Op di 29 jan. 2019 om 15:10 schreef Kars de Jong <jongk@xxxxxxxxxxxxxx>:
Hi Geert,
Op di 29 jan. 2019 om 10:45 schreef Geert Uytterhoeven <geert@xxxxxxxxxxxxxx>:
The remainder is a fix for address wrap around when there is memory located
at the end of the 32-bit address space.
That part looks OK to me, and is still applicable.
I will retest with just the last part of the patch applied.
I did that last night (using HEAD of Geert's Linux/m68k Git repo), and
it boots to the point of mounting the root fs now. There the boot
process stops, the heartbeat led blinks but there is no activity on
the network it seems.
But now I remember that the new style SDC HIL driver uses a lot of CPU
because it's polling the bus itself.
In my 2.6.20 tree I created a version that uses auto-polling (like the
old keyboard-only HIL driver), I'll see if I can use that or otherwise
test with the old keyboard driver.
Is Roman still with us? (his last message to the list appears to be
from 2008...)
Kind regards,
Kars.
