netdev - Re: [PATCH net 1/1] ethtool: Fix mod state of verbose no

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: Thu, 5 Oct 2023 10:03:49 +0200
From: Köry Maincent <kory.maincent@...tlin.com>
To: Simon Horman <horms@...nel.org>
Cc: Michal Kubecek <mkubecek@...e.cz>, "David S. Miller"
 <davem@...emloft.net>, netdev@...r.kernel.org,
 linux-kernel@...r.kernel.org, Maxime Chevallier
 <maxime.chevallier@...tlin.com>, stable@...r.kernel.org,
 thomas.petazzoni@...tlin.com, Eric Dumazet <edumazet@...gle.com>, Jakub
 Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>
Subject: Re: [PATCH net 1/1] ethtool: Fix mod state of verbose no_mask
 bitset

Hello Simon,

Thank for your review.

On Wed, 4 Oct 2023 13:07:14 +0200
Simon Horman <horms@...nel.org> wrote:

> On Tue, Oct 03, 2023 at 10:56:52AM +0200, Köry Maincent wrote:
> > From: Kory Maincent <kory.maincent@...tlin.com>
>
> > @@ -448,8 +450,11 @@ ethnl_update_bitset32_verbose(u32 *bitmap, unsigned
> > int nbits, }
> >  
> >  	no_mask = tb[ETHTOOL_A_BITSET_NOMASK];
> > -	if (no_mask)
> > -		ethnl_bitmap32_clear(bitmap, 0, nbits, mod);
> > +	if (no_mask) {
> > +		tmp = kcalloc(nbits, sizeof(u32), GFP_KERNEL);
> > +		memcpy(tmp, bitmap, nbits);  
> 
> Hi Köry,
> 
> I'm no expert on etnhl bitmaps. But the above doesn't seem correct to me.
> Given that sizeof(u32) == 4:
> 
> * The allocation is for nbits * 4 bytes
> * The copy is for its for nbits bytes
> * I believe that bitmap contains space for the value followed by a mask.
>   So it seems to me the size of bitmap, in words, is
>   DIV_ROUND_UP(nbits, 32) * 2
>   And in bytes: DIV_ROUND_UP(nbits, 32) * 16
>   But perhaps only half is needed if only the value part of tmp is used.
> 
> If I'm on the right track here I'd suggest helpers might be in order.

You are right I should use the same alloc as ethnl_update_bitset with tmp
instead of bitmap32:

        u32 small_bitmap32[ETHNL_SMALL_BITMAP_WORDS];                      
        u32 *bitmap32 = small_bitmap32; 
        if (nbits > ETHNL_SMALL_BITMAP_BITS) {                             
                unsigned int dst_words = DIV_ROUND_UP(nbits, 32);          

                bitmap32 = kmalloc_array(dst_words, sizeof(u32), GFP_KERNEL);
                if (!bitmap32)                                             
                        return -ENOMEM;                                    
        }   

But I am still wondering if it needs to be double as you said for the size of
the value followed by the mask. Not sure about it, as ethnl_update_bitset does
not do it. 

Regards,