[PATCH] Avoid deadlock in malloc on backtrace

[Florian Weimer]

On 02/24/2015 01:03 PM, Siddhesh Poyarekar wrote:
> On Tue, Feb 24, 2015 at 12:55:59PM +0100, Florian Weimer wrote:
>> I think we should really consider removal.  Most distributions will
>> have core dump catchers once they start using a glibc version derived
>> from master.
> 
> I don't need a lot of convincing for this, but I guess we need more
> consensus.  Does anybody think that removing the backtrace and maps is
> a bad idea?

Previous bugs filed for this:

https://sourceware.org/bugzilla/show_bug.cgi?id=12189 (CVE-2010-3192)
https://sourceware.org/bugzilla/show_bug.cgi?id=16159

The second bug appears to be the same thing you are fixing, Siddhesh.

-- 
Florian Weimer / Red Hat Product Security