From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Fri, 28 Feb 2003 16:40:19 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Fri, 28 Feb 2003 16:39:49 -0500 Received: from e32.co.us.ibm.com ([32.97.110.130]:1157 "EHLO e32.co.us.ibm.com") by vger.kernel.org with ESMTP id ; Fri, 28 Feb 2003 16:38:15 -0500 Date: Fri, 28 Feb 2003 13:38:51 -0800 From: "Martin J. Bligh" To: Sam Ravnborg cc: linux-kernel Subject: Re: [Bug 423] New: make -j X bzImage gives a warning Message-ID: <362820000.1046468331@flay> In-Reply-To: <20030228212504.GA21843@mars.ravnborg.org> References: <347860000.1046465385@flay> <20030228212504.GA21843@mars.ravnborg.org> X-Mailer: Mulberry/2.1.2 (Linux/x86) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > On Fri, Feb 28, 2003 at 12:49:45PM -0800, Martin J. Bligh wrote: >> http://bugme.osdl.org/show_bug.cgi?id=423 >> >> Summary: make -j X bzImage gives a warning >> Kernel Version: 2.5.63 >> Status: NEW >> Severity: low >> Owner: zippel@linux-m68k.org > > Feel free to bug me with kbuild related issues. > In this area Roman is 'only' taking care of kconfig & related issues AFAIK. > >> make -j X bzImage gives a warning: >> >> make[1]: warning: jobserver unavailable: using -j1. Add `+' to parent make >> rule. >> >> Can we get rid of this one way or the other? > > I have tried before - no luck. > This one happens due to a $(Q)$(MAKE) used as part of a $(if > construct in the top-level Makefile. > See around line 335 - 345. > It requires more than trivial changes to get rid of this one. This bit? define rule_vmlinux__ set -e $(if $(filter .tmp_kallsyms%,$^),, echo ' Generating build number' . $(src)/scripts/mkversion > .tmp_version mv -f .tmp_version .version $(Q)$(MAKE) $(build)=init ) $(call cmd,vmlinux__) echo 'cmd_$@ := $(cmd_vmlinux__)' > $(@D)/.$(@F).cmd endef Is it possible to define $(SINGLE_MAKE) or something as well as $(MAKE), where it strips the -j, and use that in the instances where it can't take a -j? Would be nice to get rid of that one last warning ... ;-) M.