From mboxrd@z Thu Jan 1 00:00:00 1970 From: Gabriel C Subject: Re: linux-next: UML build error Date: Wed, 21 May 2008 22:20:39 +0200 Message-ID: <48348417.9090103@googlemail.com> References: <48345B31.50904@googlemail.com> <20080521200515.GB12373@c2.user-mode-linux.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Return-path: Received: from gv-out-0910.google.com ([216.239.58.188]:18710 "EHLO gv-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758518AbYEUUU5 (ORCPT ); Wed, 21 May 2008 16:20:57 -0400 Received: by gv-out-0910.google.com with SMTP id e6so913380gvc.37 for ; Wed, 21 May 2008 13:20:55 -0700 (PDT) In-Reply-To: <20080521200515.GB12373@c2.user-mode-linux.org> Sender: linux-next-owner@vger.kernel.org List-ID: To: Jeff Dike Cc: Linux Kernel list , linux-next@vger.kernel.org, user-mode-linux-devel@lists.sourceforge.net Jeff Dike wrote: > On Wed, May 21, 2008 at 07:26:09PM +0200, Gabriel C wrote: >> CC arch/um/sys-x86_64/ksyms.o >> arch/um/sys-x86_64/ksyms.c:5: error: '__memcpy' undeclared here (not in a function) >> arch/um/sys-x86_64/ksyms.c:5: warning: type defaults to 'int' in declaration of '__memcpy' > > Does the patch below help? There's gcc-version-dependent logic in > include/asm-x86/string_64.h which decides whether to declare memcpy or > __memcpy and it looks like you're hitting that. Yes is fixes this error , however I get now another one but this one seems to be an compiler bug. ... CC arch/um/sys-x86_64/um_module.o LD arch/um/sys-x86_64/subarch.o LD arch/um/sys-x86_64/built-in.o LD arch/x86/crypto/built-in.o CC kernel/sched.o In file included from kernel/sched.c:1842: kernel/sched_idletask.c: In function 'check_preempt_curr_idle': kernel/sched_idletask.c:20: internal compiler error: in cgraph_estimate_size_after_inlining, at ipa-inline.c:188 ... This is gcc 4.3.0 ( with backports of some known fixed bugs ). I try to build latest snapshot of gcc 4.3* later on today and try with this one. > > Jeff > Gabriel