[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <alpine.LSU.2.20.1706281134380.30709@cbobk.fhfr.pm>
Date: Wed, 28 Jun 2017 11:40:06 +0200 (CEST)
From: Jiri Kosina <jikos@...nel.org>
To: Oleg Nesterov <oleg@...hat.com>
cc: tip-bot for Michal Hocko <tipbot@...or.com>,
linux-tip-commits@...r.kernel.org, torvalds@...ux-foundation.org,
mingo@...nel.org, hpa@...or.com, mhocko@...e.com,
tglx@...utronix.de, davej@...emonkey.org.uk, peterz@...radead.org,
linux-kernel@...r.kernel.org
Subject: Re: [tip:x86/mm] x86/mmap, ASLR: Do not treat unlimited-stack tasks
as legacy mmap
On Tue, 27 Jun 2017, Oleg Nesterov wrote:
> Perhaps it makes sense to reset RLIMITs on suid exec (say, if
> bprm->per_clear is not zero) ? Yes, it is not clear how should we define
> SANE_RLIMITS_FOR_SUID, and this should probably depend on sysctl, etc.
Hmm, this should be an userspace-defined policy. On a 'standard'
(PAM-based) system, I think a sane expectation would be to get the same
limits as the ones enforced by pam_limits configuration, but syncing those
with kernel feels awkward.
Thanks,
--
Jiri Kosina
SUSE Labs
Powered by blists - more mailing lists