Messages in this thread | | | Date | Thu, 25 Nov 1999 00:02:23 +0100 (CET) | From | Ingo Molnar <> | Subject | Re: [patch] 2.3.29 alpha updates |
| |
On Wed, 24 Nov 1999, Andrea Arcangeli wrote:
> I cleanedup the mem init stuff and I made the allocation of memory > completly based on the memory layout provided by the bios. So far part of > the dyamic memory allocation worked assuming there was some not well > defined available RAM after the kernel memory. Now on alpha nothing works > by magic anymore.
well, this is probably mostly semantics anyway - because the boot loading process (which loads the kernel image into RAM) doesnt analyze the BIOS bootmap either (on x86, i dont know about Alpha). We basically can assume that if a kernel image fits into some piece of memory, then the bootmap fits there as well. The bootmap on a box with 1GB RAM is 32k on x86, and 16k on the Alpha, so we are not talking big memory here.
the real problem was never the bootmap but mem_map[]: much bigger, and potentially problematic. This is mainly why bootmem.c got written.
-- mingo
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |