Re: [PATCH] NX protection for kernel data : fix 32 bits S3 suspend

[castet . matthieu]

Quoting Ingo Molnar <mingo@xxxxxxx>:

>
> * matthieu castet <castet.matthieu@xxxxxxx> wrote:
>
> >  static inline int is_kernel_text(unsigned long addr)
> >  {
> > +#if defined(CONFIG_X86_32) && defined(CONFIG_ACPI_SLEEP)
> > +	/*
> > +	 * We need to make the wakeup trampoline in first 1MB !NX
> > +	 */
> > +	if (addr >= PAGE_OFFSET && addr <= (PAGE_OFFSET + (1<<20)))
> > +		return 1;
> > +#endif
>
> That's pretty ugly. Why not use set_memory_x()/set_memory_nx(), and only for
> the
> trampoline itself? Does the whole 1MB need to be marked X?
The previous code was doing that.

If you prefer I can revert to the old code :

static inline int is_kernel_text(unsigned long addr)
{
    if (addr >= PAGE_OFFSET && addr <= (unsigned long)__init_end)
        return 1;
    return 0;
}

The proper fix will be done in 2.6.29, when S3 resume 32 bit trampoline will be
merged with 64 bit (ie call smp trampoline) [1]

>
> Same goes for the Xen fix.
>
The Xen fix reverts code to old behavior.
We can't assume that all data/bss will be RW. And I see no way to detect witch
page in data/bss we should not force to RW in static protection.
So we assume the code that change protection of data/bss knows what it is doing
(that what was doing old code).

Matthieu


[1]http://marc.info/?l=linux-kernel&m=129616541603610&w=2
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/