linux-kernel - Re: [PATCH] decompress_bunzip2: fix sizeof type in start

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]

Message-ID: <6a73bb96-1390-ecca-2429-12b71e2aefc2@redhat.com>
Date:   Mon, 13 Jul 2020 12:27:02 -0700
From:   Tom Rix <trix@...hat.com>
To:     hpa@...or.com, alain@...ff.lu
Cc:     linux-kernel@...r.kernel.org
Subject: Re: [PATCH] decompress_bunzip2: fix sizeof type in start_bunzip


On 7/12/20 3:21 PM, hpa@...or.com wrote:
> On July 12, 2020 8:12:43 AM PDT, Tom Rix <trix@...hat.com> wrote:
>> On 7/12/20 6:09 AM, H. Peter Anvin wrote:
>>> On 2020-07-12 05:59, trix@...hat.com wrote:
>>>> From: Tom Rix <trix@...hat.com>
>>>>
>>>> clang static analysis flags this error
>>>>
>>>> lib/decompress_bunzip2.c:671:13: warning: Result of 'malloc' is
>> converted
>>>>   to a pointer of type 'unsigned int', which is incompatible with
>> sizeof
>>>>   operand type 'int' [unix.MallocSizeof]
>>>>         bd->dbuf = large_malloc(bd->dbufSize * sizeof(int));
>>>>                    ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>>>
>>>> Reviewing the bunzip_data structure, the element dbuf is type
>>>>
>>>> 	/* Intermediate buffer and its size (in bytes) */
>>>> 	unsigned int *dbuf, dbufSize;
>>>>
>>>> So change the type in sizeof to 'unsigned int'
>>>>
>>> You must be kidding.
>>>
>>> If you want to change it, change it to sizeof(bd->dbuf) instead, but
>> this flag
>>> is at least in my opinion a total joke. For sizeof(int) !=
>> sizeof(unsigned
>>> int) is beyond bizarre, no matter how stupid the platform.
>> Using the actual type is more correct that using a type of the same
>> size.
>>
>> trix
>>
>>> 	-hpa
>>>
> "More correct?" All it is is more verbose.
>
> Using the sizeof of the actual object at least adds some actual safety.

Sorry, I am being pedantic, I mean anything that produces the correct assembly is correct. But there are different path to being correct.  The path I was suggesting to follow the type of the element/final pointer when allocating an memory.

large_malloc(bd->dbufSize * sizeof(*bd->dbuf)) would also work

I will respin.

trix