linux-kernel - Re: [PATCH] mdacon: fix redefinition of 'scr

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]

Date:   Mon, 06 Sep 2021 13:44:21 +0000
From:   "Jackie Liu" <liu.yun@...ux.dev>
To:     "Greg KH" <gregkh@...uxfoundation.org>
Cc:     linux-kernel@...r.kernel.org
Subject: Re: [PATCH] mdacon: fix redefinition of 'scr_memsetw'

Hi, Greg.


September 6, 2021 6:46 PM, "Greg KH" <gregkh@...uxfoundation.org> 写到:

> On Mon, Sep 06, 2021 at 06:20:30PM +0800, Jackie Liu wrote:
> 
>> From: Jackie Liu <liuyun01@...inos.cn>
>> 
>> CONFIG_VGA_CONSOLE=n and CONFIG_MDA_CONSOLE=n will cause vt_buffer.h not
>> include <asm/vga.h>.
>> 
>> But if we set CONFIG_MDA_CONSOLE=m, mdacon.c include <linux/vt_buffer.h>
>> is in front of include <asm/vga.h>. VT_BUF_HAVE_MEMSETW is not defined,
>> so vt_buffer.h will define a scr_memsetw, after that, vga.h also define
>> a scr_memsetw, so the repeated definition of scr_memsetw appears, builds
>> error.
>> 
>> We only need to make vt_buffer.h also contain vga.h when
>> CONFIG_MDA_CONSOLE=m. This problem can be fixed.
>> 
>> BTW, mdacon.c no need to include vga.h forcibly, let vt_buffer.h do it.
>> 
>> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
> 
> As this has always been an issue, how is this really a "fix"?

The problem has always existed, but there will be no compilation errors.
The driver needs to include <asm/vga.h> at the beginning if we want build
CONFIG_MDA_CONSOLE=m

After ac036f9570a2, redefined 'scr_memsetw'. but they are not arm and x86 architectures,
so it's not very important. It can also be Fixes: ac036f9570a2 ("vga: optimise console scrolling")

> 
> How can the above config options be set in this manner, do you have to
> do it manually or does a system really need that configuration?

This is a problem reported by my CI, when doing make randomconfigs. Of course,
this is a very tiny fix.

> 
>> Cc: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
>> Cc: linux-fbdev@...r.kernel.org
>> Cc: linux-kernel@...r.kernel.org
>> Signed-off-by: Jackie Liu <liuyun01@...inos.cn>
>> ---
>> drivers/video/console/mdacon.c | 1 -
>> include/linux/vt_buffer.h | 2 +-
>> 2 files changed, 1 insertion(+), 2 deletions(-)
>> 
>> diff --git a/drivers/video/console/mdacon.c b/drivers/video/console/mdacon.c
>> index ef29b321967f..5898d01bc492 100644
>> --- a/drivers/video/console/mdacon.c
>> +++ b/drivers/video/console/mdacon.c
>> @@ -42,7 +42,6 @@
>> #include <linux/init.h>
>> 
>> #include <asm/io.h>
>> -#include <asm/vga.h>
>> 
>> static DEFINE_SPINLOCK(mda_lock);
>> 
>> diff --git a/include/linux/vt_buffer.h b/include/linux/vt_buffer.h
>> index 848db1b1569f..3e71f879e7c0 100644
>> --- a/include/linux/vt_buffer.h
>> +++ b/include/linux/vt_buffer.h
>> @@ -16,7 +16,7 @@
>> 
>> #include <linux/string.h>
>> 
>> -#if defined(CONFIG_VGA_CONSOLE) || defined(CONFIG_MDA_CONSOLE)
>> +#if defined(CONFIG_VGA_CONSOLE) || defined(CONFIG_MDA_CONSOLE) ||
>> defined(CONFIG_MDA_CONSOLE_MODULE)
> 
> If this really is needed, please use the correct define macro that
> catches both of these options.

Do you have any suggestions? 

I saw that this file is also written like this：
./arch/arm/mach-pxa/spitz.c:455:#if defined(CONFIG_LEDS_GPIO) || defined(CONFIG_LEDS_GPIO_MODULE)

> 
> thanks,
> 
> greg k-h

Best Regards.
Jackie Liu