This follows the ARM change c0177800
("ARM: 6379/1: Assume new page cache pages have dirty D-cache") for the
same rationale:

    There are places in Linux where writes to newly allocated page
    cache pages happen without a subsequent call to flush_dcache_page()
    (several PIO drivers including USB HCD). This patch changes the
    meaning of PG_arch_1 to be PG_dcache_clean and always flush the
    D-cache for a newly mapped page in update_mmu_cache().

This addresses issues seen with executing binaries from MMC, in
addition to some of the other HCDs that don't explicitly do cache
management for their pipe-in buffers.
Requested-by: NYoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

GCC's __builtin_prefetch() was introduced a long time ago, all
supported GCC versions have it. So this patch is to use it for
implementing the prefetch on SH2A and SH4.

The current  prefetch implementation is almost equivalent with
__builtin_prefetch.
The third parameter in the __builtin_prefetch is the locality
that it's not supported on SH architectures.  It has been set
to three and it should be verified if it's suitable for SH2A
as well. I didn't test on this architecture.

The builtin usage should be more efficient that an __asm__
because less barriers, and because the compiler doesn't see the
inst as a "black box" allowing better code generation.

This has been already done on other architectures (see the commit:
0453fb3c).

Many thanks to Christian Bruel <christain.bruel@st.com> for his
support on evaluate the impact of the gcc built-in on SH4 arch.

No regressions found while testing with LMbench on STLinux targets.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NStuart Menefy <stuart.menefy@st.com>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This patch fixes the following error obtained when
compile the Kernel with the VSYSCALL support enabled:

  SYSCALL arch/sh/kernel/vsyscall/vsyscall-trapa.so
  sh4-linux/bin/ld: error in arch/sh/kernel/vsyscall/vsyscall-trapa.o(.eh_frame);
  no .eh_frame_hdr table will be created.
  AS      arch/sh/kernel/vsyscall/vsyscall-syscall.o

Disassembling the vsyscall-trapa.o code, with this fix applied,
we get the __kernel_vsyscall symbol defined; otherwise it was
missing.

Disassembly of section .text:

00000000 <__kernel_vsyscall>:
   0:	10 00       	.word 0x0010
   2:	00 00       	.word 0x0000
[snip]
0000040 <__kernel_sigreturn>:
  40:	05 93       	mov.w	4e <__kernel_sigreturn+0xe>,r3	! 77
  42:	10 c3       	trapa	#16
[snip]

00000060 <__kernel_rt_sigreturn>:
  60:	05 93       	mov.w	6e <__kernel_rt_sigreturn+0xe>,r3	! ad
  62:	10 c3       	trapa	#16
[snip]
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Reviewed-by: NCarmelo Amoroso <carmelo.amoroso@st.com>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Calling sys_cacheflush with ICACHE we can direclty flush
the icache without invoking the flush_cache_all function.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NCarmelo Amoroso <carmelo.amoroso@st.com>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Now that clk_set_rate_ex() is gone, there is also no way to get at rate
setting algo id, which is now also completely unused. Kill it off before
new clock ops start using it.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

These clocks are currently only used inside one .c file and are not
declared in any headers, therefore having them global is useless.
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Given that __in_29bit_mode() is a constant for the non-PMB case, we can
simply use the PMB-facing version of phys_addr_mask() and drop the other
variants.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Now that nommu selects 32BIT we run in to the situation where SH-2A
supports an uncached identity mapping by way of the BSC, while the SH-2
does not. This provides stubs for the PC manglers and tidies up some of
the system*.h mess in the process.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

The nommu code has regressed somewhat in that 29BIT gets set for the
SH-2/2A configs regardless of the fact that they are really 32BIT sans
MMU or PMB. This does a bit of tidying to get nommu properly selecting
32BIT as it was before.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

There was a leftover inw() used here that really just wants to be a
__raw_readw() instead. Convert it over.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

MRSHPC is wholly unused here, no need to trap it specially. If support is
added in the future it can be taken care of via platform data like on the
others.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Platform data takes care of all of these these days, kill them off.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Trivial shuffling and tidying.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

The only reason this board needs to do PIO trapping is for ethernet,
which happens to follow the same scheme as its bigger brother the
edosk7760. With ethernet properly supported through the platform device,
we can kill off the left over PIO abortion.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

The PIO trapping was only for MRSHPC and the SMC ethernet. Given that the
SMC ethernet is already properly handled and that nothing is using the
MRSHPC, none of this is needed.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This code has been untouched since it was merged many years ago, and has
severely bitrotted since, suggesting that the board has no real users
left. Notice of intent to remove has been sent out over the last few
years, with no takers. Kill it off.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Only the secureedge5410 was ever supported by this code, so make the
board specification explicit rather than perpetuating a mach group.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

None of these PIO routines do anything other than basic error checking,
get rid of them and use the generic fallbacks.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

The microdev only has to contend with silly PIO mangling on anything
within the SuperIO range. As each of the SuperIO modules is already
speciail cased, we just shift that logic over to the ioport map.

With microdev PCI never being merged (and being fudamentally broken in
hardware), and the ethernet chip only doing 16-bit accesses already,
there's no need to maintain any of the extra special casing. Kill it all
off.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

This fixes up the __cpu_disable() path's IRQ migration for the
irq_data changes.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Remove unnecessary castings and get rid of dummy pointer in favor of
offsetof() macro in ptrace_32.c. Also use temporary variables and
break long lines in order to improve readability.
Signed-off-by: NNamhyung Kim <namhyung@gmail.com>
Cc: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Fix up the arguments to arch_ptrace() to take account of the fact that
@addr and @data are now unsigned long rather than long as of a preceding
patch in this series.
Signed-off-by: NNamhyung Kim <namhyung@gmail.com>
Cc: <linux-arch@vger.kernel.org>
Acked-by: NRoland McGrath <roland@redhat.com>
Acked-by: NDavid Howells <dhowells@redhat.com>
Acked-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Acked-by: NDavid S. Miller <davem@davemloft.net>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Presently with hardware counter support disabled the backtrace op never
gets initialized. This is a regression over the previous behaviour, so
simply add it back in.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

For the backtrace code its assumed that the stack pointer is 32-bits,
which is not the case with the sh64 registers. Use the shared
kernel_stack_pointer() helper to get at the actual register, which
already takes care of the necessary typecasting.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

op_name_from_perf_id() currently returns a local variable, which isn't
terribly productive. As we only handle a single PMU case for now, simply
allocate and free the string from the arch init/exit context and have
op_name_from_perf_id() hand back the cached string.

This also takes UTS_MACHINE in to account, given that we build for
multiple architectures.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Implement get_user_pages_fast without locking in the fastpath on sh.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Now that sh64 has grown extended page flag support we finally have a free
bit for _PAGE_SPECIAL. Wire it up.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Now that all of the controllers have been fixed up, we can finally select
GENERIC_HARDIRQS_NO_DEPRECATED.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Presently the irq chip is found through the irq_desc, but as this is
going away convert over to an irq_data lookup instead.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

irq_32.c was updated for the new API, while irq_64.c was overlooked. This
syncs them up and gets things building again.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>

Signed-off-by: NPaul Mundt <lethal@linux-sh.org>