From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============8014729579097620413==" MIME-Version: 1.0 From: Rusty Russell To: lkp@lists.01.org Subject: Re: [PATCH] drm/radeon: Try to init amdkfd only if 64 bit kernel Date: Wed, 07 Jan 2015 09:28:31 +1030 Message-ID: <87mw5v4iug.fsf@rustcorp.com.au> In-Reply-To: List-Id: --===============8014729579097620413== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Kees Cook writes: > On Sun, Jan 4, 2015 at 8:28 PM, Rusty Russell w= rote: >> Kees, as far as I can tell you need another 0-terminated vmlinux.relocs >> section for weak symbols. These should not be relocated if already 0. > > A few questions: > > Why doesn't this break on 32-bit without kASLR? 32-bit does relocation > by default, even without CONFIG_RANDOMIZE_BASE. Well, the offset was 0 until I removed CONFIG_HIBERNATE. > Are there any symbols that are NULL that aren't weak? I'd expect all > strong symbols to have non-zero offsets, but I must be > misunderstanding something here. I don't think there would be. Anyway, you might be able to filter them out in x86/tools/relocs itself. Cheers, Rusty. --===============8014729579097620413==-- From mboxrd@z Thu Jan 1 00:00:00 1970 From: Rusty Russell Subject: Re: [LKP] [PATCH] drm/radeon: Try to init amdkfd only if 64 bit kernel Date: Wed, 07 Jan 2015 09:28:31 +1030 Message-ID: <87mw5v4iug.fsf@rustcorp.com.au> References: <1419246673-7222-1-git-send-email-oded.gabbay@amd.com> <20141222184940.GB17192@tassilo.jf.intel.com> <20141222190041.GC17192@tassilo.jf.intel.com> <54986EA2.106@amd.com> <874msmt1hi.fsf@rustcorp.com.au> <549A85DE.1040006@amd.com> <87r3v96ebd.fsf@rustcorp.com.au> Mime-Version: 1.0 Content-Type: text/plain Return-path: In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org To: Kees Cook Cc: Oded Gabbay , Andi Kleen , Alex Deucher , "H. Peter Anvin" , Dana Elifaz , LKML , Maling list - DRI developers , Alexander Deucher , LKP ML List-Id: dri-devel@lists.freedesktop.org Kees Cook writes: > On Sun, Jan 4, 2015 at 8:28 PM, Rusty Russell wrote: >> Kees, as far as I can tell you need another 0-terminated vmlinux.relocs >> section for weak symbols. These should not be relocated if already 0. > > A few questions: > > Why doesn't this break on 32-bit without kASLR? 32-bit does relocation > by default, even without CONFIG_RANDOMIZE_BASE. Well, the offset was 0 until I removed CONFIG_HIBERNATE. > Are there any symbols that are NULL that aren't weak? I'd expect all > strong symbols to have non-zero offsets, but I must be > misunderstanding something here. I don't think there would be. Anyway, you might be able to filter them out in x86/tools/relocs itself. Cheers, Rusty.