When the function tracer starts modifying the code via breakpoints
it sets a variable (modifying_ftrace_code) to inform the breakpoint
handler to call the ftrace int3 code.

But there's no synchronization between setting this code and the
handler, thus it is possible for the handler to be called on another
CPU before it sees the variable. This will cause a kernel crash as
the int3 handler will not know what to do with it.

I originally added smp_mb()'s to force the visibility of the variable
but H. Peter Anvin suggested that I just make it atomic.

[ Added comments as suggested by Peter Zijlstra ]
Suggested-by: NH. Peter Anvin <hpa@zytor.com>
Signed-off-by: NSteven Rostedt <rostedt@goodmis.org>

Signed-off-by: NRichard Kuo <rkuo@codeaurora.org>

Commit d065bd81
(mm: retry page fault when blocking on disk transfer) and
commit 37b23e05
(x86,mm: make pagefault killable)

The above commits introduced changes into the x86 pagefault handler
for making the page fault handler retryable as well as killable.

These changes reduce the mmap_sem hold time, which is crucial
during OOM killer invocation.

Port these changes to hexagon.
Signed-off-by: NKautuk Consul <consul.kautuk@gmail.com>
Signed-off-by: NRichard Kuo <rkuo@codeaurora.org>

This function is only used in the same file, and the other similar
functions in this file are also static.
Signed-off-by: NJulia Lawall <Julia.Lawall@lip6.fr>
[rkuo@codeaurora.org: adjusted patch to apply to latest]
Signed-off-by: NRichard Kuo <rkuo@codeaurora.org>

"make versioncheck" points out that arch/hexagon/include/asm/spinlock_types.h
does not need to include version.h .
A quick look at the file seems to confirm its findings, so here's a patch that
removes the include.
Signed-off-by: NJesper Juhl <jj@chaosbits.net>
Signed-off-by: NRichard Kuo <rkuo@codeaurora.org>

 Use resource_size function on resource object
 instead of explicit computation.

 The semantic patch that makes this change is available
 in scripts/coccinelle/api/resource_size.cocci.

 More information about semantic patching is available at
 http://coccinelle.lip6.fr/Signed-off-by: NThomas Meyer <thomas@m3y3r.de>
Signed-off-by: NRichard Kuo <rkuo@codeaurora.org>

At present reserving the IRLs in the IRQ bitmap in addition to the
dropping of the legacy IRQ pre-allocation prevent IRL IRQs from being
allocated for the x3proto board.

The only reason to permit reservations was to lock down possible hardware
vectors prior to dynamic IRQ scanning, but this doesn't matter much given
that the hardware controller configuration is sorted before we get around
to doing any dynamic IRQ allocation anyways. Beyond that, all of the
tables are __init annotated, so quite a bit more work would need to be
done to support reconfiguring things like IRL controllers on the fly,
much more than would ever make it worth the hassle.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This is a partial revert of

	15ed103a ("edac: Fix spelling errors")
	6997991a ("mips: Fix printk typos in arc/mips")

which change code that doesn't exist any more in edac/mips trees.
Reported-by: NStephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

setting CONFIG_IRQ_ALL_CPUS distributes IRQs to CPUs only when
the number of online CPUs equals NR_CPUS.  See commit
280ff974 "sparc64: fix and
optimize irq distribution" for more details.

Using the online mask fixes IRQ-to-CPU distribution on systems
that boot with less than NR_CPUS.
Signed-off-by: NKim Phillips <kim.phillips@freescale.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

This reverts commit 1b788400.

It causes oopses when passed incorrect arguments and has a
design fault using IPIs with interrupts disabled.
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
---

--
Signed-off-by: NAnshuman Khandual <khandual@linux.vnet.ibm.com>
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

never used...
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

It's already called if TIF_NOTIFY_RESUME is set, so we only
need to add the actual work.  Note that checking for RESTORE_SIGMASK
was not needed - set_restore_sigmask() also sets SIGPENDING, so
we never RESTORE_SIGMASK without SIGPENDING.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

TIF_NOTIFY_RESUME added (as bit 5).  That way nommu glue needs no changes at
all; mmu one needs just to replace jmi do_signal_return to jne do_signal_return
There we have flags shifted up, until bit 6 (SIGPENDING) is in MSBit; instead
of checking that MSBit is set (jmi) we check that MSBit or something below it
is set (jne); bits 0..4 are never set, so that's precisely "bit 6 or bit 5 is
set".

Usual handling of NOTIFY_RESUME/SIGPENDING is done in do_notify_resume(); glue
calls it instead of do_signal().
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

It used to be true, until 2.1.78 (14 years ago) when we switched to
do_sigaction()...
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

... and struct old_sigaction never used
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

* move force_sigsegv() (from setup...frame()) and clearing RESTART_SIGMASK
(from do_signal()) into hanlde_signal()
* get rid of handle_signal() return value and oldset argument
* checking for TIF_SIGPENDING is enough; set_restart_sigmask() sets this
one as well.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

... especially since we don't have the right k_sigaction here,
so resetting sa_handler won't work.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

It's been a dead code since commit 571202f5
Author: Michal Simek <monstr@monstr.eu>
Date:   Fri Dec 11 12:54:04 2009 +0100

    microblaze: Remove rt_sigsuspend wrapper

    Generic rt_sigsuspend syscalls doesn't need any asm wrapper.

but that commit has only removed it from entry.S, missing one in entry-nommu.S.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

... and resetting sa_handler in local copy filled by get_signal_to_deliver()
is obviously pointless anyway.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Complete the move of sh64 to it, trim the crap from prototypes,
tidy up a bit.  Infrastructure in do_signal() had already been
there, in signal_64 as well as in signal_32 (where it was already
used).
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Optimize RAID5 xor checksumming by taking advantage of
256-bit YMM registers introduced in AVX.
Signed-off-by: NJim Kukunas <james.t.kukunas@linux.intel.com>
Signed-off-by: NNeilBrown <neilb@suse.de>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>