From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ingo Molnar Subject: Re: linux-next: tracing tree build failure Date: Tue, 10 Mar 2009 10:10:10 +0100 Message-ID: <20090310091010.GD3097@elte.hu> References: <20090310115509.328da465.sfr@canb.auug.org.au> <49B5C733.6080005@kernel.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mx3.mail.elte.hu ([157.181.1.138]:33776 "EHLO mx3.mail.elte.hu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753399AbZCJJK1 (ORCPT ); Tue, 10 Mar 2009 05:10:27 -0400 Content-Disposition: inline In-Reply-To: <49B5C733.6080005@kernel.org> Sender: linux-next-owner@vger.kernel.org List-ID: To: Tejun Heo Cc: Stephen Rothwell , Thomas Gleixner , "H. Peter Anvin" , linux-next@vger.kernel.org, KOSAKI Motohiro * Tejun Heo wrote: > Stephen Rothwell wrote: > > Hi all, > > > > Today's linux-next build (x86_64 allmodconfig) failed like this: > > > > kernel/trace/trace_functions_graph.c: In function 'graph_trace_close': > > kernel/trace/trace_functions_graph.c:836: error: implicit declaration of function 'percpu_free' > > > > The direct cause is commit 422d3c7a577b15e1384c9d4e72a9540896b685fa > > ("tracing: current tip/master can't enable ftrace") from the tracing tree > > which exposed an interaction between commit > > f2a8205c4ef1af917d175c36a4097ae5587791c8 ("percpu: kill percpu_alloc() > > and friends") from the tip-core tree and commit > > 9005f3ebebfcfe9ccd731d16c468907a35ac1f9a ("tracing/function-graph-tracer: > > various fixes and features") from the tracing tree. > > > > I have reverted commit 422d3c7a577b15e1384c9d4e72a9540896b685fa for > > today. (As a side note, that commit has no Signed-off-by ...) > > Just in case someone doesn't know yet. The remedy is using > free_percpu() instead of percpu_free(). That's exactly how it was resolved more than two weeks ago in tip:master - and has been resolved in tip:master since then. I'm glad to hear about new bugs, but the amount of false positives from linux-next is wasting a lot of time - the duplication rate is above 90% which is way too much. tip:master is a full, well-tested integration of all those branches. It would be nice if Stephen investigated tip:master before reporting such solved-long-ago interactions. Ingo