From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pf0-f193.google.com ([209.85.192.193]:36808 "EHLO mail-pf0-f193.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750908AbdE3Dzt (ORCPT ); Mon, 29 May 2017 23:55:49 -0400 Date: Tue, 30 May 2017 13:55:36 +1000 From: Nicholas Piggin Subject: Re: [PATCH] kbuild: provide THIN_ARCHIVES option for all architectures Message-ID: <20170530135536.4d6a5f74@roar.ozlabs.ibm.com> In-Reply-To: References: <20170529081103.29999-1-npiggin@gmail.com> <20170530091343.1f98b24b@roar.ozlabs.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kbuild-owner@vger.kernel.org List-ID: To: Linus Torvalds Cc: "linux-arch@vger.kernel.org" , Linux Kbuild mailing list , Stephen Rothwell On Mon, 29 May 2017 20:22:41 -0700 Linus Torvalds wrote: > On Mon, May 29, 2017 at 4:13 PM, Nicholas Piggin wrote: > > > > *thin* archives does require binutils 2.19. That's nearly 10 years > > old, but we advertise 2.12 minimum at the moment. > > I think we're ok with upgrading that. We've been allowing some > ridiculously old tools at times. I think "feature is 10 years old and > actively used by other projects" (Qt uses it from some googling, for > example) is more than sufficient. Well we might have to. I couldn't make thick archives work easily because we have multiple levels of built-in.o, and with thick you get an archive of an archive which the linker does not understand. Fixing that would mean unpacking the archive and repacking it for every built-in.o input you link -- too painful. I suspect Google uses the format internally too because they added the initial support. https://sourceware.org/ml/binutils/2008-03/msg00150.html Support has been pervasively added to binutils -- nm, objdump, ld, etc. I think it falls into the category of well supported. > > If somebody has user space older than that, they presumably aren't > building new kernels either. Yeah I think it's reasonable to drop support for such old toolchain if we have a reason for it. I'll send a patch to 0day and if that passes, I'll see if it's something Stephen would put into linux-next. Thanks, Nick