From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 71816 invoked by alias); 9 Nov 2016 23:52:16 -0000 Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: libc-alpha-owner@sourceware.org Received: (qmail 71804 invoked by uid 89); 9 Nov 2016 23:52:15 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-1.9 required=5.0 tests=AWL,BAYES_00,RCVD_IN_DNSWL_NONE,SPF_PASS,URIBL_RED autolearn=ham version=3.3.2 spammy=mathfiniteh, UD:math-finite.h, math-finite.h, involves X-HELO: relay1.mentorg.com Date: Wed, 09 Nov 2016 23:52:00 -0000 From: Joseph Myers To: "Gabriel F. T. Gomes" CC: Subject: Re: [PATCH 0/8] More float128 declarations In-Reply-To: Message-ID: References: <1478716859-3246-1-git-send-email-gftg@linux.vnet.ibm.com> User-Agent: Alpine 2.20 (DEB 67 2015-01-07) MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" X-ClientProxiedBy: svr-ies-mbx-01.mgc.mentorg.com (139.181.222.1) To svr-ies-mbx-01.mgc.mentorg.com (139.181.222.1) X-SW-Source: 2016-11/txt/msg00370.txt.bz2 On Wed, 9 Nov 2016, Joseph Myers wrote: > This seems inconsistent about whether the recently added TS 18661-1 > functions are covered or not. I'd expect all of the following to have > version map entries: __iseqsigf128 totalorderf128 totalordermagf128 > getpayloadf128 canonicalizef128, and, where the definitions aren't > type-generic, to have function definitions for float128. Also, this inconsistency suggests to me that the patches haven't been testing in conjunction with subsequent patches that actually enable the new functions and their associated tests, because the tests would have failed to link if some functions were absent. Anything that involves adding new files / declarations / macros for each new function should probably best have its final review as part of the patch series that actually enables the new functions and associated tests for at least one system, as otherwise it's hard to have confidence in the completeness and correctness of the changes. (That is, while it's possible to spot some issues without such testing, final approval of such patches should wait until the code for the new types has gone through testing. Whereas e.g. a conversion of math-finite.h to use macros to declare functions could be done in a way involving no float128 conditionals at all, so entirely suitable to review by itself, with the float128 patches then just adapting the macro definitions to handle float128.) -- Joseph S. Myers joseph@codesourcery.com