Skip to contentSkip to conversation

Truoble with 8-pixel screen fonts and 2.6 kernels (2.6.9, 2.6.10)

DISCUSSION Alexander E. Patrakov 1 participant 2 messages Jan 12, 2005 – Jan 13, 2005

lore
Share & tools
lore
#0 lore raw

(This report is on behalf of Declan Moriarty)

To reproduce the problem, install the "kbd" package, and from the first virtual console execute the following command:

setfont lat1-08 # or any other 8-pixel font

Then switch to the second console and see that the cursor disappeared. Moreover, the following escape sequences function improperly:

echo -e '\033[?Xc' where X is a digit

Even more, from the second console this command

setfont lat1-16

gives 43 screen lines with only upper halves of characters (bottoms are cut, and I suspect this also happens with the cursor).

Looks like a bug in the character cell height logic. Could you please fix it or at least point me to some overview of the relevant source files?

Signature · 1 line
Alexander E. Patrakov
#1 lore raw

Re: Trouble with 8-pixel screen fonts and 2.6 kernels (2.6.9, 2.6.10)

Quoted 6 lines · Alexander E. Patrakov

(This report is on behalf of Declan Moriarty)

To reproduce the problem, install the "kbd" package, and from the first virtual console execute the following command:

setfont lat1-08 # or any other 8-pixel font

Forgot to say that this bug is reproducible only on non-framebuffer console.

Quoted 7 lines

Then switch to the second console and see that the cursor disappeared. Moreover, the following escape sequences function improperly:

echo -e '\033[?Xc' where X is a digit

Even more, from the second console this command

Quoted 7 lines

setfont lat1-16

gives 43 screen lines with only upper halves of characters (bottoms are cut, and I suspect this also happens with the cursor).

Looks like a bug in the character cell height logic. Could you please fix it or at least point me to some overview of the relevant source files?

This bug seems to be Radeon-specific (doesn't exist on ATI Mach64). But even on Radeons, it can't be reproduced in 2.4 kernels - so it's a kernel bug.

Signature · 1 line
Alexander E. Patrakov