[coreboot] [v2] r4238 - in trunk/coreboot-v2/src/northbridge: amd/amdfam10 amd/gx1 amd/gx2 amd/lx intel/e7501 intel/e7520 intel/e7525 intel/i3100 intel/i440bx intel/i82810 intel/i82830 intel/i855gme intel/i855pm via/cn700 via/cx700 via/vt8601 via/vt8623

Luc Verhaegen libv at skynet.be
Sat May 16 13:06:22 CEST 2009


On Sat, May 16, 2009 at 08:34:12AM +0200, Patrick Georgi wrote:
> Am 16.05.2009 04:01, schrieb Kevin O'Connor:
> >>+
> >>+#if HAVE_HIGH_TABLES == 1
> >>+		high_tables_base = (tolmk - HIGH_TABLES_SIZE) * 1024;
> >>+		high_tables_size = HIGH_TABLES_SIZE* 1024;
> >>+		printk_debug("tom: %lx, high_tables_base: %llx, 
> >>high_tables_size: %llx\n", tomk*1024, high_tables_base, high_tables_size);
> >>+#endif
> >>+
> >
> >The above is not correct.  It places the high tables over the vga
> >memory.
> Shouldn't VGA memory be excluded from tolmk, too?

IT should already have been excluded from there. On the cle266 (vt8263), 
it is counted off right before this blob.

This makes me wonder... why is this HIGH_TABLES stuff done in the ram 
controller code and not general? Is this just a transitional phase and 
will this be done generally?

Luc Verhaegen.




More information about the coreboot mailing list