> [The bootloader is] run in physical mode, and writing to non-existent
> memory doesn't appear to produce any sort of bus fault--

This matches my experience, and applies to even a fully running OS.  (I
hacked on the kernel so that /dev/mem allowed access to the first 64M
of physical space regardless, and found that I could access nonexistent
memory just fine, provided I didn't mind not necessarily getting back
what I wrote.)

> the bootloader appears to fail with an illegal instruction after
> jumping to an area with no RAM.  Given this, one could write a test
> pattern into each 4 MB memory range above the first until one reads
> back correctly, and then relocate there.

Are there models of SPARC on which the PROM does not provide an
"available memory" map?  I would hesitate to make it do its own memory
testing, especially since the ROM code steals a small amount of memory
somewhere for its own purposes, and if we trash that things are liable
to break - and the only reliable way to know where it is is to ask the
code that set it up.

					der Mouse

			       mouse@rodents.montreal.qc.ca
		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B