Ah, I see. Many thanks for the explanation.
Sunjeet
On 2012-04-05, at 7:36 AM, Jakob Bohm wrote:
> On 4/5/2012 5:36 AM, Sunjeet Singh wrote:
>>> ENGINE_free all of the structural references you got, including the one
>>> for the ENGINE you're using
>> And so in order to free all structural references of an ENGINE * e, I could
>> just do this?-
>>
>> while( e->struct_ref> 0 ) {
>> ENGINE_free( e );
>> }
>>
>> And to keep all but one functional reference to e, I could do this?-
>>
>> while(e->funct_ref> 1 ) {
>> ENGINE_finish( e );
>> }
>>
>>> I don't think there is any cost to having extra references (of either kind)
>>> to an ENGINE, as long as all of the references are freed when you are done.
>> Once again, free all functional references and structural references if any
>> and then call the ENGINE_cleanup() function ?
>>
> No, you got the concept of reference counting all wrong!
>
> You are not supposed to artificially remove references you did not
> make, as that may crash the code that made that reference for its
> own use.
>
> The rules are much simpler:
>
> If you store some kind of explicit pointer or handle to the engine
> in your code, call the applicable ENGINE_ function to prevent other
> code from unloading the engine while you use it. When you remove
> or zero out that pointer or handle, call the opposite ENGINE_
> function so others can now free it when they no longer need it
> either. If your code happens to hold the last reference when
> making the call, the unloading that was prevented by your reference
> will happen at that time.
>
> Therefore:
>
> If someone else artificially decrements all the usage counts to 0
> while you are using the pointer or handle, your code is likely to
> crash when the engine disappears unexpectedly.
>
> If someone else artificially decrements all the usage counts to 1
> while two or more parts of your code are using pointers or handles
> to the engine, then when one part of your code frees its reference,
> it will unload the engine prematurely, thus crashing the other part
> that was still using its reference.
>
> Now swap the roles of you and someone else (the Golden Rule) and it
> becomes clear that your proposed code is likely to cause some other
> "random" piece of code to crash at some "random" time after you
> decrement the counter to 1.
>
> So don't do this.
>
> --
> Jakob Bohm, CIO, partner, WiseMo A/S. http://www.wisemo.com
> Transformervej 29, 2730 Herlev, Denmark. direct: +45 31 13 16 10
> <call:+4531131610>
> This message is only for its intended recipient, delete if misaddressed.
> WiseMo - Remote Service Management for PCs, Phones and Embedded
> ______________________________________________________________________
> OpenSSL Project http://www.openssl.org
> User Support Mailing List openssl-users@openssl.org
> Automated List Manager majord...@openssl.org
______________________________________________________________________
OpenSSL Project http://www.openssl.org
User Support Mailing List openssl-users@openssl.org
Automated List Manager majord...@openssl.org
