linux-kernel - Re: Re: [PATCH RESEND] arm64: add alignment fault hanling

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

Message-id: <1241568967.925441455536689920.JavaMail.weblogic@epmlwas09b>
Date:	Mon, 15 Feb 2016 11:44:54 +0000 (GMT)
From:	EunTaik Lee <eun.taik.lee@...sung.com>
To:	Catalin Marinas <catalin.marinas@....com>
Cc:	"will.deacon@....com" <will.deacon@....com>,
	"vladimir.murzin@....com" <vladimir.murzin@....com>,
	"suzuki.poulose@....com" <suzuki.poulose@....com>,
	"riandrews@...roid.com" <riandrews@...roid.com>,
	"james.morse@....com" <james.morse@....com>,
	"salyzyn@...roid.com" <salyzyn@...roid.com>,
	"Dave.Martin@....com" <Dave.Martin@....com>,
	"linux-arm-kernel@...ts.infradead.org" 
	<linux-arm-kernel@...ts.infradead.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: Re: [PATCH RESEND] arm64: add alignment fault hanling

> How do you end up with Device nGnRnE in user space? I thought we should
> have got some guard page.
I guess it is a device file that is mmaped in the user space and later remapped in the kernel
 using pgprot_noncached(). 
I can't find the code that inserts a guard page in between vma's can you please help me find that code?

> do_strncpy_from_user() relies on unsafe_get_user() not being able to
> read 8 bytes. The problem now is that it doesn't get a page fault but an
> alignment one is isn't handled.
Yes.

> Why addr >= TASK_SIZE? addr here should be the fault address, not pc.
You're right. I should have used user_mode(regs).

> The simplest would be to use do_bad_area() here without any additional
> function.
Yes, I should've probably used do_bad_area(). I just wanted to do the minimal stuff.