On Tue, 23 Aug 2011 01:29:06 +0900
Akinobu Mita <akinobu.mita@xxxxxxxxx> wrote:
> This adds support for highmem pages poisoning and verification to the
> debug-pagealloc feature for no-architecture support.
>
> ...
>
> static void poison_page(struct page *page)
> {
> void *addr;
> -
> - if (PageHighMem(page)) {
> - poison_highpage(page);
> - return;
> + unsigned long flags;
> + bool highmem = PageHighMem(page);
> +
> + if (highmem) {
> + local_irq_save(flags);
> + addr = kmap_atomic(page);
> + } else {
> + addr = page_address(page);
> }
> set_page_poison(page);
> - addr = page_address(page);
> memset(addr, PAGE_POISON, PAGE_SIZE);
> +
> + if (highmem) {
> + kunmap_atomic(addr);
> + local_irq_restore(flags);
> + }
> }
This seems more complicated than is needed. Couldn't we just do
static void poison_page(struct page *page)
{
void *addr;
preempt_disable();
addr = kmap_atomic(page);
set_page_poison(page);
memset(addr, PAGE_POISON, PAGE_SIZE);
kunmap_atomic(addr);
preempt_enable();
}
?
> + addr = kmap_atomic(page);
That reminds me - we need to convert every "kmap_atomic(p, foo)" to
"kmap_atomic(p)" then remove the kmap_atomic back-compatibility macro.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>