On the r2p* and r3p* versions of the Cortex-A9, a speculative memory
access may cause a page table walk which starts prior to an ASID switch
but completes afterwards. This can populate the micro-TLB with a stale
entry which may be hit with the new ASID.
This workaround places two dsb instructions in the mm switching code so
that no page table walks can cross the ASID switch.
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
PL310 implements the Clean & Invalidate by Way L2 cache maintenance
operation (offset 0x7FC). This operation runs in background so that
PL310 can handle normal accesses while it is in progress. Under very
rare circumstances, due to this erratum, write data can be lost when
PL310 treats a cacheable write transaction during a Clean & Invalidate
by Way operation.
Workaround:
Disable Write-Back and Cache Linefill (Debug Control Register)
Clean & Invalidate by Way (0x7FC)
Re-enable Write-Back and Cache Linefill (Debug Control Register)
This patch also removes any OMAP dependency on PL310 Errata's
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Move L1_CACHE_SHIFT related options together, rather than spreading them
across two separate Kconfig files.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
In commit e616c59140, highmem support was
deactivated for SMP platforms without hardware TLB ops broadcast because
usage of kmap_high_get() requires that IRQs be disabled when kmap_lock
is locked which is incompatible with the IPI mechanism used by the
software TLB ops broadcast invoked through flush_all_zero_pkmaps().
The reason for kmap_high_get() is to ensure that the currently kmap'd
page usage count does not decrease to zero while we're using its
existing virtual mapping in an atomic context. With a VIVT cache this
is essential to do due to cache coherency issues, but with a VIPT cache
this is only an optimization so not to pay the price of establishing a
second mapping if an existing one can be used. However, on VIPT
platforms without hardware TLB maintenance we can give up on that
optimization in order to be able to use highmem.
From ARMv7 onwards the TLB ops are broadcasted in hardware, so let's
disable ARCH_NEEDS_KMAP_HIGH_GET only when CONFIG_SMP and
CONFIG_CPU_TLB_V6 are defined.
Signed-off-by: Nicolas Pitre <nicolas.pitre@linaro.org>
Tested-by: Saeed Bishara <saeed.bishara@gmail.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Achieve better usage of the DMA coherent region by doing top-down
allocation rather than bottom up. If we ask for a 128kB allocation,
this will be aligned to 128kB and satisfied from the very bottom
address. If we then ask for a 600kB allocation, this will be aligned
to 1MB, and we will have a 896kB hole.
Performing top-down allocation resolves this by allocating the 128kB
at the very top, and then the 600kB can come in below it without any
unnecessary wastage.
This problem was reported by Janusz Krzysztofik, who had 2 x 128kB +
1 x 640kB allocations which wouldn't fit into 1MB.
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This adds core support for saving and restoring CPU coprocessor
registers for suspend/resume support. This contains support for suspend
with ARM920, ARM926, SA11x0, PXA25x, PXA27x, PXA3xx, V6 and V7 CPUs.
Tested on Assabet and Tegra 2.
Tested-by: Colin Cross <ccross@android.com>
Tested-by: Kukjin Kim <kgene.kim@samsung.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
On versions of the Cortex-A9 prior to r3p0, an interrupted ICIALLUIS
operation may prevent the completion of a following broadcasted
operation if the second operation is received by a CPU before the
ICIALLUIS has completed, potentially leading to corrupted entries in
the cache or TLB.
This workaround sets a bit in the diagnostic register of the Cortex-A9,
causing CP15 maintenance operations to be uninterruptible.
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
The effect of cache sync operation is to drain the store buffer and
wait for all internal buffers to be empty. In normal conditions, store
buffer is able to merge the normal memory writes within its 32-byte
data buffers. Due to this erratum present in r3p0, the effect of cache
sync operation on the store buffer still remains when the operation
completes. This means that the store buffer is always asked to drain
and this prevents it from merging any further writes.
This can severely affect performance on the write traffic esp. on
Normal memory NC one.
The proposed workaround is to replace the normal offset of cache sync
operation(0x730) by another offset targeting an unmapped PL310
register 0x740.
Signed-off-by: srinidhi kasagar <srinidhi.kasagar@stericsson.com>
Acked-by: Linus Walleij <linus.walleij@stericsson.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
SWP emulation requires that CPU domain support is disabled in order to
work safely. Make that explicit in the kernel configuration to prevent
illegal configurations being generated.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
OMAP2 (armv6) and MX3 turn off support for the V6K instructions, which
when they include support for SMP kernels means that the resulting
kernel is unsafe on SMP and can result in corrupted filesystems as we
end up using unsafe bitops.
Re-enable the use of V6K instructions on such kernels, and let such
kernels running on V6 CPUs eat undefined instruction faults which will
be much safer than filesystem corruption. Next merge window we can fix
this properly (as it requires a much bigger set of changes.)
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Disable the initrd if the passed address already overlaps the reserved
region. This avoids oopses on Netwinders when NeTTrom tells the kernel
that an initrd is located at mem+4MB, but this overlaps the BSS,
resulting in the kernels in-use BSS being freed.
This should be applied to v2.6.37-stable.
Cc: <stable@kernel.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
* 'fixes' of master.kernel.org:/home/rmk/linux-2.6-arm:
ARM: fix missing branch in __error_a
ARM: fix /proc/$PID/stack on SMP
ARM: Fix build regression on SA11x0, PXA, and H720x targets
ARM: 6625/1: use memblock memory regions for "System RAM" I/O resources
ARM: fix wrongly patched constants
ARM: 6624/1: fix dependency for CONFIG_SMP_ON_UP
ARM: 6623/1: Thumb-2: Fix out-of-range offset for Thumb-2 in proc-v7.S
ARM: 6622/1: fix dma_unmap_sg() documentation
ARM: 6621/1: bitops: remove condition code clobber for CLZ
ARM: 6620/1: Change misleading warning when CONFIG_CMDLINE_FORCE is used
ARM: 6619/1: nommu: avoid mapping vectors page when !CONFIG_MMU
ARM: sched_clock: make minsec argument to clocks_calc_mult_shift() zero
ARM: sched_clock: allow init_sched_clock() to be called early
ARM: integrator: fix compile warning in cpu.c
ARM: 6616/1: Fix ep93xx-fb init/exit annotations
ARM: twd: fix display of twd frequency
ARM: udelay: prevent math rounding resulting in short udelays
* 'omap-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap-2.6: (27 commits)
omap4: Fix ULPI PHY init for ES1.0 SDP
omap3: beaglexm: fix power on of DVI
omap3: igep3: Add omap_reserve functionality
omap3: beaglexm: fix DVI reset GPIO
omap3: beaglexm: fix EHCI power up GPIO dir
omap3: igep2: Add keypad support
omap3: igep3: Fix IGEP module second MMC channel power supply
omap3: igep3: Add USB EHCI support for IGEP module
omap3: clocks: Fix build error 'CK_3430ES2' undeclared here
arm: omap4: pandaboard: turn on PHY reference clock at init
omap2plus: prm: Trvial build break fix for undefined reference to 'omap2_prm_read_mod_reg'
omap2plus: voltage: Trivial linking fix for 'EINVAL' undeclared
omap2plus: voltage: Trivial linking fix 'undefined reference'
omap2plus: voltage: Trivial warning fix 'no return statement'
omap2plus: clockdomain: Trivial fix for build break because of clktrctrl_mask
arm: omap: gpio: don't access irq_desc array directly
omap2+: pm_bus: make functions used as pointers as static
OMAP: GPIO: fix _set_gpio_triggering() for OMAP2+
OMAP2+: TWL: include pm header for init protos
OMAP2+: TWL: make conversion routines static
...
Fix up conflicts in arch/arm/mach-omap2/board-omap3beagle.c ("DVI reset
GPIO" vs "use generic DPI panel driver")
Commit d30e45e (ARM: pgtable: switch order of Linux vs hardware page tables)
introduced a pre-increment addressing offset which is out of range for
Thumb-2. Thumb-2 only permits offsets <256. So split the intruction in
two for Thumb-2.
Signed-off-by: Dave Martin <dave.martin@linaro.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
pte alloc routines must wait for split_huge_page if the pmd is not present
and not null (i.e. pmd_trans_splitting). The additional branches are
optimized away at compile time by pmd_trans_splitting if the config option
is off. However we must pass the vma down in order to know the anon_vma
lock to wait for.
[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The kerneldoc for this function is at odds with the DMA-API
document, which holds, so fix it.
Signed-off-by: Linus Walleij <linus.walleij@stericsson.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This option uses LDREXB/STREXB to emulate SWPB but these instructions
are not supported on all the ARMv6 processors.
Reported-by: Anand Gadiyar <gadiyar@ti.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Leif Lindholm <Leif.Lindholm@arm.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Add ARM support for the DMA debug infrastructure, which allows the
DMA API usage to be debugged.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Replace the page_to_dma() and dma_to_page() macros with their PFN
equivalents. This allows us to map parts of memory which do not have
a struct page allocated to them to bus addresses. This will be used
internally by dma_alloc_coherent()/dma_alloc_writecombine().
Build tested on Versatile, OMAP1, IOP13xx and KS8695.
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Conflicts:
MAINTAINERS
arch/arm/mach-omap2/pm24xx.c
drivers/scsi/bfa/bfa_fcpim.c
Needed to update to apply fixes for which the old branch was too
outdated.
The hardware page tables use an XN bit 'execute never'. Historically,
we've had a Linux 'execute allow' bit, in the positive sense. Get rid
of this artifact as future hardware will continue to have the XN sense.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
FIRST_USER_PGD_NR is now unnecessary, as this has been replaced by
FIRST_USER_ADDRESS except in the architecture code. Fix up the last
usage of FIRST_USER_PGD_NR, and remove the definition.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Remove some knowledge of our 2-level page table layout from the
identity mapping code - we assume that a step size of PGDIR_SIZE will
allow us to step over all entries. While this is true today, it won't
be true in the near future.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
We have two places where we create identity mappings - one when we bring
secondary CPUs online, and one where we setup some mappings for soft-
reboot. Combine these two into a single implementation. Also collect
the identity mapping deletion function.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This switches the ordering of the Linux vs hardware page tables in
each page, thereby eliminating some of the arithmetic in the page
table walks. As we now place the Linux page table at the beginning
of the page, we can deal with the offset in the pgt by simply masking
it away, along with the other control bits.
This also makes the arithmetic all be positive, rather than a mixture.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Since commit 3e4d3af501 "mm: stack based kmap_atomic()", it is actively
wrong to rely on fixed kmap type indices (namely KM_L2_CACHE) as
kmap_atomic() totally ignores them and a concurrent instance of it may
happily reuse any slot for any purpose. Because kmap_atomic() is now
able to deal with reentrancy, we can get rid of the ad hoc mapping here.
While the code is made much simpler, there is a needless cache flush
introduced by the usage of __kunmap_atomic(). It is not clear if the
performance difference to remove that is worth the cost in code
maintenance (I don't think there are that many highmem users on that
platform anyway) but that should be reconsidered when/if someone cares
enough to do some measurements.
Signed-off-by: Nicolas Pitre <nicolas.pitre@linaro.org>
Since commit 3e4d3af501 "mm: stack based kmap_atomic()", it is actively
wrong to rely on fixed kmap type indices (namely KM_L2_CACHE) as
kmap_atomic() totally ignores them and a concurrent instance of it may
happily reuse any slot for any purpose. Because kmap_atomic() is now
able to deal with reentrancy, we can get rid of the ad hoc mapping here,
and we even don't have to disable IRQs anymore (highmem case).
While the code is made much simpler, there is a needless cache flush
introduced by the usage of __kunmap_atomic(). It is not clear if the
performance difference to remove that is worth the cost in code
maintenance (I don't think there are that many highmem users on that
platform if at all anyway).
Signed-off-by: Nicolas Pitre <nicolas.pitre@linaro.org>
Since commit 3e4d3af501 "mm: stack based kmap_atomic()", it is no longer
necessary to carry an ad hoc version of kmap_atomic() added in commit
7e5a69e83b "ARM: 6007/1: fix highmem with VIPT cache and DMA" to cope
with reentrancy.
In fact, it is now actively wrong to rely on fixed kmap type indices
(namely KM_L1_CACHE) as kmap_atomic() totally ignores them now and a
concurrent instance of it may reuse any slot for any purpose.
Signed-off-by: Nicolas Pitre <nicolas.pitre@linaro.org>
Since CPU_PJ4 is shared between PXA95x and MMP2, select CPU_PJ4 in MMP2
configuration.
Signed-off-by: Haojian Zhuang <haojian.zhuang@marvell.com>
Signed-off-by: Eric Miao <eric.y.miao@gmail.com>
The core of PXA955 is PJ4. Add new PJ4 support. And add new macro
CONFIG_PXA95x.
Signed-off-by: Haojian Zhuang <haojian.zhuang@marvell.com>
Signed-off-by: Eric Miao <eric.y.miao@gmail.com>
Cache ownership must be acquired by reading/writing data from the
cache line to make cache operation have the desired effect on the
SMP MPCore CPU. However, the ownership is never acquired in the
v6_dma_inv_range function when cleaning the first line and
flushing the last one, in case the address is not aligned
to D_CACHE_LINE_SIZE boundary.
Fix this by reading/writing data if needed, before performing
cache operations.
While at it, fix v6_dma_flush_range to prevent RWFO outside
the buffer.
Cc: stable@kernel.org
Signed-off-by: Valentine Barshak <vbarshak@mvista.com>
Signed-off-by: George G. Davis <gdavis@mvista.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
The current implementation of the v7_coherent_*_range function assumes
that the D and I cache lines have the same size, which is incorrect
architecturally. This patch adds the icache_line_size macro which reads
the CTR register. The main loop in v7_coherent_*_range is split in two
independent loops or the D and I caches. This also has the performance
advantage that the DSB is moved outside the main loop.
Reported-by: Kevin Sapp <ksapp@quicinc.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
The current implementation of the dcache_line_size macro reads the L1
cache size from the CCSIDR register. This, however, is not guaranteed to
be the smallest cache line in the cache hierarchy. The patch changes to
the macro to use the more architecturally correct CTR register.
Reported-by: Kevin Sapp <ksapp@quicinc.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Directives such as .long and .word do not magically cause the
assembler location counter to become aligned in gas. As a result,
using these directives in code sections can result in misaligned
data words when building a Thumb-2 kernel (CONFIG_THUMB2_KERNEL).
This is a Bad Thing, since the ABI permits the compiler to assume
that fundamental types of word size or above are word- aligned when
accessing them from C. If the data is not really word-aligned,
this can cause impaired performance and stray alignment faults in
some circumstances.
In general, the following rules should be applied when using data
word declaration directives inside code sections:
* .quad and .double:
.align 3
* .long, .word, .single, .float:
.align (or .align 2)
* .short:
No explicit alignment required, since Thumb-2
instructions are always 2 or 4 bytes in size.
immediately after an instruction.
In this specific case, we can achieve the desired alignment by
forcing a 32-bit branch instruction using the W() macro, since the
assembler location counter is already 32-bit aligned in this case.
Reviewed-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Dave Martin <dave.martin@linaro.org>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Remove knowledge of the 2-level wrapping in pgd_free(), and use the
pXd_none_or_clear_bad() macros when checking the entries.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Russell King