Re: Re[6]: [Etherboot-developers] RFC: Safe booting concept
Brought to you by:
marty_connor,
stefanhajnoczi
|
From: <ebi...@ln...> - 2003-04-29 02:39:54
|
ke...@us... (Ken Yap) writes: > >I'd like to know which RAM regions I can use safely during etherboot > >runs. I cannot use -DRELOCATE at the moment as this crashes my virtual > >machine with the lance driver, but one day it probably should work. > >Can I just get myself e.g. 0x30000-0x307ff for buffer purposes? > > Why not a normal BSS buffer within Etherboot itself? A major point of relocation is that etherboot occupies no memory < 1MB, as that memory is in high demand. So a normal buffer allocated like: static unsigned char buffer[1024]; should work, and because it has no initializer gcc will put it in the bss section so it won't consume memory until runtime. Keeping the memory in etherboot will mean that etherboot will verify that it is not stomped by a loaded image. Anything like a static location is up for grabs by the image that is being loaded. When you get something working I will find it interesting. This definitely calls for an extension to mkelf or similar. Eric |