[cfe-dev] Why clang needs to fork into itself?

[Philip Reames]

On 1/28/14 12:23 PM, Yuri wrote:
> On 01/28/2014 12:11, Ted Kremenek wrote:
>> Clang forking itself remains quite useful for getting test cases for 
>> crashers.  If the forked clang crashes the parent clang process tries 
>> to generate a preprocessed source to serve as a test case.  We have 
>> found this to be invaluable for users to file useful test cases.
>
> This is not the point. The point is why make it always watch for crash 
> when it happens very rarely? If the crash occurred, one can always 
> rerun with the special option and get the same report.
Crashes tend to happen rarely, in the field, and not for developers with 
debuggers ready.  That's why you need all the help debugging them you 
can get.  :)

Frankly, I don't understand the concern with the mechanism to start 
with.  Given that several mechanisms for debugging have already been 
discussed on this thread, what is your motivation for removing it? 
Unless you have a very good one, I would be highly reluctant to consider 
removing the fork guard.

Philip