incubator-nuttx/drivers/video
chao an 4f7f70598a drviers/video: driver lock should in pairs
Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-03 14:43:12 +01:00
..
vnc tools: replace INCDIR to Makefile variable 2022-11-03 19:59:55 +08:00
Kconfig drivers/video: Update API to dump register for debug 2022-10-26 03:57:26 +08:00
Make.defs tools: replace INCDIR to Makefile variable 2022-11-03 19:59:55 +08:00
README.max7456
fb.c
isx012.c drivers/video/isx012: Add missing FAR qualifier 2022-10-26 03:57:26 +08:00
isx012_range.h
isx012_reg.h drivers/video: Fix bug that is_available() is false 2022-05-30 20:38:47 +03:00
isx019.c drivers/video/isx019: Use clock function instead of gettimeofday 2022-10-26 03:57:26 +08:00
isx019_range.h drivers: video: add isx019 camera sensor 2022-05-31 19:52:33 +08:00
isx019_reg.h drivers: video: add isx019 camera sensor 2022-05-31 19:52:33 +08:00
max7456.c Call nxsem_destroy or nxmutex_destry in the error path 2022-10-30 13:56:52 +01:00
ov2640.c
video.c drviers/video: driver lock should in pairs 2022-11-03 14:43:12 +01:00
video_framebuff.c Replace nxsem API when used as a lock with nxmutex API 2022-10-17 15:59:46 +09:00
video_framebuff.h Replace nxsem API when used as a lock with nxmutex API 2022-10-17 15:59:46 +09:00

README.max7456

drivers/video/README.max7456

23 March 2019
Bill Gatliff <bgat@billgatliff.com>

The code in max7456.[ch] is a preliminary device driver for the MAX7456 analog
on-screen-display generator. This SPI slave chip is a popular feature in many
embedded devices due its low cost and power requirements. In particular, you
see it a lot on drone flight-management units.

I use the term "preliminary" because at present, only the most rudimentary
capabilities of the chip are supported:

 * chip reset and startup
 * read and write low-level chip control registers (DEBUG mode only)
 * write CA (Character Address) data to the chip's framebuffer memory

Some key missing features are, in no particular order:

 * VSYNC and HSYNC synchronization (prevents flicker)
 * ability to update NVM (define custom character sets)

If you have a factory-fresh chip, then the datasheet shows you what the factory
character data set looks like. If you've used the chip in other scenarios,
i.e. with Betaflight or similar, then your chip will almost certainly have had
the factory character data replaced with something application-specific.

Either way, you'll probably want to update your character set before long. I
should probably get that working, unless you want to take a look at it
yoruself...

The max7456_register() function starts things rolling. The omnibusf4 target
device provides an example (there may be others by the time you read this).

In normal use, the driver creates a set of interfaces under /dev, i.e.:

/dev/osd0/fb
/dev/osd0/raw   (*)
/dev/osd0/vsync (*)

* - not yet implemented

By writing character data to the "fb" interface, you'll see data appear on the
display. NOTE that the data you write is NOT, for example, ASCII text: it is
the addresses of the characters in the chip's onboard character map.

For example, if entry 42 in your onboard character map is a bitmap that looks
like "H", then when you write the ASCII "*" (decimal 42, hex 2a), you'll see
that "H" appear on your screen.

If you build the code with the DEBUG macro defined, you will see a bunch more interfaces:

/dev/osd0/VM0
/dev/osd0/VM1
/dev/osd/DMM
...
...

These are interfaces to the low-level chip registers, which can be read and/or
written to help you figure out what's going on inside the chip. They're
probably more useful for me than you, but there they are in case I'm wrong
about that.

b.g.