Re: [PATCH 0/6] slab: Provide full coverage for __alloc_size attribute

From: Kees Cook <keescook@chromium.org>
To: Arnd Bergmann <arnd@arndb.de>
Cc: "Conor.Dooley" <conor.dooley@microchip.com>,
	Vlastimil Babka <vbabka@suse.cz>, David Gow <davidgow@google.com>,
	Christoph Lameter <cl@linux.com>,
	Pekka Enberg <penberg@kernel.org>,
	David Rientjes <rientjes@google.com>,
	Joonsoo Kim <iamjoonsoo.kim@lge.com>,
	Roman Gushchin <roman.gushchin@linux.dev>,
	Hyeonggon Yoo <42.hyeyoo@gmail.com>,
	Rasmus Villemoes <rasmus.villemoes@prevas.dk>,
	Guenter Roeck <linux@roeck-us.net>,
	Andy Shevchenko <andriy.shevchenko@intel.com>,
	Paolo Abeni <pabeni@redhat.com>,
	Geert Uytterhoeven <geert@linux-m68k.org>,
	Nathan Chancellor <nathan@kernel.org>,
	Nick Desaulniers <ndesaulniers@google.com>,
	Tom Rix <trix@redhat.com>,
	linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	linux-hardening@vger.kernel.org, llvm@lists.linux.dev
Subject: Re: [PATCH 0/6] slab: Provide full coverage for __alloc_size attribute
Date: Thu, 1 Dec 2022 09:15:34 -0800	[thread overview]
Message-ID: <202212010915.ADB7D298@keescook> (raw)
In-Reply-To: <63ced4a4-33b9-4c03-8d7a-987ceb1fc2c7@app.fastmail.com>

On Tue, Nov 29, 2022 at 01:33:03PM +0100, Arnd Bergmann wrote:
> On Tue, Nov 29, 2022, at 13:24, Conor Dooley wrote:
> > On Tue, Nov 01, 2022 at 03:33:08PM -0700, Kees Cook wrote:
> >> Hi,
> >> 
> >> This is a series to work around a deficiency in GCC (>=11) and Clang
> >> (<16) where the __alloc_size attribute does not apply to inlines. :(
> >> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=96503
> >> 
> >> This manifests as reduced overflow detection coverage for many allocation
> >> sites under CONFIG_FORTIFY_SOURCE=y, where the allocation size was
> >> not actually being propagated to __builtin_dynamic_object_size(). In
> >> addition to working around the issue, expand use of __alloc_size (and
> >> __realloc_size) to more places and provide KUnit tests to validate all
> >> the covered allocator APIs.
> >
> > Hello Kees!
> >
> > It would appear that one of the macros you've added here is doing Bad
> > Things^TM to allmodconfig on RISC-V since the 22nd:
> >
> > ../lib/fortify_kunit.c: In function 'alloc_size_kmalloc_const_test':
> > ../lib/fortify_kunit.c:140:1: error: the frame size of 2384 bytes is 
> > larger than 2048 bytes [-Werror=frame-larger-than=]
> >   140 | }                                                               
> >         \
> >       | ^
> > ../lib/fortify_kunit.c:209:1: note: in expansion of macro 
> > 'DEFINE_ALLOC_SIZE_TEST_PAIR'
> >   209 | DEFINE_ALLOC_SIZE_TEST_PAIR(kmalloc)
> >       | ^~~~~~~~~~~~~~~~~~~~~~~~~~~
> > cc1: all warnings being treated as errors
> >
> > CONFIG_GCC_VERSION=110100
> > CONFIG_AS_VERSION=23700
> > CONFIG_LD_VERSION=23700
> >
> > The report came out of my CI (which I should have passed on sooner) so
> > I do not have anything other than stderr - I can get you anything else
> > you'd like/need though if you LMK.
> 
> There is generally a conflict between kunit and the structleak
> gcc plugin, I think the Makefile needs a line like
> 
> CFLAGS_fortify_kunit.o += $(DISABLE_STRUCTLEAK_PLUGIN)

Thanks for the report! I've taken Anders's patch for this now.

-- 
Kees Cook