Commit 5c0484e2 ('powerpc: Endian safe trampoline') resulted in
losing proper alignment of the spinlock variables used when booting
secondary CPUs, causing some quite odd issues with failing to boot on
PA Semi-based systems.

This showed itself on ppc64_defconfig, but not on pasemi_defconfig,
so it had gone unnoticed when I initially tested the LE patch set.

Fix is to add explicit alignment instead of relying on good luck. :)

[ It appears that there is a different issue with PA Semi systems
  however this fix is definitely correct so applying anyway -- BenH
]

Fixes: 5c0484e2 ('powerpc: Endian safe trampoline')
Reported-by: NChristian Zigotzky <chzigotzky@xenosoft.de>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=67811Signed-off-by: NOlof Johansson <olof@lixom.net>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

p_end is an 8 byte value embedded in the text section. This means it
is only 4 byte aligned when it should be 8 byte aligned. Fix this
by adding an explicit alignment.

This fixes an issue where POWER7 little endian builds with
CONFIG_RELOCATABLE=y fail to boot.
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Prevent ioda_eeh_hub_diag() from clobbering itself when called by supplying
a per-PHB buffer for P7IOC hub diagnostic data.  Take care to inform OPAL of
the correct size for the buffer.

[Small style change to the use of sizeof -- BenH]
Signed-off-by: NBrian W Hart <hartb@linux.vnet.ibm.com>
Acked-by: NGavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

PHB diagnostic buffer may be smaller than PAGE_SIZE, especially when
PAGE_SIZE > 4KB.
Signed-off-by: NBrian W Hart <hartb@linux.vnet.ibm.com>
Acked-by: NGavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

The powerpc 64-bit __copy_tofrom_user() function uses shifts to handle
unaligned invocations.  However, these shifts were designed for
big-endian systems: On little-endian systems, they must shift in the
opposite direction.

This commit relies on the C preprocessor to insert the correct shifts
into the assembly code.

[ This is a rare but nasty LE issue. Most of the time we use the POWER7
optimised __copy_tofrom_user_power7 loop, but when it hits an exception
we fall back to the base __copy_tofrom_user loop. - Anton ]
Signed-off-by: NPaul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

The generic put_unaligned/get_unaligned macros were made endian-safe by
calling the appropriate endian dependent macros based on the endian type
of the powerpc processor.
Signed-off-by: NRajesh B Prathipati <rprathip@linux.vnet.ibm.com>
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

In EXCEPTION_PROLOG_COMMON() we check to see if the stack pointer (r1)
is valid when coming from the kernel.  If it's not valid, we die but
with a nice oops message.

Currently we allocate a stack frame (subtract INT_FRAME_SIZE) before we
check to see if the stack pointer is negative.  Unfortunately, this
won't detect a bad stack where r1 is less than INT_FRAME_SIZE.

This patch fixes the check to compare the modified r1 with
-INT_FRAME_SIZE.  With this, bad kernel stack pointers (including NULL
pointers) are correctly detected again.

Kudos to Paulus for finding this.
Signed-off-by: NMichael Neuling <mikey@neuling.org>
cc: stable@vger.kernel.org
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

We are passing pointers to the firmware for reads, we need to properly
convert the result as OPAL is always BE.
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

opal_xscom_read uses a pointer to return the data so we need
to byteswap it on LE builds.
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

A couple more device tree properties that need byte swapping.
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

The MSI code is miscalculating quotas in little endian mode.
Add required byteswaps to fix this.

Before we claimed a quota of 65536, after the patch we
see the correct value of 256.
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

We need to byteswap ibm,pcie-link-speed-stats.
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

The NVRAM code has a number of endian issues. I noticed a very
confused error log count:

RTAS: 100663330 -------- RTAS event begin --------

100663330 == 0x06000022. 0x6 LE error logs and 0x22 BE error logs.

The pstore code has similar issues - if we write an oops in one
endian and attempt to read it in another we get junk.

Make both of these formats big endian, and byteswap as required.
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Some obvious issues:

cat /proc/ppc64/lparcfg
...
partition_id=16777216
...
partition_potential_processors=268435456
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

cpu_to_core_id() is missing a byteswap:

cat /sys/devices/system/cpu/cpu63/topology/core_id
201326592
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

During on LE boot we see:

    Partition configured for 1073741824 cpus, operating system maximum is 2048.

Clearly missing a byteswap here.
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

There is a bug in using ptrace to access FPRs via PTRACE_PEEKUSR /
PTRACE_POKEUSR. In effect, trying to access any of the FPRs always
really accesses FPR0, which does seriously break debugging :-)

The problem seems to have been introduced by commit 3ad26e5c
(Merge branch 'for-kvm' into next).

[ It is indeed a merge conflict between Paul's FPU/VSX state rework
and my LE patches - Anton ]
Signed-off-by: NUlrich Weigand <Ulrich.Weigand@de.ibm.com>
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

The current logic sets the kdump base to min of 2G or ppc64_rma_size/2.
On PowerNV kernel the first memory block 'memory@0' can be very large,
equal to the DIMM size with ppc64_rma_size value capped to 1G. Hence on
PowerNV, kdump base is set to 512M resulting kdump to fail while allocating
paca array. This is because, paca need its memory from RMA region capped
at 256M (see allocate_pacas()).

This patch lowers the kdump base cap to 128M so that kdump kernel can
successfully get memory below 256M for paca allocation.
Signed-off-by: NMahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

I have recently found out that no iommu_groups could be found under
/sys/ on a P8. That prevents PCI passthrough from working.

During my investigation, I found out there seems to be a missing
iommu_register_group for PHB3. The following patch seems to fix the
problem. After applying it, I see iommu_groups under
/sys/kernel/iommu_groups/, and can also bind vfio-pci to an adapter,
which gives me a device at /dev/vfio/.
Signed-off-by: NThadeu Lima de Souza Cascardo <cascardo@linux.vnet.ibm.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

The bestcomm driver has been moved to drivers/dma, so to select
this driver by default additionally CONFIG_DMADEVICES has to be
enabled. Currently it is not enabled in the config despite existing
CONFIG_PPC_BESTCOMM=y in the config files. Fix it.
Signed-off-by: NAnatolij Gustschin <agust@denx.de>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

At least some distros expect it these days; turn it on. Also, random
churn from doing a savedefconfig for the first time in a year or so.
Signed-off-by: NOlof Johansson <olof@lixom.net>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

The pseudo palette color entries need to be ajusted for little
endian.

This patch byteswaps the values in the pseudo palette depending
on the host endian order and the screen depth.
Signed-off-by: NCédric Le Goater <clg@fr.ibm.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

The "screen" properties : depth, width, height, linebytes need
to be converted to the host endian order when read from the device
tree.

The offb_init_palette_hacks() routine also made assumption on the
host endian order.
Signed-off-by: NCédric Le Goater <clg@fr.ibm.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

In pte_alloc_one(), pgtable_page_ctor() is passed an address that has
not been converted by page_address() to the newly allocated PTE page.

When the PTE is freed, __pte_free_tlb() calls pgtable_page_dtor()
with an address to the PTE page that has been converted by page_address().
The mismatch in the PTE's page address causes pgtable_page_dtor() to access
invalid memory, so resources for that PTE (such as the page lock) is not
properly cleaned up.

On PPC32, only SMP kernels are affected.

On PPC64, only SMP kernels with 4K page size are affected.

This bug was introduced by commit d614bb04
"powerpc: Move the pte free routines from common header".

On a preempt-rt kernel, a spinlock is dynamically allocated for each
PTE in pgtable_page_ctor().  When the PTE is freed, calling
pgtable_page_dtor() with a mismatched page address causes a memory leak,
as the pointer to the PTE's spinlock is bogus.

On mainline, there isn't any immediately obvious symptoms, but the
problem still exists here.

Fixes: d614bb04 "powerpc: Move the pte free routes from common header"
Cc: Paul Mackerras <paulus@samba.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: linux-stable <stable@vger.kernel.org> # v3.10+
Signed-off-by: NHong H. Pham <hong.pham@windriver.com>
Reviewed-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Allocate enough memory for the ocm_block structure, not just a pointer
to it.
Signed-off-by: NIlia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

A kernel configured with PPC_EARLY_DEBUG_BOOTX=y but PPC_PMAC=n and
PPC_MAPLE=n will fail to link:

  btext.c:(.text+0x2d0fc): undefined reference to `.rmci_off'
  btext.c:(.text+0x2d214): undefined reference to `.rmci_on'

Fix it by making the build of rmci_on/off() depend on
PPC_EARLY_DEBUG_BOOTX, which also enable the only code that uses them.
Signed-off-by: NMichael Ellerman <mpe@ellerman.id.au>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Anatolij Gustschin says:

<<
Please pull a device tree fix for v3.13. The booting on mpc512x
is broken since v3.13-rc1, this patch repairs it.
>>

the 'soc' node in the common .dtsi for MPC5121 has an '#interrupt-cells'
property although this node is not an interrupt controller

remove this erroneously placed property because starting with v3.13-rc1
lookup and resolution of 'interrupts' specs for peripherals gets misled,
emits 'no irq domain found' WARN() messages and breaks the boot process

  irq: no irq domain found for /soc@80000000 !
  ------------[ cut here ]------------
  WARNING: at drivers/of/platform.c:171
  Modules linked in:
  CPU: 0 PID: 1 Comm: swapper Tainted: G        W    3.13.0-rc1-00001-g8a66234 #8
  task: df823bb0 ti: df834000 task.ti: df834000
  NIP: c02b5190 LR: c02b5180 CTR: c01cf4e0
  REGS: df835c50 TRAP: 0700   Tainted: G        W     (3.13.0-rc1-00001-g8a66234)
  MSR: 00029032 <EE,ME,IR,DR,RI>  CR: 229a9d42  XER: 20000000

  GPR00: c02b5180 df835d00 df823bb0 00000000 00000000 df835b18 ffffffff 00000308
  GPR08: c0479cc0 c0480000 c0479cc0 00000308 00000308 00000000 c00040fc 00000000
  GPR16: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 df850880
  GPR24: df84d670 00000000 00000001 df8561a0 dffffccc df85089c 00000020 00000001
  NIP [c02b5190] of_device_alloc+0xf4/0x1a0
  LR [c02b5180] of_device_alloc+0xe4/0x1a0
  Call Trace:
  [df835d00] [c02b5180] of_device_alloc+0xe4/0x1a0 (unreliable)
  [df835d50] [c02b5278] of_platform_device_create_pdata+0x3c/0xc8
  [df835d70] [c02b53fc] of_platform_bus_create+0xf8/0x170
  [df835dc0] [c02b5448] of_platform_bus_create+0x144/0x170
  [df835e10] [c02b55a8] of_platform_bus_probe+0x98/0xe8
  [df835e30] [c0437508] mpc512x_init+0x28/0x1c4
  [df835e70] [c0435de8] ppc_init+0x4c/0x60
  [df835e80] [c0003b28] do_one_initcall+0x150/0x1a4
  [df835ef0] [c0432048] kernel_init_freeable+0x114/0x1c0
  [df835f30] [c0004114] kernel_init+0x18/0x124
  [df835f40] [c000e910] ret_from_kernel_thread+0x5c/0x64
  Instruction dump:
  409effd4 57c9103a 57de2834 7c89f050 7f83e378 7c972214 7f45d378 48001f55
  7c63d278 7c630034 5463d97e 687a0001 <0f1a0000> 2f990000 387b0010 939b0098
  ---[ end trace 2257f10e5a20cbdd ]---

  ...
  irq: no irq domain found for /soc@80000000 !
  fsl-diu-fb 80002100.display: could not get DIU IRQ
  fsl-diu-fb: probe of 80002100.display failed with error -22
  irq: no irq domain found for /soc@80000000 !
  mpc512x_dma 80014000.dma: Error mapping IRQ!
  mpc512x_dma: probe of 80014000.dma failed with error -22
  ...
  irq: no irq domain found for /soc@80000000 !
  fs_enet: probe of 80002800.ethernet failed with error -22
  ...
  irq: no irq domain found for /soc@80000000 !
  mpc5121-rtc 80000a00.rtc: mpc5121_rtc_probe: could not request irq: 0
  mpc5121-rtc: probe of 80000a00.rtc failed with error -22
  ...

Cc: Anatolij Gustschin <agust@denx.de>
Cc: linuxppc-dev@lists.ozlabs.org
Cc: devicetree@vger.kernel.org
Signed-off-by: NGerhard Sittig <gsi@denx.de>
Signed-off-by: NAnatolij Gustschin <agust@denx.de>

We are missing building windfarm_max6690_sensor.o when building
CONFIG_WINDFARM_RM31. Usually all the windfarm drivers are built
and thus this isn't a problem but some more "tailored" setups
(Gentoo ?) building only that driver are not working because
the require sensor module is missing.
Reported-by: NStanislav Ponomarev <devhexorg@gmail.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

for tmp_part->header.name:
    it is "Terminating null required only for names < 12 chars".
    so need to limit the %.12s for it in printk

  additional info:

    %12s  limit the width, not for the original string output length
          if name length is more than 12, it still can be fully displayed.
          if name length is less than 12, the ' ' will be filled before name.

    %.12s truly limit the original string output length (precision)
Signed-off-by: NChen Gang <gang.chen@asianux.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

In a recent patch:
  commit c13f20ac
  Author: Michael Neuling <mikey@neuling.org>
  powerpc/signals: Mark VSX not saved with small contexts

We fixed an issue but an improved solution was later discussed after the patch
was merged.

Firstly, this patch doesn't handle the 64bit signals case, which could also hit
this issue (but has never been reported).

Secondly, the original patch isn't clear what MSR VSX should be set to.  The
new approach below always clears the MSR VSX bit (to indicate no VSX is in the
context) and sets it only in the specific case where VSX is available (ie. when
VSX has been used and the signal context passed has space to provide the
state).

This reverts the original patch and replaces it with the improved solution.  It
also adds a 64 bit version.
Signed-off-by: NMichael Neuling <mikey@neuling.org>
Cc: stable@vger.kernel.org
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

When CONFIG_SPARSEMEM_VMEMMAP option is used in kernel, makedumpfile fails
to filter vmcore dump as it fails to do vmemmap translations. So far
dump filtering on ppc64 never had to deal with vmemmap addresses seperately
as vmemmap regions where mapped in zone normal. But with the inclusion of
CONFIG_SPARSEMEM_VMEMMAP config option in kernel, this vmemmap address
translation support becomes necessary for dump filtering. For vmemmap adress
translation, few kernel symbols are needed by dump filtering tool. This patch
adds those symbols to vmcoreinfo, which a dump filtering tool can use for
filtering the kernel dump. Tested this changes successfully with makedumpfile
tool that supports vmemmap to physical address translation outside zone normal.

[ Removed unneeded #ifdef as suggested by Michael Ellerman --BenH ]
Signed-off-by: NHari Bathini <hbathini@linux.vnet.ibm.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Stephen reported a failure in an allyesconfig build.
CONFIG_CPU_LITTLE_ENDIAN=y gets set but his toolchain is not
new enough to support little endian. We really want to
default to a big endian build; Ben suggested using a choice
which defaults to CPU_BIG_ENDIAN.
Signed-off-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Currently if I cross build TAGS or cscope from x86 I get this:
  % make ARCH=powerpc TAGS
  gcc-4.8.real: error: unrecognized command line option ‘-mbig-endian’
  GEN     TAGS
  %

I'm not setting CROSS_COMPILE= as logically I shouldn't need to and I
haven't needed to in the past when building TAGS or cscope.  Also, the
above completess correct as the error is not fatal to the build.

This was caused by:
    commit d72b0801
    Author: Ian Munsie <imunsie@au1.ibm.com>
    powerpc: Add ability to build little endian kernels

The below fixes this by testing for the -mbig-endian option before
adding it.

I've not done the same thing in the little endian case as if
-mlittle-endian doesn't exist, we probably want to fail quickly as you
probably have an old big endian compiler.
Signed-off-by: NMichael Neuling <mikey@neuling.org>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Scott wrote:

<<
The corenet64 patch fixes a regression introduced in 3.13-rc1 (commit
ef1313de, "powerpc: Add VMX optimised xor
for RAID5").

The 8xx patch fixes a regression introduced in 3.12 (commit
beb2dc0a, "powerpc: Convert some
mftb/mftbu into mfspr").

The other two patches are fixes for minor, long standing bugs.
>>

And in flush_hugetlb_page(), don't check whether vma is NULL after
we've already dereferenced it.

This was found by Dan using static analysis as described here:
https://lists.ozlabs.org/pipermail/linuxppc-dev/2013-November/113161.html

We currently get away with this because the callers that currently pass
NULL for vma seem to be 32-bit-only (e.g. highmem, and
CONFIG_DEBUG_PGALLOC in pgtable_32.c)  Hugetlb is currently 64-bit only,
so we never saw a NULL vma here.
Signed-off-by: NScott Wood <scottwood@freescale.com>
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>

These lines were inoperative for four years, which puts some doubt into
their importance, and it's possible the fixed version will regress, but
at the very least they should be removed instead.
Signed-off-by: NAdam Borowski <kilobyte@angband.pl>
Signed-off-by: NScott Wood <scottwood@freescale.com>

Commit beb2dc0a breaks the MPC8xx which
seems to not support using mfspr SPRN_TBRx instead of mftb/mftbu
despite what is written in the reference manual.

This patch reverts to the use of mftb/mftbu when CONFIG_8xx is
selected.
Signed-off-by: NChristophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: NScott Wood <scottwood@freescale.com>

If CONFIG_ALTIVEC is enabled for CoreNet64, and if we also
select CONFIG_E{5,6}500_CPU this may introduce -mcpu=e500mc64
into $CFLAGS. But Altivec option not allowed with e500mc64,
then some compiling errors occur like this:

      CC      arch/powerpc/lib/xor_vmx.o
    arch/powerpc/lib/xor_vmx.c:1:0: error: AltiVec not supported in this target
    make[1]: *** [arch/powerpc/lib/xor_vmx.o] Error 1
    make: *** [arch/powerpc/lib] Error 2

So we should restrict e500mc64 in altivec scenario.
Signed-off-by: NTiejun Chen <tiejun.chen@windriver.com>
Signed-off-by: NScott Wood <scottwood@freescale.com>