Re: [PATCH] x86/head_32: Fix overflow warning with 32-bit binutils

[Ingo Molnar]

* Borislav Petkov <bp@xxxxxxxxx> wrote:

> On Fri, Jun 19, 2015 at 10:10:44AM +0200, Ingo Molnar wrote:
> > 
> > * Borislav Petkov <bp@xxxxxxxxx> wrote:
> > 
> > > From: Borislav Petkov <bp@xxxxxxx>
> > > 
> > > When building the kernel with 32-bit binutils built with support only
> > > for the i386 target, we get the following warning:
> > > 
> > >   arch/x86/kernel/head_32.S:66: Warning: shift count out of range (32 is not between 0 and 31)
> > > 
> > > The problem is that in that case, binutils' internal type representation
> > > is 32-bit wide and the shift range overflows.
> > > 
> > > In order to fix this, manipulate the shift expression which creates the
> > > 4GiB constant to not overflow the shift count.
> > > 
> > > Reported-and-tested-by: Enrico Mioso <mrkiko.rs@xxxxxxxxx>
> > > Suggested-by: Michael Matz <matz@xxxxxxx>
> > > Signed-off-by: Borislav Petkov <bp@xxxxxxx>
> > > ---
> > >  arch/x86/kernel/head_32.S | 4 ++--
> > >  1 file changed, 2 insertions(+), 2 deletions(-)
> > > 
> > > diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S
> > > index 53eeb226657c..c601d1de5ae5 100644
> > > --- a/arch/x86/kernel/head_32.S
> > > +++ b/arch/x86/kernel/head_32.S
> > > @@ -63,8 +63,8 @@
> > >  #endif
> > >  
> > >  /* Number of possible pages in the lowmem region */
> > > -LOWMEM_PAGES = (((1<<32) - __PAGE_OFFSET) >> PAGE_SHIFT)
> > > -	
> > > +LOWMEM_PAGES = (((2<<31) - __PAGE_OFFSET) >> PAGE_SHIFT)
> > > +
> > 
> > Hm, so if internally GAS uses 64-bit types, couldn't we 'make sure' it's a 64-bit 
> > type, by using something like 31LL? Assuming GAS understands that?
> 
> I don't think it does.
> 
> This fix is only when gas is compiled with the i386 target which uses solely 
> 32-bit sized types for internal representation. The warning won't fire on gas 
> built with multiple targets support even if you build with -m32.

Ugh, nasty.

> > Or if the internal representation is 32 bits, then 2<<31 is just a fancy way of 
> > saying '0', right?
> > 
> > So this could be written as:
> > 
> >   LOWMEM_PAGES = (-__PAGE_OFFSET >> PAGE_SHIFT)
> > 
> > or:
> > 
> >   LOWMEM_PAGES = ((0 - __PAGE_OFFSET) >> PAGE_SHIFT)
> > 
> > right?
> 
> Yes, you can do that only when the internal representation is a 32-bit
> type. If it is not and you still do an -m32 build, you then want the
> 4Gib value there and not a 0:
> 
> arch/x86/kernel/head_32.S: Assembler messages:
> arch/x86/kernel/head_32.S:227: Error: value of 18014395285110787 too large for field of 4 bytes at 146

Grumble. I guess the workaround is OK then because I cannot think of any cleaner 
solution - but I'd suggest to put a comment there at minimum, to explain what it's 
about.

... and chances are that GAS might start warning about 2<<31 in the future as 
well.

Thanks,

	Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
Please read the FAQ at  http://www.tux.org/lkml/