Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Mon, 19 Jun 2017 10:12:22 +0300
From: Igor Stoppa <igor.stoppa@...wei.com>
To: Laura Abbott <labbott@...hat.com>, <keescook@...omium.org>,
        <mhocko@...nel.org>, <jmorris@...ei.org>
CC: <penguin-kernel@...ove.SAKURA.ne.jp>, <paul@...l-moore.com>,
        <sds@...ho.nsa.gov>, <casey@...aufler-ca.com>, <hch@...radead.org>,
        <linux-security-module@...r.kernel.org>, <linux-mm@...ck.org>,
        <linux-kernel@...r.kernel.org>, <kernel-hardening@...ts.openwall.com>
Subject: Re: [PATCH 2/4] Protectable Memory Allocator

On 09/06/17 21:56, Laura Abbott wrote:
> On 06/07/2017 05:35 AM, Igor Stoppa wrote:

[...]

> The pool logic looks remarkably similar to genalloc (lib/genalloc.c).
> It's not a perfect 1-to-1 mapping but it's close enough to be worth
> a look.

Indeed. I have prepared a new incarnation of pmalloc, based on genalloc.
There are a couple of things that I would like to adjust in genalloc,
but I'll discuss this in the new submission.

>> +
>> +const char msg[] = "Not a valid Pmalloc object.";
>> +const char *__pmalloc_check_object(const void *ptr, unsigned long n)
>> +{
>> +	unsigned long p;
>> +
>> +	p = (unsigned long)ptr;
>> +	n = p + n - 1;
>> +	for (; (PAGE_MASK & p) <= (PAGE_MASK & n); p += PAGE_SIZE) {
>> +		if (is_vmalloc_addr((void *)p)) {
>> +			struct page *page;
>> +
>> +			page = vmalloc_to_page((void *)p);
>> +			if (!(page && PagePmalloc(page)))
>> +				return msg;
>> +		}
> 
> Should this be an error if is_vmalloc_addr returns false?

Yes, if this function is called, at least the beginning of the range
*is* a vmalloc address and therefore the rest should be a vmalloc
address as well.

thanks, igor

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.