linux-kernel - Re: [PATCH] x86/lib: Fix overflow of variable m when val >= 1410065408

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

Date:   Thu, 2 Nov 2023 17:25:29 +0000
From:   "Colin King (gmail)" <colin.i.king@...il.com>
To:     Dave Hansen <dave.hansen@...el.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org,
        "H . Peter Anvin" <hpa@...or.com>
Cc:     kernel-janitors@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] x86/lib: Fix overflow of variable m when val >=
 1410065408

On 02/11/2023 16:38, Dave Hansen wrote:
> On 11/1/23 08:32, Colin Ian King wrote:
> ...
>>   int num_digits(int val)
>>   {
>> -	int m = 10;
>> +	long m = 10;
>>   	int d = 1;
>>   
>>   	if (val < 0) {
> 
> Isn't this still broken on 32-bit where sizeof(long) == sizeof(int)?
> Seems like we need 'm' to be able to hold values that are ~10x larger
> than 'val' if we need this to work for the entire int range.

Good point, long long is required for 32 bit,

sizes:
arch     int   long   long long
i386     4     4      8
x86_64   4     8      8

I'll send a V2.


> 
> Also, performance doesn't matter here at *all* with the current use in
> a couple of printk()'s.  Just making 'm' 'long long' or u64 probably be
> just fine.