Jordan Crouse wrote:
Having a copious amount of time on my hands right now, I had occasion to review the coreboot v2 multiboot patch from some time back.
http://www.coreboot.org/pipermail/coreboot/2008-September/039364.html
It looks technically correct, and it compiles and works. I did see one issue:
Here is the memory map from the coreboot table by way of coreinfo:
-- Memory Map -- Table: 0000000000000000 - 0000000000000fff RAM: 0000000000001000 - 00000000000effff Table: 00000000000f0000 - 00000000000fffff RAM: 0000000000100000 - 0000000007ffffff
And the multiboot map, also courtesy of coreinfo:
-- Memory Map -- Reserved: 0000000000000000 - 0000000000000527 Reserved: 00000000000f0000 - 00000000000f0457 RAM: 0000000000000528 - 00000000000effff RAM: 00000000000f0458 - 0000000007ffffff
Multiboot is being rather generous in giving back memory to the system - I think that the two table should match - or failing that, at least multiboot should round up to the nearest 4k page. V3 has the same issue.
Other then that, I'm cool with acking this.
Unless I hear otherwise, I _am_ going to ack it with the minor modification of rounding up the memory regions to the nearest 4k boundary.
Jordan