linux-kernel - Re: [GIT PULL] x86/platform updates for v5.10

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <335ef705-d395-1260-b668-2edab8bf32ff@hpe.com>
Date:   Mon, 12 Oct 2020 15:00:05 -0700
From:   Mike Travis <mike.travis@....com>
To:     Borislav Petkov <bp@...e.de>,
        Linus Torvalds <torvalds@...ux-foundation.org>
Cc:     x86-ml <x86@...nel.org>, lkml <linux-kernel@...r.kernel.org>
Subject: Re: [GIT PULL] x86/platform updates for v5.10



On 10/12/2020 2:56 PM, Borislav Petkov wrote:
> On Mon, Oct 12, 2020 at 02:46:10PM -0700, Linus Torvalds wrote:
>> At least "git grep" only shows two assignments to it.
>>
>> Of course, that would miss any cases that play games with preprocessor
>> token pasting etc, so it's not entirely meaningful, but it's certainly
>> a hint..
> 
>  From a quick staring at gcc asm, it looks write only. And gcc didn't
> warn because it optimized that assignment away completely AFAICT.
> 
>> And yes, I expect that the fix is to just make it "unsigned long", but
>> if it truly isn't actually used, maybe removal is better.
> 
> Yeah, below is a proper patch which builds fine with gcc and clang-10.
> You guys have fun - I'm going to bed. :-)
> 
> ---
> From: Mike Travis <mike.travis@....com>
> Date: Mon, 12 Oct 2020 23:46:34 +0200
> Subject: [PATCH] x86/platform/uv: Correct uvh_nmi_mmrx_mask's type
> 
> Clang rightfully warns:
> 
>    arch/x86/platform/uv/uv_nmi.c:250:23: warning: implicit conversion
>    from 'unsigned long' to 'int' changes value from 1152921504606846976
>    to 0 [-Wconstant-conversion]
>                    uvh_nmi_mmrx_mask = UVH_EVENT_OCCURRED0_EXTIO_INT0_MASK;
>                                    ~ ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 
> Make the variable unsigned long.
> 
>   [ bp: Productize it. ]

Thanks, I will look at it (and test it on hardware and the UV5 simulator 
to make sure it's correct.)
> 
> Signed-off-by: Mike Travis <mike.travis@....com>
> Signed-off-by: Borislav Petkov <bp@...e.de>
> ---
>   arch/x86/platform/uv/uv_nmi.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/arch/x86/platform/uv/uv_nmi.c b/arch/x86/platform/uv/uv_nmi.c
> index 0f5cbcf0da63..8566730f154d 100644
> --- a/arch/x86/platform/uv/uv_nmi.c
> +++ b/arch/x86/platform/uv/uv_nmi.c
> @@ -59,7 +59,7 @@ DEFINE_PER_CPU(struct uv_cpu_nmi_s, uv_cpu_nmi);
>   static unsigned long uvh_nmi_mmrx;		/* UVH_EVENT_OCCURRED0/1 */
>   static unsigned long uvh_nmi_mmrx_clear;	/* UVH_EVENT_OCCURRED0/1_ALIAS */
>   static int uvh_nmi_mmrx_shift;			/* UVH_EVENT_OCCURRED0/1_EXTIO_INT0_SHFT */
> -static int uvh_nmi_mmrx_mask;			/* UVH_EVENT_OCCURRED0/1_EXTIO_INT0_MASK */
> +static unsigned long uvh_nmi_mmrx_mask;			/* UVH_EVENT_OCCURRED0/1_EXTIO_INT0_MASK */
>   static char *uvh_nmi_mmrx_type;			/* "EXTIO_INT0" */
>   
>   /* Non-zero indicates newer SMM NMI handler present */
>