Bug#392078: cannot boot 2.6.18-1 on sunblade 100

[Jurij Smakov <jurij@wooyd.org>]

Hi Jim,

I'm fairly certain that all the problems arise due to the fact that 
the address of the clock registers on your machine is calculated 
incorrectly for some reason in 2.6.18, but correctly in 2.6.16. If we 
can confirm it, that would be a good starting point in understanding 
of what went wrong.

In 2.6.16 the calculation of this address happened in time.c (all the 
files mentioned are in arch/sparc64/kernel directory) . In 2.6.18 it 
happens in function build_device_resources() in of_device.c, as far as 
I can tell. Later the value obtained there is used in time.c. It would 
be great if you could print out few addresses both in 2.6.16 and 
2.6.18, so we could compare. For 2.6.18 you could insert a line like

prom_printf("mstk48t59_regs=%p mstk48t02_regs=%p regs=%p\n", mstk48t59_regs, mstk48t02_regs, regs);

around line 808 in time.c (right after these variables have been set). 
Similarly, for 2.6.16 you could insert a line

prom_printf("mstk48t59_regs=%p mstk48t02_regs=%p\n", mstk48t59_regs, mstk48t02_regs);

around line 895 in time.c. If the values do not match, we'll at least 
know what to look for.

Best regards,
-- 
Jurij Smakov                                           jurij@wooyd.org
Key: http://www.wooyd.org/pgpkey/                      KeyID: C99E03CC