[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <45087800-218a-7ff5-22c0-d0a5bfea5001@gmail.com>
Date: Mon, 26 Feb 2018 09:46:43 -0800
From: J Freyensee <why2jjj.linux@...il.com>
To: Igor Stoppa <igor.stoppa@...wei.com>, david@...morbit.com,
willy@...radead.org, keescook@...omium.org, mhocko@...nel.org
Cc: labbott@...hat.com, linux-security-module@...r.kernel.org,
linux-mm@...ck.org, linux-kernel@...r.kernel.org,
kernel-hardening@...ts.openwall.com
Subject: Re: [PATCH 2/7] genalloc: selftest
On 2/26/18 4:11 AM, Igor Stoppa wrote:
>
> On 24/02/18 00:42, J Freyensee wrote:
>>> + locations[action->location] = gen_pool_alloc(pool, action->size);
>>> + BUG_ON(!locations[action->location]);
>> Again, I'd think it through if you really want to use BUG_ON() or not:
>>
>> https://lwn.net/Articles/13183/
>> https://lkml.org/lkml/2016/10/4/1
> Is it acceptable to display only a WARNing, in case of risking damaging
> a mounted filesystem?
That's a good question. Based upon those articles, 'yes'. But it seems
like a 'darned-if-you-do, darned-if-you-don't' question as couldn't you
also corrupt a mounted filesystem by crashing the kernel, yes/no?
If you really want a system crash, maybe just do a panic() like
filesystems also use?
>
> --
> igor
Powered by blists - more mailing lists