Now that the KVM/arm code knows about affinity, remove the hard
limit of 4 vcpus per VM.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

The L2CTLR register contains the number of CPUs in this cluster.

Make sure the register content is actually relevant to the vcpu
that is being configured by computing the number of cores that are
part of its cluster.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

In order to be able to support more than 4 A7 or A15 CPUs,
we need to fix the MPIDR computing to reflect the fact that
both A15 and A7 can only exist in clusters of at most 4 CPUs.

Fix the MPIDR computing to allow virtual clusters to be exposed
to the guest.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

Support transparent huge pages in KVM/ARM and KVM/ARM64.  The
transparent_hugepage_adjust is not very pretty, but this is also how
it's solved on x86 and seems to be simply an artifact on how THPs
behave.  This should eventually be shared across architectures if
possible, but that can always be changed down the road.
Acked-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

Support huge pages in KVM/ARM and KVM/ARM64.  The pud_huge checking on
the unmap path may feel a bit silly as the pud_huge check is always
defined to false, but the compiler should be smart about this.

Note: This deals only with VMAs marked as huge which are allocated by
users through hugetlbfs only.  Transparent huge pages can only be
detected by looking at the underlying pages (or the page tables
themselves) and this patch so far simply maps these on a page-by-page
level in the Stage-2 page tables.

Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Russell King <rmk+kernel@arm.linux.org.uk>
Acked-by: NCatalin Marinas <catalin.marinas@arm.com>
Acked-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

Update comments to reflect what is really going on and add the TWE bit
to the comments in kvm_arm.h.

Also renames the function to kvm_handle_wfx like is done on arm64 for
consistency and uber-correctness.
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

On an (even slightly) oversubscribed system, spinlocks are quickly
becoming a bottleneck, as some vcpus are spinning, waiting for a
lock to be released, while the vcpu holding the lock may not be
running at all.

This creates contention, and the observed slowdown is 40x for
hackbench. No, this isn't a typo.

The solution is to trap blocking WFEs and tell KVM that we're
now spinning. This ensures that other vpus will get a scheduling
boost, allowing the lock to be released more quickly. Also, using
CONFIG_HAVE_KVM_CPU_RELAX_INTERCEPT slightly improves the performance
when the VM is severely overcommited.

Quick test to estimate the performance: hackbench 1 process 1000

2xA15 host (baseline):	1.843s

2xA15 guest w/o patch:	2.083s
4xA15 guest w/o patch:	80.212s
8xA15 guest w/o patch:	Could not be bothered to find out

2xA15 guest w/ patch:	2.102s
4xA15 guest w/ patch:	3.205s
8xA15 guest w/ patch:	6.887s

So we go from a 40x degradation to 1.5x in the 2x overcommit case,
which is vaguely more acceptable.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

Update comments to reflect what is really going on and add the TWE bit
to the comments in kvm_arm.h.

Also renames the function to kvm_handle_wfx like is done on arm64 for
consistency and uber-correctness.
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

On an (even slightly) oversubscribed system, spinlocks are quickly
becoming a bottleneck, as some vcpus are spinning, waiting for a
lock to be released, while the vcpu holding the lock may not be
running at all.

This creates contention, and the observed slowdown is 40x for
hackbench. No, this isn't a typo.

The solution is to trap blocking WFEs and tell KVM that we're
now spinning. This ensures that other vpus will get a scheduling
boost, allowing the lock to be released more quickly. Also, using
CONFIG_HAVE_KVM_CPU_RELAX_INTERCEPT slightly improves the performance
when the VM is severely overcommited.

Quick test to estimate the performance: hackbench 1 process 1000

2xA15 host (baseline):	1.843s

2xA15 guest w/o patch:	2.083s
4xA15 guest w/o patch:	80.212s
8xA15 guest w/o patch:	Could not be bothered to find out

2xA15 guest w/ patch:	2.102s
4xA15 guest w/ patch:	3.205s
8xA15 guest w/ patch:	6.887s

So we go from a 40x degradation to 1.5x in the 2x overcommit case,
which is vaguely more acceptable.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

The KVM_HPAGE_DEFINES are a little artificial on ARM, since the huge
page size is statically defined at compile time and there is only a
single huge page size.

Now when the main kvm code relying on these defines has been moved to
the x86 specific part of the world, we can get rid of these.
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: NGleb Natapov <gleb@redhat.com>

This patch adds support for running Cortex-A7 guests on Cortex-A7 hosts.

As Cortex-A7 is architecturally compatible with A15, this patch is largely just
generalising existing code. Areas where 'implementation defined' behaviour
is identical for A7 and A15 is moved to allow it to be used by both cores.

The check to ensure that coprocessor register tables are sorted correctly is
also moved in to 'common' code to avoid each new cpu doing its own check
(and possibly forgetting to do so!)
Signed-off-by: NJonathan Austin <jonathan.austin@arm.com>
Acked-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

The T{0,1}SZ fields of TTBCR are 3 bits wide when using the long descriptor
format. Likewise, the T0SZ field of the HTCR is 3-bits. KVM currently
defines TTBCR_T{0,1}SZ as 3, not 7.

The T0SZ mask is used to calculate the value for the HTCR, both to pick out
TTBCR.T0SZ and mask off the equivalent field in the HTCR during
read-modify-write. The incorrect mask size causes the (UNKNOWN) reset value
of HTCR.T0SZ to leak in to the calculated HTCR value. Linux will hang when
initializing KVM if HTCR's reset value has bit 2 set (sometimes the case on
A7/TC2)

Fixing T0SZ allows A7 cores to boot and T1SZ is also fixed for completeness.
Signed-off-by: NJonathan Austin <jonathan.austin@arm.com>
Acked-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

KVM does not have a notion of multiple clusters for CPUs, just a linear
array of CPUs. When using a system with cores in more than one cluster, the
current method for calculating the virtual MPIDR will leak the (physical)
cluster information into the virtual MPIDR. One effect of this is that
Linux under KVM fails to boot multiple CPUs that aren't in the 0th cluster.

This patch does away with exposing the real MPIDR fields in favour of simply
using the virtual CPU number (but preserving the U bit, as before).
Signed-off-by: NJonathan Austin <jonathan.austin@arm.com>
Acked-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

For implementing CPU=host, we need a mechanism for querying
preferred VCPU target type on underlying Host.

This patch implements KVM_ARM_PREFERRED_TARGET vm ioctl which
returns struct kvm_vcpu_init instance containing information
about preferred VCPU target type and target specific features
available for it.
Signed-off-by: NAnup Patel <anup.patel@linaro.org>
Signed-off-by: NPranavkumar Sawargaonkar <pranavkumar@linaro.org>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

This patch implements kvm_vcpu_preferred_target() function for
KVM ARM which will help us implement KVM_ARM_PREFERRED_TARGET ioctl
for user space.
Signed-off-by: NAnup Patel <anup.patel@linaro.org>
Signed-off-by: NPranavkumar Sawargaonkar <pranavkumar@linaro.org>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

Very minor typo in comments of inject_abt() when we update fault status
register for injecting prefetch abort.
Signed-off-by: NAnup Patel <anup.patel@linaro.org>
Signed-off-by: NPranavkumar Sawargaonkar <pranavkumar@linaro.org>
Signed-off-by: NChristoffer Dall <christoffer.dall@linaro.org>

Currently all clkdev registration use "cpufreq-cpu0.0" as dev_id
for cpu clock which refers to virtual platform device. It needs to
be "cpu0" instead which is actual cpu0 device id.

This patch changes the dev_id from "cpufreq-cpu0.0" to "cpu0".
Reported-and-tested-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Cc: Shawn Guo <shawn.guo@linaro.org>
Cc: Magnus Damm <damm@opensource.se>
Signed-off-by: NSudeep KarkadaNagesha <sudeep.karkadanagesha@arm.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Currently all clkdev registration use "cpufreq-cpu0.0" as dev_id
for cpu clock which refers to virtual platform device. It needs to
be "cpu0" instead which is actual cpu0 device id.

This patch changes the dev_id from "cpufreq-cpu0.0" to "cpu0".
Reported-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Tested-by: NShawn Guo <shawn.guo@linaro.org>
Signed-off-by: NSudeep KarkadaNagesha <sudeep.karkadanagesha@arm.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Commit cdc58d60 "cpufreq: imx6q-cpufreq:
remove device tree parsing for cpu nodes" assumed the pdev->dev is set to
cpu0 device in the platform code. But it actually points to the virtual
cpufreq-cpu0 platform device which is not present in the device tree.
Most of the information needed by cpufreq is stored in cpu0 DT node.
So cpu_dev must point to cpu0 device.

This patch fixes the wrong assignment to cpu_dev.
Reported-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Tested-by: NShawn Guo <shawn.guo@linaro.org>
Signed-off-by: NSudeep KarkadaNagesha <sudeep.karkadanagesha@arm.com>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

This is just a standard board for the Ux500, include it in the
v7 multiplatform defconfig.
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NOlof Johansson <olof@lixom.net>

Fix trivial typo in name.
Signed-off-by: NPhil Carmody <phil.carmody@partner.samsung.com>
Signed-off-by: NTony Lindgren <tony@atomide.com>

Corrected the functions spelling mistake in the OMAP4 SMP source file.
Signed-off-by: NAnoop Thomas Mathew <atm@profoundis.com>
Signed-off-by: NTony Lindgren <tony@atomide.com>

We call cpu_cluster_pm_enter for dev->cpu == 0 only, but
cpu_cluster_pm_exit called without that check.

Because of that unhandled page fault may happen:

[    3.803405] Unable to handle kernel paging request at virtual address 00002500
[    3.810974] pgd = c0004000
[    3.813812] [00002500] *pgd=00000000
[    3.817596] Internal error: Oops: 5 [#1] SMP ARM
[    3.822418] Modules linked in:
[    3.825653] CPU: 1 PID: 0 Comm: swapper/1 Not tainted 3.11.0-rc6+ #21
[    3.832397] task: ed86ef40 ti: ed896000 task.ti: ed896000
[    3.838073] PC is at irq_notifier+0x234/0x25c
[    3.842651] LR is at irq_notifier+0x218/0x25c
[    3.847229] pc : [<c0029ed8>]    lr : [<c0029ebc>]    psr: 80000193
[    3.847229] sp : ed897ee8  ip : 00000005  fp : 00000001
[    3.859283] r10: c0b395f0  r9 : c0b30594  r8 : c0b8c2ac
[    3.864776] r7 : ffffffff  r6 : 00000000  r5 : 00000005  r4 : 00000000
[    3.871643] r3 : 00002500  r2 : 00000000  r1 : 00000005  r0 : 44302244
[    3.878479] Flags: Nzcv  IRQs off  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
[    3.886260] Control: 10c5387d  Table: 8000404a  DAC: 00000015
[    3.892272] Process swapper/1 (pid: 0, stack limit = 0xed896240)
[    3.898590] Stack: (0xed897ee8 to 0xed898000)
[    3.903167] 7ee0:                   c0979c3a 00000001 ed897ef8 ed896000 c0014f7c 00000000
[    3.911743] 7f00: 00000005 00000000 ffffffff c0b8c2ac c0b395f0 c077c04c c0c94b48 c0b3953c
[    3.920318] 7f20: c0bcd928 00000002 c0b39524 c00cfad8 00000000 ffffffff 00000000 c00cfb10
[    3.928924] 7f40: c14e62c0 c002c1c8 c002c0ac c14e62c0 00000002 e251c37d 00000000 c0b39548
[    3.937499] 7f60: c0b395f0 c05a1bc4 e251c37d 00000000 00000005 c05a3870 edc90380 edc90380
[    3.946105] 7f80: edc90394 c14e62c0 c0b39548 00000002 c0784064 c05a3c78 c0b395e0 c14e62c0
[    3.954681] 7fa0: 00000002 c0b39548 c0bc9db8 00000000 00000001 c05a1dc0 ed896000 00000015
[    3.963287] 7fc0: c0bc9db8 ed896000 8000406a c0b30594 c0784064 c000e504 00000746 c007a528
[    3.971862] 7fe0: 00000001 0000001d 600001d3 c0bcc004 00000000 800086c4 ee0aa6a7 d2aabaa9
[    3.980499] [<c0029ed8>] (irq_notifier+0x234/0x25c) from [<c077c04c>] (notifier_call_chain+0x38/0x68)
[    3.990173] [<c077c04c>] (notifier_call_chain+0x38/0x68) from [<c00cfad8>] (cpu_pm_notify+0x20/0x38)
[    3.999786] [<c00cfad8>] (cpu_pm_notify+0x20/0x38) from [<c00cfb10>] (cpu_cluster_pm_exit+0x20/0x50)
[    4.009399] [<c00cfb10>] (cpu_cluster_pm_exit+0x20/0x50) from [<c002c1c8>] (omap_enter_idle_coupled+0x11c/0x14c)
[    4.020111] [<c002c1c8>] (omap_enter_idle_coupled+0x11c/0x14c) from [<c05a1bc4>] (cpuidle_enter_state+0x40/0xec)
[    4.030822] [<c05a1bc4>] (cpuidle_enter_state+0x40/0xec) from [<c05a3c78>] (cpuidle_enter_state_coupled+0x1f4/0x240)
[    4.041870] [<c05a3c78>] (cpuidle_enter_state_coupled+0x1f4/0x240) from [<c05a1dc0>] (cpuidle_idle_call+0x150/0x228)
[    4.052947] [<c05a1dc0>] (cpuidle_idle_call+0x150/0x228) from [<c000e504>] (arch_cpu_idle+0x8/0x38)
[    4.062499] [<c000e504>] (arch_cpu_idle+0x8/0x38) from [<c007a528>] (cpu_startup_entry+0x178/0x1e4)
[    4.071990] [<c007a528>] (cpu_startup_entry+0x178/0x1e4) from [<800086c4>] (0x800086c4)
[    4.080383] Code: e5922288 03a03b0a 13a03c25 e0823003 (e5932000)
[    4.086791] ---[ end trace d83954a84a6fa69e ]---

It is supposed that sar_base is initialized in irq_save_context, which
is called on CPU_CLUSTER_PM_ENTER notification. If this notification
has been missed and CPU_CLUSTER_PM_EXIT is received sar_base is NULL.

Fix it by calling CPU_CLUSTER_PM_{ENTER,EXIT} under the same condition.
Signed-off-by: NVladimir Murzin <murzin.v@gmail.com>
Acked-by: NKevin Hilman <khilman@linaro.org>
Signed-off-by: NTony Lindgren <tony@atomide.com>

When CONFIG_ARM_LPAE=y the following build warning is generated:

arch/arm/mach-omap2/gpmc.c:1495:4: warning: format '%x' expects argument of type 'unsigned int', but argument 4 has type 'resource_size_t' [-Wformat]

According to Documentation/printk-formats.txt '%pa' can be used to properly
print 'resource_size_t'.
Reported-by: NKevin Hilman <khilman@linaro.org>
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Acked-by: NKevin Hilman <khilman@linaro.org>
Signed-off-by: NTony Lindgren <tony@atomide.com>

In case of error, the function omap_device_alloc() returns ERR_PTR()
and never returns NULL. The NULL test in the return value check
should be replaced with IS_ERR().
Signed-off-by: NWei Yongjun <yongjun_wei@trendmicro.com.cn>
Acked-by: NKevin Hilman <khilman@linaro.org>
Signed-off-by: NTony Lindgren <tony@atomide.com>

Looks like we still have the legacy clock alias name for
omap4 GPMC (General Purpose Memory Controller), so let's
fix it for the device tree naming. There's no need to keep
the legacy naming as omap4 is DT only nowadays.

Without this fix we get the following error while booting:

[    0.440399] omap-gpmc 50000000.gpmc: error: clk_get
Reported-by: NOlof Johansson <olof@lixom.net>
Cc: stable@vger.kernel.org # v3.11
Signed-off-by: NTony Lindgren <tony@atomide.com>

Zaurus collie contains 2 LH28F640BFHE-PTTL90 (64M 4Mx16) and
at the moment cfi will not detect the collie NOR.
In the meanwhile we can revert to the jedec-probe map which has been
fixed with following commit:

mtd: jedec_probe: fix LH28F640BF definition
fe2f4c8e

Somehow this is unsatisfactory because the flash is mounted READ ONLY
(as from factory, with a RO cramfs)
Signed-off-by: NAndrea Adami <andrea.adami@gmail.com>
Signed-off-by: NOlof Johansson <olof@lixom.net>

Right now the U300 submenus are showcased for everyone even if
we're not on v5 multiplatforms. Hide this in the multiplatform
configuration properly.

Cc: arm@kernel.org
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NOlof Johansson <olof@lixom.net>

Add pinmux configuration for MCBSP2 connected to the TDM interface.
With this configuration the Headset modules works as expected.
Signed-off-by: NEnric Balletbo i Serra <eballetbo@gmail.com>
Acked-by: NJavier Martinez Canillas <javier@dowhile0.org>
Signed-off-by: NBenoit Cousson <bcousson@baylibre.com>

Commit 76787b3b (ARM: OMAP2+: Remove board-4430sdp.c)
removed legacy booting in favor of device tree based booting
for 4430sdp. That caused the WLAN to stop working as the
related .dts entries fell through the cracks.

I don't have the "1283 PG 2.21 connectivity device" on my 4430sdp,
but the earlier version of this patch was tested by Luciano
Coelho. This version has left out the input logic for MMC CLK
line compared to the earlier version as that is not bidirectional,
and should be safe to do.

Cc: Rajendra Nayak <rnayak@ti.com>
Cc: Luciano Coelho <luca@coelho.fi>
Cc: Ruslan Bilovol <ruslan.bilovol@ti.com>
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBenoit Cousson <bcousson@baylibre.com>

Commit b42b9181 (ARM: OMAP2+: Remove board-omap4panda.c)
removed legacy booting in favor of device tree based booting
for pandaboard. That caused the WLAN to stop working as the
related .dts entries fell through the cracks.

The legacy muxing was setting pulls for GPIO 48 and 49, so let's
keep that behaviour for now to avoid further regressions for
BT and FM. Also input logic was enabled for MMC CLK line, but
I've verified that the input logic we don't need enabled for
CLK line as it's not bidirectional.

Also, we want to use non-removable instead of ti,non-removable
as the ti,non-removable also sets no_regulator_off_init which
is really not what we want as then wl12xx won't get powered
up and down which is needed for resetting it.

Note that looks like the WLAN interface fails to come up after
a warm reset, but that most likely was also happening with
the legacy booting and needs a separate fix.

Cc: Paolo Pisati <p.pisati@gmail.com>
Cc: Rajendra Nayak <rnayak@ti.com>
Cc: Luciano Coelho <luca@coelho.fi>
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NBenoit Cousson <bcousson@baylibre.com>

Without this, legacy platforms that can boot with a multiplatform
kernel but that need the DTB to be appended, won't have a way to pass
firmware-set bootargs to the kernel.

This is needed to boot multi_v7_defconfig on snowball, for instance.
Signed-off-by: NOlof Johansson <olof@lixom.net>

This fixes a multiplatform regression on the Ux500.

When compiling the Ux500 platforms in multiplatform configurations
both PL310_ERRATA_588369 and PL310_ERRATA_727915 would crash the
platform when trying to launch the init process.

The Ux500 cannot access the debug registers of the PL310, it will
just crash if you try this. So disable this by setting the debug
callback to NULL when initializing the l2x0 on this platform.

Cc: Lee Jones <lee.jones@linaro.org>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NOlof Johansson <olof@lixom.net>

Fix the DTS data for ocp2scp node by adding the
missing reg property.
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NKishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: NBenoit Cousson <bcousson@baylibre.com>

USB3 block has a 64KiB space, another 64KiB is
used for the wrapper.

Without this change, resource_size() will get
confused and driver won't probe because size
will be negative.
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NKishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: NBenoit Cousson <bcousson@baylibre.com>

The BeagleBone Black is basically a regular BeagleBone with eMMC and
HDMI added, so create a common dtsi both can use.

IMPORTANT: booting the existing am335x-bone.dts will blow up the HDMI
transceiver after a dozen boots with an uSD card inserted because LDO
will be at 3.3V instead of 1.8.

MMC support for AM335x still isn't in, so only the LDO change has been
added.
Signed-off-by: NKoen Kooi <koen@dominion.thruhere.net>
Tested-by: NTom Rini <trini@ti.com>
Tested-by: NMatt Porter <matt.porter@linaro.org>
Acked-by: NKevin Hilman <khilman@linaro.org>
Tested-by: NKevin Hilman <khilman@linaro.org>
Tested-by: NJoel Fernandes <joelf@ti.com>
Signed-off-by: NBenoit Cousson <bcousson@baylibre.com>

The beagle and beagle-xm entries were inside the same double quote.
Split them to have two distinct entries.
Signed-off-by: NRobert Nelson <robertcnelson@gmail.com>
Signed-off-by: NBenoit Cousson <bcousson@baylibre.com>

The patch "0b7a76aa ARM: dts: imx6q{dl}: add DTE pads for uart"
adds the DTE pads for uart. For PAD_EIM_D29, the offset of the
Pad Mux register should be 0x0c8, not 0x0c4.

This patch fixes it.
Signed-off-by: NHuang Shijie <b32955@freescale.com>
Signed-off-by: NShawn Guo <shawn.guo@linaro.org>

The L2 cache controller(PL310) version on the i.MX6D/Q is r3p1-50rel0
The L2 cache controller(PL310) version on the i.MX6DL/SOLO/SL is r3p2
But according to ARM PL310 errata: 752271
ID: 752271: Double linefill feature can cause data corruption
Fault Status: Present in: r3p0, r3p1, r3p1-50rel0. Fixed in r3p2
Workaround: The only workaround to this erratum is to disable the
double linefill feature. This is the default behavior.

without this patch, you will meet the following error when run the
memtester application at: http://pyropus.ca/software/memtester/

FAILURE: 0x00100000 != 0x00200000 at offset 0x01365664.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365668.
FAILURE: 0x00100000 != 0x00200000 at offset 0x0136566c.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365670.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365674.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365678.
Signed-off-by: NJason Liu <r64343@freescale.com>
Signed-off-by: NShawn Guo <shawn.guo@linaro.org>

Commit 718a3500 (ARM: i.MX51: Add PATA support) adds pata support to the
imx51.dtsi file and is using clock 161. The problem is that the right
clock is 172, according to commit 5d530bb0 (ARM: i.MX5: Add PATA and SRTC
clocks).  Using the clock 172 makes things work again (and kills a nasty
system freeze).
Tested-by: NSteev Klimaszewski <steev@gentoo.org>
Signed-off-by: NArnaud Patard <arnaud.patard@rtp-net.org>
Signed-off-by: NShawn Guo <shawn.guo@linaro.org>