Re: [Xen-devel] [PATCH] x86: refine link time stub area related assertion

[Wei Liu <wl@xxxxxxx>]

On Wed, Jan 15, 2020 at 02:36:24PM +0000, Andrew Cooper wrote:
> On 15/01/2020 10:26, Jan Beulich wrote:
> > While it has been me to introduce this, the use of | there has become
> > (and perhaps was from the very beginning) misleading. Rather than
> > avoiding the right side of it when linking the xen.efi intermediate file
> > at a different base address, make the expression cope with that case,
> > thus verifying placement on every step.
> >
> > Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
> 
> Acked-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx> as this is simply a
> rearranging, but...
> 
> >
> > --- a/xen/arch/x86/xen.lds.S
> > +++ b/xen/arch/x86/xen.lds.S
> > @@ -351,8 +351,8 @@ SECTIONS
> >    .comment 0 : { *(.comment) }
> >  }
> >  
> > -ASSERT(__image_base__ > XEN_VIRT_START |
> > -       __2M_rwdata_end <= XEN_VIRT_END - NR_CPUS * PAGE_SIZE,
> > +ASSERT(__2M_rwdata_end <= XEN_VIRT_END - XEN_VIRT_START + __XEN_VIRT_START 
> > -
> > +                          NR_CPUS * PAGE_SIZE,
> 
> ... doesn't this want a stubs_per_page term?  We don't have 4k per cpu.
> 
> Wei: FYI, if you do introduce executable fixmap entries, this ASSERT()
> wants adjusting by NR_X_FIXMAP * PAGE_SIZE as well.

It is a bit problematic to do that if NR_X_FIXMAP is calculated from the
last enum element (see FIXADDR_SIZE)

I can expose a constant for the largest possible numbers of executable
fixed mappings to put in the ASSERT.

Wei.

> 
> ~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel