2007-06-29 09:08:21

by Dave Young

Subject: [PATCH] ioremap: fix iounmap numpages

Hi,
The second parameter of change_page_attr in iounmap is wrong, it should be (p->size - 1) >> PAGE_SHIFT

Signed-off-by: Dave Young <[email protected]>
---

diff -upr linux/arch/i386/mm/ioremap.c linux.new/arch/i386/mm/ioremap.c
--- linux/arch/i386/mm/ioremap.c 2007-06-29 16:48:40.000000000 +0000
+++ linux.new/arch/i386/mm/ioremap.c 2007-06-29 16:50:09.000000000 +0000
@@ -196,7 +196,7 @@ void iounmap(volatile void __iomem *addr
/* Reset the direct mapping. Can block */
if ((p->flags >> 20) && p->phys_addr < virt_to_phys(high_memory) - 1) {
change_page_attr(virt_to_page(__va(p->phys_addr)),
- p->size >> PAGE_SHIFT,
+ (p->size - 1) >> PAGE_SHIFT,
PAGE_KERNEL);
global_flush_tlb();
}

Regards
dave

2007-06-29 12:37:29

by Jeremy Fitzhardinge

Subject: Re: [PATCH] ioremap: fix iounmap numpages

Dave Young wrote:
> Hi,
> The second parameter of change_page_attr in iounmap is wrong, it should be (p->size - 1) >> PAGE_SHIFT
>

Why's that? Isn't p->size always going to be a pagesize multiple; in
which case, why would you want to change_page_attr on n-1 pages?

Are you seeing a problem that this patch fixes?

J