From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:41961) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1gC7iQ-0005QT-Pe for qemu-devel@nongnu.org; Mon, 15 Oct 2018 14:35:43 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1gC7iO-0002GY-00 for qemu-devel@nongnu.org; Mon, 15 Oct 2018 14:35:42 -0400 Received: from mail-pf1-x444.google.com ([2607:f8b0:4864:20::444]:37349) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1gC7iM-0002FV-LE for qemu-devel@nongnu.org; Mon, 15 Oct 2018 14:35:39 -0400 Received: by mail-pf1-x444.google.com with SMTP id j23-v6so10115716pfi.4 for ; Mon, 15 Oct 2018 11:35:37 -0700 (PDT) Sender: Richard Henderson References: <20181005154910.3099-1-alex.bennee@linaro.org> <20181005154910.3099-9-alex.bennee@linaro.org> <07e7c534-c9b4-4d07-6f1a-797b2904d9e4@linaro.org> <87ftx73whg.fsf@linaro.org> From: Richard Henderson Message-ID: <974684e3-afa7-2834-d26c-a6edabb570cb@twiddle.net> Date: Mon, 15 Oct 2018 11:35:32 -0700 MIME-Version: 1.0 In-Reply-To: <87ftx73whg.fsf@linaro.org> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 8bit Subject: Re: [Qemu-devel] [RFC PATCH 08/21] accel/tcg/cputlb: convert remaining tlb_debug() to trace events List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: =?UTF-8?Q?Alex_Benn=c3=a9e?= , Richard Henderson Cc: qemu-devel@nongnu.org, Peter Crosthwaite , cota@braap.org, Pavel.Dovgaluk@ispras.ru, Paolo Bonzini , vilanova@ac.upc.edu On 10/15/18 11:17 AM, Alex Bennée wrote: > > Richard Henderson writes: > >> On 10/5/18 8:48 AM, Alex Bennée wrote: >>> To avoid overly long lines in trace events I've split the event into >>> tlb_set_page and tlb_set_page_attrs. >> >> I don't think this split is a good idea. It allows data from different threads >> to be interleaved. > > Won't we soon have a lock serialising this? For each individual trace entry, surely. If you split the trace in two, you leave a race in the middle. > It was mainly to shut-up checkpatch because the trace-events don't seem > to do multi-line. I would just ignore that. r~