[Gc] Re[2]: [bdwgc] boehm-gc incompatible with aslr on darwin11 (#13)

Ivan Maidanski ivmai at mail.ru
Wed Feb 22 12:36:21 PST 2012


Hi Patrick and nicity,

Patrick -
No, thank you - AICS, it is a false alarm, I fixed it a year ago without any knowledge of ASLR (stack pointer hard-coding is a bad thing, and once one man (it was Morten Hermansen) had called GC_INIT on iPad device from non-primordial thread and GC had crushed then - after that I've fixed the code to use pthread_get_stackaddr_np). See part of commit faef04e (related to STACKBOTTOM) - https://github.com/ivmai/bdwgc/commit/faef04e7cb3741163dfdf65900ef5d2a0530be0f

nicity -
Am I right that you do not have such problem with (or haven't tested) BDWGC snapshot (or 7.2alpha6 tarball)?

Regards.

22 02 2012, 20:53 nicity <reply+i-3329737-a1dea316d11d14c478bc86091887fd24684337c1-460469 at reply.github.com>:
> Per garbage collection implementation fix for D language / MacOSX Lion (https://github.com/D-Programming-Language/druntime/pull/43) :
> When ASLR is enabled, the stack address of Mac OS X is no longer fixed at 0xc000_0000 — actually, often higher than this
> pthread_get_stackaddr_np() can be used to retrieve the stack start
> 
> ---
> Reply to this email directly or view it on GitHub:
> https://github.com/ivmai/bdwgc/issues/13#issuecomment-4108155
>


23 02 2012, 00:16 Patrick Marlier <patrick.marlier at gmail.com>:
> Hi Ivan and Jack,
> 
> Unfortunately, I can only have an access to a darwin10 and I cannot
> reproduce the bug with it. I will try to ask around me if somebody has a
> darwin11. So when I get access to one, I will definitely try to have a
> look at it.
> 
> Regards,
> --
> Patrick.
> 
> On 02/22/2012 06:46 AM, Ivan Maidanski wrote:
> > Hi,
> >
> > Could somebody propose a workaround? May be Patrick Marlier could help.
> >
> > Regards.
> >
> > 22 02 2012, 13:11 nicity<reply+i-3329737-a1dea316d11d14c478bc86091887fd24684337c1-460469 at reply.github.com>:
> >> GC_mark_from crashes on executable built on MacOSX Lion, which has ASLR option on
> >> Please see detailed description at
> >> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52179
> >>
> >> ---
> >> Reply to this email directly or view it on GitHub:
> >> https://github.com/ivmai/bdwgc/issues/13
> 
> 



More information about the Gc mailing list