On Fri, Jul 29, 2005 at 09:45:31AM -0700, ppc-dev@storix.com wrote: > > I would like to get a better understanding of why this limitation exists. This > information would be helpful to get a jump start at the problem. I suspect > that the available area of memory is only 12mb and this has been split at 6 x > 2. Since the size of the initrd can be changed easier than the kernel, it > might be helpful to incorporate some logic into the max size and calculate > the max size for the initrd based on the amount of memory left over after > bringing across the kernel. (8 + 4 or 7 + 5 etc.) Again, I am grasping at > straws and would like to get some information. basically yaboot currently just does a CLAIM of whatever size that define is set to. with OF though you cannot be guarenteed to have arbitrarily large blobs of memory available, especially since yaboot has already done a big CLAIM for its broken mallocator. for one thing you cannot even be sure where OF decides to live. bottom line we need to deal with memory in a much much more sane manner. that is going to be difficult with the current codebase thats why that has always been a 2.0 issue. -- Ethan Benson http://www.alaska.net/~erbenson/
Attachment:
pgp00004.pgp
Description: PGP signature