Re: [PATCH] i386: instead of poisoning .init zone, change protectionbits to force a fault

[Eric Dumazet]

Benjamin LaHaise a écrit :
> On Sun, Jan 29, 2006 at 09:04:44PM +0100, Eric Dumazet wrote:
> > Chasing some invalid accesses to .init zone, I found that free_init_pages() 
> > was properly freeing the pages but virtual was still usable.
>
> This change will break the large table entries up, resulting in more TLB 
> pressure and reducing performance, and so should only be activated as a 
> debug option.

Hum... yet another CONFIG option ?

Could we 'just' move rodata (because of CONFIG_DEBUG_RODATA) and .init section 
  in their own 2MB/4MB page, playing with vmlinux.lds.S ?

It would be possible to have a full hugepage readonly for rodata,  and a full 
NOPROT for .init ?

Eric
-
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/