Date: Sun, 20 May 2007 07:37:14 GMT From: Mark Linimon <linimon@FreeBSD.org> To: rea-fbsd@codelabs.ru, linimon@FreeBSD.org, freebsd-bugs@FreeBSD.org Subject: Re: kern/112771: [libc] [patch] unwrapped magic checks in the malloc code Message-ID: <200705200737.l4K7bENP036515@freefall.freebsd.org>
next in thread | raw e-mail | index | archive | help
Synopsis: [libc] [patch] unwrapped magic checks in the malloc code State-Changed-From-To: open->closed State-Changed-By: linimon State-Changed-When: Sun May 20 07:36:21 UTC 2007 State-Changed-Why: >From misfiled PR kern/112781: Date: Sat, 19 May 2007 02:05:25 -0700 From: Jason Evans <jasone@freebsd.org> I don't think it is possible for this to cause a compilation error, due to the logic that assures NDEBUG and MALLOC_DEBUG are set in a mutually coherent manner. Is there some set of compilation flags that can cause compilation to fail? Jason >From misfiled PR kern/112782: Date: Sat, 19 May 2007 13:41:42 +0400 From: Eygene Ryabinkin <rea-fbsd@codelabs.ru> Sat, May 19, 2007 at 02:05:25AM -0700, Jason Evans wrote: > I don't think it is possible for this to cause a compilation error, due > to the logic that assures NDEBUG and MALLOC_DEBUG are set in a mutually > coherent manner. Yes, you're right: I had overlooked that the NDEBUG is mangled basing on the value of the MALLOC_DEBUG. In principle, one can later do something nasty with the NDEBUG that will broke the coherency and this will result in the non-compilable code. Though it will be rather easy to find out. > Is there some set of compilation flags that can cause compilation to > fail? No, I was just digging through the code and (wrongly) assumed that an inconsistency had been found. Sorry for the noise. The PR can be closed. -- Eygene http://www.freebsd.org/cgi/query-pr.cgi?pr=112771
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200705200737.l4K7bENP036515>