Re: loading kernels with different LMA and VMA (fwd)

[Rogelio M. Serrano Jr.]

On 28 Mar 2001, Gordon Matzigkeit wrote:

> >>>>> Rogelio M Serrano writes:
> 
> 105      if (type == KERNEL_TYPE_MULTIBOOT)
> 106        entry_addr = (entry_func) pu.elf->e_entry;
> 107      else
> 108        entry_addr = (entry_func) (pu.elf->e_entry & 0xFFFFFF);
> 109
> 110      if (((int) entry_addr) < 0x100000)
> 111        errnum = ERR_BELOW_1MB;
> 
>  RMS> in line 110 if my e_entry is 0xf0100000 we should not get an
>  RMS> error right?  same with line 520: my p_addr this time is
>  RMS> 0x100000
> 
> [...]
> 
>  RMS> am i getting the error at 111?
> 
> The problem I see is that the test at 110 looks like:
> 
>       if (((int) 0xf0100000) < 0x100000)
> 
> Try this on your own, and you'll see that the problem is the (int)

yes i noticed it already. i thought of using unsigned int instead of int
but it looks like your suggestion is better anyway ill try both.
 
> cast turning your large address into a negative number.  Perhaps a
> better test would be:
> 
>       if (entry_addr < (entry_func) 0x100000)
> 
> Of course, this is all speculation, since I can't check the results
> myself at the moment.
> 
> I hope this helps you,
>
 
It does help. The remaining problem now is what physical address to use
for an entry address if the entry address that i get from the elf header
is virtual?

perhaps we use the _start? or maybe extend the multiboot_header struct to
indicate the physical entry point? 

could this be the reason why the grub implementation assumes that LMA =
VMA?
 
> -- 
>  Gordon Matzigkeit <address@hidden>  //\ I'm a FIG (http://fig.org/)
> Committed to freedom and diversity \//  Run Bash (http://fig.org/gnu/bash/)
>