linux-kernel - Re: [PATCHv2 1/2] x86/mm: Do not allow non-MAP_FIXED mapping across DEFAULT_MAP

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: <20171114192113.t7pq5p2n5emmiw2n@node.shutemov.name>
Date:   Tue, 14 Nov 2017 22:21:13 +0300
From:   "Kirill A. Shutemov" <kirill@...temov.name>
To:     Thomas Gleixner <tglx@...utronix.de>
Cc:     "Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>,
        Ingo Molnar <mingo@...hat.com>, x86@...nel.org,
        "H. Peter Anvin" <hpa@...or.com>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Andy Lutomirski <luto@...capital.net>,
        Nicholas Piggin <npiggin@...il.com>, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCHv2 1/2] x86/mm: Do not allow non-MAP_FIXED mapping across
 DEFAULT_MAP_WINDOW border

On Tue, Nov 14, 2017 at 05:01:50PM +0100, Thomas Gleixner wrote:
> On Tue, 14 Nov 2017, Kirill A. Shutemov wrote:
> > --- a/arch/x86/mm/hugetlbpage.c
> > +++ b/arch/x86/mm/hugetlbpage.c
> > @@ -166,11 +166,20 @@ hugetlb_get_unmapped_area(struct file *file, unsigned long addr,
> >  
> >  	if (addr) {
> >  		addr = ALIGN(addr, huge_page_size(h));
> > +		if (TASK_SIZE - len >= addr)
> > +			goto get_unmapped_area;
> 
> That's wrong. You got it right in arch_get_unmapped_area_topdown() ...

Ouch.

Please ignore selftest patch. I'll rework it to cover hugetlb.

> > +
> > +		/* See a comment in arch_get_unmapped_area_topdown */
> 
> This is lame, really.
> 
> > +		if ((addr > DEFAULT_MAP_WINDOW) !=
> > +				(addr + len > DEFAULT_MAP_WINDOW))
> > +			goto get_unmapped_area;
> 
> Instead of duplicating that horrible formatted condition and adding this
> lousy comment why can't you just put all of it (including the TASK_SIZE
> check) into a proper validation function and put the comment there?
> 
> The fixed up variant of your patch below does that.
> 
> Aside of that please spend a bit more time on describing things precisely
> at the technical and factual level next time. I fixed that up (once more)
> both in the comment and the changelog.
> 
> Please double check.

Works fine.

> +bool mmap_address_hint_valid(unsigned long addr, unsigned long len)
> +{
> +	if (TASK_SIZE - len < addr)
> +		return false;
> +#if CONFIG_PGTABLE_LEVELS >= 5
> +	return (addr > DEFAULT_MAP_WINDOW) == (addr + len > DEFAULT_MAP_WINDOW);

Is it micro optimization? I don't feel it necessary. It's not that hot
codepath to care about few cycles. (And one more place to care about for
boot-time switching.)

If you think it's needed, maybe IS_ENABLED() instead?

> +#else
> +	return true;
> +#endif
> +}

-- 
 Kirill A. Shutemov