Hello, I have a couple of questions about the console layer. Specifically, I'm a little confused in regards to how it maintains the frame buffer for each virtual console in vt.c. It seems to me that vc_screenbuf is an internal representation of the active frame buffer. On a VGA system, this just points into VRAM and vt.c is actually directly modifying VGA memory in a number of places via scr_writew. So I guess it is just a happy coincidence that the screenbuf format and VGA's text mode format are the same, and as a result vt.c directly modifies VGA vram with scr_writew in a few places. On other types of displays, presumably screenbuf just points somewhere in normal RAM which is then rendered by the low-level console driver into something that the hardware will understand. Is this right? Is this why vgacon doesn't need to implement putc? Is this done as a speed hack to avoid having to copy bytes around unnecessarily, or is it just how things evolved? I.E. would there be a big downside to moving the framebuffer out of VGA VRAM if it became necessary to implement a few features I'm thinking about? I would think not since fbcon already has a lot more overhead than that to put text on the screen. The idea of vt.c writing directly to VGA VRAM seems a little bit ugly, so I just wanted to make sure I was understanding everything correctly. Thanks, Steve. -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
