Re: [PATCH v2 3/3] manual: Manual update for strlcat, strlcpy, wcslcat, wclscpy

[Siddhesh Poyarekar <siddhesh@gotplt.org>]

On 2023-06-06 08:27, Florian Weimer wrote:
> * Siddhesh Poyarekar:
> 
>>> +The behavior is undefined if @var{to} or @var{from} is a null pointer,
>>> +or if the destination array's size is less than @var{size}, or if the
>>> +string @var{from} overlaps the first @var{size} bytes of the
>>> +destination array.
>>
>> Shouldn't this be undefined for all kinds of overlaps between @var{to}
>> and @var{from} and not just when the @{from} overlaps with the first
>> @var{size} bytes of @var{to}?
> 
> I don't think so.  There is no reason why the data couldn't be copied
> within the same array.  This can plausibly happen if a custom memory
> allocator is used, for example.

Uhmm, I haven't heard of the customer allocator argument being used in 
this context but I suppose it makes sense.  I'm probably just getting 
jitters about specifying behaviour to that much detail when a simple 
"behavior is undefined if FROM and TO overlap" would suffice.  I mean 
this argument could be made for memcpy or any function that assumes 
non-aliased inputs.

>> Also, perhaps s/destination array/@var{to}/ to make it clearer.
> 
> I don't think so because it's confusing whether the size refers to TO
> itself or (conceptually) to *TO.

Fair enough.

Reviewed-by: Siddhesh Poyarekar <siddhesh@sourceware.org>