Re: do_div() silently truncates "base" to 32bit

[Randy Dunlap <rdunlap@infradead.org>]

On 09/04/13 08:32, Anatol Pomozov wrote:
> Hi
> 
> On Fri, Aug 30, 2013 at 5:50 PM, Anatol Pomozov
> <anatol.pomozov@gmail.com> wrote:
>> Hi, Joe
>>
>> On Fri, Aug 30, 2013 at 4:28 PM, Joe Perches <joe@perches.com> wrote:
>>> On Fri, 2013-08-30 at 15:48 -0700, Randy Dunlap wrote:
>>>> On 08/30/13 15:14, Anatol Pomozov wrote:
>>>>> But another question still remains: why compiler does not warn that
>>>>> size truncation happens? How to prevent bugs like CFQ one in the
>>>>> future? Should we add a compile-time assert to do_div() to prevent
>>>>> passing 64 numbers in "base" macro parameter?
>>>> That sounds like a fine idea to me.
>>>
>>> Geert thought so too and submitted a patch
>>>
>>> http://www.spinics.net/lists/linux-btrfs/msg26788.html
>>
>> Thanks it works!
>>
>> block/cfq-iosched.c: In function 'cfqg_prfill_avg_queue_size':
>> block/cfq-iosched.c:4423:3: error: size of unnamed array is negative
>> make[2]: *** [block/gfq-iosched.s] Error 1
>> make[1]: *** [block/gfq-iosched.s] Error 2
>> I see a number of other truncation errors. We need to fix them as well.
> 
> I looked at the places where do_div() passes incorrect parameters and
> it is quite a lot of them. It would be better to warn users until all
> the places are fixed. Is there any BUILD_WARN_ON() macro?
> 


Not that I know of or can find.

-- 
~Randy