Subject: Kernel and compiler troubles
To: RiscBSD mailing list <port-arm32@NetBSD.ORG>
From: David Croft <david@infotrek.demon.co.uk>
List: port-arm32
Date: 11/08/1996 20:49:55
I have 66 MB of RAM, so I've been using the BSD-4634-64MB (Protagonist)
kernel. I have now tried the Voyager 4702 kernel from the FTP site, and
get as far as this on booting:

...
swap dev 1001->30965760 bytes
spl0=ffffffff splsoft=fffffe7d splbio=00108409 splnet=00100001
spltty=00100001 splclock=00100001 splimp=00100001 splhigh=00000000
clock: hz=100 stathz = 0 profhz = 0
kmem_suballoc: bad status return of 3.
panic: kmem_suballoc
Stopped at     0xf01184bc:  ldmdb r11, {r11, r13, r15}
db>

The same problem occurs with the Protagonist 4705 kernel that Scott
compiled for me at the show. (2 Nov 13:19)

Do these newer kernels not support 64MB machines, or am I yet again doing
something stupid? :-)

ARM710, 1.2-beta sets, 1.2-release bootloader, 32+32+2 RAM, etc.


On a different note, I'm having trouble with one of the programs I use for
my work. It appears to be related to the warnings I get at compile time.
These warning do not appear when compiling for x86 Linux:

run.c: In function `load_persona':
run.c:679: warning: comparison is always 1 due to limited range of data type

The message occurs several times, sometimes saying that it is always 0.
The aforementioned line reads:

if (*temp2 > -1 && *temp2 < 14)

The next one is "comparison is always 0", and the line is:

while (*temp > 13 || *temp < 0)

The variables `temp' and `temp2' aren't declared anywhere, so I guess
they're declared somewhere in the libraries? I'm not a C programmer,
can anyway suggest how to fix this?


Finally, would any like me to turn Pine into a set?
(If so, could you send me your split program please, Mark?)

Many thanks.

-- 
David Croft                     mailto:david@infotrek.demon.co.uk
Infotrek Software Developments    http://www.infotrek.demon.co.uk/