPC] 86xx: Add support for Wind River SBC8641D board
This adds support for the Wind River SBC8641D board, based
largely on the mpc86xx_hpcn support. The biggest difference is
the lack of the Uli and the i8259 cascade, which simplifies things.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Scott Wood [Tue, 15 Apr 2008 16:02:31 +0000 (11:02 -0500)]
[POWERPC] cuboot-pq2: PCI fixes
1. Detect (and bail out on) more conditions that violate the
assumptions of the setup code -- we assume in such cases that the device
tree is correct and reflects what the firmware did.
2. The inbound memory mask calculation was wrong.
Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Kumar Gala [Wed, 9 Apr 2008 21:15:40 +0000 (16:15 -0500)]
[POWERPC] Make Book-E debug handling SMP safe
global_dbcr0 needs to be a per cpu set of save areas instead of a single
global on all processors.
Also, we switch to using DBCR0_IDM to determine if the user space app is
being debugged as its a more consistent way. In the future we should
support features like hardware breakpoint and watchpoints which will
have DBCR0_IDM set but not necessarily DBCR0_IC (single step).
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Anton Vorontsov [Tue, 11 Mar 2008 17:24:21 +0000 (20:24 +0300)]
[POWERPC] QE: immap_qe.h should include asm/io.h
Headers should include prototypes they use, otherwise build will
break if we use it without explicitly including io.h:
CC arch/powerpc/sysdev/qe_lib/gtm.o
In file included from include/asm/qe.h:20,
from arch/powerpc/sysdev/qe_lib/gtm.c:18:
include/asm/immap_qe.h: In function ‘immrbar_virt_to_phys’:
include/asm/immap_qe.h:480: error: implicit declaration of function ‘virt_to_phys’
make[2]: *** [arch/powerpc/sysdev/qe_lib/gtm.o] Error 1
make[1]: *** [arch/powerpc/sysdev/qe_lib] Error 2
gtm.c needs qe.h (which includes immap_qe.h) to use qe_get_brg_clk().
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Anton Vorontsov [Fri, 11 Apr 2008 17:03:40 +0000 (21:03 +0400)]
[POWERPC] fsl_lbc: implement few UPM routines
Freescale UPM can be used to adjust localbus timings or to generate
orbitrary, pre-programmed "patterns" on the external Localbus signals.
This patch implements few routines so drivers could work with UPMs in
safe and generic manner.
So far there is just one user of these routines: Freescale UPM NAND
driver.
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Anton Vorontsov [Tue, 11 Mar 2008 17:23:28 +0000 (20:23 +0300)]
[POWERPC] fsl_elbc_nand: factor out localbus defines
This is needed to support other localbus peripherals, such as
NAND on FSL UPM.
Signed-off-by: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Anton Vorontsov [Fri, 11 Apr 2008 16:06:54 +0000 (20:06 +0400)]
[POWERPC] QE: UCC nodes cleanup
- get rid of `model = "UCC"' in the ucc nodes
It isn't used anywhere, so remove it. If we'll ever need something
like this, we'll use compatible property instead.
- replace last occurrences of device-id with cell-index.
Drivers are modified for backward compatibility's sake.
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Acked-by: Timur Tabi <timur@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
[POWERPC] CPM2: Reset the CPM when early debugging is not enabled.
Similarly to what is done for PQ1-based platforms, this patch resets the
PQ2 Communication Processor Module in cpm2_reset() when early debugging is
not enabled. This helps avoiding conflicts when the boot loader configured
the CPM in an unexpected way.
Signed-off-by: Laurent Pinchart <laurentp@cse-semaphore.com> Acked-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
[POWERPC] ep8248e: Reference SMC parameter RAM base in the device tree.
This patch modifies the Embedded Planet EP8248E device tree to reference the
SMC paramater RAM base register instead of the parameter RAM allocated by the
boot loader.
The cpm_uart driver will allocate parameter RAM itself, making the serial port
initialisation independent of the boot loader.
The patch adds the parameter RAM allocated by the boot loader in the CPM muram
node, making it available to the kernel.
Signed-off-by: Laurent Pinchart <laurentp@cse-semaphore.com> Acked-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
[POWERPC] cpm-serial: Relocate CPM buffer descriptors and SMC parameter ram.
This patch relocates the buffer descriptors and the SMC parameter RAM at the
end of the first CPM muram chunk, as described in the device tree. This allows
device trees to stop excluding SMC parameter ram allocated by the boot loader
from the CPM muram node.
Signed-off-by: Laurent Pinchart <laurentp@cse-semaphore.com> Acked-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
[POWERPC] Add bootwrapper function to get virtual reg from the device tree.
This patch adds a new generic device tree processing function that retrieves
virtual reg addresses from the device tree to the bootwrapper code. It also
updates the bootwrapper code to use the new function.
dt_get_virtual_reg() retrieves the virtual reg addresses from the
"virtual-reg" property. If the property can't be found, it uses the "reg"
property and walks the tree to translate it to absolute addresses.
Signed-off-by: Laurent Pinchart <laurentp@cse-semaphore.com> Acked-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
[POWERPC] cpm_uart: Allocate DPRAM memory for SMC ports on CPM2-based platforms.
This patch allocates parameter RAM for SMC serial ports without relying on
previous initialisation by a boot loader or a wrapper layer.
SMC parameter RAM on CPM2-based platforms can be allocated anywhere in the
general-purpose areas of the dual-port RAM. The current code relies on the
boot loader to allocate a section of general-purpose CPM RAM and gets the
section address from the device tree.
This patch modifies the device tree address usage to reference the SMC
parameter RAM base pointer instead of a pre-allocated RAM section and
allocates memory from the CPM dual-port RAM when initialising the SMC port.
CPM1-based platforms are not affected.
Signed-off-by: Laurent Pinchart <laurentp@cse-semaphore.com> Acked-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Kumar Gala [Wed, 9 Apr 2008 11:06:11 +0000 (06:06 -0500)]
[POWERPC] Rework Book-E debug exception handling
The architecture allows for "Book-E" style debug interrupts to either go
to critial interrupts of their own debug interrupt level. To allow for
a dynamic kernel to support machines of either type we want to be able to
compile in the interrupt handling code for both exception levels.
Towards this goal we renamed the debug handling macros to specify the
interrupt level in their name (DEBUG_CRIT_EXCEPTION/DebugCrit and
DEBUG_DEBUG_EXCEPTION/DebugDebug).
Additionally, on the Freescale Book-e parts we expanded the exception
stacks to cover the maximum case of needing three exception stacks (normal,
machine check and debug).
There is some kernel text space optimization to be gained if a kernel is
configured for a specific Freescale implementation but we aren't handling
that now to allow for the single kernel image support.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Kumar Gala [Wed, 16 Apr 2008 13:19:36 +0000 (23:19 +1000)]
[POWERPC] Remove unused machine call outs
When we moved to arch/powerpc we actively tried to avoid using the
ppc_md.setup_io_mappings(). Currently no board ports use it so let's
remove it to avoid any new boards using it.
Also, remove early_serial_map() since we don't even have a call out for
it in arch/powerpc.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Michael Ellerman [Wed, 16 Apr 2008 03:51:49 +0000 (13:51 +1000)]
[POWERPC] Always add preferred consoles in platforms/pseries/lpar.c
There is logic in platforms/peries/lpars.c which checks if the user has
specified a console on the command line, and refrains from adding a
preferred console entry for the hvc/hvsi console if they have.
This trips up if you use "netconsole=foo" on the command line, and has
the result that you get _only_ the netconsole, because the hvc device is
never added as a preferred console. Worse still if you get the netconsole
configuration wrong somehow, you end up with no console at all.
As it turns out we don't need to worry about checking the command line.
If the user has specified "console=foo", then foo will be set as the
preferred console when the command line is parsed in start_kernel(), much
later than the pseries code, and so the latter setting will take effect.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] pseries/phyp dump: Reserve a variable amount of space at boot
This changes the way we calculate how much space to reserve for the
pHyp dump. Currently we reserve 256MB only. With this change, the
code first checks to see if an amount has been specified on the boot
command line with the "phyp_dump_reserve_size" option, and if so, uses
that much.
Otherwise it computes 5% of total ram and rounds it down to a multiple
of 256MB, and uses the larger of that or 256MB.
This is for large systems with a lot of memory (10GB or more). The
aim is to have more space available for the kernel on reboot on
machines with more resources. Although the dump will be collected
pretty fast and the memory released really early on allowing the
machine to have the full memory available, this alleviates any issues
that can be caused by having way too little memory on very very large
systems during those few minutes.
Signed-off-by: Manish Ahuja <mahuja@us.ibm.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Kumar Gala [Tue, 15 Apr 2008 19:52:30 +0000 (05:52 +1000)]
[POWERPC] Cleanup pgtable-ppc32.h
* Removed defines KERNEL_PGD_PTRS & USER_PGD_PTRS since they aren't
used anywhere
* Changed pmd_page macro to use pfn_to_page so we get proper behavior
if ARCH_PFN_OFFSET is set as well if we use a different memory model
on ppc32.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Kumar Gala [Tue, 15 Apr 2008 19:52:29 +0000 (05:52 +1000)]
[POWERPC] bootwrapper: Use physical address in PHDR for uImage
Now that we properly set the physical address in the program header of the
vmlinux ELF we can extract it to properly set the load and entry point for
u-boot uImages. Before we always hard coded the load & entry point to 0.
However there are situations that the kernel may be built with a non-zero
physical address.
We use objdump to extract the PHDR. We assume that there is only one
PHDR in the vmlinux of type LOAD.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Kumar Gala [Tue, 15 Apr 2008 19:52:27 +0000 (05:52 +1000)]
[POWERPC] Move phys_addr_t definition into asm/types.h
Moved phys_addr_t out of mmu-*.h and into asm/types.h so we can use it in
places that before would have caused recursive includes.
For example to use phys_addr_t in <asm/page.h> we would have included
<asm/mmu.h> which would have possibly included <asm/mmu-hash64.h> which
includes <asm/page.h>. Wheeee recursive include.
CONFIG_PHYS_64BIT is a bit counterintuitive in light of ppc64 systems
and thus the config option is only used for ppc32 systems with >32-bit
physical addresses (44x, 85xx, 745x, etc.).
Signed-off-by: Kumar Gala <galak@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Kumar Gala [Tue, 15 Apr 2008 19:52:26 +0000 (05:52 +1000)]
[POWERPC] Clean up some linker and symbol usage
* PAGE_OFFSET is not always the start of code, use _stext instead.
* grab PAGE_SIZE and KERNELBASE from asm/page.h like ppc64 does. Makes the
code a bit more common and provide a single place to manipulate the
defines for things like kdump.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Kumar Gala [Tue, 15 Apr 2008 19:52:22 +0000 (05:52 +1000)]
[POWERPC] Introduce lowmem_end_addr to distinguish from total_lowmem
total_lowmem represents the amount of low memory, not the physical
address that low memory ends at. If the start of memory is at 0 it
happens that total_lowmem can be used as both the size and the address
that lowmem ends at (or more specifically one byte beyond the end).
To make the code a bit more clear and deal with the case when the start of
memory isn't at physical 0, we introduce lowmem_end_addr that represents
one byte beyond the last physical address in the lowmem region.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Kumar Gala [Tue, 15 Apr 2008 19:52:21 +0000 (05:52 +1000)]
[POWERPC] Remove and replace uses of PPC_MEMSTART with memstart_addr
A number of users of PPC_MEMSTART (40x, ppc_mmu_32) can just always
use 0 as we don't support booting these kernels at non-zero physical
addresses since their exception vectors must be at 0 (or 0xfffx_xxxx).
For the sub-arches that support relocatable interrupt vectors
(book-e), it's reasonable to have memory start at a non-zero physical
address. For those cases use the variable memstart_addr instead of
the #define PPC_MEMSTART since the only uses of PPC_MEMSTART are for
initialization and in the future we can set memstart_addr at runtime
to have a relocatable kernel.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
David Woodhouse [Sun, 13 Apr 2008 16:52:38 +0000 (02:52 +1000)]
[POWERPC] Efika: Really, don't pretend to be CHRP
Fedora 9 works on Efika without the separate 'device-tree supplement',
thanks to the kernel's own fixups. With one exception -- because 'CHRP'
still appears on the 'machine:' line in /proc/cpuinfo, the installer
misdetects the platform and misconfigures yaboot, putting it into a PReP
boot partition instead of in the /boot filesystem where the Efika's
firmware could find it.
The kernel's fixups for Efika already correct one instance of 'chrp', in
the 'device_type' property. This fixes it in the 'CODEGEN,description'
property too, since that's what's exposed to userspace in /proc/cpuinfo.
Signed-off-by: David Woodhouse <dwmw2@infradead.org> Acked-by: Grant Likely <grant.likely@secretlab.ca> Signed-off-by: Paul Mackerras <paulus@samba.org>
Jochen Friedrich [Fri, 11 Apr 2008 19:22:35 +0000 (05:22 +1000)]
[POWERPC] i2c: OF helpers for the i2c API
This implements various helpers to support OF bindings for the i2c
API.
Signed-off-by: Jochen Friedrich <jochen@scram.de> Acked-by: David S. Miller <davem@davemloft.net> Acked-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Paul Mackerras <paulus@samba.org>
Anton Vorontsov [Fri, 11 Apr 2008 13:06:36 +0000 (23:06 +1000)]
[POWERPC] Implement support for the GPIO LIB API
This implements support for the GPIO LIB API. Two calls are still
unimplemented though: irq_to_gpio and gpio_to_irq.
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Acked-by: Grant Likely <grant.likely@secretlab.ca> Signed-off-by: Paul Mackerras <paulus@samba.org>
Anton Vorontsov [Fri, 11 Apr 2008 13:06:45 +0000 (23:06 +1000)]
[POWERPC] OF helpers for the GPIO API
This implements various helpers to support OF bindings for the GPIO
LIB API.
Previously this was PowerPC specific, but it seems this code isn't
arch-dependent anyhow, so let's place it into of/.
SPARC will not see this addition yet, real hardware seem to not use
GPIOs at all. But this might change:
http://www.leox.org/docs/faq_MLleon.html
"16-bit I/O port" sounds promising. :-)
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Acked-by: Grant Likely <grant.likely@secretlab.ca> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] properly declare onstack completion in iseries veth
The iSeries veth driver uses an on-stack struct completion that
it initializes using the COMPLETION_INITIALIZER instead of
COMPLETION_INITIALIZER_ONSTACK macro, causing problems with
lockdep.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
This fixes the handling of the preempt count when switching
interrupt stacks so that HW interrupt properly get the softirq
mask copied over from the previous stack.
It also initializes the softirq stack preempt_count to 0 instead
of SOFTIRQ_OFFSET, like x86, as __do_softirq() does the increment,
and we hit some lockdep checks if we have it twice.
That means we do run for a little while off the softirq stack
with the preempt-count set to 0, which could be deadly if we
try to take a softirq at that point, however we do so with
interrupts disabled, so I think we are ok.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Currently, we initialize the "current" pointer in the PACA (which
is used by the "current" macro in the kernel) before calling
setup_system(). That means that early_setup() is called with
current still "NULL" which is -not- a good idea. It happens to
work so far but breaks with lockdep when early code calls printk.
This changes it so that all PACAs are statically initialized with
__current pointing to the init task. For non-0 CPUs, this is fixed
up before use.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Ishizaki Kou [Mon, 17 Mar 2008 10:05:18 +0000 (21:05 +1100)]
[POWERPC] hvcbeat: Fix buffer manipulation
This fixes a potential bug at drivers/char/hvc_beat.c.
- hvc_put_term_char routine will decrement "rest" variable twice,
and forget to advance "buf" pointer by "nlen" bytes.
This bug was not hit previously because the output handler in
drivers/char/hvc_console.c splits given output into 16 bytes
at maximum.
Reported-by: Timur Tabi <timur@freescale.com> Signed-off-by: Kou Ishizaki <kou.ishizaki@toshiba.co.jp> Acked-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Paul Mackerras <paulus@samba.org>
Jerone Young [Fri, 11 Apr 2008 19:03:33 +0000 (05:03 +1000)]
[POWERPC] 4xx: Add idle wait support for 44x platforms
This changes the cpu_idle loop for 44x platforms to utilize the Wait Enable
feature of the CPU. This helps virtulization solutions know when the guest
Linux kernel is in an idle state.
A command line option called "idle" is also added to allow people to change
the idle loop back to the original variation. This is done by setting
"idle=spin" on the kernel command line.
Signed-off-by: Jerone Young <jyoung5@us.ibm.com> Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
Paul Mackerras [Sat, 12 Apr 2008 05:20:59 +0000 (15:20 +1000)]
[LMB] Restructure allocation loops to avoid unsigned underflow
There is a potential bug in __lmb_alloc_base where we subtract `size'
from the base address of a reserved region without checking whether
the subtraction could wrap around and produce a very large unsigned
value. In fact it probably isn't possible to hit the bug in practice
since it would only occur in the situation where we can't satisfy the
allocation request and there is a reserved region starting at 0.
This fixes the potential bug by breaking out of the loop when we get
to the point where the base of the reserved region is less than the
size requested. This also restructures the loop to be a bit easier to
follow.
The same logic got copied into lmb_alloc_nid_unreserved, so this makes
a similar change there. Here the bug is more likely to be hit because
the outer loop (in lmb_alloc_nid) goes through the memory regions in
increasing order rather than decreasing order as __lmb_alloc_base
does, and we are therefore more likely to hit the case where we are
testing against a reserved region with a base address of 0.
Paul Mackerras [Sat, 12 Apr 2008 05:20:59 +0000 (15:20 +1000)]
[LMB] Fix some whitespace and other formatting issues, use pr_debug
This makes no semantic changes. It fixes the whitespace and formatting
a bit, gets rid of a local DBG macro and uses the equivalent pr_debug
instead, and restructures one while loop that had a function call and
assignment in the condition to be a bit more readable. Some comments
about functions being called with relocation disabled were also removed
as they would just be confusing to most readers now that the code is
in lib/.
David S. Miller [Mon, 24 Mar 2008 09:50:48 +0000 (20:50 +1100)]
[LMB] Add lmb_alloc_nid()
A variant of lmb_alloc() that tries to allocate memory on a specified
NUMA node 'nid' but falls back to normal lmb_alloc() if that fails.
The caller provides a 'nid_range' function pointer which assists the
allocator. It is given args 'start', 'end', and pointer to integer
'this_nid'.
It places at 'this_nid' the NUMA node id that corresponds to 'start',
and returns the end address within 'start' to 'end' at which memory
assosciated with 'nid' ends.
This callback allows a platform to use lmb_alloc_nid() in just
about any context, even ones in which early_pfn_to_nid() might
not be working yet.
This function will be used by the NUMA setup code on sparc64, and also
it can be used by powerpc, replacing it's hand crafted
"careful_allocation()" function in arch/powerpc/mm/numa.c
If x86 ever converts it's NUMA support over to using the LMB helpers,
it can use this too as it has something entirely similar.
Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Paul Mackerras <paulus@samba.org>
Stephen Rothwell [Mon, 14 Apr 2008 02:00:47 +0000 (12:00 +1000)]
[POWERPC] remove include of asm/of_device.h from pmi.h
pmi.h does not diectly reference anything in of_device.h and of the two
files that include asm/pmi.h, one includes of_device.h and the other
includes of_platform.h (which includes of_device.h).
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Paul Mackerras <paulus@samba.org>
Stephen Rothwell [Thu, 10 Apr 2008 06:43:47 +0000 (16:43 +1000)]
[POWERPC] iSeries: Make iseries_reg_save private to iSeries
Now that we have the alpaca, the reg_save_ptr is no longer needed in the
paca. Eradicate all global uses of it and make it static in the iSeries
lpardata.c
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Paul Mackerras <paulus@samba.org>
Stephen Rothwell [Thu, 10 Apr 2008 06:39:18 +0000 (16:39 +1000)]
[POWERPC] iSeries: Use alternate paca structure for booting
The iSeries HV only needs the first two fields of the paca statically
initialised, so create an alternate paca that contains only those and
switch to our real paca immediately after boot.
This is in order to make the 1024 cpu patches easier since they will no
longer have to statically initialise the pacas for iSeries.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Paul Mackerras <paulus@samba.org>
The prpmc2800 platform requires a zImage formatted file with an
embedded dtb file. Rename the requested boot image file to
dtbImage.prpmc2800.
Signed-off-by: Dale Farnsworth <dale@farnsworth.org> Acked-by: Mark A. Greer <mgreer@mvista.com> Acked-by: Grant Likely <grant.likely@secretlab.ca> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] mv643xx_eth: Prepare to support multiple silicon blocks
The mv643xx_eth driver is being modified to support multiple instances
of the ethernet silicon block on the same platform. Each block contains
a single register bank containing the registers for up to three ports
interleaved within that bank. This patch updates the PowerPC OF to
platform_device glue code to support multiple silicon blocks, each
with up to three ethernet ports. The main difference is that we now
allow multiple mv64x60_shared platform_devices to be registered and
we provide each port platform_device with a pointer to its associated
shared platform_device. The pointer will not be used until the
mv643xx_eth driver changes are committed.
Signed-off-by: Dale Farnsworth <dale@farnsworth.org> Acked-by: Mark Greer <mgreer@mvista.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Mark A. Greer [Mon, 7 Apr 2008 22:10:34 +0000 (08:10 +1000)]
[POWERPC] prpmc2800: Clean up dts properties
Remove several unused (or software config only) properties.
Rename marvel node to "system-controller". Also, rename the
"block-index" property to "cell-index" to conform to current
practice.
Signed-off-by: Mark A. Greer <mgreer@mvista.com> Signed-off-by: Dale Farnsworth <dale@farnsworth.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
Compatible names should refer to a specific version of the hardware,
without wildcards. Change each instance of mv64x60 to mv64360, which
is the oldest version we currently support.
Signed-off-by: Mark A. Greer <mgreer@mvista.com> Signed-off-by: Dale Farnsworth <dale@farnsworth.org> Signed-off-by: Paul Mackerras <paulus@samba.org>
[POWERPC] prpmc2800: Fix frequencies in prpmc2800.dts
After the conversion to dts v1 format, seeing the frequencies
in decimal made it obvious that some of them had been
incorrectly truncated. This fixes them. Note that the PCI
frequency comes from a different source and is documented
as 66MHz, so it was left at 66000000.
Signed-off-by: Dale Farnsworth <dale@farnsworth.org> Acked-by: Mark A. Greer <mgreer@mvista.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
Stephen Rothwell [Fri, 11 Apr 2008 01:11:56 +0000 (11:11 +1000)]
[POWERPC] Fix MAX_ORDER config problem
The allyesconfig (among others) build was giving this:
In file included from include/linux/gfp.h:4,
from include/linux/slab.h:14,
from include/linux/percpu.h:5,
from include2/asm/time.h:18,
from include2/asm/cputime.h:26,
from include/linux/sched.h:67,
from
arch/powerpc/kernel/asm-offsets.c:17:
include/linux/mmzone.h:791:2: error: #error Allocator MAX_ORDER exceeds SECTION_SIZE
Kconfig options are order depenendent, so move the setting of
FORCE_MAX_ZONEORDER to after the setting of PPC_64K_PAGES. Also add an
explicit !PPC_64K_PAGES.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Paul Mackerras <paulus@samba.org>
Paul Mackerras [Mon, 14 Apr 2008 03:59:02 +0000 (13:59 +1000)]
[POWERPC] Fix handling of unrecoverable SLB miss interrupts
If an SLB miss interrupt happens while the RI bit of MSR is zero, we
can't just return, because RI being zero indicates that SRR0/SRR1
potentially had live values in them, and the process of taking an
interrupt overwrites them.
This should never happen, but if it does, we try to print a nice oops
message. That doesn't work, however, because the code at unrecov_slb
assumes that the MMU has been turned on, but we call it with the MMU
off (and have done so since the SLB miss handler was rewritten to run
without turning the MMU on) -- except on iSeries, where everything runs
with the MMU on.
This fixes it by adding the necessary code to turn the MMU on if
necessary.
Merge branch 'docs' of git://git.lwn.net/linux-2.6
* 'docs' of git://git.lwn.net/linux-2.6:
Add additional examples in Documentation/spinlocks.txt
Move sched-rt-group.txt to scheduler/
Documentation: move rpc-cache.txt to filesystems/
Documentation: move nfsroot.txt to filesystems/
Spell out behavior of atomic_dec_and_lock() in kerneldoc
Fix a typo in highres.txt
Fixes to the seq_file document
Fill out information on patch tags in SubmittingPatches
Add the seq_file documentation
Mark Fasheh [Thu, 10 Apr 2008 20:55:21 +0000 (13:55 -0700)]
Add additional examples in Documentation/spinlocks.txt
Checkpatch will throw an error if code doesn't use the correct initializers
for static spinlocks:
ERROR: Use of SPIN_LOCK_UNLOCKED is deprecated: see Documentation/spinlocks.txt
This is fine, but Documentation/spinlocks.txt isn't very clear on how to
_use_ the new initializers for static variables. To save people time in the
future, I added two small examples of how to fix old-style static
initializers to be more lockdep friendly.
Signed-off-by: Mark Fasheh <mfasheh@suse.com> Signed-off-by: Jonathan Corbet <corbet@lwn.net>
projects / linux-2.6 / log