Re: [PATCH] access gdtdesc on segment 0 unconditionally (Re: [PATCH] i386-qemu port)

[Robert Millan]

On Mon, Jun 22, 2009 at 03:39:03PM -0400, Pavel Roskin wrote:
> On Mon, 2009-06-22 at 11:52 +0200, Robert Millan wrote: 
> > > Since %cs is pointing to the code, it should be possible to point it to
> > > gdtdesc.  They should be nearby.
> > 
> > It is nearby, but the address reference for `gdtdesc' is absolute, NOT
> > relative to %cs.  Of course, when %cs is 0 that's no problem.  But in my
> > case I can't set %cs to 0 because my code is above 0x10000.
> 
> I think we can remove ADDR32 from the command.  I tried that on i386-pc
> and it works in qemu and on real hardware.  I don't see any need to use
> a 32-bit address in the lgdt command.

This won't build.  I don't think it's possible to use relative addresses
with this particular instruction.  "DATA32 lgdt %cs:gdtdesc" results in:

  boot_img-boot_i386_qemu_boot.o: In function `real_to_prot':
  (.text+0x64): relocation truncated to fit: R_386_16 against `.text'

What's the problem with removing %cs?  It's presence there is bogus.  It
*seems* to indicate gdtdesc is a segment-relative reference, but in fact
it's not, and it just happens to work because %cs was set to 0.

Note: "ADDR32 lgdt %cs:gdtdesc" builds, but generates an absolute address
too.

-- 
Robert Millan

  The DRM opt-in fallacy: "Your data belongs to us. We will decide when (and
  how) you may access your data; but nobody's threatening your freedom: we
  still allow you to remove your data and not access it at all."