From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from smtp.duncanthrax.net ([89.31.1.170]:35385 "EHLO smtp.duncanthrax.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726840AbfGARJD (ORCPT ); Mon, 1 Jul 2019 13:09:03 -0400 Date: Mon, 1 Jul 2019 19:08:53 +0200 From: Sven Schnelle Subject: Re: [PATCH RFC] generic ELF support for kexec Message-ID: <20190701170853.GD19243@t470p.stackframe.org> References: <20190625185433.GA10934@t470p.stackframe.org> <87o92isbxg.fsf@concordia.ellerman.id.au> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <87o92isbxg.fsf@concordia.ellerman.id.au> Sender: linux-s390-owner@vger.kernel.org List-ID: To: Michael Ellerman Cc: kexec@lists.infradead.org, linux-s390@vger.kernel.org, deller@gmx.de, linuxppc-dev@lists.ozlabs.org Hi Michael, On Fri, Jun 28, 2019 at 12:04:11PM +1000, Michael Ellerman wrote: > Sven Schnelle writes: > https://github.com/linuxppc/wiki/wiki/Booting-with-Qemu > > But I'm not sure where you get a version of kexec that uses kexec_file(). kexec-tools HEAD supports it, so that's not a problem. > > If that change is acceptable i would finish the patch and submit it. I think > > best would be to push this change through Helge's parisc tree, so we don't > > have any dependencies to sort out. > > That will work for you but could cause us problems if we have any > changes that touch that code. > > It's easy enough to create a topic branch with just that patch that both > of us merge. What should be the base branch for that patch? Christophe suggested the powerpc/merge branch? > > #include > > #include > > #include > > @@ -31,540 +29,6 @@ > > #include > > #include > > > > -#define PURGATORY_STACK_SIZE (16 * 1024) > > This is unused AFAICS. We should probably remove it explicitly rather > than as part of this patch. I have one patch right now. If wanted i can split up all the changes suggested during the review into smaller pieces, whatever you prefer. > Or that. > > > +#include > > +#include > > + > > +#define elf_addr_to_cpu elf64_to_cpu > > Why are we doing that rather than just using elf64_to_cpu directly? > > > +#ifndef Elf_Rel > > +#define Elf_Rel Elf64_Rel > > +#endif /* Elf_Rel */ > > And that? Don't know - ask the PPC people :-) Regards Sven