[parisc-linux] Today's boot experience on a 735

[John David Anglin]

> The _end symbol is not a reliable indicator of where
> the kernel's memory ends.  And the hpux
> bootloader doesn't provide the SOM header from whence this could
> theoretically be calculated.  I haven't tried out the _end trick
> posted last week, but it looks like it'll work (don't call the symbol
> _end if you try it).  As long as your kernel links, the guess is
> probably OK for now.

Ok, I see the problem now.  The section $DLT$ contains a bunch of
undefined common symbols as well as _end.  For example,

boot_cpu_data       |       416|undef |common |$CODE$

The _end defined by the linker is the end of the bss section but not
the end of last section of the kernel.

Why is boot_cpu_data in $CODE$ rather than $DATA$?  There may be
some incorrect .section/.subsection directives somewhere (probably
in the include files).

--
J. David Anglin                                  dave.anglin@nrc.ca
National Research Council of Canada              (613) 990-0752 (FAX: 952-6605)