Ensure that cache operations complete before returning from
mips_cache_reset by placing a completion barrier (sync instruction)
before the return. Without this there is no guarantee that the cache ops
will complete before any subsequent memory accesses, since they are
indexed cache ops & thus not implicitly ordered with memory accesses.

Signed-off-by: Paul Burton <paul.burton@imgtec.com>

Writing to the coprocessor 0 TagLo registers introduces an execution
hazard in that we need that write to complete before any cache
instructions execute. Ensure that hazard is cleared by inserting an ehb
instruction between the TagLo writes & cache op loop.

Signed-off-by: Paul Burton <paul.burton@imgtec.com>

MIPS Linux expects the bootloader to leave the boot CPU a member of the
coherent domain when running on a system with a CM, and we will need to
do so if we wish to make use of IOCUs to have cache-coherent DMA in
U-Boot (and on some systems there is no choice in that matter). When a
CM is present, join the coherent domain after completing cache
initialisation.

Signed-off-by: Paul Burton <paul.burton@imgtec.com>

This patch adds support for initialising & maintaining L2 caches on MIPS
systems. The L2 cache configuration may be advertised through either
coprocessor 0 or the MIPS Coherence Manager depending upon the system,
and support for both is included.

If the L2 can be bypassed then we bypass it early in boot & initialise
the L1 caches first, such that we can start making use of the L1
instruction cache as early as possible. Otherwise we initialise the L2
first such that the L1s have no opportunity to generate access to the
uninitialised L2.

Signed-off-by: Paul Burton <paul.burton@imgtec.com>

Define names for registers holding cache sizes throughout
mips_cache_reset, in order to make the code easier to read & allow for
changing register assignments more easily.

Signed-off-by: Paul Burton <paul.burton@imgtec.com>

Enable use of the instruction cache immediately after it has been
initialised. This will only take effect if U-Boot was linked to run from
kseg0 rather than kseg1, but when this is the case the data cache
initialisation code will run cached & thus significantly faster.

Signed-off-by: Paul Burton <paul.burton@imgtec.com>

Allow L1 Icache & L1 Dcache line size to be specified separately, since
there's no architectural mandate that they be the same. The
[id]cache_line_size functions are tidied up to take advantage of the
fact that the Kconfig entries are always present to simply check them
for zero rather than needing to #ifdef on their presence.

Signed-off-by: Paul Burton <paul.burton@imgtec.com>
[removed CONFIG_SYS_CACHELINE_SIZE in include/configs/pic32mzdask.h]
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>

Move details of the L1 cache line sizes & total sizes into Kconfig,
defaulting to 0. A new CONFIG_SYS_CACHE_SIZE_AUTO Kconfig entry is
introduced to allow platforms to select auto-detection of cache sizes,
and it defaults to being enabled if none of the cache sizes are set by
the configuration (ie. sizes are all the default 0), and code is
adjusted to #ifdef on that rather than on the definition of the sizes
(which will always be defined even if 0).

Signed-off-by: Paul Burton <paul.burton@imgtec.com>

In MIPS assembly there have historically been 2 variants of immediate
addition - the standard "addi" which traps if an overflow occurs, and
the unchecked "addiu" which does not trap on overflow. In release 6 of
the MIPS architecture the trapping variants of immediate addition &
subtraction have been removed. In preparation for supporting MIPSr6,
stop using the trapping instructions from assembly & switch to their
unchecked variants.

Signed-off-by: Paul Burton <paul.burton@imgtec.com>

Update definitions for processor, registers as well as assemby
macros.

Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>