[Bug nptl/12416] Dynamic linker incorrectly reduces the stack size when making it executable

["siddhesh at redhat dot com" <sourceware-bugzilla at sourceware dot org>]

http://sourceware.org/bugzilla/show_bug.cgi?id=12416

--- Comment #6 from Siddhesh Poyarekar <siddhesh at redhat dot com> 2012-05-28 05:26:59 UTC ---
(In reply to comment #5)
> Thanks for improving the situation, but this is a real problem for programs
> that do conservative stack scanning such as those that depends on boehm-gc or
> mono.

How so? pthread_getattr_np will return the same stack boundaries and you should
not have to query it all the time if you're only interested in the underflow
end. If you're interested in the overflow end then you have no choice other
than calling pthread_getattr_np all the time because the overflow end may
change as extra vmas get mapped in below it.

> It still has not been answered why shrinking the stack is needed.

It is more the case that adding extra code to make sure that the entire vma is
covered is not demonstrated to be useful at all. If you can provide an
unambiguous reason as to why that is required, then we could consider going
through that extra effort.

I'd suggest you start a discussion on this on the developer mailing list
(libc-alpha at sourceware.org) so that everyone can pitch in on this.

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.