Re: POD function static initialization from function return, thread safe?

"Thomas Richter" <thor@[EMAIL PROTECTED]
> wrote in message 
news:gc7q8k$ult$1@[EMAIL PROTECTED]
> Chris M. Thomasson wrote:
>>
>>
>> do you think that:
>>
>> PTHREAD_MUTEX_INITIALIZER
>>
>> is busted?
>
> It's the matter of the compiler vendor to ensure that this works as 
> intended.

Yes; POSIX compliant compiler along with a POSIX library is the only way
to 
ensure correctness.




> Furthermore, as long as a variable has global static storage, it might
be 
> initialized "at compile time" by placing the right data into the data 
> section of the executable, and at loading time, the problem does not 
> exist.

impl detail indeed.