linux-kernel - Re: [PATCH 1/2] mm: zswap.c: add xarray tree to zswap

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] [day] [month] [year] [list]

Message-ID: <CAJD7tkbKZqOwkDgNUQ7+XCHs_Uqb=gbCk7L=Es4sR1T3wWG+gA@mail.gmail.com>
Date: Fri, 19 Jan 2024 14:08:04 -0800
From: Yosry Ahmed <yosryahmed@...gle.com>
To: Chris Li <chriscli@...gle.com>
Cc: Matthew Wilcox <willy@...radead.org>, Andrew Morton <akpm@...ux-foundation.org>, 
	linux-kernel@...r.kernel.org, linux-mm@...ck.org, 
	Wei Xu <weixugc@...gle.com>, Yu Zhao <yuzhao@...gle.com>, 
	Greg Thelen <gthelen@...gle.com>, Chun-Tse Shao <ctshao@...gle.com>, 
	Suren Baghdasaryan <surenb@...gle.com>, 
	Brain Geffon <bgeffon@...gle.com>, Minchan Kim <minchan@...nel.org>, Michal Hocko <mhocko@...e.com>, 
	Mel Gorman <mgorman@...hsingularity.net>, Huang Ying <ying.huang@...el.com>, 
	Nhat Pham <nphamcs@...il.com>, Johannes Weiner <hannes@...xchg.org>, Kairui Song <kasong@...cent.com>, 
	Zhongkun He <hezhongkun.hzk@...edance.com>, Kemeng Shi <shikemeng@...weicloud.com>, 
	Barry Song <v-songbaohua@...o.com>, "Liam R. Howlett" <Liam.Howlett@...cle.com>, 
	Joel Fernandes <joel@...lfernandes.org>, Chengming Zhou <zhouchengming@...edance.com>
Subject: Re: [PATCH 1/2] mm: zswap.c: add xarray tree to zswap

On Fri, Jan 19, 2024 at 2:05 PM Chris Li <chriscli@...gle.com> wrote:
>
> Hi Yosry,
>
> On Fri, Jan 19, 2024 at 1:41 PM Yosry Ahmed <yosryahmed@...gle.com> wrote:
> > > if your code is naturally structured so that you delete the entries
> > > after freeing them, you have no reason to call xa_destroy().
> >
> > Thanks for elaborating. Based on this, I believe doing xas_for_each()
> > { zswap_free_entry(); }; xa_destroy(); is both closer to the current
> > code structure and more efficient.
> >
> Can't do that in this case though.  Because you get the RCU read lock
> on the tree.
> Other threads can still lookup the xarray (also RCU read lock) and get
> a pointer to the already freed memory.

During swapoff no one should be trying to swapin or swapout, where
would the lookups come from?