Re: Endianness issues using the framebuffer

Interesting idea. Is this a pure software application? The writing suggest 
you modified the kernel. I haven't looked at the code.


On Mon, 27 Jun 2005, James Lehman wrote:

> Check this out!
> 
> EZFB
> http://www.akrobiz.com/ezfb/
> 
> Please let me know what you think!
> 
> ~James.   :o)
> 
> 
> 
> 
> ----- Original Message ----- 
> From: "Roger Leigh" <[email protected]>
> To: <[email protected]>
> Sent: Monday, June 27, 2005 4:46 PM
> Subject: [Linux-fbdev-users] Endianness issues using the framebuffer
> 
> 
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> >
> > Hi folks,
> >
> > I'm looking into writing a program which will use the framebuffer to
> > display rendered fonts and some graphics.  However, I've not yet found
> > much in the way of programming documentation.  Is there anything you
> > would recommend?
> >
> > (I've looked at SDL, directfb and GGI sources, but they are very
> > complex.)
> >
> > Using the comments in <linux/fb.h> and some simple examples I found,
> > I've written the sample code below.  However, none of the examples I
> > saw made any use of the fb_bitfield information.  I ran into problems
> > when I tried my (i386 tested) code on a PPC.  After fixing the code,
> > it still doesn't work.  I have the bizarre situation that if I run it
> > under X11 (using XF86_FBDev), I see the glyphs rendered fine, but if I
> > run it on the console the colours are mucked up, and I can't see any
> > change in the bitfield data.
> >
> > Are there any examples of painting simple RGB pixel data onto any
> > arbitrary framebuffer device which also works on big and little endian
> > systems.  I'm sure this is fairly simple once you know how!
> >
> > Additionally, is there any documentation about how colourmaps and the
> > other framebuffer features work?
> >
> >
> > Many thanks,
> > Roger
> >
> >
> > This program simply renders each ASCII character in turn and paints
> > it.  You might need to change the fb device and font path.  It's
> > currently hard-coded for RBG LCDs, but is simple to change for a CRT.
> >
> > Compile with
> > c99 -o ft-lcd ft-lcd.c `freetype-config --cflags --libs`
> >
> > #include <unistd.h>
> > #include <stdio.h>
> > #include <stdlib.h>
> > #include <fcntl.h>
> > #include <linux/fb.h>
> > #include <sys/ioctl.h>
> > #include <sys/mman.h>
> >
> > #include <ft2build.h>
> > #include FT_FREETYPE_H
> >
> > void fail (const char *reason)
> > {
> >   fprintf (stderr, "%s\n", reason);
> >   exit (EXIT_FAILURE);
> > }
> >
> > int
> > main()
> > {
> >   int fbfd = 0;
> >   struct fb_var_screeninfo vinfo;
> >   struct fb_fix_screeninfo finfo;
> >   long int screensize = 0;
> >   char *fbp = 0;
> >   int x = 0, y = 0;
> >   long int location = 0;
> >
> >   // Open the file for reading and writing
> >   fbfd = open("/dev/fb/0", O_RDWR);
> >   if (fbfd < 0)
> >     fail("Error: cannot open framebuffer device.");
> >
> >   // Get fixed screen information
> >   if (ioctl(fbfd, FBIOGET_FSCREENINFO, &finfo))
> >     fail("Error reading fixed information.");
> >
> >   // Get variable screen information
> >   if (ioctl(fbfd, FBIOGET_VSCREENINFO, &vinfo))
> >     fail("Error reading variable information.");
> >
> >   printf("%ux%u (%ux%u), %ubpp\n", vinfo.xres, vinfo.yres,
> > vinfo.xres_virtual, vinfo.yres_virtual, vinfo.bits_per_pixel );
> >   printf("offsets: x=%u, y=%u\n", vinfo.xoffset, vinfo.yoffset);
> >   printf("red: o=%u, l=%u, mr=%u\n", vinfo.red.offset, vinfo.red.length,
> vinfo.red.msb_right);
> >   printf("green: o=%u, l=%u, mr=%u\n", vinfo.green.offset,
> vinfo.green.length, vinfo.green.msb_right);
> >   printf("blue: o=%u, l=%u, mr=%u\n", vinfo.blue.offset,
> vinfo.blue.length, vinfo.blue.msb_right);
> >   printf("transp: o=%u, l=%u, mr=%u\n", vinfo.transp.offset,
> vinfo.transp.length, vinfo.transp.msb_right);
> >
> >   unsigned int red_max = 1 << vinfo.red.length;
> >   unsigned int green_max = 1 << vinfo.green.length;
> >   unsigned int blue_max = 1 << vinfo.blue.length;
> >   unsigned int transp_max = 1 << vinfo.transp.length;
> >
> >   unsigned int red_mask = 0;
> >   for (unsigned int i=0; i<vinfo.red.length; ++i)
> >     red_mask |= 1 << (vinfo.red.offset + i);
> >   unsigned int green_mask = 0;
> >   for (unsigned int i=0; i<vinfo.green.length; ++i)
> >     green_mask |= 1 << (vinfo.green.offset + i);
> >   unsigned int blue_mask = 0;
> >   for (unsigned int i=0; i<vinfo.blue.length; ++i)
> >     blue_mask |= 1 << (vinfo.blue.offset + i);
> >   unsigned int transp_mask = 0;
> >   for (unsigned int i=0; i<vinfo.transp.length; ++i)
> >     transp_mask |= 1 << (vinfo.transp.offset + i);
> >
> >
> >   // Figure out the size of the screen in bytes
> >   screensize = vinfo.xres * vinfo.yres * vinfo.bits_per_pixel / 8;
> >
> >   // Map the device to memory
> >   fbp = (char *)mmap(0, screensize, PROT_READ | PROT_WRITE, MAP_SHARED,
> >      fbfd, 0);
> >   if ((int)fbp == -1)
> >     fail("Error: failed to map framebuffer device to memory.\n");
> >
> >   // Now setup FreeType
> >   FT_Library library;
> >   FT_Face face;
> >
> >   int error = FT_Init_FreeType( &library );
> >   if ( error )
> >     {
> >       fail ("Failed to initialise FreeType");
> >     }
> >
> >   error = FT_New_Face( library,
> "/usr/share/fonts/truetype/ttf-bitstream-vera/VeraMono.ttf", 0, &face );
> >   if ( error == FT_Err_Unknown_File_Format )
> >     {
> >       fail ("Unsupported font format");
> >     }
> >   else if ( error )
> >     {
> >       fail ("Error reading font file");
> >     }
> >
> >   error = FT_Set_Char_Size ( face,
> >      0,
> >      12 * 64,
> >      100,
> >      100 );
> >   if (error)
> >     {
> >       fail ("Error setting char size");
> >     }
> >
> >   // Display each ASCII printable char
> >   for (FT_ULong n = 33; n < 127; ++n)
> >     {
> >       FT_UInt glyph_index = FT_Get_Char_Index (face, n);
> >
> >       FT_Load_Glyph (face,
> >      glyph_index,
> >      FT_LOAD_DEFAULT|FT_LOAD_TARGET_LCD);
> >
> >       error = FT_Render_Glyph (face->glyph,
> >        FT_RENDER_MODE_LCD);
> >
> >       if (error)
> > {
> >   fail ("Error rendering glyph");
> > }
> >
> >       if (n == '_')
> > continue;
> >
> >       FT_Bitmap *bitmap = &face->glyph->bitmap;
> >
> >       //      printf ("Glyph '%c': %d * %d\n", (char) n, bitmap->width,
> bitmap->rows);
> >
> >       //printf ("Bitmap top: %d\nBitmap left: %d\n",
> face->glyph->bitmap_top,
> >       //      face->glyph->bitmap_left);
> >
> >       for (y=0; y<200; ++y)
> > for (x=0; x<200; ++x)
> >   {
> >     location = ((x + 0)+vinfo.xoffset) *
> >       (vinfo.bits_per_pixel/8) +
> >       ((y + 0)+vinfo.yoffset) * finfo.line_length;
> >
> >     if ( vinfo.bits_per_pixel == 32 ) {
> >       unsigned int *pixel = (int *) (fbp + location);
> >       *pixel = 0;
> >     }
> >   }
> >
> >       for (y=0; y<bitmap->rows; ++y)
> > for (x=0; x< (bitmap->width / 3); ++x)
> >   {
> >     location = ((x + face->glyph->bitmap_left)+vinfo.xoffset) *
> >       (vinfo.bits_per_pixel/8) +
> >       ((y + face->glyph->bitmap_top)+vinfo.yoffset) * finfo.line_length;
> >
> >     unsigned int r = *(bitmap->buffer + (y * bitmap->pitch) + (x*3 + 0));
> >     //     r *= (red_max /8);
> >     unsigned int g = *(bitmap->buffer + (y * bitmap->pitch) + (x*3 + 1));
> >     //     g *= (red_max /8);
> >     unsigned int b = *(bitmap->buffer + (y * bitmap->pitch) + (x*3 + 2));
> >     //     b *= (red_max /8);
> >
> >     // printf("r=%u, g=%u, b=%u\n", r, g, b);
> >
> >     if ( vinfo.bits_per_pixel == 32 ) {
> >       unsigned int *pixel = (int *) (fbp + location);
> >       *pixel = 0;
> >
> >       // This works when I run under X11_FBDev (i.e. the
> >       // colours are correct), but if I run on the console, I
> >       // get a mess and mostly blue colours...
> >
> >       *pixel |= ((r << vinfo.red.offset)   & red_mask);
> >       *pixel |= ((g << vinfo.green.offset) & green_mask);
> >       *pixel |= ((r << vinfo.blue.offset)  & blue_mask);
> >
> >     }
> >   }
> >
> >       sleep (1);
> >
> >     }
> >
> >   return 0;
> > }
> >
> >
> > - -- 
> > Roger Leigh
> >                 Printing on GNU/Linux?  http://gimp-print.sourceforge.net/
> >                 Debian GNU/Linux        http://www.debian.org/
> >                 GPG Public Key: 0x25BFB848.  Please sign and encrypt your
> mail.
> > -----BEGIN PGP SIGNATURE-----
> > Version: GnuPG v1.4.1 (GNU/Linux)
> > Comment: Processed by Mailcrypt 3.5.8 <http://mailcrypt.sourceforge.net/>
> >
> > iD8DBQFCwGWTVcFcaSW/uEgRAkt1AJ9n98Nx0sFOCSQz6FKJ4mftcnijLACgiW6A
> > Ew9wJRanhXh419I9x6PRgzU=
> > =2YCz
> > -----END PGP SIGNATURE-----
> >
> >
> > -------------------------------------------------------
> > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
> > from IBM. Find simple to follow Roadmaps, straightforward articles,
> > informative Webcasts and more! Get everything you need to get up to
> > speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> > _______________________________________________
> > Linux-fbdev-users mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/linux-fbdev-users
> >
> 
> 
> 
> -------------------------------------------------------
> SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
> from IBM. Find simple to follow Roadmaps, straightforward articles,
> informative Webcasts and more! Get everything you need to get up to
> speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> _______________________________________________
> Linux-fbdev-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/linux-fbdev-users
> 


-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click