Use a tab on second and subsequent lines of multiline #if's, for
consistency with the next commit.
Signed-off-by: NKevin Cernekee <cernekee@gmail.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/2954/Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Remove the CONFIG_SOC_AU1X00 Kconfig symbol since its job can also be done
by MACH_ALCHEMY, now renamed to MIPS_ALCHEMY.
Signed-off-by: NManuel Lauss <manuel.lauss@googlemail.com>
To: Linux-MIPS <linux-mips@linux-mips.org>
Patchwork: https://patchwork.linux-mips.org/patch/1461/Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Quoting from Loongson2FUserGuide.pdf:

5.22.1 Hazards
The processor detects most of the pipeline hazards in hardware, including
CP0 hazards and load hazards. No NOP instructions are required to correct
instruction sequences.
Signed-off-by: NZhang Le <r0bertz@gentoo.org>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

The Alchemy manuals state:

"All pipeline hazards and dependencies are enforced by hardware interlocks
 so that any sequence of instructions is guaranteed to execute correctly.
 Therefore, it is not necessary to pad legacy MIPS hazards (such as
 load delay slots and coprocessor accesses) with NOPs."

Run-tested on Au12x0, without any ill effects.
Signed-off-by: NManuel Lauss <mano@roarinelk.homelinux.net>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Current VR5500 processor support lacks of some functions which are
expected to be configured/synthesized on arch initialization.

Here're some VR5500A spec notes:

* All execution hazards are handled in hardware.

* Once VR5500A stops the operation of the pipeline by WAIT instruction,
  it could return from the standby mode only when either a reset, NMI
  request, or all enabled interrupts is/are detected.  In other words,
  if interrupts are disabled by Status.IE=0, it keeps in standby mode
  even when interrupts are internally asserted.

  Notes on WAIT: The operation of the processor is undefined if WAIT
  insn is in the branch delay slot.  The operation is also undefined
  if WAIT insn is executed when Status.EXL and Status.ERL are set to 1.

* VR5500A core only implements the Load prefetch.

With these changes, it boots fine.
Signed-off-by: NShinya Kuribayashi <shinya.kuribayashi@necel.com>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

For Cavium CPU, we treat the same as R10000, in that all hazards
are dealt with in hardware.
Signed-off-by: NTomaso Paoletti <tpaoletti@caviumnetworks.com>
Signed-off-by: NPaul Gortmaker <Paul.Gortmaker@windriver.com>
Signed-off-by: NDavid Daney <ddaney@caviumnetworks.com>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Tested with Malta; inflates malta_defconfig by 3932 bytes.  Ideally there
should be additional configuration to allow getting rid of this overhead
but that would be too much complexity at this stage of the release cycle.
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Following a strict interpretation the empty definition of irq_enable_hazard
has always been a bug - but an intentional one because it didn't bite.
This has now changed, for uniprocessor kernels mm/slab.c:do_drain()

[...]
        on_each_cpu(do_drain, cachep, 1, 1);
        check_irq_on();
[...]

may be compiled into a mtc0 c0_status; mfc0 c0_status sequence resulting
in a back-to-back hazard.
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Following a strict interpretation the empty definition of irq_enable_hazard
has always been a bug - but an intentional one because it didn't bite.
This has now changed, for uniprocessor kernels mm/slab.c:do_drain()

[...]
        on_each_cpu(do_drain, cachep, 1, 1);
        check_irq_on();
[...]

may be compiled into a mtc0 c0_status; mfc0 c0_status sequence resulting
in a back-to-back hazard.
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Move FPU hazard handling to hazards.h and provide proper support for
MIPSR2 processors
Signed-off-by: NChris Dearman <chris@mips.com>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

In the the sequence:
        ei
        ..
        mfc0    $x, $status

the mfc0 may not see the SR_IE bit set. This was a deliberate bug in the
kernel code because we knew this was a safe thing to do on all R2 silicon
so far but new silicon is changing this.
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

NEC VR4111 and VR4121 need one more nop with mtc0_tlbw_hazard().
Signed-off-by: NYoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Mostly based on patch by Chris Dearman and cleanups from Yoichi.
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

    
The SB1 core has a three cycle interrupt disable hazard but we were
wrongly treating it as fully interlocked.
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

    
Gcc has a tradition of misscompiling the previous construct using the
address of a label as argument to inline assembler.  Gas otoh has the
annoying difference between la and dla which are only usable for 32-bit
rsp. 64-bit code, so can't be used without conditional compilation.
The alterantive is switching the assembler to 64-bit code which happens
to work right even for 32-bit code ...
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

back mtc0 / mfc0 pairs from the same coprocessor register.
Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.

Let it rip!