From mboxrd@z Thu Jan 1 00:00:00 1970 From: Willem Jan Withagen Subject: Re: Has rgw recently switched to deque usuage?? Date: Sun, 21 Feb 2016 01:49:05 +0100 Message-ID: <56C90981.9020906@digiware.nl> References: <56C87B29.5040005@digiware.nl> <1565267399.44124691.1456003985235.JavaMail.zimbra@redhat.com> <56C8E590.7090609@digiware.nl> <1135436239.44164632.1456013615024.JavaMail.zimbra@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Return-path: Received: from smtp.digiware.nl ([31.223.170.169]:13710 "EHLO smtp.digiware.nl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752002AbcBUCAg (ORCPT ); Sat, 20 Feb 2016 21:00:36 -0500 In-Reply-To: <1135436239.44164632.1456013615024.JavaMail.zimbra@redhat.com> Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Matt Benjamin Cc: Ceph Development On 21-2-2016 01:13, Matt Benjamin wrote: > Hi Willem, > > That's good news, but massive header re-ordering sounds suspicious. Likely, there are just missing unix headers. Well perhaps massive is the wrong word. And Clang is a lot more picky than GCC.... Major problem is that compiling rgw modules complains with just about all modules: In file included from test/rgw/test_rgw_period_history.cc:14: ./rgw/rgw_period_history.h:24:16: error: use of undeclared identifier 'RGWPeriod' std::deque periods; ^ Which is relieved by including: #include "rgw_rados.h" #include "rgw_period_puller.h" #include "rgw_period_history.h" #include "rgw_metadata.h" In all complaining modules. And needed to fix including ordering, where std::queue was called on things not yet fully defined. same for the modules in src/test.... --WjW > > Matt > > ----- Original Message ----- >> From: "Willem Jan Withagen" >> To: "Matt Benjamin" >> Cc: "Ceph Development" >> Sent: Saturday, February 20, 2016 5:15:44 PM >> Subject: Re: Has rgw recently switched to deque usuage?? >> >> On 20-2-2016 22:33, Matt Benjamin wrote: >>> Hi Willem, >>> >>> A number of new features have merged, sure, and they make use of new >>> primitives. I haven't compiled with clang recently, mainly due to >>> issues linking with gcc-compiled leveldb and rocksdb on Linux >>> platforms. For now, maybe you can send more specific info. >>> >>> (I'll be able to try-build on FreeBSD at some point, probably, but >>> not in the next day or so due to travel.) >> >> Hi Matt, >> >> It turns out to be all about the order of includes and that Clang not >> always sees eye to eye with GCC. >> Right now I'm sort of reordering a lot includes rados/rgw headers in the >> rgw/*cc files. >> Probably too much reordering but first I want it to get it to compile. >> :) After that lets see what was really needed. >> >> Save travels, >> --WjW >> >>> Matt >>> >>> ----- Original Message ----- >>>> From: "Willem Jan Withagen" To: "Ceph >>>> Development" Sent: Saturday, February >>>> 20, 2016 9:41:45 AM Subject: Has rgw recently switched to deque >>>> usuage?? >>>> >>>> Hi, >>>> >>>> With the last rebase I'm getting compilation errors on rgw stuff. >>>> So I was wondering if deque usuage got included recently, or that >>>> I'm running into a case where Clang doesn't like the way some code >>>> is used. And GCC just compiles fine... >>>> >>>> Any answer would be appriciated >>>> >>>> --WjW -- To unsubscribe from this list: send the line "unsubscribe >>>> ceph-devel" in the body of a message to majordomo@vger.kernel.org >>>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>>> >>> >> >> >