linux-kernel - [RFC] discard task stack pages instead of pageout into swap partition

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Fri, 12 Dec 2014 11:30:53 +0800
From:	"Wang, Yalin" <Yalin.Wang@...ymobile.com>
To:	'Konstantin Khlebnikov' <koct9i@...il.com>,
	"Kirill A. Shutemov" <kirill@...temov.name>,
	Andrew Morton <akpm@...ux-foundation.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	"linux-mm@...ck.org" <linux-mm@...ck.org>,
	"linux-arm-kernel@...ts.infradead.org" 
	<linux-arm-kernel@...ts.infradead.org>,
	"n-horiguchi@...jp.nec.com" <n-horiguchi@...jp.nec.com>
Subject: [RFC] discard task stack pages instead of pageout into swap
 partition

Hi,

I am think of discard stack pages if the old
Page is under the stack pointer(Assume stack grow down)
of the task, This page don't need pageout, we can free it directly,
When the task need it again, we just use a zero page to
Map, it is safe for stack .

But I don't know how to implement it,
And is there some issue if do like this ?

The following is pseudo code to explain my ideas.
Any comments are appreciated !
Thanks
---
diff --git a/mm/vmscan.c b/mm/vmscan.c
index dcb4707..52e8314 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -962,6 +962,12 @@ static unsigned long shrink_page_list(struct list_head *page_list,
 			; /* try to reclaim the page below */
 		}
 
+		if (page_vm_flags(page) & (VM_GROWSUP | VM_GROWSDOWN) &&
+				PageAnon(page) && !PageSwapCache(page)) {
+			if (page_task_is_sleep(page) && task_sp > page->index) {
+				zap_page_range(vma, page->index, PAGE_SIZE);
+			}
+		}
 		/*
 		 * Anonymous process memory has backing store?
 		 * Try to allocate it some swap space here.