Fwd: [PATCH] tdestroy() should allow passing 'NULL' for 'freefct' callback

[Carlos O'Donell]

On 06/30/2013 03:01 PM, Ondřej Bílka wrote:
> On Wed, Jun 26, 2013 at 07:57:29AM -0400, Carlos O'Donell wrote:
>> On 06/25/2013 05:32 PM, Roland McGrath wrote:
>>> This is a user-visible change, so needs a BZ.  Probably this should use a
>>> new symbol version.  The compat routine need not behave any differently.
>>> But when someone builds a new program that might rely on the new behavior,
>>> they should get a symbol version dependency that ensures that attempting to
>>> run that program against an older library will give a coherent
>>> binary-compatibility error rather than just crashing in tdestroy.
>>
>> Roland,
>>
>> Thanks for the review.
>>
>> It has a user-visible BZ. It's BZ #15009, as written into the ChangeLog?
>>
>> Versioning is an interesting point I hadn't considered. In the past I'd 
>> say we haven't always done this kind of bumping of the symbol version 
>> to ensure considerate error messages. Though I agree with you, bumping 
>> the symbol version is the right thing to do here.
>>
>> I should really keep my eye out for more of these kinds of issues. It's
>> my fault that I missed this.
>>
>> Ondrej,
>>
>> Could you help with adding the appropriate symbol versioning to the patch?
>>
> Sorry, now I have limited internet access so I cannot look into this until
> wednesday.
> 

That's fine. This isn't urgent.

Cheers,
Carlos.