Messages in this thread | | | From | "Petr Vandrovec" <> | Date | Wed, 9 Aug 2000 15:17:09 MET-1 | Subject | Re: Console fix |
| |
On 9 Aug 00 at 9:08, James Simmons wrote: > > Hmm.I'm looking into vgacon and I just cannot get idea why ordering > > of vgacon_switch and vga_set_palette can matter, as vgacon_switch is > > simple memcpy() and vga_set_palette just programs palette... Looking > > at code again... Oops. Hey, James made a bug... Code in console.c > > should read > > [snip] > > That is what we had before. Linus pointed out the current code in > redraw_screen is correct. Con_switch might change the video mode (fbcon) > which has a impact on how the palette is set. This is true even for > vgacon. You don't want to be setting the color palette via vgacon when > switching back to X. That is also why set_palette is inside the block.
set_palette itself takes care of vcmode != KD_GRAPHICS for X. I do not see any relation between whether console driver updated screen itself and updating palette, so I believe that console.c in 2.4.0-test6-pre9 is correct...
> What was causing the problem was we where not setting the palette for vga > text mode when we where switching away from the X server.
I think that both fixes are needed. With console.c from test6-pre8 "echo ^[]P0FF0000" behaves a bit unexpected during VT switches, as ^^ ESC palette is not reprogrammed... (tested with plain test5 + vgacon; test6-pre9 + vgacon or fbcon does it right). Best regards, Petr Vandrovec vandrove@vc.cvut.cz
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |