nanogui: Correct value for SCREEN_PIXTYPE when building mwin for framebuffer
Subject:
Re: [nanogui] Correct value for SCREEN_PIXTYPE when building mwin for framebuffer
From:
"Greg Haerr" ####@####.####
Date:
15 Feb 2003 21:38:57 -0000
Message-Id: <0f0a01c2d539$95788170$6401a8c0@gregnewport>
Patrick,
Understanding SCREEN_PIXTYPE is a bit convuluted,
I must admit. Basically, the SCREEN_PIXTYPE define
is only used in order to generate smaller storage sizes for the
MWPIXELVAL, which is used to hold pixels when using
GrReadArea and some other functions. Microwindows will
always calculate the pixel values correctly regardless of the
setting of SCREEN_PIXTYPE. That's why I recommend
using MWPF_TRUECOLOR0888, which allows the binary
to run on any framebuffer, since it allocates a 32-bit MWPIXELVAL.
If you set SCREEN_PIXTYPE to 565 or 555, it will generate
an 16-bit storage location for all pixel values internally. But
theses are only really used by GrArea, GrReadArea and a few
other functions that pack/unpack pixels into arrays.
Only for those designs where truely the smallest storage is
desired should SCREEN_PIXTYPE be used.
Hope this clarifies some things,
Greg
I am somewhat uncertain how to set the SCREEN_PIXTYPE parameter in the
Microwindows configuration file when building Microwindows for a framebuffer
system.
The documentation in the framebuffer configuration files (either
Configs/config.fb or Configs/config.fb2) states:
# If using Linux framebuffer, set to MWPF_TRUECOLOR0888, and use fbset.
Does this mean use fbset to set the framebuffer to have a pixel depth of 32
bits (e.g. invoking something like fbset -depth 32 at the linux prompt)?
Suppose I am developing for deployment on a 16-bpp framebuffer. Then should
I set SCREEN_PIXTYPE to
MWPF_TRUECOLOR565
or
MWPF_TRUECOLOR555
in the config file?
Thank you.
Many regards,
pat
