linux-kernel - [PATCH 0/1] fs: cap max_files at LONG

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-next>] [day] [month] [year] [list]

Message-Id: <20181017231149.26302-1-christian@brauner.io>
Date:   Thu, 18 Oct 2018 01:11:48 +0200
From:   Christian Brauner <christian@...uner.io>
To:     linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Cc:     viro@...iv.linux.org.uk, keescook@...omium.org,
        Christian Brauner <christian@...uner.io>
Subject: [PATCH 0/1] fs: cap max_files at LONG_MAX

Hey,

This is a follow-up to [1] and [2] that caps the file-max sysctl at
LONG_MAX and rejects any larger values with EINVAL because of the way
percup counters work. This intends to align the kernel side of files_max
handling as well by capping the maximum number of files at LONG_MAX
until such time as we have the need and the precision for percpu
unsigned long counters.

Note, that in a quick discussion with Tejun investigating percpu unsigned
long counters it became obvious that technically, both LONG_MAX and
ULONG_MAX should be unreachable. Given that an fd is usually larger than 2
bytes one would likely run out of address space before even reaching the
LONG_MAX limit.

Al, I suggested this patch in [3] but I didn't see you reply so I'm not
sure whether you would be in favor of this or not.

Thanks!
Christian

[1]: https://lkml.org/lkml/2018/10/17/74
[2]: https://lkml.org/lkml/2018/10/17/184
[3]: https://lkml.org/lkml/2018/10/17/555

Christian Brauner (1):
  fs: cap max_files at LONG_MAX

 fs/file_table.c | 8 ++++++++
 1 file changed, 8 insertions(+)

-- 
2.19.1