lkml.org 
[lkml]   [2020]   [Apr]   [28]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRE: [PATCH v2 3/6] ima: Fix ima digest hash table key calculation
Date
> -----Original Message-----
> From: David Laight [mailto:David.Laight@ACULAB.COM]
> Sent: Monday, April 27, 2020 4:28 PM
> To: Roberto Sassu <roberto.sassu@huawei.com>; zohar@linux.ibm.com;
> rgoldwyn@suse.de
> Cc: linux-integrity@vger.kernel.org; linux-security-module@vger.kernel.org;
> linux-kernel@vger.kernel.org; Silviu Vlasceanu
> <Silviu.Vlasceanu@huawei.com>; Krzysztof Struczynski
> <krzysztof.struczynski@huawei.com>; stable@vger.kernel.org
> Subject: RE: [PATCH v2 3/6] ima: Fix ima digest hash table key calculation
>
> From: Roberto Sassu
> > Sent: 27 April 2020 13:51
> ...
> > > > -static inline unsigned long ima_hash_key(u8 *digest)
> > > > +static inline unsigned int ima_hash_key(u8 *digest)
> > > > {
> > > > - return hash_long(*digest, IMA_HASH_BITS);
> > > > + return (*(unsigned int *)digest % IMA_MEASURE_HTABLE_SIZE);
> > >
> > > That almost certainly isn't right.
> > > It falls foul of the *(integer_type *)ptr being almost always wrong.
> >
> > I didn't find the problem. Can you please explain?
>
> The general problem with *(int_type *)ptr is that it does completely
> the wrong thing if 'ptr' is the address of a larger integer type on
> a big-endian system.
> You may also get a misaligned access trap.
>
> In this case I guess that digest is actually u8[SHA1_DIGEST_SIZE].
> Maybe what you should return is:
> (digest[0] | digest[1] << 8) % IMA_MEASURE_HTABLE_SIZE;
> and comment that there is no point taking a hash of part of
> a SHA1 digest.

Ok, thanks.

Roberto

HUAWEI TECHNOLOGIES Duesseldorf GmbH, HRB 56063
Managing Director: Li Peng, Li Jian, Shi Yanli


> David
>
> -
> Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes,
> MK1 1PT, UK
> Registration No: 1397386 (Wales)
\
 
 \ /
  Last update: 2020-04-28 09:20    [W:0.232 / U:0.248 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site