linux-kernel - Re: [PATCH] ARM: mmu: fix access to illegal address when using earlycon & memblock=debug

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]

Message-ID: <0fa192d1-b6cc-4a8a-c319-36471cba79e8@allwinnertech.com>
Date:   Mon, 18 Apr 2022 23:08:09 +0800
From:   Victor Hassan <victor@...winnertech.com>
To:     Linus Walleij <linus.walleij@...aro.org>
Cc:     linux@...linux.org.uk, rmk+kernel@...linux.org.uk,
        yanfei.xu@...driver.com, ardb@...nel.org, tglx@...utronix.de,
        mirq-linux@...e.qmqm.pl, arnd@...db.de,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
        allwinner-opensource-support@...winnertech.com
Subject: Re: [PATCH] ARM: mmu: fix access to illegal address when using
 earlycon & memblock=debug

On 18/04/2022 07:21, Linus Walleij wrote:
> On Wed, Mar 16, 2022 at 3:33 AM Victor Hassan <victor@...winnertech.com> wrote:
> 
>> earlycon uses fixmap to create a memory map,
>> So we need to close earlycon before closing fixmap,
>> otherwise printk will access illegal addresses.
>> After creating a new memory map, we open earlycon again.
>>
>> Signed-off-by: Victor Hassan <victor@...winnertech.com>
> 
> I think noone really noticed this because everyone on Arm systems
> use CONFIG_DEBUG_LL, and that makes printascii hammer out
> stuff on the console very early, it even accounts for whether we have
> MMU on or not.
> 
Hi Linus,
Thank you for the reply. I used earlycon, in early_fixmap_shutdown, the 
base address of earlycon is in the critical stage of release and 
reassignment, so early_fixmap_shutdown -> create_mapping should not call 
earlycon in this process, and create_mapping has a lot of conditions 
that trigger print, memblock=debug just makes it easier to expose problems.

> How are you using this on Arm even? What system and what serial
> driver?
I'm using serial driver 8250 on arm32, with cmdline: memblock=debug.
CONFIG_SERIAL_8250=y
CONFIG_SERIAL_8520_CONSOLE=y
> 
> That said, it looks correct.
> Acked-by: Linus Walleij <linus.walleij@...aro.org>
Thank you.
> 
> Yours,
> Linus Walleij

Sincerely,
Victor Hassan