From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Date: Mon, 27 Oct 2008 16:29:24 +0000 Subject: Re: [PATCH] kvm: external module: fix unifdef problem Message-Id: <4905EC64.7090603@redhat.com> List-Id: References: <1224871854.9634.50.camel@localhost.localdomain> In-Reply-To: <1224871854.9634.50.camel@localhost.localdomain> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: kvm-ppc@vger.kernel.org Christian Ehrhardt wrote: > Hollis and I discussed how to continue on that. > Atm the upstream code is broken for powerpc and your response is too > vague for me to extend our patch in some way. > So which way should we go? Will you apply (or should I resubmit ?) > Hollis patch for now to fix upstream for powerpc. And we/you extend it > later or what else would you prefer? My thinking is that we convert CONFIG_* to __i386__, __x86_64__, __powerpc__, and the like, using the hack-modules pass in 'make sync'. The advantages of that are: - 'make sync' headers are architecture independent. Currently once you run 'make sync', the headers become tied to the arch, even the ones in include/linux - The headers are standalone -- no need for -DCONFIG_* in anything that includes it - We already mangle things, so it's not a big change I think Xiantao was going to try it, but if not, then please do it yourself. I could too, of course, but I prefer someone that can test on non-x86 do it. -- error compiling committee.c: too many arguments to function