Re: [PATCH] libxl: handle null lists in libxl_string_list_length

[Andrew Cooper <andrew.cooper3@citrix.com>]

On 27/09/13 13:20, Matthew Daley wrote:
> On Sat, Sep 28, 2013 at 12:08 AM, Boris Ostrovsky
> <boris.ostrovsky@oracle.com> wrote:
>> ----- mattjd@gmail.com wrote:
>>
>>> After commit b0be2b12 ("libxl: fix libxl_string_list_length and its
>>> only
>>> caller") libxl_string_list_length no longer handles null (empty)
>>> lists. Fix
>>> so they are handled, returning length 0.
>>>
>>> While at it, remove the unneccessary undereferenced null pointer
>>> check
>> Are you sure this check should be removed? This routine can be called
>> from anywhere (at least within libxl it seems) and one day someone will
>> call it with NULL argument.
>>
>> I'd probably leave this check in.
> I would argue that any such invocation would be an error by the caller
> and should fail noisily, similar to how passing NULL into strlen
> should not return 0. libxl_{string,key_value}_list_dispose similarly
> assumes non-NULL pointers, FWIW.
>
> Ian C., do you have an opinion either way?
>
> - Matthew

I would agree that any passing of NULL is a caller error.  Possibly an
explicit check and abort()? If it is going to be noisy, we should be
nice and help out the debugger.

~Andrew