This funcionality already exist in private dvb-hisi module (for tuners), so only will compile if not exist dvb-hisi module.

Removed all unecessary data from Hisilicon MSP/COMMON.

[ADDED] Missing Files.
[ADDED] Some Kernel Modules.
[ADDED] Atto DVB Tuners.
[ADDED] Atto RC support of models Pixel, Pixel Premium and Turing.
[ADDED] Atto RC WakeUp Support.
[ADDED] Descrambler Proc Support: /proc/msp/demux_key
[FIXED] Compilation for SoC Hi3798Cv200 / Hi3798Mv200.
[FIXED] I2C GPIO Support for Hi3798Mv200.
[FIXED] Correct GPU Driver for Hi3798Cv200 / Hi3798Mv200.
[FIXED] Demux on dvb-core for Compatibility with Enigma2.
[FIXED] USB 3.0 Warnings.
[FIXED] FUSE to Support 12MB of Data Transfer.
[FIXED] Bluetooth Controllers Connect.
[FIXED] Config Size of CFG_HI_HD0_FB_VRAM_SIZE / CFG_HI_SD0_FB_VRAM_SIZE.
[FIXED] HI_AO Errors in BOOT.

Example:

[01:44:49:241 ERROR-HI_AO]:AO_SND_SetVolume[1485]:NULL pointer
[01:44:49:248 ERROR-HI_AO]:AO_SND_SetVolume[1485]:NULL pointer
[01:44:49:255 ERROR-HI_AO]:AO_SND_SetVolume[1485]:NULL pointer
[01:44:49:263 ERROR-HI_AO]:AO_SND_SetVolume[1485]:NULL pointer
[01:44:49:270 ERROR-HI_AO]:AO_SND_SetMute[1291]:NULL pointer

With this option enabled, Debian system will be unable to access network
as normal, due to socket permission failure.  Let's disable it, since
the target users of this kernel branch and defconfig is Debian system.
Signed-off-by: NShawn Guo <shawn.guo@linaro.org>

Upstream kernel and U-Boot name the board dts as hi3798cv200-poplar.
Let's rename it for alignment.
Signed-off-by: NShawn Guo <shawn.guo@linaro.org>

It adjust kernel options for booting linaro-stretch-developer debian
system into console.  The MT76X2U BT/Wifi support gets disabled along
the way, since we did not really verify it on debian system.
Signed-off-by: NShawn Guo <shawn.guo@linaro.org>

HDMI_TEE_SUPPORT is never set anywhere, so the code that depends on
it can be deleted.
Signed-off-by: NAlex Elder <elder@linaro.org>

CFG_HIFB_PROC_SUPPORT is unconditionally set to be defined in build
commands in "drivers/msp/hifb/Makefile".  Make conditional code
that's based on that symbol be unconditional, and delete its
definition.
Signed-off-by: NAlex Elder <elder@linaro.org>

HDMI_HDCP_SUPPORT is not set for Poplar, so the small bit of code
that's conditional on it being defined can be deleted.
Signed-off-by: NAlex Elder <elder@linaro.org>

New versions of GCC report many new errors.  One of those reports
when the range of possible inputs passed to printf()-like functions
could lead to formatted output consuming more than the space available
in the target buffer.  I examined the locations that were causing
these warnings, and have concluded they are conditions that will not
happen.
Signed-off-by: NAlex Elder <elder@linaro.org>

Fix a spot in "drv_display.c" that uses the array count rather than
the size in bytes of a struct field as the size argument for a
memset() call.
Signed-off-by: NAlex Elder <elder@linaro.org>

Fix a spot in "drv_gpio_i2c.c" that uses the array count rather than
the size in bytes of a struct field as the size argument for a
memset() call.
Signed-off-by: NAlex Elder <elder@linaro.org>

Fix a spot in "hiaiao_dma.c" that uses the array count rather than
the size in bytes of a struct field as the size argument for a
memset() call.
Signed-off-by: NAlex Elder <elder@linaro.org>

Fix places in "sta_cfg.c" that use the array count rather than the
size in bytes of a struct field as the size argument for memset()
calls.
Signed-off-by: NAlex Elder <elder@linaro.org>

A "while" loop was comparing character pointer against '\0' rather
than derereferencing that pointer.  The effect is an infinite loop
(so it seems likely it's never actually used).  But regardless, it's
a bug.
Signed-off-by: NAlex Elder <elder@linaro.org>

A few more symbols are not defined for Poplar, and as a result,
blocks of code and/or entire files can just be removed.
Signed-off-by: NAlex Elder <elder@linaro.org>

In "drivers/msp/base/Makefile", CA_FRAMEWORK_V100_SUPPORT is set
to be define for builds unconditionally.  Delete the two spots of
code that are present when that is *not* defined, and eliminate
the defintion from the Makefile.  A corresponding header can be
deleted as a result.
Signed-off-by: NAlex Elder <elder@linaro.org>

A number of files have become essentially empty, containing no code
and possibly consisting only of #includes of other files.  Delete
the files that aren't needed.  In a couple cases, ensure the deleted
files are no longer used in an #include.
Signed-off-by: NAlex Elder <elder@linaro.org>
					    `

We currently aren't providing a formal user space API for the
HiSilicon code.  It's likely that header files found under
drivers/sdk are at least part of that, but at some point we need to
do a better job of really exposing the interface shared between the
kernel and user space.

Until we have that, there's no reason to protect against compilation
by a C++ compiler, certainly in header files not clearly meant to be
exported there.

So get rid of these "ifdef __cpluplus" blocks at the beginning and
end of C source files that contain them.
Signed-off-by: NAlex Elder <elder@linaro.org>

There is a coding idiom used to support C++ compilers parsing C code
(including header files) that changes some details affecting linkage
of symbols found within the affected file.

Basically no kernel C code uses this.  The kernel is compiled with the
C compiler, not C++.  And until we have an header file in the kernel
source that's exported to user space, there's no reason to protect
against compilation by a C++ compiler.

So get rid of these "ifdef __cpluplus" blocks at the beginning and
end of C source files that contain them.
Signed-off-by: NAlex Elder <elder@linaro.org>

Delete a sort of unrelated set definitions for symbols that are
never used in the kernel code.
Signed-off-by: NAlex Elder <elder@linaro.org>

HI_LOADER_APPLOADER is not defined for Poplar.  Make the code that
depends on it (not being set) be built unconditionally.
Signed-off-by: NAlex Elder <elder@linaro.org>

The symbol JPGE_SUPPORT defined in "drivers/msp/Makefile" is used
only once, 8 lines below its definition.  Just get rid of that
symbol and build the "jpge" directory unconditionally.
Signed-off-by: NAlex Elder <elder@linaro.org>

HiSilicon added the "-follow" flag in the clean targets in the top
level Makefile, presumably because they used symbolic links to point
at the code under drivers/common and drivers/msp (and a few others).
We don't do that any more, so we don't need the "-follow" flag.
Signed-off-by: NAlex Elder <elder@linaro.org>

Kbuild defines the $(src) variable to be a relative path to the
directory that contains the current Makefile.  There remained
a definition of DRV_SRC_BASE in "drivers/msp/keyled/Makefile"
that spells out the relative path explicitly.  Just use $(src)
using it in place of $(DRV_SRC_BASE).
Signed-off-by: NAlex Elder <elder@linaro.org>

Delete some more definitions under "drivers/msp/include" that
are never used.
Signed-off-by: NAlex Elder <elder@linaro.org>

There are a lot of definitions of under "drivers/msp/include" that
are never used.  The ones deleted here are just the result of sort
of randomly poking around in a few files, and deleting things that
are never used.  There could easily be a lot more patches like this.
Signed-off-by: NAlex Elder <elder@linaro.org>

We only support 64-bit operation for Poplar, and AARCH64 CPUs are
always expected to be SMP.  Assuming that allows us to delete some
conditional code.
Signed-off-by: NAlex Elder <elder@linaro.org>

We only support 64-bit mode on Poplar.  So we can eliminate
conditional code that's dependent on that in drivers/common and
drivers/msp.  We don't touch the Mali drivers located under
"drivers/msp/drivers/msp/gpu/mali700" however.
Signed-off-by: NAlex Elder <elder@linaro.org>

VFMW_VPSS_BYPASS_EN is not defined for Poplar, and a lot of
conditional code can be deleted as a result.

Signed-off-by: Alex Elder <elder@linaro.org> `

There is a set of symbols defined in "drv_vdec_ext.h" used for
conditionally building somecode.  We know the values assigned to
these symbols (or that they are defined), so we can make the code
unconditional and delete the block of symbols.

The symbols are:
    PRE_ALLOC_VDEC_VDH_MMZ is not defined
    PRE_ALLOC_VDEC_SCD_MMZ is 0
    PRE_ALLOC_VDEC_ESBUF_MMZ is 0
Signed-off-by: NAlex Elder <elder@linaro.org>

There is a set of symbols defined in "vpss_common.h" used for
conditionally building VPSS code.  We know the values assigned to
these symbols (or that they are defined), so we can make the code
unconditional and delete the block of symbols.

They are:
    ZME_2L_TEST is defined
    FB_DBG is 0
    DEF_VPSS_STATIC is 0
    DEF_VPSS_DEBUG is 1
    DEF_VPSS_HIGH_4K is 1
    DEF_VPSS_ANDROID_BYPASS is 1
    VPSS_ISLOSS_CMP is HI_FALSE
    DEF_VPSS_LOGIC_WORK is 1
Signed-off-by: NAlex Elder <elder@linaro.org>