linux-kernel - Re: [RFC][PATCHSET] epoll cleanups

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:   Sun, 4 Oct 2020 15:15:56 +0100
From:   Al Viro <viro@...iv.linux.org.uk>
To:     Matthew Wilcox <willy@...radead.org>
Cc:     Linus Torvalds <torvalds@...ux-foundation.org>,
        linux-kernel@...r.kernel.org, linux-fsdevel@...r.kernel.org,
        Marc Zyngier <maz@...nel.org>
Subject: Re: [RFC][PATCHSET] epoll cleanups

On Sun, Oct 04, 2020 at 01:13:29PM +0100, Matthew Wilcox wrote:

> Have you considered just storing a pointer to each struct file in an
> epoll set in an XArray?  Linked lists suck for modern CPUs, and there'd
> be no need to store any additional data in each struct file.  Using
> xa_alloc() to store the pointer and throw away the index the pointer
> got stored at would leave you with something approximating a singly
> linked list, except it's an array.  Which does zero memory allocations
> for a single entry and will then allocate a single node for your first
> 64 entries.

Won't work - those struct file can get freed while we are collecting the
set/allocating epitem/calling ->poll()/etc.