From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail.kernel.org ([198.145.29.99]:47840 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1729589AbeKIIBI (ORCPT ); Fri, 9 Nov 2018 03:01:08 -0500 Date: Thu, 8 Nov 2018 17:23:30 -0500 From: Steven Rostedt To: Tzvetomir Stoyanov Cc: "linux-trace-devel@vger.kernel.org" Subject: Re: tep_*host_bigendian() Message-ID: <20181108172330.2182459e@gandalf.local.home> In-Reply-To: References: <20181108133053.6eb22596@gandalf.local.home> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-trace-devel-owner@vger.kernel.org List-ID: On Thu, 8 Nov 2018 19:40:32 +0000 Tzvetomir Stoyanov wrote: > Hi Steven, > Ok, I''ll rename tep_is_file_bigendian() to tep_file_bigendian(). > But about tep_is_host_bigendian() - I think it makes sense to remove completely > tep_is_host_bigendian() / tep_set_host_bigendian() APIs, what do you think ? > The tep_set_host_bigendian() API is used only in perf code, to set the > order in tep handler - and it uses tep_host_bigendian() to get the > local host byte order. > I think it is safe to the remove tep_set_host_bigendian() API and > replace tep_is_host_bigendian() with tep_host_bigendian(). I'm fine with this, as long as we don't calculate the endianess each time. We still need the tep->host_bigendian setting. But we can set it when we allocate a tep handle. -- Steve > > On Thu, Nov 8, 2018 at 8:30 PM Steven Rostedt wrote: > > > > Hi Tzvetomir, > > > > I was looking at the code for tep_is_file_bigendian(), > > tep_is_host_bigendian() and tep_host_bigendian(). > > > > The tep_host_bigendian() checks if the current machine that is > > executing is bigendian or not. > > > > tep_is_host_bigendian() looks at the pevent to see what it has, which > > should be the same as tep_host_bigendian(). > > > > tep_is_file_bigendian() is if the raw file format is bigendian or not. > > If the file does not mach the host, byte swapping needs to be made. > > > > Looking at this deeper, I think the names are backwards. > > > > Let's rename tep_host_bigendian() to tep_is_host_bigendian() > > rename tep_is_host_bigendian() to tep_host_bigendian() > > rename tep_is_file_bigendian() to tep_file_bigendian() > > > > > > The "is_host" to me sounds more like we are testing the actual host, > > where as just saying 'tep_host_bigendian' sounds to me that it's what > > the pevent has the host set as. > > > > -- Steve > > >