> Rather than debating about how various platforms handle malformed 
> format strings for strftime(), and whether or not they crash, we 
> should simply prevent the native strftime() function from seeing 
> them in the first place.

Of course the check needs to be restored. I wasn't suggesting to go back to the default IPH to have it terminate the process. That won't help, anyway. The potential over-read is in the for loop, before calling strftime. I just thought the context for the accidental removal was relevant. Care needs to be taken when removing checks that were put in place to avoid the default IPH, because those checks may serve a dual purpose in the surrounding code.

The patch also fixes the AIX/Sun code, which needs to be applied to 3.4 as well. I don't think this issue applies to 3.2 and 3.3.